{
  "best_metric": 0.03486395254731178,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold5/checkpoint-1950",
  "epoch": 4.9971563981042655,
  "eval_steps": 50,
  "global_step": 3295,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015165876777251185,
      "grad_norm": 5.7730183601379395,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 1.2893,
      "step": 10
    },
    {
      "epoch": 0.03033175355450237,
      "grad_norm": 5.499680042266846,
      "learning_rate": 6.060606060606061e-06,
      "loss": 1.2625,
      "step": 20
    },
    {
      "epoch": 0.04549763033175355,
      "grad_norm": 4.859879493713379,
      "learning_rate": 9.090909090909091e-06,
      "loss": 1.1694,
      "step": 30
    },
    {
      "epoch": 0.06066350710900474,
      "grad_norm": 2.21482253074646,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.7344,
      "step": 40
    },
    {
      "epoch": 0.07582938388625593,
      "grad_norm": 1.7206193208694458,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.5262,
      "step": 50
    },
    {
      "epoch": 0.07582938388625593,
      "eval_loss": 0.3221642076969147,
      "eval_runtime": 156.8442,
      "eval_samples_per_second": 7.479,
      "eval_steps_per_second": 7.479,
      "step": 50
    },
    {
      "epoch": 0.0909952606635071,
      "grad_norm": 1.7212742567062378,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.2913,
      "step": 60
    },
    {
      "epoch": 0.1061611374407583,
      "grad_norm": 1.2331215143203735,
      "learning_rate": 2.1212121212121215e-05,
      "loss": 0.2245,
      "step": 70
    },
    {
      "epoch": 0.12132701421800948,
      "grad_norm": 1.7794642448425293,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.1663,
      "step": 80
    },
    {
      "epoch": 0.13649289099526066,
      "grad_norm": 1.1978228092193604,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.1061,
      "step": 90
    },
    {
      "epoch": 0.15165876777251186,
      "grad_norm": 1.6290844678878784,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.0888,
      "step": 100
    },
    {
      "epoch": 0.15165876777251186,
      "eval_loss": 0.08710210025310516,
      "eval_runtime": 156.9155,
      "eval_samples_per_second": 7.475,
      "eval_steps_per_second": 7.475,
      "step": 100
    },
    {
      "epoch": 0.16682464454976303,
      "grad_norm": 0.869321882724762,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.108,
      "step": 110
    },
    {
      "epoch": 0.1819905213270142,
      "grad_norm": 1.4029194116592407,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.0797,
      "step": 120
    },
    {
      "epoch": 0.1971563981042654,
      "grad_norm": 1.8435430526733398,
      "learning_rate": 3.939393939393939e-05,
      "loss": 0.0795,
      "step": 130
    },
    {
      "epoch": 0.2123222748815166,
      "grad_norm": 1.3845666646957397,
      "learning_rate": 4.242424242424243e-05,
      "loss": 0.0671,
      "step": 140
    },
    {
      "epoch": 0.22748815165876776,
      "grad_norm": 1.9482444524765015,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.0821,
      "step": 150
    },
    {
      "epoch": 0.22748815165876776,
      "eval_loss": 0.08279303461313248,
      "eval_runtime": 157.0629,
      "eval_samples_per_second": 7.468,
      "eval_steps_per_second": 7.468,
      "step": 150
    },
    {
      "epoch": 0.24265402843601896,
      "grad_norm": 1.3024379014968872,
      "learning_rate": 4.848484848484849e-05,
      "loss": 0.0895,
      "step": 160
    },
    {
      "epoch": 0.25781990521327014,
      "grad_norm": 1.7065895795822144,
      "learning_rate": 5.151515151515152e-05,
      "loss": 0.0961,
      "step": 170
    },
    {
      "epoch": 0.2729857819905213,
      "grad_norm": 0.6068471074104309,
      "learning_rate": 5.4545454545454546e-05,
      "loss": 0.0622,
      "step": 180
    },
    {
      "epoch": 0.2881516587677725,
      "grad_norm": 0.9339556694030762,
      "learning_rate": 5.757575757575758e-05,
      "loss": 0.0615,
      "step": 190
    },
    {
      "epoch": 0.3033175355450237,
      "grad_norm": 0.84297114610672,
      "learning_rate": 6.060606060606061e-05,
      "loss": 0.0718,
      "step": 200
    },
    {
      "epoch": 0.3033175355450237,
      "eval_loss": 0.06267111003398895,
      "eval_runtime": 157.0961,
      "eval_samples_per_second": 7.467,
      "eval_steps_per_second": 7.467,
      "step": 200
    },
    {
      "epoch": 0.3184834123222749,
      "grad_norm": 0.9764337539672852,
      "learning_rate": 6.363636363636364e-05,
      "loss": 0.0662,
      "step": 210
    },
    {
      "epoch": 0.33364928909952607,
      "grad_norm": 0.5065613389015198,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0817,
      "step": 220
    },
    {
      "epoch": 0.34881516587677724,
      "grad_norm": 1.0672837495803833,
      "learning_rate": 6.96969696969697e-05,
      "loss": 0.0537,
      "step": 230
    },
    {
      "epoch": 0.3639810426540284,
      "grad_norm": 0.5060169696807861,
      "learning_rate": 7.272727272727273e-05,
      "loss": 0.0495,
      "step": 240
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 0.7706537246704102,
      "learning_rate": 7.575757575757576e-05,
      "loss": 0.0622,
      "step": 250
    },
    {
      "epoch": 0.3791469194312796,
      "eval_loss": 0.05433733016252518,
      "eval_runtime": 157.0809,
      "eval_samples_per_second": 7.467,
      "eval_steps_per_second": 7.467,
      "step": 250
    },
    {
      "epoch": 0.3943127962085308,
      "grad_norm": 1.116126537322998,
      "learning_rate": 7.878787878787879e-05,
      "loss": 0.0608,
      "step": 260
    },
    {
      "epoch": 0.409478672985782,
      "grad_norm": 1.0627580881118774,
      "learning_rate": 8.181818181818183e-05,
      "loss": 0.0473,
      "step": 270
    },
    {
      "epoch": 0.4246445497630332,
      "grad_norm": 0.4674009084701538,
      "learning_rate": 8.484848484848486e-05,
      "loss": 0.0543,
      "step": 280
    },
    {
      "epoch": 0.43981042654028435,
      "grad_norm": 1.030065894126892,
      "learning_rate": 8.787878787878789e-05,
      "loss": 0.0684,
      "step": 290
    },
    {
      "epoch": 0.4549763033175355,
      "grad_norm": 0.37593016028404236,
      "learning_rate": 9.090909090909092e-05,
      "loss": 0.0509,
      "step": 300
    },
    {
      "epoch": 0.4549763033175355,
      "eval_loss": 0.05037209019064903,
      "eval_runtime": 157.063,
      "eval_samples_per_second": 7.468,
      "eval_steps_per_second": 7.468,
      "step": 300
    },
    {
      "epoch": 0.47014218009478675,
      "grad_norm": 0.4311511516571045,
      "learning_rate": 9.393939393939395e-05,
      "loss": 0.0548,
      "step": 310
    },
    {
      "epoch": 0.48530805687203793,
      "grad_norm": 0.5414978265762329,
      "learning_rate": 9.696969696969698e-05,
      "loss": 0.044,
      "step": 320
    },
    {
      "epoch": 0.500473933649289,
      "grad_norm": 0.4985443651676178,
      "learning_rate": 0.0001,
      "loss": 0.0589,
      "step": 330
    },
    {
      "epoch": 0.5156398104265403,
      "grad_norm": 0.3067478537559509,
      "learning_rate": 9.999719336268101e-05,
      "loss": 0.0536,
      "step": 340
    },
    {
      "epoch": 0.5308056872037915,
      "grad_norm": 1.05351722240448,
      "learning_rate": 9.998877376581251e-05,
      "loss": 0.0576,
      "step": 350
    },
    {
      "epoch": 0.5308056872037915,
      "eval_loss": 0.05062875524163246,
      "eval_runtime": 157.1043,
      "eval_samples_per_second": 7.466,
      "eval_steps_per_second": 7.466,
      "step": 350
    },
    {
      "epoch": 0.5459715639810426,
      "grad_norm": 0.3739762306213379,
      "learning_rate": 9.997474215462472e-05,
      "loss": 0.0498,
      "step": 360
    },
    {
      "epoch": 0.5611374407582939,
      "grad_norm": 0.4102906286716461,
      "learning_rate": 9.995510010438337e-05,
      "loss": 0.032,
      "step": 370
    },
    {
      "epoch": 0.576303317535545,
      "grad_norm": 0.4799945652484894,
      "learning_rate": 9.992984982021295e-05,
      "loss": 0.0669,
      "step": 380
    },
    {
      "epoch": 0.5914691943127962,
      "grad_norm": 0.6564961671829224,
      "learning_rate": 9.9898994136849e-05,
      "loss": 0.0677,
      "step": 390
    },
    {
      "epoch": 0.6066350710900474,
      "grad_norm": 0.7356056571006775,
      "learning_rate": 9.986253651832005e-05,
      "loss": 0.066,
      "step": 400
    },
    {
      "epoch": 0.6066350710900474,
      "eval_loss": 0.04723864048719406,
      "eval_runtime": 157.0307,
      "eval_samples_per_second": 7.47,
      "eval_steps_per_second": 7.47,
      "step": 400
    },
    {
      "epoch": 0.6218009478672986,
      "grad_norm": 0.2881098985671997,
      "learning_rate": 9.982048105755859e-05,
      "loss": 0.0492,
      "step": 410
    },
    {
      "epoch": 0.6369668246445498,
