{
  "best_metric": 0.034580811858177185,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold2/checkpoint-1250",
  "epoch": 5.0,
  "eval_steps": 50,
  "global_step": 3225,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015503875968992248,
      "grad_norm": 5.518270492553711,
      "learning_rate": 3.0959752321981426e-06,
      "loss": 1.3797,
      "step": 10
    },
    {
      "epoch": 0.031007751937984496,
      "grad_norm": 5.734408378601074,
      "learning_rate": 6.191950464396285e-06,
      "loss": 1.277,
      "step": 20
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 3.4467711448669434,
      "learning_rate": 9.287925696594429e-06,
      "loss": 1.1112,
      "step": 30
    },
    {
      "epoch": 0.06201550387596899,
      "grad_norm": 2.214425563812256,
      "learning_rate": 1.238390092879257e-05,
      "loss": 0.7071,
      "step": 40
    },
    {
      "epoch": 0.07751937984496124,
      "grad_norm": 1.933417558670044,
      "learning_rate": 1.5479876160990712e-05,
      "loss": 0.4179,
      "step": 50
    },
    {
      "epoch": 0.07751937984496124,
      "eval_loss": 0.34340164065361023,
      "eval_runtime": 153.6608,
      "eval_samples_per_second": 7.464,
      "eval_steps_per_second": 7.464,
      "step": 50
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 1.1443437337875366,
      "learning_rate": 1.8575851393188857e-05,
      "loss": 0.3034,
      "step": 60
    },
    {
      "epoch": 0.10852713178294573,
      "grad_norm": 1.6830190420150757,
      "learning_rate": 2.1671826625387e-05,
      "loss": 0.2055,
      "step": 70
    },
    {
      "epoch": 0.12403100775193798,
      "grad_norm": 1.6667404174804688,
      "learning_rate": 2.476780185758514e-05,
      "loss": 0.1391,
      "step": 80
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 1.0958671569824219,
      "learning_rate": 2.7863777089783283e-05,
      "loss": 0.1286,
      "step": 90
    },
    {
      "epoch": 0.15503875968992248,
      "grad_norm": 1.1141353845596313,
      "learning_rate": 3.0959752321981425e-05,
      "loss": 0.0933,
      "step": 100
    },
    {
      "epoch": 0.15503875968992248,
      "eval_loss": 0.08204219490289688,
      "eval_runtime": 153.6607,
      "eval_samples_per_second": 7.464,
      "eval_steps_per_second": 7.464,
      "step": 100
    },
    {
      "epoch": 0.17054263565891473,
      "grad_norm": 2.2812178134918213,
      "learning_rate": 3.4055727554179566e-05,
      "loss": 0.1012,
      "step": 110
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 2.2276628017425537,
      "learning_rate": 3.7151702786377715e-05,
      "loss": 0.0923,
      "step": 120
    },
    {
      "epoch": 0.20155038759689922,
      "grad_norm": 1.0762747526168823,
      "learning_rate": 4.024767801857585e-05,
      "loss": 0.1105,
      "step": 130
    },
    {
      "epoch": 0.21705426356589147,
      "grad_norm": 1.2775787115097046,
      "learning_rate": 4.3343653250774e-05,
      "loss": 0.0983,
      "step": 140
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.8789132237434387,
      "learning_rate": 4.6439628482972134e-05,
      "loss": 0.0662,
      "step": 150
    },
    {
      "epoch": 0.23255813953488372,
      "eval_loss": 0.0667666420340538,
      "eval_runtime": 153.5687,
      "eval_samples_per_second": 7.469,
      "eval_steps_per_second": 7.469,
      "step": 150
    },
    {
      "epoch": 0.24806201550387597,
      "grad_norm": 0.9150206446647644,
      "learning_rate": 4.953560371517028e-05,
      "loss": 0.079,
      "step": 160
    },
    {
      "epoch": 0.26356589147286824,
      "grad_norm": 1.2402392625808716,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.0753,
      "step": 170
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.6241523027420044,
      "learning_rate": 5.5727554179566566e-05,
      "loss": 0.0691,
      "step": 180
    },
    {
      "epoch": 0.29457364341085274,
      "grad_norm": 1.7111248970031738,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.0706,
      "step": 190
    },
    {
      "epoch": 0.31007751937984496,
      "grad_norm": 1.1303967237472534,
      "learning_rate": 6.191950464396285e-05,
      "loss": 0.0584,
      "step": 200
    },
    {
      "epoch": 0.31007751937984496,
      "eval_loss": 0.058870695531368256,
      "eval_runtime": 153.1035,
      "eval_samples_per_second": 7.492,
      "eval_steps_per_second": 7.492,
      "step": 200
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.660438597202301,
      "learning_rate": 6.501547987616098e-05,
      "loss": 0.0659,
      "step": 210
    },
    {
      "epoch": 0.34108527131782945,
      "grad_norm": 0.8467134237289429,
      "learning_rate": 6.811145510835913e-05,
      "loss": 0.0563,
      "step": 220
    },
    {
      "epoch": 0.35658914728682173,
      "grad_norm": 0.5430559515953064,
      "learning_rate": 7.120743034055728e-05,
      "loss": 0.06,
      "step": 230
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.5925267338752747,
      "learning_rate": 7.430340557275543e-05,
      "loss": 0.0844,
      "step": 240
    },
    {
      "epoch": 0.3875968992248062,
      "grad_norm": 0.6996983289718628,
      "learning_rate": 7.739938080495357e-05,
      "loss": 0.0666,
      "step": 250
    },
    {
      "epoch": 0.3875968992248062,
      "eval_loss": 0.05274031311273575,
      "eval_runtime": 153.2155,
      "eval_samples_per_second": 7.486,
      "eval_steps_per_second": 7.486,
      "step": 250
    },
    {
      "epoch": 0.40310077519379844,
      "grad_norm": 0.6012355089187622,
      "learning_rate": 8.04953560371517e-05,
      "loss": 0.0531,
      "step": 260
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.5030301809310913,
      "learning_rate": 8.359133126934985e-05,
      "loss": 0.0542,
      "step": 270
    },
    {
      "epoch": 0.43410852713178294,
      "grad_norm": 0.5277416706085205,
      "learning_rate": 8.6687306501548e-05,
      "loss": 0.0835,
      "step": 280