      "grad_norm": 0.4902149736881256,
      "learning_rate": 9.977283247594166e-05,
      "loss": 0.0419,
      "step": 420
    },
    {
      "epoch": 0.6521327014218009,
      "grad_norm": 0.4223526418209076,
      "learning_rate": 9.971959612276076e-05,
      "loss": 0.062,
      "step": 430
    },
    {
      "epoch": 0.6672985781990521,
      "grad_norm": 0.49113377928733826,
      "learning_rate": 9.966077797462129e-05,
      "loss": 0.0485,
      "step": 440
    },
    {
      "epoch": 0.6824644549763034,
      "grad_norm": 1.1138062477111816,
      "learning_rate": 9.959638463477165e-05,
      "loss": 0.0593,
      "step": 450
    },
    {
      "epoch": 0.6824644549763034,
      "eval_loss": 0.04495476186275482,
      "eval_runtime": 157.1855,
      "eval_samples_per_second": 7.463,
      "eval_steps_per_second": 7.463,
      "step": 450
    },
    {
      "epoch": 0.6976303317535545,
      "grad_norm": 0.36527907848358154,
      "learning_rate": 9.952642333236186e-05,
      "loss": 0.0497,
      "step": 460
    },
    {
      "epoch": 0.7127962085308057,
      "grad_norm": 0.4513853192329407,
      "learning_rate": 9.945090192163202e-05,
      "loss": 0.0488,
      "step": 470
    },
    {
      "epoch": 0.7279620853080568,
      "grad_norm": 0.4938872456550598,
      "learning_rate": 9.936982888103051e-05,
      "loss": 0.0418,
      "step": 480
    },
    {
      "epoch": 0.7431279620853081,
      "grad_norm": 0.40595391392707825,
      "learning_rate": 9.928321331226219e-05,
      "loss": 0.0577,
      "step": 490
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 0.4399415850639343,
      "learning_rate": 9.919106493926655e-05,
      "loss": 0.0316,
      "step": 500
    },
    {
      "epoch": 0.7582938388625592,
      "eval_loss": 0.0444808192551136,
      "eval_runtime": 157.1449,
      "eval_samples_per_second": 7.464,
      "eval_steps_per_second": 7.464,
      "step": 500
    },
    {
      "epoch": 0.7734597156398104,
      "grad_norm": 0.6373604536056519,
      "learning_rate": 9.909339410712612e-05,
      "loss": 0.0495,
      "step": 510
    },
    {
      "epoch": 0.7886255924170616,
      "grad_norm": 0.21484985947608948,
      "learning_rate": 9.8990211780905e-05,
      "loss": 0.0399,
      "step": 520
    },
    {
      "epoch": 0.8037914691943128,
      "grad_norm": 0.6859444379806519,
      "learning_rate": 9.888152954441785e-05,
      "loss": 0.0545,
      "step": 530
    },
    {
      "epoch": 0.818957345971564,
      "grad_norm": 0.3476887345314026,
      "learning_rate": 9.876735959892953e-05,
      "loss": 0.0465,
      "step": 540
    },
    {
      "epoch": 0.8341232227488151,
      "grad_norm": 0.1995847523212433,
      "learning_rate": 9.864771476178522e-05,
      "loss": 0.0318,
      "step": 550
    },
    {
      "epoch": 0.8341232227488151,
      "eval_loss": 0.047177962958812714,
      "eval_runtime": 157.1722,
      "eval_samples_per_second": 7.463,
      "eval_steps_per_second": 7.463,
      "step": 550
    },
    {
      "epoch": 0.8492890995260663,
      "grad_norm": 0.33388960361480713,
      "learning_rate": 9.852260846497153e-05,
      "loss": 0.0421,
      "step": 560
    },
    {
      "epoch": 0.8644549763033176,
      "grad_norm": 0.19013848900794983,
      "learning_rate": 9.839205475360851e-05,
      "loss": 0.044,
      "step": 570
    },
    {
      "epoch": 0.8796208530805687,
      "grad_norm": 0.5361432433128357,
      "learning_rate": 9.825606828437291e-05,
      "loss": 0.0501,
      "step": 580
    },
    {
      "epoch": 0.8947867298578199,
      "grad_norm": 0.2981376051902771,
      "learning_rate": 9.811466432385267e-05,
      "loss": 0.0535,
      "step": 590
    },
    {
      "epoch": 0.909952606635071,
      "grad_norm": 0.31023114919662476,
      "learning_rate": 9.796785874683314e-05,
      "loss": 0.0436,
      "step": 600
    },
    {
      "epoch": 0.909952606635071,
      "eval_loss": 0.044910140335559845,
      "eval_runtime": 157.0856,
      "eval_samples_per_second": 7.467,
      "eval_steps_per_second": 7.467,
      "step": 600
    },
    {
      "epoch": 0.9251184834123223,
      "grad_norm": 0.19991198182106018,
      "learning_rate": 9.781566803451475e-05,
      "loss": 0.0385,
      "step": 610
    },
    {
      "epoch": 0.9402843601895735,
      "grad_norm": 0.3173238933086395,
      "learning_rate": 9.765810927266281e-05,
      "loss": 0.0326,
      "step": 620
    },
    {
      "epoch": 0.9554502369668246,
      "grad_norm": 0.46283426880836487,
      "learning_rate": 9.749520014968934e-05,
      "loss": 0.0468,
      "step": 630
    },
    {
      "epoch": 0.9706161137440759,
      "grad_norm": 0.1699405312538147,
      "learning_rate": 9.732695895466735e-05,
      "loss": 0.0454,
      "step": 640
    },
    {
      "epoch": 0.985781990521327,
      "grad_norm": 0.2577606737613678,
      "learning_rate": 9.715340457527746e-05,
      "loss": 0.0454,
      "step": 650
    },
    {
      "epoch": 0.985781990521327,
      "eval_loss": 0.041555456817150116,
      "eval_runtime": 157.0425,
      "eval_samples_per_second": 7.469,
      "eval_steps_per_second": 7.469,
      "step": 650
    },
    {
      "epoch": 1.000947867298578,
      "grad_norm": 0.3391583561897278,
      "learning_rate": 9.697455649568761e-05,
      "loss": 0.0432,
      "step": 660
    },
    {
      "epoch": 1.0161137440758294,
      "grad_norm": 0.37824738025665283,
      "learning_rate": 9.679043479436556e-05,
      "loss": 0.0379,
      "step": 670
    },
    {
      "epoch": 1.0312796208530806,
      "grad_norm": 0.3136765956878662,
      "learning_rate": 9.660106014182489e-05,
      "loss": 0.0368,
      "step": 680
    },
    {
      "epoch": 1.0464454976303317,
      "grad_norm": 0.3872745931148529,
      "learning_rate": 9.640645379830424e-05,
      "loss": 0.037,
      "step": 690
    },
    {
      "epoch": 1.061611374407583,
      "grad_norm": 0.20634222030639648,
      "learning_rate": 9.620663761138067e-05,
      "loss": 0.0339,
      "step": 700
    },
    {
      "epoch": 1.061611374407583,
      "eval_loss": 0.04243304207921028,
      "eval_runtime": 157.0864,
      "eval_samples_per_second": 7.467,
      "eval_steps_per_second": 7.467,
      "step": 700
    },
    {
      "epoch": 1.0767772511848341,
      "grad_norm": 0.14541089534759521,
      "learning_rate": 9.600163401351688e-05,
      "loss": 0.0305,
      "step": 710
    },
    {
      "epoch": 1.0919431279620853,
      "grad_norm": 0.3501947820186615,
      "learning_rate": 9.579146601954276e-05,
      "loss": 0.0448,
      "step": 720
    },
    {
      "epoch": 1.1071090047393364,
      "grad_norm": 0.5332701802253723,
      "learning_rate": 9.557615722407177e-05,
      "loss": 0.0392,
      "step": 730
    },
    {
      "epoch": 1.1222748815165877,
      "grad_norm": 0.33177003264427185,
      "learning_rate": 9.535573179885191e-05,
      "loss": 0.0412,
      "step": 740
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 0.7103090286254883,
      "learning_rate": 9.513021449005214e-05,
      "loss": 0.0369,
      "step": 750
    },
    {
      "epoch": 1.1374407582938388,
      "eval_loss": 0.044771771878004074,
      "eval_runtime": 157.0661,
      "eval_samples_per_second": 7.468,
      "eval_steps_per_second": 7.468,
      "step": 750
    },
    {
      "epoch": 1.15260663507109,
      "grad_norm": 0.21694399416446686,
      "learning_rate": 9.489963061548428e-05,
      "loss": 0.0334,
      "step": 760
    },
    {
      "epoch": 1.1677725118483413,
      "grad_norm": 0.23920297622680664,
      "learning_rate": 9.466400606176062e-05,
      "loss": 0.035,
      "step": 770
    },
    {
      "epoch": 1.1829383886255924,
      "grad_norm": 0.3608671724796295,
      "learning_rate": 9.442336728138779e-05,
      "loss": 0.0408,
      "step": 780
    },
    {
      "epoch": 1.1981042654028435,
      "grad_norm": 0.40870538353919983,
      "learning_rate": 9.417774128979706e-05,
      "loss": 0.0399,
      "step": 790
    },
    {
      "epoch": 1.2132701421800949,
      "grad_norm": 0.24113492667675018,
      "learning_rate": 9.39271556623114e-05,
      "loss": 0.034,
      "step": 800
    },
    {
      "epoch": 1.2132701421800949,
      "eval_loss": 0.03740939870476723,
      "eval_runtime": 157.0498,
      "eval_samples_per_second": 7.469,
      "eval_steps_per_second": 7.469,
      "step": 800
    },
    {
      "epoch": 1.228436018957346,
      "grad_norm": 0.44084838032722473,
      "learning_rate": 9.367163853104975e-05,
      "loss": 0.0331,
      "step": 810
    },
    {
      "epoch": 1.2436018957345971,
      "grad_norm": 0.32109037041664124,
      "learning_rate": 9.341121858176876e-05,
      "loss": 0.038,
      "step": 820
    },
    {
      "epoch": 1.2587677725118485,
      "grad_norm": 0.5154683589935303,
      "learning_rate": 9.314592505064239e-05,
      "loss": 0.0327,
      "step": 830
    },
    {
      "epoch": 1.2739336492890996,
      "grad_norm": 0.27967405319213867,
      "learning_rate": 9.28757877209796e-05,
      "loss": 0.033,
      "step": 840
    },
    {
      "epoch": 1.2890995260663507,
      "grad_norm": 0.48304447531700134,
      "learning_rate": 9.260083691988084e-05,
      "loss": 0.0332,
      "step": 850
    },
    {
      "epoch": 1.2890995260663507,
      "eval_loss": 0.0419432632625103,
      "eval_runtime": 156.962,
      "eval_samples_per_second": 7.473,
      "eval_steps_per_second": 7.473,
      "step": 850
    },
    {
      "epoch": 1.3042654028436018,
      "grad_norm": 0.5451303720474243,
      "learning_rate": 9.232110351483327e-05,
      "loss": 0.0496,
      "step": 860
    },
    {
      "epoch": 1.3194312796208532,
      "grad_norm": 0.3698858916759491,
      "learning_rate": 9.203661891024547e-05,
      "loss": 0.0361,
      "step": 870
    },
    {
      "epoch": 1.3345971563981043,
      "grad_norm": 0.23554925620555878,
      "learning_rate": 9.174741504392173e-05,
      "loss": 0.0308,
      "step": 880
    },
    {
      "epoch": 1.3497630331753554,
      "grad_norm": 0.27226176857948303,
      "learning_rate": 9.145352438347662e-05,
      "loss": 0.0278,
      "step": 890
    },
    {
      "epoch": 1.3649289099526065,
      "grad_norm": 0.5869334936141968,
      "learning_rate": 9.115497992268995e-05,
      "loss": 0.029,
      "step": 900
    },
    {
      "epoch": 1.3649289099526065,
      "eval_loss": 0.04283256083726883,
      "eval_runtime": 156.9509,
      "eval_samples_per_second": 7.474,
      "eval_steps_per_second": 7.474,
      "step": 900
    },
    {
      "epoch": 1.3800947867298579,
      "grad_norm": 0.338230699300766,
      "learning_rate": 9.085181517780273e-05,
      "loss": 0.0368,
      "step": 910
    },
    {
      "epoch": 1.395260663507109,
      "grad_norm": 0.38272976875305176,
      "learning_rate": 9.054406418375443e-05,
      "loss": 0.0403,
      "step": 920
    },
    {
      "epoch": 1.41042654028436,
      "grad_norm": 0.3138972520828247,
      "learning_rate": 9.023176149036203e-05,
      "loss": 0.0338,
      "step": 930
    },
    {
      "epoch": 1.4255924170616114,
      "grad_norm": 0.34866902232170105,
      "learning_rate": 8.991494215844132e-05,
      "loss": 0.0366,
      "step": 940
    },
    {
      "epoch": 1.4407582938388626,
      "grad_norm": 0.13520507514476776,
      "learning_rate": 8.959364175587069e-05,
      "loss": 0.031,
      "step": 950
    },
    {
      "epoch": 1.4407582938388626,
      "eval_loss": 0.043241605162620544,
      "eval_runtime": 157.0199,
      "eval_samples_per_second": 7.47,
      "eval_steps_per_second": 7.47,
      "step": 950
    },
    {
      "epoch": 1.4559241706161137,
      "grad_norm": 0.25074341893196106,
      "learning_rate": 8.926789635359817e-05,
      "loss": 0.0387,
      "step": 960
    },
    {
      "epoch": 1.471090047393365,
      "grad_norm": 0.29269757866859436,
      "learning_rate": 8.893774252159187e-05,
      "loss": 0.0346,
      "step": 970
    },
    {
      "epoch": 1.4862559241706161,
      "grad_norm": 0.5580613613128662,
      "learning_rate": 8.860321732473439e-05,
      "loss": 0.0339,
      "step": 980
    },
    {
      "epoch": 1.5014218009478673,
      "grad_norm": 0.4414357841014862,
      "learning_rate": 8.826435831866184e-05,
      "loss": 0.0344,
      "step": 990
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 0.16885732114315033,
      "learning_rate": 8.79212035455475e-05,
      "loss": 0.0353,
      "step": 1000
    },
    {
      "epoch": 1.5165876777251186,
      "eval_loss": 0.038838956505060196,
      "eval_runtime": 156.3851,
      "eval_samples_per_second": 7.501,
      "eval_steps_per_second": 7.501,
      "step": 1000
    },
    {
      "epoch": 1.5317535545023697,
      "grad_norm": 0.24777741730213165,
      "learning_rate": 8.757379152983103e-05,
      "loss": 0.0295,
      "step": 1010
    },
    {
      "epoch": 1.5469194312796208,
      "grad_norm": 0.3332612216472626,
      "learning_rate": 8.722216127389363e-05,
      "loss": 0.0209,
      "step": 1020
    },
    {
      "epoch": 1.5620853080568722,
      "grad_norm": 0.21875545382499695,
      "learning_rate": 8.686635225367919e-05,
      "loss": 0.0251,
      "step": 1030
    },
    {
      "epoch": 1.577251184834123,
      "grad_norm": 0.2943437099456787,
      "learning_rate": 8.650640441426274e-05,
      "loss": 0.0419,
      "step": 1040
    },
    {
      "epoch": 1.5924170616113744,
      "grad_norm": 0.12683098018169403,
      "learning_rate": 8.614235816536582e-05,
      "loss": 0.0278,
      "step": 1050
    },
    {
      "epoch": 1.5924170616113744,
      "eval_loss": 0.03803436458110809,
      "eval_runtime": 156.3868,
      "eval_samples_per_second": 7.501,
      "eval_steps_per_second": 7.501,
      "step": 1050
    },
    {
      "epoch": 1.6075829383886255,
      "grad_norm": 0.16973666846752167,
      "learning_rate": 8.577425437681994e-05,
      "loss": 0.0338,
      "step": 1060
    },
    {
      "epoch": 1.6227488151658767,
      "grad_norm": 0.39585891366004944,
      "learning_rate": 8.540213437397833e-05,
      "loss": 0.0401,
      "step": 1070
    },
    {
      "epoch": 1.637914691943128,
      "grad_norm": 0.23856215178966522,
      "learning_rate": 8.502603993307647e-05,
      "loss": 0.0301,
      "step": 1080
    },
    {
      "epoch": 1.6530805687203791,
      "grad_norm": 0.34952136874198914,
      "learning_rate": 8.464601327654207e-05,
      "loss": 0.0401,
      "step": 1090
    },
    {
      "epoch": 1.6682464454976302,
      "grad_norm": 0.2077425867319107,
      "learning_rate": 8.4262097068255e-05,
      "loss": 0.0261,
      "step": 1100
    },
    {
      "epoch": 1.6682464454976302,
      "eval_loss": 0.036638446152210236,
      "eval_runtime": 156.4879,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "step": 1100
    },
    {
      "epoch": 1.6834123222748816,
      "grad_norm": 0.3030443787574768,
      "learning_rate": 8.387433440875758e-05,
      "loss": 0.038,
      "step": 1110
    },
    {
      "epoch": 1.6985781990521327,
      "grad_norm": 0.1555524468421936,
      "learning_rate": 8.348276883041583e-05,
      "loss": 0.0265,
      "step": 1120
    },
    {
      "epoch": 1.7137440758293838,
      "grad_norm": 0.18828414380550385,
      "learning_rate": 8.308744429253238e-05,
      "loss": 0.031,
      "step": 1130
    },
    {
      "epoch": 1.7289099526066352,
      "grad_norm": 0.3289598226547241,
      "learning_rate": 8.268840517641123e-05,
      "loss": 0.0279,
      "step": 1140
    },
    {
      "epoch": 1.7440758293838863,
      "grad_norm": 0.2232438027858734,
      "learning_rate": 8.228569628037543e-05,
      "loss": 0.0294,
      "step": 1150
    },
    {
      "epoch": 1.7440758293838863,
      "eval_loss": 0.03896127641201019,
      "eval_runtime": 156.4636,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 1150
    },
    {
      "epoch": 1.7592417061611374,
      "grad_norm": 0.3435378968715668,
      "learning_rate": 8.187936281473758e-05,
      "loss": 0.038,
      "step": 1160
    },
    {
      "epoch": 1.7744075829383887,
      "grad_norm": 0.16090552508831024,
      "learning_rate": 8.146945039672446e-05,
      "loss": 0.0316,
      "step": 1170
    },
    {
      "epoch": 1.7895734597156399,
      "grad_norm": 0.20053334534168243,
      "learning_rate": 8.105600504535567e-05,
      "loss": 0.0211,
      "step": 1180
    },
    {
      "epoch": 1.804739336492891,
      "grad_norm": 0.5578015446662903,
      "learning_rate": 8.06390731762773e-05,
      "loss": 0.0373,
      "step": 1190
    },
    {
      "epoch": 1.8199052132701423,
      "grad_norm": 0.2711741626262665,
      "learning_rate": 8.021870159655109e-05,
      "loss": 0.0337,
      "step": 1200
    },
    {
      "epoch": 1.8199052132701423,
      "eval_loss": 0.036644402891397476,
      "eval_runtime": 156.4568,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 1200
    },
    {
      "epoch": 1.8350710900473932,
      "grad_norm": 0.3220699727535248,
      "learning_rate": 7.979493749939955e-05,
      "loss": 0.0278,
      "step": 1210
    },
    {
      "epoch": 1.8502369668246446,
      "grad_norm": 0.47723233699798584,
      "learning_rate": 7.93678284589079e-05,
      "loss": 0.0286,
      "step": 1220
    },
    {
      "epoch": 1.865402843601896,
      "grad_norm": 0.21900814771652222,
      "learning_rate": 7.893742242468301e-05,
      "loss": 0.0296,
      "step": 1230
    },
    {
      "epoch": 1.8805687203791468,
      "grad_norm": 0.40307942032814026,
      "learning_rate": 7.850376771647038e-05,
      "loss": 0.041,
      "step": 1240
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 0.2520243525505066,
      "learning_rate": 7.806691301872958e-05,
      "loss": 0.0359,
      "step": 1250
    },
    {
      "epoch": 1.8957345971563981,