    },
    {
      "epoch": 0.4496124031007752,
      "grad_norm": 1.0271929502487183,
      "learning_rate": 8.978328173374613e-05,
      "loss": 0.0561,
      "step": 290
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.5835898518562317,
      "learning_rate": 9.287925696594427e-05,
      "loss": 0.0448,
      "step": 300
    },
    {
      "epoch": 0.46511627906976744,
      "eval_loss": 0.052115585654973984,
      "eval_runtime": 153.2036,
      "eval_samples_per_second": 7.487,
      "eval_steps_per_second": 7.487,
      "step": 300
    },
    {
      "epoch": 0.4806201550387597,
      "grad_norm": 0.5142839550971985,
      "learning_rate": 9.597523219814242e-05,
      "loss": 0.0597,
      "step": 310
    },
    {
      "epoch": 0.49612403100775193,
      "grad_norm": 0.6132373213768005,
      "learning_rate": 9.907120743034056e-05,
      "loss": 0.0406,
      "step": 320
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.37188827991485596,
      "learning_rate": 9.999856438185238e-05,
      "loss": 0.0499,
      "step": 330
    },
    {
      "epoch": 0.5271317829457365,
      "grad_norm": 0.4470522701740265,
      "learning_rate": 9.999153298122152e-05,
      "loss": 0.0558,
      "step": 340
    },
    {
      "epoch": 0.5426356589147286,
      "grad_norm": 0.36978986859321594,
      "learning_rate": 9.997864293614054e-05,
      "loss": 0.0474,
      "step": 350
    },
    {
      "epoch": 0.5426356589147286,
      "eval_loss": 0.048954565078020096,
      "eval_runtime": 153.2276,
      "eval_samples_per_second": 7.486,
      "eval_steps_per_second": 7.486,
      "step": 350
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.38897690176963806,
      "learning_rate": 9.995989575722902e-05,
      "loss": 0.0462,
      "step": 360
    },
    {
      "epoch": 0.5736434108527132,
      "grad_norm": 0.4426390826702118,
      "learning_rate": 9.993529364152018e-05,
      "loss": 0.063,
      "step": 370
    },
    {
      "epoch": 0.5891472868217055,
      "grad_norm": 0.42807143926620483,
      "learning_rate": 9.990483947220319e-05,
      "loss": 0.0606,
      "step": 380
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.45578905940055847,
      "learning_rate": 9.986853681828546e-05,
      "loss": 0.0531,
      "step": 390
    },
    {
      "epoch": 0.6201550387596899,
      "grad_norm": 0.6233142614364624,
      "learning_rate": 9.982638993417425e-05,
      "loss": 0.0546,
      "step": 400
    },
    {
      "epoch": 0.6201550387596899,
      "eval_loss": 0.0430627167224884,
      "eval_runtime": 153.1943,
      "eval_samples_per_second": 7.487,
      "eval_steps_per_second": 7.487,
      "step": 400
    },
    {
      "epoch": 0.6356589147286822,
      "grad_norm": 0.380621999502182,
      "learning_rate": 9.977840375917817e-05,
      "loss": 0.0509,
      "step": 410
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.3944797217845917,
      "learning_rate": 9.972458391692827e-05,
      "loss": 0.06,
      "step": 420
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.4984765350818634,
      "learning_rate": 9.966493671471904e-05,
      "loss": 0.0507,
      "step": 430
    },
    {
      "epoch": 0.6821705426356589,
      "grad_norm": 0.3995707929134369,
      "learning_rate": 9.959946914276922e-05,
      "loss": 0.0597,
      "step": 440
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.5828382968902588,
      "learning_rate": 9.952818887340257e-05,
      "loss": 0.0432,
      "step": 450
    },
    {
      "epoch": 0.6976744186046512,
      "eval_loss": 0.03933678939938545,
      "eval_runtime": 153.1698,
      "eval_samples_per_second": 7.488,
      "eval_steps_per_second": 7.488,
      "step": 450
    },
    {
      "epoch": 0.7131782945736435,
      "grad_norm": 0.37284985184669495,
      "learning_rate": 9.945110426014878e-05,
      "loss": 0.0521,
      "step": 460
    },
    {
      "epoch": 0.7286821705426356,
      "grad_norm": 0.48646485805511475,
      "learning_rate": 9.936822433676444e-05,
      "loss": 0.0431,
      "step": 470
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.342167466878891,
      "learning_rate": 9.927955881617444e-05,
      "loss": 0.0555,
      "step": 480
    },
    {
      "epoch": 0.7596899224806202,
      "grad_norm": 0.2680320143699646,
      "learning_rate": 9.918511808933358e-05,
      "loss": 0.0435,
      "step": 490
    },
    {
      "epoch": 0.7751937984496124,
      "grad_norm": 0.30043846368789673,
      "learning_rate": 9.908491322400885e-05,
      "loss": 0.0526,
      "step": 500
    },
    {
      "epoch": 0.7751937984496124,
      "eval_loss": 0.04014807939529419,
      "eval_runtime": 153.0483,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 500
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.13588973879814148,
      "learning_rate": 9.897895596348247e-05,
      "loss": 0.0466,
      "step": 510
    },
    {
      "epoch": 0.8062015503875969,
      "grad_norm": 0.5235325694084167,
      "learning_rate": 9.886725872517552e-05,
      "loss": 0.0425,
      "step": 520
    },
    {
      "epoch": 0.8217054263565892,
      "grad_norm": 0.32793891429901123,
      "learning_rate": 9.874983459919277e-05,
      "loss": 0.0495,
      "step": 530
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.3768357038497925,
      "learning_rate": 9.862669734678867e-05,
      "loss": 0.0371,
      "step": 540
    },
    {
      "epoch": 0.8527131782945736,
      "grad_norm": 0.29829540848731995,
      "learning_rate": 9.849786139875452e-05,
      "loss": 0.0506,
      "step": 550
    },
    {
      "epoch": 0.8527131782945736,
      "eval_loss": 0.0400104746222496,
      "eval_runtime": 153.0848,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 550
    },
    {
      "epoch": 0.8682170542635659,
      "grad_norm": 0.5966230630874634,
      "learning_rate": 9.836334185372738e-05,
      "loss": 0.0531,
      "step": 560
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.36028406023979187,
      "learning_rate": 9.822315447642056e-05,
      "loss": 0.0434,
      "step": 570
    },
    {
      "epoch": 0.8992248062015504,
      "grad_norm": 0.5609809756278992,
      "learning_rate": 9.807731569577615e-05,
      "loss": 0.0372,
      "step": 580
    },
    {
      "epoch": 0.9147286821705426,
      "grad_norm": 1.050203800201416,
      "learning_rate": 9.792584260303964e-05,
      "loss": 0.0397,
      "step": 590
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 1.292805552482605,
      "learning_rate": 9.776875294975698e-05,
      "loss": 0.0622,
      "step": 600
    },
    {
      "epoch": 0.9302325581395349,
      "eval_loss": 0.04188257083296776,
      "eval_runtime": 153.0764,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 600
    },
    {
      "epoch": 0.9457364341085271,
      "grad_norm": 0.429943323135376,
      "learning_rate": 9.760606514569416e-05,
      "loss": 0.0454,
      "step": 610
    },
    {
      "epoch": 0.9612403100775194,
      "grad_norm": 0.2638005316257477,
      "learning_rate": 9.743779825667984e-05,
      "loss": 0.0413,
      "step": 620
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.4185207784175873,
      "learning_rate": 9.726397200237085e-05,
      "loss": 0.0404,
      "step": 630
    },
    {
      "epoch": 0.9922480620155039,
      "grad_norm": 0.2455182522535324,
      "learning_rate": 9.708460675394126e-05,
      "loss": 0.0549,
      "step": 640
    },
    {
      "epoch": 1.0077519379844961,
      "grad_norm": 0.16050851345062256,
      "learning_rate": 9.689972353169499e-05,
      "loss": 0.0363,
      "step": 650
    },
    {
      "epoch": 1.0077519379844961,
      "eval_loss": 0.03797760233283043,
      "eval_runtime": 153.0835,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 650
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.6184531450271606,
      "learning_rate": 9.67093440026024e-05,
      "loss": 0.0361,
      "step": 660
    },
    {
      "epoch": 1.0387596899224807,
      "grad_norm": 0.26317697763442993,
      "learning_rate": 9.651349047776112e-05,
      "loss": 0.0465,
      "step": 670
    },
    {
      "epoch": 1.054263565891473,
      "grad_norm": 0.32697394490242004,
      "learning_rate": 9.631218590978126e-05,
      "loss": 0.0278,
      "step": 680
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.22065386176109314,
      "learning_rate": 9.610545389009562e-05,
      "loss": 0.0408,
      "step": 690
    },
    {
      "epoch": 1.0852713178294573,
      "grad_norm": 0.3184354901313782,
      "learning_rate": 9.589331864619486e-05,
      "loss": 0.032,
      "step": 700
    },
    {
      "epoch": 1.0852713178294573,
      "eval_loss": 0.03769562765955925,
      "eval_runtime": 153.1203,
      "eval_samples_per_second": 7.491,
      "eval_steps_per_second": 7.491,
      "step": 700
    },
    {
      "epoch": 1.1007751937984496,
      "grad_norm": 0.363477885723114,
      "learning_rate": 9.567580503878833e-05,
      "loss": 0.0285,
      "step": 710
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.2624370753765106,
      "learning_rate": 9.545293855889043e-05,
      "loss": 0.035,
      "step": 720
    },
    {
      "epoch": 1.1317829457364341,
      "grad_norm": 0.27370238304138184,
      "learning_rate": 9.522474532483337e-05,
      "loss": 0.0371,
      "step": 730
    },
    {
      "epoch": 1.1472868217054264,
      "grad_norm": 0.3022075295448303,
      "learning_rate": 9.499125207920622e-05,
      "loss": 0.0443,
      "step": 740
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.22131796181201935,
      "learning_rate": 9.475248618572096e-05,
      "loss": 0.0361,
      "step": 750
    },
    {
      "epoch": 1.1627906976744187,
      "eval_loss": 0.04346157982945442,
      "eval_runtime": 153.1316,
      "eval_samples_per_second": 7.49,
      "eval_steps_per_second": 7.49,
      "step": 750
    },
    {
      "epoch": 1.178294573643411,
      "grad_norm": 0.5927603244781494,
      "learning_rate": 9.45084756260055e-05,
      "loss": 0.0396,
      "step": 760
    },
    {
      "epoch": 1.193798449612403,
      "grad_norm": 0.12391407042741776,
      "learning_rate": 9.425924899632458e-05,
      "loss": 0.0336,
      "step": 770
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.1478697508573532,
      "learning_rate": 9.400483550422845e-05,
      "loss": 0.0299,
      "step": 780
    },
    {
      "epoch": 1.2248062015503876,
      "grad_norm": 0.4696236848831177,
      "learning_rate": 9.374526496512992e-05,
      "loss": 0.0434,
      "step": 790
    },
    {
      "epoch": 1.2403100775193798,
      "grad_norm": 0.35958048701286316,
      "learning_rate": 9.348056779881025e-05,
      "loss": 0.0256,
      "step": 800
    },
    {
      "epoch": 1.2403100775193798,
      "eval_loss": 0.03653433918952942,
      "eval_runtime": 152.9162,
      "eval_samples_per_second": 7.501,
      "eval_steps_per_second": 7.501,
      "step": 800
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.42049720883369446,
      "learning_rate": 9.321077502585417e-05,
      "loss": 0.0277,
      "step": 810
    },
    {
      "epoch": 1.2713178294573644,
      "grad_norm": 0.3786177933216095,
      "learning_rate": 9.293591826401451e-05,
      "loss": 0.0389,
      "step": 820
    },
    {
      "epoch": 1.2868217054263567,
      "grad_norm": 0.4121144115924835,
      "learning_rate": 9.265602972450673e-05,
      "loss": 0.0368,
      "step": 830
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.43422478437423706,
      "learning_rate": 9.237114220823413e-05,
      "loss": 0.0324,
      "step": 840
    },
    {
      "epoch": 1.3178294573643412,
      "grad_norm": 0.18510602414608002,
      "learning_rate": 9.208128910194377e-05,
      "loss": 0.0361,
      "step": 850
    },
    {
      "epoch": 1.3178294573643412,
      "eval_loss": 0.0356990210711956,
      "eval_runtime": 152.9527,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 850