      "eval_loss": 0.035044267773628235,
      "eval_runtime": 156.512,
      "eval_samples_per_second": 7.495,
      "eval_steps_per_second": 7.495,
      "step": 1250
    },
    {
      "epoch": 1.9109004739336493,
      "grad_norm": 0.20044682919979095,
      "learning_rate": 7.762690737516846e-05,
      "loss": 0.0376,
      "step": 1260
    },
    {
      "epoch": 1.9260663507109004,
      "grad_norm": 0.6365770101547241,
      "learning_rate": 7.718380018323742e-05,
      "loss": 0.0354,
      "step": 1270
    },
    {
      "epoch": 1.9412322274881517,
      "grad_norm": 0.22586770355701447,
      "learning_rate": 7.673764118858371e-05,
      "loss": 0.0256,
      "step": 1280
    },
    {
      "epoch": 1.9563981042654028,
      "grad_norm": 0.6756897568702698,
      "learning_rate": 7.628848047946675e-05,
      "loss": 0.0455,
      "step": 1290
    },
    {
      "epoch": 1.971563981042654,
      "grad_norm": 0.4071941077709198,
      "learning_rate": 7.583636848113483e-05,
      "loss": 0.0358,
      "step": 1300
    },
    {
      "epoch": 1.971563981042654,
      "eval_loss": 0.03631194680929184,
      "eval_runtime": 156.5217,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 1300
    },
    {
      "epoch": 1.9867298578199053,
      "grad_norm": 0.4594072997570038,
      "learning_rate": 7.538135595016423e-05,
      "loss": 0.0367,
      "step": 1310
    },
    {
      "epoch": 2.001895734597156,
      "grad_norm": 0.1991509646177292,
      "learning_rate": 7.492349396876096e-05,
      "loss": 0.0366,
      "step": 1320
    },
    {
      "epoch": 2.0170616113744075,
      "grad_norm": 0.08454001694917679,
      "learning_rate": 7.4462833939026e-05,
      "loss": 0.0217,
      "step": 1330
    },
    {
      "epoch": 2.032227488151659,
      "grad_norm": 0.03855079039931297,
      "learning_rate": 7.399942757718455e-05,
      "loss": 0.0207,
      "step": 1340
    },
    {
      "epoch": 2.0473933649289098,
      "grad_norm": 0.6126855611801147,
      "learning_rate": 7.35333269077802e-05,
      "loss": 0.0233,
      "step": 1350
    },
    {
      "epoch": 2.0473933649289098,
      "eval_loss": 0.03633115813136101,
      "eval_runtime": 156.5043,
      "eval_samples_per_second": 7.495,
      "eval_steps_per_second": 7.495,
      "step": 1350
    },
    {
      "epoch": 2.062559241706161,
      "grad_norm": 0.1471548080444336,
      "learning_rate": 7.306458425783426e-05,
      "loss": 0.0189,
      "step": 1360
    },
    {
      "epoch": 2.0777251184834125,
      "grad_norm": 0.2453272044658661,
      "learning_rate": 7.25932522509713e-05,
      "loss": 0.0195,
      "step": 1370
    },
    {
      "epoch": 2.0928909952606634,
      "grad_norm": 0.24803659319877625,
      "learning_rate": 7.211938380151133e-05,
      "loss": 0.0146,
      "step": 1380
    },
    {
      "epoch": 2.1080568720379147,
      "grad_norm": 0.22643889486789703,
      "learning_rate": 7.164303210852934e-05,
      "loss": 0.0218,
      "step": 1390
    },
    {
      "epoch": 2.123222748815166,
      "grad_norm": 0.2256947159767151,
      "learning_rate": 7.116425064988286e-05,
      "loss": 0.0157,
      "step": 1400
    },
    {
      "epoch": 2.123222748815166,
      "eval_loss": 0.03634479641914368,
      "eval_runtime": 156.5505,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 1400
    },
    {
      "epoch": 2.138388625592417,
      "grad_norm": 0.2964327931404114,
      "learning_rate": 7.068309317620827e-05,
      "loss": 0.023,
      "step": 1410
    },
    {
      "epoch": 2.1535545023696683,
      "grad_norm": 0.19321192800998688,
      "learning_rate": 7.019961370488645e-05,
      "loss": 0.0203,
      "step": 1420
    },
    {
      "epoch": 2.1687203791469196,
      "grad_norm": 0.14514701068401337,
      "learning_rate": 6.971386651397849e-05,
      "loss": 0.0247,
      "step": 1430
    },
    {
      "epoch": 2.1838862559241705,
      "grad_norm": 0.3884860873222351,
      "learning_rate": 6.922590613613211e-05,
      "loss": 0.0241,
      "step": 1440
    },
    {
      "epoch": 2.199052132701422,
      "grad_norm": 0.12207037210464478,
      "learning_rate": 6.873578735245961e-05,
      "loss": 0.0227,
      "step": 1450
    },
    {
      "epoch": 2.199052132701422,
      "eval_loss": 0.035737331956624985,
      "eval_runtime": 157.0653,
      "eval_samples_per_second": 7.468,
      "eval_steps_per_second": 7.468,
      "step": 1450
    },
    {
      "epoch": 2.2142180094786728,
      "grad_norm": 0.16073821485042572,
      "learning_rate": 6.824356518638775e-05,
      "loss": 0.0158,
      "step": 1460
    },
    {
      "epoch": 2.229383886255924,
      "grad_norm": 0.25047633051872253,
      "learning_rate": 6.774929489748052e-05,
      "loss": 0.0202,
      "step": 1470
    },
    {
      "epoch": 2.2445497630331754,
      "grad_norm": 0.16967254877090454,
      "learning_rate": 6.725303197523548e-05,
      "loss": 0.0216,
      "step": 1480
    },
    {
      "epoch": 2.259715639810427,
      "grad_norm": 0.3594084680080414,
      "learning_rate": 6.675483213285412e-05,
      "loss": 0.0202,
      "step": 1490
    },
    {
      "epoch": 2.2748815165876777,
      "grad_norm": 0.20273494720458984,
      "learning_rate": 6.625475130098728e-05,
      "loss": 0.018,
      "step": 1500
    },
    {
      "epoch": 2.2748815165876777,
      "eval_loss": 0.03894480690360069,
      "eval_runtime": 157.4005,
      "eval_samples_per_second": 7.452,
      "eval_steps_per_second": 7.452,
      "step": 1500
    },
    {
      "epoch": 2.290047393364929,
      "grad_norm": 0.19824734330177307,
      "learning_rate": 6.575284562145593e-05,
      "loss": 0.0336,
      "step": 1510
    },
    {
      "epoch": 2.30521327014218,
      "grad_norm": 0.43414950370788574,
      "learning_rate": 6.524917144094851e-05,
      "loss": 0.0218,
      "step": 1520
    },
    {
      "epoch": 2.3203791469194313,
      "grad_norm": 0.24347089231014252,
      "learning_rate": 6.474378530469509e-05,
      "loss": 0.0275,
      "step": 1530
    },
    {
      "epoch": 2.3355450236966826,
      "grad_norm": 0.14786885678768158,
      "learning_rate": 6.42367439501193e-05,
      "loss": 0.0189,
      "step": 1540
    },
    {
      "epoch": 2.3507109004739335,
      "grad_norm": 0.31469598412513733,
      "learning_rate": 6.372810430046862e-05,
      "loss": 0.0237,
      "step": 1550
    },
    {
      "epoch": 2.3507109004739335,
      "eval_loss": 0.03542237728834152,
      "eval_runtime": 157.6197,
      "eval_samples_per_second": 7.442,
      "eval_steps_per_second": 7.442,
      "step": 1550
    },
    {
      "epoch": 2.365876777251185,
      "grad_norm": 0.23118460178375244,
      "learning_rate": 6.321792345842402e-05,
      "loss": 0.0176,
      "step": 1560
    },
    {
      "epoch": 2.381042654028436,
      "grad_norm": 0.31388726830482483,
      "learning_rate": 6.270625869968906e-05,
      "loss": 0.0196,
      "step": 1570
    },
    {
      "epoch": 2.396208530805687,
      "grad_norm": 0.17334702610969543,
      "learning_rate": 6.219316746656007e-05,
      "loss": 0.021,
      "step": 1580
    },
    {
      "epoch": 2.4113744075829384,
      "grad_norm": 0.11117410659790039,
      "learning_rate": 6.167870736147713e-05,
      "loss": 0.0234,
      "step": 1590
    },
    {
      "epoch": 2.4265402843601898,
      "grad_norm": 0.38643473386764526,
      "learning_rate": 6.116293614055744e-05,
      "loss": 0.025,
      "step": 1600
    },
    {
      "epoch": 2.4265402843601898,
      "eval_loss": 0.03646675497293472,
      "eval_runtime": 157.8176,
      "eval_samples_per_second": 7.433,
      "eval_steps_per_second": 7.433,
      "step": 1600
    },
    {
      "epoch": 2.4417061611374407,
      "grad_norm": 0.3526012897491455,
      "learning_rate": 6.06459117071113e-05,
      "loss": 0.0141,
      "step": 1610
    },
    {
      "epoch": 2.456872037914692,
      "grad_norm": 0.18626412749290466,
      "learning_rate": 6.012769210514146e-05,
      "loss": 0.0229,
      "step": 1620
    },
    {
      "epoch": 2.472037914691943,
      "grad_norm": 0.2591479420661926,
      "learning_rate": 5.9608335512826915e-05,
      "loss": 0.0184,
      "step": 1630
    },
    {
      "epoch": 2.4872037914691942,
      "grad_norm": 0.42772844433784485,
      "learning_rate": 5.908790023599144e-05,
      "loss": 0.0225,
      "step": 1640
    },
    {
      "epoch": 2.5023696682464456,
      "grad_norm": 0.3181972801685333,
      "learning_rate": 5.856644470155781e-05,
      "loss": 0.015,
      "step": 1650
    },
    {
      "epoch": 2.5023696682464456,
      "eval_loss": 0.039200831204652786,
      "eval_runtime": 158.1097,
      "eval_samples_per_second": 7.419,
      "eval_steps_per_second": 7.419,
      "step": 1650
    },
    {
      "epoch": 2.517535545023697,
      "grad_norm": 0.16040761768817902,
      "learning_rate": 5.8044027450988546e-05,