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.5393287539482117,
      "learning_rate": 9.17865043743138e-05,
      "loss": 0.0408,
      "step": 860
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.10767626017332077,
      "learning_rate": 9.148682257197245e-05,
      "loss": 0.0276,
      "step": 870
    },
    {
      "epoch": 1.3643410852713178,
      "grad_norm": 0.737008810043335,
      "learning_rate": 9.118227881544967e-05,
      "loss": 0.0485,
      "step": 880
    },
    {
      "epoch": 1.37984496124031,
      "grad_norm": 0.6310213804244995,
      "learning_rate": 9.087290879506104e-05,
      "loss": 0.0324,
      "step": 890
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.1790456771850586,
      "learning_rate": 9.055874876672519e-05,
      "loss": 0.0428,
      "step": 900
    },
    {
      "epoch": 1.3953488372093024,
      "eval_loss": 0.03691842034459114,
      "eval_runtime": 152.9673,
      "eval_samples_per_second": 7.498,
      "eval_steps_per_second": 7.498,
      "step": 900
    },
    {
      "epoch": 1.4108527131782946,
      "grad_norm": 0.21055588126182556,
      "learning_rate": 9.023983554771492e-05,
      "loss": 0.0314,
      "step": 910
    },
    {
      "epoch": 1.4263565891472867,
      "grad_norm": 0.3077942728996277,
      "learning_rate": 8.991620651234242e-05,
      "loss": 0.0287,
      "step": 920
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.5952494740486145,
      "learning_rate": 8.958789958757928e-05,
      "loss": 0.0325,
      "step": 930
    },
    {
      "epoch": 1.4573643410852712,
      "grad_norm": 0.37136706709861755,
      "learning_rate": 8.925495324861177e-05,
      "loss": 0.034,
      "step": 940
    },
    {
      "epoch": 1.4728682170542635,
      "grad_norm": 0.4016863703727722,
      "learning_rate": 8.89174065143318e-05,
      "loss": 0.0423,
      "step": 950
    },
    {
      "epoch": 1.4728682170542635,
      "eval_loss": 0.036706067621707916,
      "eval_runtime": 152.9951,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 950
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.3515150547027588,
      "learning_rate": 8.857529894276412e-05,
      "loss": 0.0351,
      "step": 960
    },
    {
      "epoch": 1.503875968992248,
      "grad_norm": 0.3203086853027344,
      "learning_rate": 8.822867062643059e-05,
      "loss": 0.032,
      "step": 970
    },
    {
      "epoch": 1.5193798449612403,
      "grad_norm": 0.22203227877616882,
      "learning_rate": 8.78775621876514e-05,
      "loss": 0.0348,
      "step": 980
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.3345692753791809,
      "learning_rate": 8.752201477378459e-05,
      "loss": 0.03,
      "step": 990
    },
    {
      "epoch": 1.550387596899225,
      "grad_norm": 0.2074439525604248,
      "learning_rate": 8.716207005240382e-05,
      "loss": 0.0298,
      "step": 1000
    },
    {
      "epoch": 1.550387596899225,
      "eval_loss": 0.03816875442862511,
      "eval_runtime": 153.003,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 1000
    },
    {
      "epoch": 1.5658914728682172,
      "grad_norm": 0.49668076634407043,
      "learning_rate": 8.679777020641525e-05,
      "loss": 0.0411,
      "step": 1010
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.3684081435203552,
      "learning_rate": 8.6429157929114e-05,
      "loss": 0.0428,
      "step": 1020
    },
    {
      "epoch": 1.5968992248062015,
      "grad_norm": 0.39902442693710327,
      "learning_rate": 8.60562764191808e-05,
      "loss": 0.0352,
      "step": 1030
    },
    {
      "epoch": 1.6124031007751938,
      "grad_norm": 0.27167969942092896,
      "learning_rate": 8.567916937561944e-05,
      "loss": 0.0338,
      "step": 1040
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.14951859414577484,
      "learning_rate": 8.529788099263554e-05,
      "loss": 0.0357,
      "step": 1050
    },
    {
      "epoch": 1.627906976744186,
      "eval_loss": 0.03658704832196236,
      "eval_runtime": 153.0001,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 1050
    },
    {
      "epoch": 1.6434108527131783,
      "grad_norm": 0.22118541598320007,
      "learning_rate": 8.49124559544573e-05,
      "loss": 0.0297,
      "step": 1060
    },
    {
      "epoch": 1.6589147286821704,
      "grad_norm": 0.32734546065330505,
      "learning_rate": 8.452293943009889e-05,
      "loss": 0.0299,
      "step": 1070
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.25646790862083435,
      "learning_rate": 8.412937706806692e-05,
      "loss": 0.0374,
      "step": 1080
    },
    {
      "epoch": 1.689922480620155,
      "grad_norm": 0.3148137331008911,
      "learning_rate": 8.373181499101077e-05,
      "loss": 0.0355,
      "step": 1090
    },
    {
      "epoch": 1.7054263565891472,
      "grad_norm": 0.48944637179374695,
      "learning_rate": 8.333029979031737e-05,
      "loss": 0.0271,
      "step": 1100
    },
    {
      "epoch": 1.7054263565891472,
      "eval_loss": 0.037469059228897095,
      "eval_runtime": 152.9713,
      "eval_samples_per_second": 7.498,
      "eval_steps_per_second": 7.498,
      "step": 1100
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.32591354846954346,
      "learning_rate": 8.292487852065104e-05,
      "loss": 0.0478,
      "step": 1110
    },
    {
      "epoch": 1.7364341085271318,
      "grad_norm": 0.4463321566581726,
      "learning_rate": 8.251559869443897e-05,
      "loss": 0.0342,
      "step": 1120
    },
    {
      "epoch": 1.751937984496124,
      "grad_norm": 0.35801300406455994,
      "learning_rate": 8.210250827630313e-05,
      "loss": 0.0328,
      "step": 1130
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.2991816997528076,
      "learning_rate": 8.168565567743924e-05,
      "loss": 0.029,
      "step": 1140
    },
    {
      "epoch": 1.7829457364341086,
      "grad_norm": 0.3584236204624176,
      "learning_rate": 8.126508974994321e-05,
      "loss": 0.0325,
      "step": 1150
    },
    {
      "epoch": 1.7829457364341086,
      "eval_loss": 0.03696885332465172,
      "eval_runtime": 152.9978,
      "eval_samples_per_second": 7.497,
      "eval_steps_per_second": 7.497,
      "step": 1150
    },
    {
      "epoch": 1.7984496124031009,
      "grad_norm": 0.31954172253608704,
      "learning_rate": 8.084085978108611e-05,
      "loss": 0.0202,
      "step": 1160
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.2720739543437958,
      "learning_rate": 8.041301548753804e-05,
      "loss": 0.0299,
      "step": 1170
    },
    {
      "epoch": 1.8294573643410854,
      "grad_norm": 0.2808537781238556,
      "learning_rate": 7.998160700954165e-05,
      "loss": 0.0248,
      "step": 1180
    },
    {
      "epoch": 1.8449612403100775,
      "grad_norm": 0.42915990948677063,
      "learning_rate": 7.954668490503621e-05,
      "loss": 0.0362,
      "step": 1190
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.29753658175468445,
      "learning_rate": 7.910830014373237e-05,
      "loss": 0.0328,
      "step": 1200
    },
    {
      "epoch": 1.8604651162790697,
      "eval_loss": 0.03458752855658531,
      "eval_runtime": 153.173,
      "eval_samples_per_second": 7.488,
      "eval_steps_per_second": 7.488,
      "step": 1200
    },
    {
      "epoch": 1.875968992248062,
      "grad_norm": 0.14288680255413055,
      "learning_rate": 7.866650410113907e-05,
      "loss": 0.0288,
      "step": 1210
    },
    {
      "epoch": 1.8914728682170543,
      "grad_norm": 0.40016353130340576,
      "learning_rate": 7.822134855254263e-05,
      "loss": 0.038,
      "step": 1220
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.3772751986980438,
      "learning_rate": 7.777288566693905e-05,
      "loss": 0.0316,
      "step": 1230
    },
    {
      "epoch": 1.9224806201550386,
      "grad_norm": 0.419696182012558,
      "learning_rate": 7.732116800092018e-05,
      "loss": 0.0345,
      "step": 1240
    },
    {
      "epoch": 1.937984496124031,
      "grad_norm": 0.34854450821876526,
      "learning_rate": 7.686624849251453e-05,
      "loss": 0.0373,
      "step": 1250
    },
    {
      "epoch": 1.937984496124031,
      "eval_loss": 0.034580811858177185,
      "eval_runtime": 153.0298,
      "eval_samples_per_second": 7.495,
      "eval_steps_per_second": 7.495,
      "step": 1250
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.27317097783088684,
      "learning_rate": 7.640818045498324e-05,
      "loss": 0.0427,
      "step": 1260
    },
    {
      "epoch": 1.9689922480620154,
      "grad_norm": 0.1250971257686615,
      "learning_rate": 7.59470175705722e-05,
      "loss": 0.0353,
      "step": 1270
    },
    {
      "epoch": 1.9844961240310077,
      "grad_norm": 0.33024606108665466,
      "learning_rate": 7.548281388422088e-05,
      "loss": 0.036,
      "step": 1280
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.241469606757164,
      "learning_rate": 7.50156237972286e-05,
      "loss": 0.0282,
      "step": 1290
    },
    {
      "epoch": 2.0155038759689923,
      "grad_norm": 0.2277582883834839,
      "learning_rate": 7.454550206087921e-05,
      "loss": 0.0219,
      "step": 1300
    },
    {
      "epoch": 2.0155038759689923,
      "eval_loss": 0.03509150817990303,
      "eval_runtime": 153.0277,
      "eval_samples_per_second": 7.495,
      "eval_steps_per_second": 7.495,
      "step": 1300
    },
    {
      "epoch": 2.0310077519379846,
      "grad_norm": 0.23277609050273895,
      "learning_rate": 7.407250377002451e-05,
      "loss": 0.0196,
      "step": 1310
    },
    {
      "epoch": 2.046511627906977,
      "grad_norm": 0.23019228875637054,
      "learning_rate": 7.359668435662757e-05,
      "loss": 0.0237,
      "step": 1320
    },
    {
      "epoch": 2.062015503875969,
      "grad_norm": 0.31736159324645996,
      "learning_rate": 7.311809958326659e-05,
      "loss": 0.017,
      "step": 1330
    },
    {
      "epoch": 2.0775193798449614,
      "grad_norm": 0.2850285768508911,
      "learning_rate": 7.263680553659983e-05,
      "loss": 0.0198,
      "step": 1340
    },
    {
      "epoch": 2.0930232558139537,
      "grad_norm": 0.49550530314445496,
      "learning_rate": 7.21528586207927e-05,
      "loss": 0.0179,
      "step": 1350
    },
    {
      "epoch": 2.0930232558139537,
      "eval_loss": 0.038030773401260376,
      "eval_runtime": 153.0983,
      "eval_samples_per_second": 7.492,
      "eval_steps_per_second": 7.492,
      "step": 1350
    },
    {
      "epoch": 2.108527131782946,
      "grad_norm": 0.2637981176376343,
      "learning_rate": 7.166631555090759e-05,
      "loss": 0.0242,
      "step": 1360
    },
    {
      "epoch": 2.124031007751938,
      "grad_norm": 0.3910108506679535,
      "learning_rate": 7.117723334625731e-05,
      "loss": 0.0204,
      "step": 1370
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 0.37799614667892456,
      "learning_rate": 7.068566932372278e-05,
      "loss": 0.0197,
      "step": 1380
    },
    {
      "epoch": 2.1550387596899223,
      "grad_norm": 0.23790650069713593,
      "learning_rate": 7.019168109103599e-05,
      "loss": 0.0173,
      "step": 1390
    },
    {
      "epoch": 2.1705426356589146,
      "grad_norm": 0.23004008829593658,
      "learning_rate": 6.96953265400287e-05,
      "loss": 0.018,
      "step": 1400
    },
    {
      "epoch": 2.1705426356589146,
      "eval_loss": 0.03982286527752876,
      "eval_runtime": 153.0702,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 1400
    },
    {
      "epoch": 2.186046511627907,
      "grad_norm": 0.36417263746261597,
      "learning_rate": 6.9196663839848e-05,
      "loss": 0.0157,
      "step": 1410
    },
    {
      "epoch": 2.201550387596899,
      "grad_norm": 0.4610263705253601,
      "learning_rate": 6.869575143013934e-05,
      "loss": 0.0164,
      "step": 1420
    },
    {
      "epoch": 2.2170542635658914,