      "loss": 0.0222,
      "step": 1660
    },
    {
      "epoch": 2.532701421800948,
      "grad_norm": 0.5829377770423889,
      "learning_rate": 5.752070713371371e-05,
      "loss": 0.0176,
      "step": 1670
    },
    {
      "epoch": 2.547867298578199,
      "grad_norm": 0.18790672719478607,
      "learning_rate": 5.699654250054662e-05,
      "loss": 0.0295,
      "step": 1680
    },
    {
      "epoch": 2.56303317535545,
      "grad_norm": 0.3851199746131897,
      "learning_rate": 5.647159239708809e-05,
      "loss": 0.0255,
      "step": 1690
    },
    {
      "epoch": 2.5781990521327014,
      "grad_norm": 0.20358450710773468,
      "learning_rate": 5.5945915757120146e-05,
      "loss": 0.0265,
      "step": 1700
    },
    {
      "epoch": 2.5781990521327014,
      "eval_loss": 0.036025311797857285,
      "eval_runtime": 158.4247,
      "eval_samples_per_second": 7.404,
      "eval_steps_per_second": 7.404,
      "step": 1700
    },
    {
      "epoch": 2.5933649289099527,
      "grad_norm": 0.28103622794151306,
      "learning_rate": 5.5419571595989825e-05,
      "loss": 0.0239,
      "step": 1710
    },
    {
      "epoch": 2.6085308056872036,
      "grad_norm": 0.1263958066701889,
      "learning_rate": 5.4892619003983734e-05,
      "loss": 0.0176,
      "step": 1720
    },
    {
      "epoch": 2.623696682464455,
      "grad_norm": 0.4374920427799225,
      "learning_rate": 5.436511713969428e-05,
      "loss": 0.0232,
      "step": 1730
    },
    {
      "epoch": 2.6388625592417063,
      "grad_norm": 0.30128213763237,
      "learning_rate": 5.383712522337817e-05,
      "loss": 0.0193,
      "step": 1740
    },
    {
      "epoch": 2.654028436018957,
      "grad_norm": 0.06995616108179092,
      "learning_rate": 5.3308702530308076e-05,
      "loss": 0.0181,
      "step": 1750
    },
    {
      "epoch": 2.654028436018957,
      "eval_loss": 0.03681906685233116,
      "eval_runtime": 158.5593,
      "eval_samples_per_second": 7.398,
      "eval_steps_per_second": 7.398,
      "step": 1750
    },
    {
      "epoch": 2.6691943127962086,
      "grad_norm": 0.2425563931465149,
      "learning_rate": 5.2779908384118025e-05,
      "loss": 0.0188,
      "step": 1760
    },
    {
      "epoch": 2.68436018957346,
      "grad_norm": 0.41001105308532715,
      "learning_rate": 5.22508021501434e-05,
      "loss": 0.0179,
      "step": 1770
    },
    {
      "epoch": 2.699526066350711,
      "grad_norm": 0.09781570732593536,
      "learning_rate": 5.1721443228756284e-05,
      "loss": 0.0155,
      "step": 1780
    },
    {
      "epoch": 2.714691943127962,
      "grad_norm": 0.2674010694026947,
      "learning_rate": 5.119189104869683e-05,
      "loss": 0.0234,
      "step": 1790
    },
    {
      "epoch": 2.729857819905213,
      "grad_norm": 0.31364890933036804,
      "learning_rate": 5.066220506040148e-05,
      "loss": 0.0228,
      "step": 1800
    },
    {
      "epoch": 2.729857819905213,
      "eval_loss": 0.03799626976251602,
      "eval_runtime": 158.697,
      "eval_samples_per_second": 7.391,
      "eval_steps_per_second": 7.391,
      "step": 1800
    },
    {
      "epoch": 2.7450236966824644,
      "grad_norm": 0.2892131805419922,
      "learning_rate": 5.013244472932872e-05,
      "loss": 0.0218,
      "step": 1810
    },
    {
      "epoch": 2.7601895734597157,
      "grad_norm": 0.3578006625175476,
      "learning_rate": 4.960266952928316e-05,
      "loss": 0.0235,
      "step": 1820
    },
    {
      "epoch": 2.775355450236967,
      "grad_norm": 0.3018692135810852,
      "learning_rate": 4.907293893573867e-05,
      "loss": 0.0193,
      "step": 1830
    },
    {
      "epoch": 2.790521327014218,
      "grad_norm": 0.2626613974571228,
      "learning_rate": 4.8543312419161396e-05,
      "loss": 0.0126,
      "step": 1840
    },
    {
      "epoch": 2.8056872037914693,
      "grad_norm": 0.42779844999313354,
      "learning_rate": 4.8013849438333165e-05,
      "loss": 0.0224,
      "step": 1850
    },
    {
      "epoch": 2.8056872037914693,
      "eval_loss": 0.03656047210097313,
      "eval_runtime": 158.855,
      "eval_samples_per_second": 7.384,
      "eval_steps_per_second": 7.384,
      "step": 1850
    },
    {
      "epoch": 2.82085308056872,
      "grad_norm": 0.41182562708854675,
      "learning_rate": 4.748460943367643e-05,
      "loss": 0.0208,
      "step": 1860
    },
    {
      "epoch": 2.8360189573459715,
      "grad_norm": 0.0459756925702095,
      "learning_rate": 4.695565182058113e-05,
      "loss": 0.0155,
      "step": 1870
    },
    {
      "epoch": 2.851184834123223,
      "grad_norm": 0.22212868928909302,
      "learning_rate": 4.642703598273431e-05,
      "loss": 0.0147,
      "step": 1880
    },
    {
      "epoch": 2.8663507109004738,
      "grad_norm": 0.37007784843444824,
      "learning_rate": 4.589882126545352e-05,
      "loss": 0.0168,
      "step": 1890
    },
    {
      "epoch": 2.881516587677725,
      "grad_norm": 0.13385622203350067,
      "learning_rate": 4.537106696902425e-05,
      "loss": 0.0162,
      "step": 1900
    },
    {
      "epoch": 2.881516587677725,
      "eval_loss": 0.035648539662361145,
      "eval_runtime": 159.0125,
      "eval_samples_per_second": 7.377,
      "eval_steps_per_second": 7.377,
      "step": 1900
    },
    {
      "epoch": 2.8966824644549765,
      "grad_norm": 0.33246728777885437,
      "learning_rate": 4.484383234204266e-05,
      "loss": 0.0192,
      "step": 1910
    },
    {
      "epoch": 2.9118483412322274,
      "grad_norm": 0.27486807107925415,
      "learning_rate": 4.4317176574763935e-05,
      "loss": 0.0234,
      "step": 1920
    },
    {
      "epoch": 2.9270142180094787,
      "grad_norm": 0.10704711079597473,
      "learning_rate": 4.3791158792457334e-05,
      "loss": 0.0193,
      "step": 1930
    },
    {
      "epoch": 2.94218009478673,
      "grad_norm": 0.7915742993354797,
      "learning_rate": 4.3265838048768334e-05,
      "loss": 0.0298,
      "step": 1940
    },
    {
      "epoch": 2.957345971563981,
      "grad_norm": 0.38838285207748413,
      "learning_rate": 4.274127331908915e-05,
      "loss": 0.0284,
      "step": 1950
    },
    {
      "epoch": 2.957345971563981,
      "eval_loss": 0.03486395254731178,
      "eval_runtime": 159.3462,
      "eval_samples_per_second": 7.361,
      "eval_steps_per_second": 7.361,
      "step": 1950
    },
    {
      "epoch": 2.9725118483412323,
      "grad_norm": 0.25486910343170166,
      "learning_rate": 4.221752349393758e-05,
      "loss": 0.0236,
      "step": 1960
    },
    {
      "epoch": 2.987677725118483,
      "grad_norm": 0.44221231341362,
      "learning_rate": 4.1694647372345886e-05,
      "loss": 0.0271,
      "step": 1970
    },
    {
      "epoch": 3.0028436018957345,
      "grad_norm": 0.05971141532063484,
      "learning_rate": 4.117270365525946e-05,
      "loss": 0.0226,
      "step": 1980
    },
    {
      "epoch": 3.018009478672986,
      "grad_norm": 0.17855848371982574,
      "learning_rate": 4.065175093894694e-05,
      "loss": 0.0133,
      "step": 1990
    },
    {
      "epoch": 3.0331753554502368,
      "grad_norm": 0.1266752928495407,
      "learning_rate": 4.013184770842167e-05,
      "loss": 0.0069,
      "step": 2000
    },
    {
      "epoch": 3.0331753554502368,
      "eval_loss": 0.03510182350873947,
      "eval_runtime": 159.5736,
      "eval_samples_per_second": 7.351,
      "eval_steps_per_second": 7.351,
      "step": 2000
    },
    {
      "epoch": 3.048341232227488,
      "grad_norm": 0.20517046749591827,
      "learning_rate": 3.9613052330876064e-05,
      "loss": 0.0134,
      "step": 2010
    },
    {
      "epoch": 3.0635071090047394,
      "grad_norm": 0.06853601336479187,
      "learning_rate": 3.909542304912881e-05,
      "loss": 0.0118,
      "step": 2020
    },
    {
      "epoch": 3.0786729857819903,
      "grad_norm": 0.29589417576789856,
      "learning_rate": 3.857901797508628e-05,
      "loss": 0.0092,
      "step": 2030
    },
    {
      "epoch": 3.0938388625592417,
      "grad_norm": 0.24239981174468994,
      "learning_rate": 3.80638950832186e-05,
      "loss": 0.0101,
      "step": 2040
    },
    {
      "epoch": 3.109004739336493,
      "grad_norm": 0.15986329317092896,
      "learning_rate": 3.7550112204051014e-05,
      "loss": 0.0065,
      "step": 2050
    },
    {
      "epoch": 3.109004739336493,
      "eval_loss": 0.03937237337231636,
      "eval_runtime": 159.802,
      "eval_samples_per_second": 7.34,
      "eval_steps_per_second": 7.34,
      "step": 2050
    },
    {
      "epoch": 3.124170616113744,
      "grad_norm": 0.2081887423992157,
      "learning_rate": 3.703772701767167e-05,
      "loss": 0.0087,
      "step": 2060
    },
    {
      "epoch": 3.1393364928909953,
      "grad_norm": 0.19468900561332703,
      "learning_rate": 3.652679704725596e-05,
      "loss": 0.0088,
      "step": 2070
    },
    {
      "epoch": 3.1545023696682466,
      "grad_norm": 0.2196388840675354,
      "learning_rate": 3.601737965260882e-05,
      "loss": 0.0142,
      "step": 2080
    },
    {
      "epoch": 3.1696682464454975,
      "grad_norm": 0.35034120082855225,
      "learning_rate": 3.550953202372503e-05,
      "loss": 0.0102,
      "step": 2090
    },
    {
      "epoch": 3.184834123222749,
      "grad_norm": 0.13505339622497559,
      "learning_rate": 3.500331117436895e-05,
      "loss": 0.0069,
      "step": 2100
    },
    {
      "epoch": 3.184834123222749,
      "eval_loss": 0.04376286268234253,
      "eval_runtime": 159.8812,
      "eval_samples_per_second": 7.337,
      "eval_steps_per_second": 7.337,
      "step": 2100
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.24390918016433716,
      "learning_rate": 3.4498773935673626e-05,
      "loss": 0.0097,
      "step": 2110
    },
    {
      "epoch": 3.215165876777251,
      "grad_norm": 0.2587122917175293,
      "learning_rate": 3.399597694976081e-05,
      "loss": 0.0066,
      "step": 2120
    },
    {
      "epoch": 3.2303317535545024,
      "grad_norm": 0.09908251464366913,
      "learning_rate": 3.349497666338187e-05,
      "loss": 0.0088,
      "step": 2130
    },
    {
      "epoch": 3.2454976303317538,
      "grad_norm": 0.29646173119544983,
      "learning_rate": 3.299582932158085e-05,
      "loss": 0.006,
      "step": 2140
    },
    {
      "epoch": 3.2606635071090047,
      "grad_norm": 0.09596337378025055,
      "learning_rate": 3.2498590961379996e-05,
      "loss": 0.006,
      "step": 2150
    },
    {
      "epoch": 3.2606635071090047,
      "eval_loss": 0.044551946222782135,
      "eval_runtime": 159.7698,
      "eval_samples_per_second": 7.342,
      "eval_steps_per_second": 7.342,
      "step": 2150
    },
    {
      "epoch": 3.275829383886256,
      "grad_norm": 0.19044414162635803,
      "learning_rate": 3.200331740548887e-05,
      "loss": 0.0079,
      "step": 2160
    },
    {
      "epoch": 3.2909952606635073,
      "grad_norm": 0.40478262305259705,
      "learning_rate": 3.1510064256037274e-05,
      "loss": 0.0097,
      "step": 2170
    },
    {
      "epoch": 3.3061611374407582,
      "grad_norm": 0.3682207763195038,
      "learning_rate": 3.1018886888333065e-05,
      "loss": 0.0077,
      "step": 2180
    },
    {
      "epoch": 3.3213270142180096,
      "grad_norm": 0.3857901096343994,
      "learning_rate": 3.052984044464548e-05,
      "loss": 0.0118,
      "step": 2190
    },
    {
      "epoch": 3.3364928909952605,
      "grad_norm": 0.2310226857662201,
      "learning_rate": 3.0042979828014496e-05,
      "loss": 0.0088,
      "step": 2200
    },
    {
      "epoch": 3.3364928909952605,
      "eval_loss": 0.04364483430981636,
      "eval_runtime": 159.855,
      "eval_samples_per_second": 7.338,
      "eval_steps_per_second": 7.338,
      "step": 2200
    },
    {
      "epoch": 3.351658767772512,
      "grad_norm": 0.3267790675163269,
      "learning_rate": 2.9558359696087152e-05,
      "loss": 0.0112,
      "step": 2210
    },
    {
      "epoch": 3.366824644549763,
      "grad_norm": 0.41595402359962463,
      "learning_rate": 2.9076034454981332e-05,
      "loss": 0.009,
      "step": 2220
    },
    {
      "epoch": 3.381990521327014,
      "grad_norm": 0.0655096098780632,
      "learning_rate": 2.8596058253177932e-05,
      "loss": 0.0106,
      "step": 2230
    },
    {
      "epoch": 3.3971563981042654,
      "grad_norm": 0.3616434931755066,
      "learning_rate": 2.811848497544175e-05,
      "loss": 0.0114,
      "step": 2240
    },
    {
      "epoch": 3.4123222748815167,
      "grad_norm": 0.13481537997722626,
      "learning_rate": 2.764336823677216e-05,
      "loss": 0.0056,
      "step": 2250
    },
    {
      "epoch": 3.4123222748815167,
      "eval_loss": 0.04239945113658905,
      "eval_runtime": 160.0687,
      "eval_samples_per_second": 7.328,
      "eval_steps_per_second": 7.328,
      "step": 2250
    },
    {
      "epoch": 3.4274881516587676,
      "grad_norm": 0.4856715798377991,
      "learning_rate": 2.717076137638388e-05,
      "loss": 0.0083,
      "step": 2260
    },
    {
      "epoch": 3.442654028436019,
      "grad_norm": 0.49463364481925964,
      "learning_rate": 2.6700717451719066e-05,
      "loss": 0.0076,
      "step": 2270
    },
    {
      "epoch": 3.4578199052132703,
      "grad_norm": 0.1439306139945984,
      "learning_rate": 2.6233289232490447e-05,
      "loss": 0.0062,
      "step": 2280
    },
    {
      "epoch": 3.472985781990521,
      "grad_norm": 0.07714873552322388,
      "learning_rate": 2.5768529194757474e-05,
      "loss": 0.0099,
      "step": 2290
    },
    {
      "epoch": 3.4881516587677726,
      "grad_norm": 0.24623507261276245,
      "learning_rate": 2.5306489515034713e-05,
      "loss": 0.0086,
      "step": 2300
    },
    {
      "epoch": 3.4881516587677726,
      "eval_loss": 0.04448748379945755,
      "eval_runtime": 160.0261,
      "eval_samples_per_second": 7.33,
      "eval_steps_per_second": 7.33,
      "step": 2300
    },
    {
      "epoch": 3.5033175355450235,
      "grad_norm": 0.3714815676212311,
      "learning_rate": 2.484722206443455e-05,
      "loss": 0.0162,
      "step": 2310
    },
    {
      "epoch": 3.518483412322275,
      "grad_norm": 0.03868235647678375,
      "learning_rate": 2.4390778402843605e-05,
      "loss": 0.0064,
      "step": 2320
    },
    {
      "epoch": 3.533649289099526,
      "grad_norm": 0.3829624056816101,
      "learning_rate": 2.3937209773134472e-05,
      "loss": 0.0053,
      "step": 2330
    },
    {
      "epoch": 3.5488151658767775,
      "grad_norm": 0.1869666576385498,
      "learning_rate": 2.3486567095412864e-05,
      "loss": 0.0091,
      "step": 2340
    },
    {
      "epoch": 3.5639810426540284,
      "grad_norm": 0.12133553624153137,
      "learning_rate": 2.3038900961301053e-05,
      "loss": 0.0058,
      "step": 2350
    },
    {
      "epoch": 3.5639810426540284,
      "eval_loss": 0.04382281377911568,
      "eval_runtime": 160.2079,
      "eval_samples_per_second": 7.322,
      "eval_steps_per_second": 7.322,
      "step": 2350
    },
    {
      "epoch": 3.5791469194312797,
      "grad_norm": 0.4621219336986542,
      "learning_rate": 2.259426162825819e-05,
      "loss": 0.0083,
      "step": 2360
    },
    {
      "epoch": 3.5943127962085306,
      "grad_norm": 0.045843854546546936,
      "learning_rate": 2.215269901393805e-05,
      "loss": 0.0073,
      "step": 2370
    },
    {
      "epoch": 3.609478672985782,
      "grad_norm": 0.31095975637435913,
      "learning_rate": 2.1714262690585192e-05,
      "loss": 0.0089,
      "step": 2380
    },
    {
      "epoch": 3.6246445497630333,
      "grad_norm": 0.6029720306396484,
      "learning_rate": 2.1279001879469424e-05,
      "loss": 0.0154,
      "step": 2390
    },
    {
      "epoch": 3.639810426540284,
      "grad_norm": 0.28314200043678284,
      "learning_rate": 2.084696544536025e-05,
      "loss": 0.0092,
      "step": 2400
    },
    {
      "epoch": 3.639810426540284,
      "eval_loss": 0.042627111077308655,
      "eval_runtime": 160.2463,
      "eval_samples_per_second": 7.32,
      "eval_steps_per_second": 7.32,
      "step": 2400
    },
    {
      "epoch": 3.6549763033175355,
      "grad_norm": 0.0889393612742424,
      "learning_rate": 2.0418201891040778e-05,
      "loss": 0.01,
      "step": 2410
    },
    {
      "epoch": 3.670142180094787,
      "grad_norm": 0.041327815502882004,
      "learning_rate": 1.9992759351862772e-05,
      "loss": 0.0057,
      "step": 2420
    },
    {
      "epoch": 3.6853080568720378,
      "grad_norm": 0.3208984434604645,
      "learning_rate": 1.9570685590342463e-05,
      "loss": 0.006,
      "step": 2430
    },
    {
      "epoch": 3.700473933649289,
      "grad_norm": 0.38806548714637756,
      "learning_rate": 1.9152027990798748e-05,
      "loss": 0.0043,
      "step": 2440
    },
    {
      "epoch": 3.7156398104265405,
      "grad_norm": 0.5781579613685608,
      "learning_rate": 1.8736833554033263e-05,
      "loss": 0.0082,
      "step": 2450
    },
    {
      "epoch": 3.7156398104265405,
      "eval_loss": 0.04327119514346123,
      "eval_runtime": 160.272,
      "eval_samples_per_second": 7.319,
      "eval_steps_per_second": 7.319,
      "step": 2450
    },
    {
      "epoch": 3.7308056872037914,
      "grad_norm": 0.31082025170326233,
      "learning_rate": 1.8325148892054093e-05,
      "loss": 0.0086,
      "step": 2460
    },
    {
      "epoch": 3.7459715639810427,
      "grad_norm": 0.10387292504310608,
      "learning_rate": 1.7917020222842662e-05,
      "loss": 0.006,
      "step": 2470
    },
    {
      "epoch": 3.7611374407582936,
      "grad_norm": 0.5228342413902283,
      "learning_rate": 1.751249336516513e-05,
      "loss": 0.0081,
      "step": 2480
    },
    {
      "epoch": 3.776303317535545,
      "grad_norm": 0.5724895596504211,