      "grad_norm": 0.2598446011543274,
      "learning_rate": 6.819264801419771e-05,
      "loss": 0.0305,
      "step": 1430
    },
    {
      "epoch": 2.2325581395348837,
      "grad_norm": 0.16078923642635345,
      "learning_rate": 6.768741255208813e-05,
      "loss": 0.0195,
      "step": 1440
    },
    {
      "epoch": 2.248062015503876,
      "grad_norm": 0.43891769647598267,
      "learning_rate": 6.718010425373599e-05,
      "loss": 0.0203,
      "step": 1450
    },
    {
      "epoch": 2.248062015503876,
      "eval_loss": 0.038189034909009933,
      "eval_runtime": 153.0558,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 1450
    },
    {
      "epoch": 2.2635658914728682,
      "grad_norm": 0.329623281955719,
      "learning_rate": 6.667078257198796e-05,
      "loss": 0.0188,
      "step": 1460
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.4400428831577301,
      "learning_rate": 6.615950719564466e-05,
      "loss": 0.0163,
      "step": 1470
    },
    {
      "epoch": 2.294573643410853,
      "grad_norm": 0.45587706565856934,
      "learning_rate": 6.56463380424655e-05,
      "loss": 0.0295,
      "step": 1480
    },
    {
      "epoch": 2.310077519379845,
      "grad_norm": 0.2078193873167038,
      "learning_rate": 6.51313352521468e-05,
      "loss": 0.0157,
      "step": 1490
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.44753116369247437,
      "learning_rate": 6.46145591792738e-05,
      "loss": 0.0257,
      "step": 1500
    },
    {
      "epoch": 2.3255813953488373,
      "eval_loss": 0.040530748665332794,
      "eval_runtime": 153.0525,
      "eval_samples_per_second": 7.494,
      "eval_steps_per_second": 7.494,
      "step": 1500
    },
    {
      "epoch": 2.3410852713178296,
      "grad_norm": 0.1375536173582077,
      "learning_rate": 6.409607038624759e-05,
      "loss": 0.0165,
      "step": 1510
    },
    {
      "epoch": 2.356589147286822,
      "grad_norm": 0.30814239382743835,
      "learning_rate": 6.357592963618761e-05,
      "loss": 0.028,
      "step": 1520
    },
    {
      "epoch": 2.3720930232558137,
      "grad_norm": 0.363112211227417,
      "learning_rate": 6.305419788581067e-05,
      "loss": 0.0211,
      "step": 1530
    },
    {
      "epoch": 2.387596899224806,
      "grad_norm": 0.07269316166639328,
      "learning_rate": 6.253093627828724e-05,
      "loss": 0.0195,
      "step": 1540
    },
    {
      "epoch": 2.4031007751937983,
      "grad_norm": 0.41061273217201233,
      "learning_rate": 6.200620613607597e-05,
      "loss": 0.0165,
      "step": 1550
    },
    {
      "epoch": 2.4031007751937983,
      "eval_loss": 0.0382014699280262,
      "eval_runtime": 153.069,
      "eval_samples_per_second": 7.493,
      "eval_steps_per_second": 7.493,
      "step": 1550
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.37901002168655396,
      "learning_rate": 6.148006895373706e-05,
      "loss": 0.0202,
      "step": 1560
    },
    {
      "epoch": 2.434108527131783,
      "grad_norm": 0.23691922426223755,
      "learning_rate": 6.0952586390725644e-05,
      "loss": 0.0181,
      "step": 1570
    },
    {
      "epoch": 2.449612403100775,
      "grad_norm": 0.40029415488243103,
      "learning_rate": 6.042382026416563e-05,
      "loss": 0.0179,
      "step": 1580
    },
    {
      "epoch": 2.4651162790697674,
      "grad_norm": 0.27985653281211853,
      "learning_rate": 5.989383254160529e-05,
      "loss": 0.0203,
      "step": 1590
    },
    {
      "epoch": 2.4806201550387597,
      "grad_norm": 0.34605786204338074,
      "learning_rate": 5.936268533375506e-05,
      "loss": 0.0212,
      "step": 1600
    },
    {
      "epoch": 2.4806201550387597,
      "eval_loss": 0.037529706954956055,
      "eval_runtime": 152.9495,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 1600
    },
    {
      "epoch": 2.496124031007752,
      "grad_norm": 0.5515694618225098,
      "learning_rate": 5.883044088720865e-05,
      "loss": 0.0262,
      "step": 1610
    },
    {
      "epoch": 2.511627906976744,
      "grad_norm": 0.1899733543395996,
      "learning_rate": 5.829716157714814e-05,
      "loss": 0.0235,
      "step": 1620
    },
    {
      "epoch": 2.5271317829457365,
      "grad_norm": 0.2844202518463135,
      "learning_rate": 5.77629099000341e-05,
      "loss": 0.0279,
      "step": 1630
    },
    {
      "epoch": 2.5426356589147288,
      "grad_norm": 0.2058931291103363,
      "learning_rate": 5.7227748466281486e-05,
      "loss": 0.0257,
      "step": 1640
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.2736983597278595,
      "learning_rate": 5.669173999292208e-05,
      "loss": 0.0315,
      "step": 1650
    },
    {
      "epoch": 2.558139534883721,
      "eval_loss": 0.03725634142756462,
      "eval_runtime": 152.9551,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 1650
    },
    {
      "epoch": 2.5736434108527133,
      "grad_norm": 0.41483446955680847,
      "learning_rate": 5.615494729625458e-05,
      "loss": 0.0232,
      "step": 1660
    },
    {
      "epoch": 2.5891472868217056,
      "grad_norm": 0.36093488335609436,
      "learning_rate": 5.561743328448296e-05,
      "loss": 0.0294,
      "step": 1670
    },
    {
      "epoch": 2.604651162790698,
      "grad_norm": 0.18070422112941742,
      "learning_rate": 5.5079260950344035e-05,
      "loss": 0.0197,
      "step": 1680
    },
    {
      "epoch": 2.62015503875969,
      "grad_norm": 0.12596550583839417,
      "learning_rate": 5.454049336372531e-05,
      "loss": 0.0197,
      "step": 1690
    },
    {
      "epoch": 2.6356589147286824,
      "grad_norm": 0.34723135828971863,
      "learning_rate": 5.4001193664273454e-05,
      "loss": 0.0155,
      "step": 1700
    },
    {
      "epoch": 2.6356589147286824,
      "eval_loss": 0.03793378174304962,
      "eval_runtime": 152.9691,
      "eval_samples_per_second": 7.498,
      "eval_steps_per_second": 7.498,
      "step": 1700
    },
    {
      "epoch": 2.6511627906976747,
      "grad_norm": 0.223840594291687,
      "learning_rate": 5.346142505399495e-05,
      "loss": 0.0173,
      "step": 1710
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.16230659186840057,
      "learning_rate": 5.292125078984925e-05,
      "loss": 0.018,
      "step": 1720
    },
    {
      "epoch": 2.682170542635659,
      "grad_norm": 0.14621299505233765,
      "learning_rate": 5.2380734176335425e-05,
      "loss": 0.0184,
      "step": 1730
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.33405810594558716,
      "learning_rate": 5.183993855807343e-05,
      "loss": 0.0238,
      "step": 1740
    },
    {
      "epoch": 2.7131782945736433,
      "grad_norm": 0.15875190496444702,
      "learning_rate": 5.1298927312380586e-05,
      "loss": 0.0188,
      "step": 1750
    },
    {
      "epoch": 2.7131782945736433,
      "eval_loss": 0.03785843402147293,
      "eval_runtime": 152.9656,
      "eval_samples_per_second": 7.498,
      "eval_steps_per_second": 7.498,
      "step": 1750
    },
    {
      "epoch": 2.7286821705426356,
      "grad_norm": 0.2579461634159088,
      "learning_rate": 5.075776384184411e-05,
      "loss": 0.0271,
      "step": 1760
    },
    {
      "epoch": 2.744186046511628,
      "grad_norm": 0.2938865125179291,
      "learning_rate": 5.021651156689094e-05,
      "loss": 0.0249,
      "step": 1770
    },
    {
      "epoch": 2.75968992248062,
      "grad_norm": 0.29400309920310974,
      "learning_rate": 4.967523391835521e-05,
      "loss": 0.0234,
      "step": 1780
    },
    {
      "epoch": 2.7751937984496124,
      "grad_norm": 0.34708184003829956,
      "learning_rate": 4.9133994330044644e-05,
      "loss": 0.0253,
      "step": 1790
    },
    {
      "epoch": 2.7906976744186047,
      "grad_norm": 0.5770057439804077,
      "learning_rate": 4.85928562313066e-05,
      "loss": 0.0195,
      "step": 1800
    },
    {
      "epoch": 2.7906976744186047,
      "eval_loss": 0.03974982351064682,
      "eval_runtime": 152.93,
      "eval_samples_per_second": 7.5,
      "eval_steps_per_second": 7.5,
      "step": 1800
    },
    {
      "epoch": 2.806201550387597,
      "grad_norm": 0.2976359724998474,
      "learning_rate": 4.8051883039594616e-05,
      "loss": 0.017,
      "step": 1810
    },
    {
      "epoch": 2.8217054263565893,
      "grad_norm": 0.3403018116950989,
      "learning_rate": 4.751113815303624e-05,
      "loss": 0.0262,
      "step": 1820
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.7027060389518738,
      "learning_rate": 4.697068494300343e-05,
      "loss": 0.0187,
      "step": 1830
    },
    {
      "epoch": 2.8527131782945734,
      "grad_norm": 0.1998434215784073,
      "learning_rate": 4.6430586746685724e-05,
      "loss": 0.0218,
      "step": 1840
    },
    {
      "epoch": 2.8682170542635657,
      "grad_norm": 0.5356177091598511,
      "learning_rate": 4.589090685966758e-05,
      "loss": 0.0213,
      "step": 1850
    },
    {
      "epoch": 2.8682170542635657,
      "eval_loss": 0.0373452827334404,
      "eval_runtime": 152.9595,
      "eval_samples_per_second": 7.499,
      "eval_steps_per_second": 7.499,
      "step": 1850
    },
    {
      "epoch": 2.883720930232558,
      "grad_norm": 0.27649322152137756,
      "learning_rate": 4.535170852851073e-05,
      "loss": 0.0331,
      "step": 1860
    },
    {
      "epoch": 2.89922480620155,
      "grad_norm": 0.22276313602924347,
      "learning_rate": 4.481305494334201e-05,
      "loss": 0.0188,
      "step": 1870
    },
    {
      "epoch": 2.9147286821705425,
      "grad_norm": 0.15337777137756348,
      "learning_rate": 4.427500923044801e-05,
      "loss": 0.0194,
      "step": 1880
    },
    {
      "epoch": 2.9302325581395348,
      "grad_norm": 0.11468372493982315,
      "learning_rate": 4.373763444487705e-05,
      "loss": 0.0187,
      "step": 1890
    },
    {
      "epoch": 2.945736434108527,
      "grad_norm": 0.31443291902542114,
      "learning_rate": 4.3200993563049725e-05,
      "loss": 0.0171,
      "step": 1900
    },
    {
      "epoch": 2.945736434108527,
      "eval_loss": 0.037420857697725296,
      "eval_runtime": 153.0053,
      "eval_samples_per_second": 7.496,
      "eval_steps_per_second": 7.496,
      "step": 1900
    },
    {
      "epoch": 2.9612403100775193,
      "grad_norm": 0.20770350098609924,
      "learning_rate": 4.266514947537839e-05,
      "loss": 0.0206,
      "step": 1910
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.20626762509346008,
      "learning_rate": 4.2130164978896916e-05,
      "loss": 0.0202,
      "step": 1920
    },
    {
      "epoch": 2.992248062015504,
      "grad_norm": 0.022488873451948166,
      "learning_rate": 4.159610276990137e-05,
      "loss": 0.0145,
      "step": 1930
    },
    {
      "epoch": 3.007751937984496,
      "grad_norm": 0.5588874220848083,
      "learning_rate": 4.106302543660235e-05,
      "loss": 0.0161,
      "step": 1940
    },
    {
      "epoch": 3.0232558139534884,
      "grad_norm": 0.13723929226398468,
      "learning_rate": 4.053099545179028e-05,
      "loss": 0.0108,
      "step": 1950
    },
    {
      "epoch": 3.0232558139534884,
      "eval_loss": 0.03898989409208298,
      "eval_runtime": 152.971,
      "eval_samples_per_second": 7.498,
      "eval_steps_per_second": 7.498,
      "step": 1950
    },
    {
      "epoch": 3.0387596899224807,
      "grad_norm": 0.20176100730895996,
      "learning_rate": 4.0000075165513845e-05,
      "loss": 0.0127,
      "step": 1960
    },
    {
      "epoch": 3.054263565891473,
      "grad_norm": 0.13127611577510834,
      "learning_rate": 3.9470326797773216e-05,
      "loss": 0.009,
      "step": 1970
    },
    {
      "epoch": 3.0697674418604652,
      "grad_norm": 0.2873622477054596,
      "learning_rate": 3.8941812431228166e-05,
      "loss": 0.0098,
      "step": 1980
    },
    {
      "epoch": 3.0852713178294575,
      "grad_norm": 0.3359582722187042,