      "learning_rate": 1.7111613733428522e-05,
      "loss": 0.0167,
      "step": 2490
    },
    {
      "epoch": 3.7914691943127963,
      "grad_norm": 0.3599369525909424,
      "learning_rate": 1.671442633258218e-05,
      "loss": 0.01,
      "step": 2500
    },
    {
      "epoch": 3.7914691943127963,
      "eval_loss": 0.042170289903879166,
      "eval_runtime": 160.3642,
      "eval_samples_per_second": 7.315,
      "eval_steps_per_second": 7.315,
      "step": 2500
    },
    {
      "epoch": 3.8066350710900476,
      "grad_norm": 0.4104222357273102,
      "learning_rate": 1.6320975753065453e-05,
      "loss": 0.0117,
      "step": 2510
    },
    {
      "epoch": 3.8218009478672985,
      "grad_norm": 0.1062493547797203,
      "learning_rate": 1.5931306165801452e-05,
      "loss": 0.0062,
      "step": 2520
    },
    {
      "epoch": 3.83696682464455,
      "grad_norm": 0.054498132318258286,
      "learning_rate": 1.554546131723848e-05,
      "loss": 0.0104,
      "step": 2530
    },
    {
      "epoch": 3.8521327014218008,
      "grad_norm": 0.29258275032043457,
      "learning_rate": 1.5163484524438516e-05,
      "loss": 0.0076,
      "step": 2540
    },
    {
      "epoch": 3.867298578199052,
      "grad_norm": 0.3757239580154419,
      "learning_rate": 1.4785418670214496e-05,
      "loss": 0.0082,
      "step": 2550
    },
    {
      "epoch": 3.867298578199052,
      "eval_loss": 0.042569711804389954,
      "eval_runtime": 160.3896,
      "eval_samples_per_second": 7.313,
      "eval_steps_per_second": 7.313,
      "step": 2550
    },
    {
      "epoch": 3.8824644549763034,
      "grad_norm": 0.1663331836462021,
      "learning_rate": 1.4411306198315777e-05,
      "loss": 0.0057,
      "step": 2560
    },
    {
      "epoch": 3.8976303317535548,
      "grad_norm": 0.1922130584716797,
      "learning_rate": 1.4041189108663421e-05,
      "loss": 0.005,
      "step": 2570
    },
    {
      "epoch": 3.9127962085308057,
      "grad_norm": 0.5593583583831787,
      "learning_rate": 1.3675108952634824e-05,
      "loss": 0.0154,
      "step": 2580
    },
    {
      "epoch": 3.927962085308057,
      "grad_norm": 0.3150312900543213,
      "learning_rate": 1.3313106828399147e-05,
      "loss": 0.01,
      "step": 2590
    },
    {
      "epoch": 3.943127962085308,
      "grad_norm": 0.3135119378566742,
      "learning_rate": 1.2955223376303205e-05,
      "loss": 0.012,
      "step": 2600
    },
    {
      "epoch": 3.943127962085308,
      "eval_loss": 0.04326267167925835,
      "eval_runtime": 160.508,
      "eval_samples_per_second": 7.308,
      "eval_steps_per_second": 7.308,
      "step": 2600
    },
    {
      "epoch": 3.9582938388625593,
      "grad_norm": 0.4532774090766907,
      "learning_rate": 1.2601498774309112e-05,
      "loss": 0.0067,
      "step": 2610
    },
    {
      "epoch": 3.9734597156398106,
      "grad_norm": 0.313206285238266,
      "learning_rate": 1.2251972733483612e-05,
      "loss": 0.008,
      "step": 2620
    },
    {
      "epoch": 3.9886255924170615,
      "grad_norm": 0.251285582780838,
      "learning_rate": 1.19066844935399e-05,
      "loss": 0.0073,
      "step": 2630
    },
    {
      "epoch": 4.003791469194312,
      "grad_norm": 0.19865204393863678,
      "learning_rate": 1.156567281843241e-05,
      "loss": 0.0105,
      "step": 2640
    },
    {
      "epoch": 4.018957345971564,
      "grad_norm": 0.2726939022541046,
      "learning_rate": 1.1228975992004842e-05,
      "loss": 0.0054,
      "step": 2650
    },
    {
      "epoch": 4.018957345971564,
      "eval_loss": 0.043674346059560776,
      "eval_runtime": 160.5744,
      "eval_samples_per_second": 7.305,
      "eval_steps_per_second": 7.305,
      "step": 2650
    },
    {
      "epoch": 4.034123222748815,
      "grad_norm": 0.03906554728746414,
      "learning_rate": 1.0896631813692377e-05,
      "loss": 0.0031,
      "step": 2660
    },
    {
      "epoch": 4.049289099526066,
      "grad_norm": 0.040388550609350204,
      "learning_rate": 1.0568677594277899e-05,
      "loss": 0.0028,
      "step": 2670
    },
    {
      "epoch": 4.064454976303318,
      "grad_norm": 0.06655081361532211,
      "learning_rate": 1.0245150151703509e-05,
      "loss": 0.0015,
      "step": 2680
    },
    {
      "epoch": 4.079620853080569,
      "grad_norm": 0.05049607902765274,
      "learning_rate": 9.926085806936918e-06,
      "loss": 0.0052,
      "step": 2690
    },
    {
      "epoch": 4.0947867298578196,
      "grad_norm": 0.018399395048618317,
      "learning_rate": 9.611520379894068e-06,
      "loss": 0.0037,
      "step": 2700
    },
    {
      "epoch": 4.0947867298578196,
      "eval_loss": 0.045233942568302155,
      "eval_runtime": 160.6326,
      "eval_samples_per_second": 7.302,
      "eval_steps_per_second": 7.302,
      "step": 2700
    },
    {
      "epoch": 4.109952606635071,
      "grad_norm": 0.16862346231937408,
      "learning_rate": 9.301489185417578e-06,
      "loss": 0.0051,
      "step": 2710
    },
    {
      "epoch": 4.125118483412322,
      "grad_norm": 0.026153651997447014,
      "learning_rate": 8.996027029312314e-06,
      "loss": 0.0017,
      "step": 2720
    },
    {
      "epoch": 4.140284360189574,
      "grad_norm": 0.2588901221752167,
      "learning_rate": 8.695168204437721e-06,
      "loss": 0.003,
      "step": 2730
    },
    {
      "epoch": 4.155450236966825,
      "grad_norm": 0.12721817195415497,
      "learning_rate": 8.398946486858029e-06,
      "loss": 0.0032,
      "step": 2740
    },
    {
      "epoch": 4.170616113744076,
      "grad_norm": 0.14331641793251038,
      "learning_rate": 8.107395132050333e-06,
      "loss": 0.0049,
      "step": 2750
    },
    {
      "epoch": 4.170616113744076,
      "eval_loss": 0.047378573566675186,
      "eval_runtime": 160.6316,
      "eval_samples_per_second": 7.302,
      "eval_steps_per_second": 7.302,
      "step": 2750
    },
    {
      "epoch": 4.185781990521327,
      "grad_norm": 0.03555993735790253,
      "learning_rate": 7.820546871171142e-06,
      "loss": 0.0031,
      "step": 2760
    },
    {
      "epoch": 4.200947867298578,
      "grad_norm": 0.09507564455270767,
      "learning_rate": 7.5384339073818145e-06,
      "loss": 0.0031,
      "step": 2770
    },
    {
      "epoch": 4.216113744075829,
      "grad_norm": 0.054082565009593964,
      "learning_rate": 7.261087912233228e-06,
      "loss": 0.0029,
      "step": 2780
    },
    {
      "epoch": 4.231279620853081,
      "grad_norm": 0.21810650825500488,
      "learning_rate": 6.988540022110235e-06,
      "loss": 0.0061,
      "step": 2790
    },
    {
      "epoch": 4.246445497630332,
      "grad_norm": 0.031083092093467712,
      "learning_rate": 6.720820834735969e-06,
      "loss": 0.0024,
      "step": 2800
    },
    {
      "epoch": 4.246445497630332,
      "eval_loss": 0.04916524887084961,
      "eval_runtime": 160.7475,
      "eval_samples_per_second": 7.297,
      "eval_steps_per_second": 7.297,
      "step": 2800
    },
    {
      "epoch": 4.2616113744075825,
      "grad_norm": 0.05597074702382088,
      "learning_rate": 6.457960405736968e-06,
      "loss": 0.0019,
      "step": 2810
    },
    {
      "epoch": 4.276777251184834,
      "grad_norm": 0.05571829900145531,
      "learning_rate": 6.199988245268778e-06,
      "loss": 0.003,
      "step": 2820
    },
    {
      "epoch": 4.291943127962085,
      "grad_norm": 0.01073633786290884,
      "learning_rate": 5.946933314703157e-06,
      "loss": 0.0032,
      "step": 2830
    },
    {
      "epoch": 4.3071090047393366,
      "grad_norm": 0.01971004158258438,
      "learning_rate": 5.698824023376531e-06,
      "loss": 0.0015,
      "step": 2840
    },
    {
      "epoch": 4.322274881516588,
      "grad_norm": 0.12915189564228058,
      "learning_rate": 5.455688225400802e-06,
      "loss": 0.0023,
      "step": 2850
    },
    {
      "epoch": 4.322274881516588,
      "eval_loss": 0.05073701962828636,
      "eval_runtime": 160.723,
      "eval_samples_per_second": 7.298,
      "eval_steps_per_second": 7.298,
      "step": 2850
    },
    {
      "epoch": 4.337440758293839,
      "grad_norm": 0.36161208152770996,
      "learning_rate": 5.217553216536098e-06,
      "loss": 0.0058,
      "step": 2860
    },
    {
      "epoch": 4.35260663507109,
      "grad_norm": 0.04417311027646065,
      "learning_rate": 4.984445731126542e-06,
      "loss": 0.0045,
      "step": 2870
    },
    {
      "epoch": 4.367772511848341,
      "grad_norm": 0.7135786414146423,
      "learning_rate": 4.756391939098853e-06,
      "loss": 0.0026,
      "step": 2880
    },
    {
      "epoch": 4.382938388625592,
      "grad_norm": 0.026783248409628868,
      "learning_rate": 4.533417443024374e-06,
      "loss": 0.0039,
      "step": 2890
    },
    {
      "epoch": 4.398104265402844,
      "grad_norm": 0.13430364429950714,
      "learning_rate": 4.315547275244769e-06,
      "loss": 0.0048,
      "step": 2900
    },
    {
      "epoch": 4.398104265402844,
      "eval_loss": 0.05119829997420311,
      "eval_runtime": 160.8774,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 2900
    },
    {
      "epoch": 4.413270142180095,
      "grad_norm": 0.3382275700569153,
      "learning_rate": 4.10280589506179e-06,
      "loss": 0.0021,
      "step": 2910
    },
    {
      "epoch": 4.4284360189573455,
      "grad_norm": 0.3276625871658325,
      "learning_rate": 3.895217185991312e-06,
      "loss": 0.003,
      "step": 2920
    },
    {
      "epoch": 4.443601895734597,
      "grad_norm": 0.22590480744838715,
      "learning_rate": 3.692804453082038e-06,
      "loss": 0.0042,
      "step": 2930
    },
    {
      "epoch": 4.458767772511848,
      "grad_norm": 0.38036713004112244,
      "learning_rate": 3.495590420299194e-06,
      "loss": 0.0035,
      "step": 2940
    },
    {
      "epoch": 4.4739336492890995,
      "grad_norm": 0.1442231684923172,
      "learning_rate": 3.3035972279733207e-06,
      "loss": 0.0073,
      "step": 2950
    },
    {
      "epoch": 4.4739336492890995,
      "eval_loss": 0.05144655704498291,
      "eval_runtime": 160.956,
      "eval_samples_per_second": 7.288,
      "eval_steps_per_second": 7.288,
      "step": 2950
    },
    {
      "epoch": 4.489099526066351,
      "grad_norm": 0.057509418576955795,
      "learning_rate": 3.116846430314796e-06,
      "loss": 0.002,
      "step": 2960
    },
    {
      "epoch": 4.504265402843602,
      "grad_norm": 0.0358334444463253,
      "learning_rate": 2.9353589929939087e-06,
      "loss": 0.0012,
      "step": 2970
    },
    {
      "epoch": 4.519431279620854,
      "grad_norm": 0.2266499549150467,
      "learning_rate": 2.7591552907872574e-06,
      "loss": 0.003,
      "step": 2980
    },
    {
      "epoch": 4.534597156398104,
      "grad_norm": 0.2837143838405609,
      "learning_rate": 2.5882551052902883e-06,
      "loss": 0.0031,
      "step": 2990
    },
    {
      "epoch": 4.549763033175355,
      "grad_norm": 0.22253455221652985,
      "learning_rate": 2.4226776226965453e-06,
      "loss": 0.0014,
      "step": 3000
    },
    {
      "epoch": 4.549763033175355,
      "eval_loss": 0.05179252475500107,
      "eval_runtime": 160.892,
      "eval_samples_per_second": 7.291,
      "eval_steps_per_second": 7.291,
      "step": 3000
    },
    {
      "epoch": 4.564928909952607,
      "grad_norm": 0.03284859284758568,
      "learning_rate": 2.262441431643697e-06,
      "loss": 0.0032,
      "step": 3010
    },
    {
      "epoch": 4.580094786729858,
      "grad_norm": 0.13714928925037384,
      "learning_rate": 2.1075645211266927e-06,
      "loss": 0.0026,
      "step": 3020
    },
    {
      "epoch": 4.595260663507109,
      "grad_norm": 0.06458429247140884,
      "learning_rate": 1.9580642784782053e-06,
      "loss": 0.0046,
      "step": 3030
    },
    {
      "epoch": 4.61042654028436,
      "grad_norm": 0.028915604576468468,
      "learning_rate": 1.813957487416651e-06,
      "loss": 0.0028,
      "step": 3040
    },
    {
      "epoch": 4.625592417061611,
      "grad_norm": 0.02218010276556015,
      "learning_rate": 1.6752603261619315e-06,
      "loss": 0.0021,
      "step": 3050
    },
    {
      "epoch": 4.625592417061611,
      "eval_loss": 0.05206415429711342,
      "eval_runtime": 160.9701,
      "eval_samples_per_second": 7.287,
      "eval_steps_per_second": 7.287,
      "step": 3050
    },
    {
      "epoch": 4.6407582938388625,
      "grad_norm": 0.04888171702623367,
      "learning_rate": 1.541988365619207e-06,
      "loss": 0.0022,
      "step": 3060
    },
    {
      "epoch": 4.655924170616114,
      "grad_norm": 0.15427792072296143,
      "learning_rate": 1.4141565676307865e-06,
      "loss": 0.002,
      "step": 3070
    },
    {
      "epoch": 4.671090047393365,
      "grad_norm": 0.047208454459905624,
      "learning_rate": 1.291779283296468e-06,
      "loss": 0.0027,
      "step": 3080
    },
    {
      "epoch": 4.6862559241706165,
      "grad_norm": 0.03485613688826561,
      "learning_rate": 1.1748702513623922e-06,
      "loss": 0.0022,
      "step": 3090
    },
    {
      "epoch": 4.701421800947867,
      "grad_norm": 0.038091257214546204,
      "learning_rate": 1.0634425966786155e-06,
      "loss": 0.0043,
      "step": 3100
    },
    {
      "epoch": 4.701421800947867,
      "eval_loss": 0.052517589181661606,
      "eval_runtime": 160.908,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "step": 3100
    },
    {
      "epoch": 4.716587677725118,
      "grad_norm": 0.21638447046279907,
      "learning_rate": 9.575088287257118e-07,
      "loss": 0.0035,
      "step": 3110
    },
    {
      "epoch": 4.73175355450237,
      "grad_norm": 0.10230497270822525,
      "learning_rate": 8.570808402103392e-07,
      "loss": 0.0035,
      "step": 3120
    },
    {
      "epoch": 4.746919431279621,
      "grad_norm": 0.08958570659160614,
      "learning_rate": 7.62169905730109e-07,
      "loss": 0.0042,
      "step": 3130
    },
    {
      "epoch": 4.762085308056872,
      "grad_norm": 0.2054518759250641,
      "learning_rate": 6.727866805078531e-07,
      "loss": 0.0018,
      "step": 3140
    },
    {
      "epoch": 4.777251184834123,
      "grad_norm": 0.46021974086761475,
      "learning_rate": 5.889411991953975e-07,
      "loss": 0.005,
      "step": 3150
    },
    {
      "epoch": 4.777251184834123,
      "eval_loss": 0.0527673102915287,
      "eval_runtime": 160.9961,
      "eval_samples_per_second": 7.286,
      "eval_steps_per_second": 7.286,
      "step": 3150
    },
    {
      "epoch": 4.792417061611374,
      "grad_norm": 0.030841287225484848,
      "learning_rate": 5.106428747470137e-07,
      "loss": 0.0012,
      "step": 3160
    },
    {
      "epoch": 4.8075829383886255,
      "grad_norm": 0.012616468593478203,
      "learning_rate": 4.37900497362681e-07,
      "loss": 0.002,
      "step": 3170
    },
    {
      "epoch": 4.822748815165877,
      "grad_norm": 0.1586543619632721,
      "learning_rate": 3.7072223350124167e-07,
      "loss": 0.0021,
      "step": 3180
    },
    {
      "epoch": 4.837914691943128,
      "grad_norm": 0.2468951940536499,
      "learning_rate": 3.0911562496358517e-07,
      "loss": 0.0043,
      "step": 3190
    },
    {
      "epoch": 4.8530805687203795,
      "grad_norm": 0.051594868302345276,
      "learning_rate": 2.53087588045986e-07,
      "loss": 0.0021,
      "step": 3200
    },
    {
      "epoch": 4.8530805687203795,
      "eval_loss": 0.05314692109823227,
      "eval_runtime": 161.0886,
      "eval_samples_per_second": 7.282,
      "eval_steps_per_second": 7.282,
      "step": 3200
    },
    {
      "epoch": 4.86824644549763,
      "grad_norm": 0.3177165687084198,
      "learning_rate": 2.026444127636029e-07,
      "loss": 0.0021,
      "step": 3210
    },
    {
      "epoch": 4.883412322274881,
      "grad_norm": 0.8044673204421997,
      "learning_rate": 1.577917621443825e-07,
      "loss": 0.003,
      "step": 3220
    },
    {
      "epoch": 4.898578199052133,
      "grad_norm": 0.15097960829734802,
      "learning_rate": 1.185346715932345e-07,
      "loss": 0.006,
      "step": 3230
    },
    {
      "epoch": 4.913744075829384,
      "grad_norm": 0.12999580800533295,
      "learning_rate": 8.48775483267783e-08,
      "loss": 0.003,
      "step": 3240
    },
    {
      "epoch": 4.928909952606635,
      "grad_norm": 0.051713138818740845,
      "learning_rate": 5.6824170878544414e-08,
      "loss": 0.0032,
      "step": 3250
    },
    {
      "epoch": 4.928909952606635,
      "eval_loss": 0.05286237224936485,
      "eval_runtime": 161.1111,
      "eval_samples_per_second": 7.281,
      "eval_steps_per_second": 7.281,
      "step": 3250
    },
    {
      "epoch": 4.944075829383886,
      "grad_norm": 0.2822915017604828,
      "learning_rate": 3.4377688674774866e-08,
      "loss": 0.0035,
      "step": 3260
    },
    {
      "epoch": 4.959241706161137,
      "grad_norm": 0.028655484318733215,
      "learning_rate": 1.7540621680850466e-08,
      "loss": 0.0028,
      "step": 3270
    },
    {
      "epoch": 4.9744075829383885,
      "grad_norm": 0.4117952883243561,
      "learning_rate": 6.314860118400434e-09,
      "loss": 0.0026,
      "step": 3280
    },
    {
      "epoch": 4.98957345971564,
      "grad_norm": 0.1054949164390564,
      "learning_rate": 7.016642530777162e-10,
      "loss": 0.0038,
      "step": 3290
    },
    {
      "epoch": 4.9971563981042655,
      "step": 3295,
      "total_flos": 8.270218162679316e+17,
      "train_loss": 0.04085336653218917,
      "train_runtime": 35847.6988,
      "train_samples_per_second": 1.472,
      "train_steps_per_second": 0.092
    }
  ],
  "logging_steps": 10,
  "max_steps": 3295,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.270218162679316e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}