      "learning_rate": 3.8414594003922515e-05,
      "loss": 0.0128,
      "step": 1990
    },
    {
      "epoch": 3.10077519379845,
      "grad_norm": 0.15983670949935913,
      "learning_rate": 3.788873330202544e-05,
      "loss": 0.0125,
      "step": 2000
    },
    {
      "epoch": 3.10077519379845,
      "eval_loss": 0.04365137219429016,
      "eval_runtime": 153.0405,
      "eval_samples_per_second": 7.495,
      "eval_steps_per_second": 7.495,
      "step": 2000
    },
    {
      "epoch": 3.116279069767442,
      "grad_norm": 0.37191739678382874,
      "learning_rate": 3.736429195259051e-05,
      "loss": 0.0059,
      "step": 2010
    },
    {
      "epoch": 3.1317829457364343,
      "grad_norm": 0.2656456232070923,
      "learning_rate": 3.684133141633358e-05,
      "loss": 0.0118,
      "step": 2020
    },
    {
      "epoch": 3.147286821705426,
      "grad_norm": 0.3109544813632965,
      "learning_rate": 3.6319912980429846e-05,
      "loss": 0.0099,
      "step": 2030
    },
    {
      "epoch": 3.1627906976744184,
      "grad_norm": 0.13874967396259308,
      "learning_rate": 3.580009775133168e-05,
      "loss": 0.0082,
      "step": 2040
    },
    {
      "epoch": 3.1782945736434107,
      "grad_norm": 0.05979770049452782,
      "learning_rate": 3.528194664760714e-05,
      "loss": 0.0046,
      "step": 2050
    },
    {
      "epoch": 3.1782945736434107,
      "eval_loss": 0.04593827575445175,
      "eval_runtime": 152.6999,
      "eval_samples_per_second": 7.511,
      "eval_steps_per_second": 7.511,
      "step": 2050
    },
    {
      "epoch": 3.193798449612403,
      "grad_norm": 0.27979713678359985,
      "learning_rate": 3.476552039280096e-05,
      "loss": 0.0095,
      "step": 2060
    },
    {
      "epoch": 3.2093023255813953,
      "grad_norm": 0.11609620600938797,
      "learning_rate": 3.42508795083181e-05,
      "loss": 0.007,
      "step": 2070
    },
    {
      "epoch": 3.2248062015503876,
      "grad_norm": 0.365993469953537,
      "learning_rate": 3.373808430633106e-05,
      "loss": 0.0054,
      "step": 2080
    },
    {
      "epoch": 3.24031007751938,
      "grad_norm": 0.8401245474815369,
      "learning_rate": 3.32271948827118e-05,
      "loss": 0.0103,
      "step": 2090
    },
    {
      "epoch": 3.255813953488372,
      "grad_norm": 0.2735046446323395,
      "learning_rate": 3.2718271109988863e-05,
      "loss": 0.0059,
      "step": 2100
    },
    {
      "epoch": 3.255813953488372,
      "eval_loss": 0.047852758318185806,
      "eval_runtime": 152.5478,
      "eval_samples_per_second": 7.519,
      "eval_steps_per_second": 7.519,
      "step": 2100
    },
    {
      "epoch": 3.2713178294573644,
      "grad_norm": 0.381459504365921,
      "learning_rate": 3.2211372630330835e-05,
      "loss": 0.0069,
      "step": 2110
    },
    {
      "epoch": 3.2868217054263567,
      "grad_norm": 0.2578096389770508,
      "learning_rate": 3.170655884855661e-05,
      "loss": 0.0124,
      "step": 2120
    },
    {
      "epoch": 3.302325581395349,
      "grad_norm": 0.1936161369085312,
      "learning_rate": 3.120388892517368e-05,
      "loss": 0.0195,
      "step": 2130
    },
    {
      "epoch": 3.317829457364341,
      "grad_norm": 0.20528022944927216,
      "learning_rate": 3.070342176944494e-05,
      "loss": 0.0146,
      "step": 2140
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.12017946690320969,
      "learning_rate": 3.0205216032484805e-05,
      "loss": 0.0088,
      "step": 2150
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 0.043197184801101685,
      "eval_runtime": 152.4637,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2150
    },
    {
      "epoch": 3.3488372093023258,
      "grad_norm": 0.15760734677314758,
      "learning_rate": 2.970933010038599e-05,
      "loss": 0.0107,
      "step": 2160
    },
    {
      "epoch": 3.3643410852713176,
      "grad_norm": 0.4512033462524414,
      "learning_rate": 2.921582208737681e-05,
      "loss": 0.0085,
      "step": 2170
    },
    {
      "epoch": 3.37984496124031,
      "grad_norm": 0.23206011950969696,
      "learning_rate": 2.872474982901081e-05,
      "loss": 0.0169,
      "step": 2180
    },
    {
      "epoch": 3.395348837209302,
      "grad_norm": 0.1863328367471695,
      "learning_rate": 2.8236170875388744e-05,
      "loss": 0.007,
      "step": 2190
    },
    {
      "epoch": 3.4108527131782944,
      "grad_norm": 0.16738107800483704,
      "learning_rate": 2.775014248441422e-05,
      "loss": 0.0074,
      "step": 2200
    },
    {
      "epoch": 3.4108527131782944,
      "eval_loss": 0.045535024255514145,
      "eval_runtime": 152.427,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 2200
    },
    {
      "epoch": 3.4263565891472867,
      "grad_norm": 0.21157847344875336,
      "learning_rate": 2.726672161508341e-05,
      "loss": 0.0106,
      "step": 2210
    },
    {
      "epoch": 3.441860465116279,
      "grad_norm": 0.2758280038833618,
      "learning_rate": 2.678596492080984e-05,
      "loss": 0.0056,
      "step": 2220
    },
    {
      "epoch": 3.4573643410852712,
      "grad_norm": 0.2630395293235779,
      "learning_rate": 2.630792874278516e-05,
      "loss": 0.0072,
      "step": 2230
    },
    {
      "epoch": 3.4728682170542635,
      "grad_norm": 0.4737556278705597,
      "learning_rate": 2.583266910337624e-05,
      "loss": 0.0084,
      "step": 2240
    },
    {
      "epoch": 3.488372093023256,
      "grad_norm": 0.3222001791000366,
      "learning_rate": 2.5360241699559816e-05,
      "loss": 0.0105,
      "step": 2250
    },
    {
      "epoch": 3.488372093023256,
      "eval_loss": 0.04931580275297165,
      "eval_runtime": 152.3529,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 2250
    },
    {
      "epoch": 3.503875968992248,
      "grad_norm": 0.12883061170578003,
      "learning_rate": 2.4890701896395146e-05,
      "loss": 0.0066,
      "step": 2260
    },
    {
      "epoch": 3.5193798449612403,
      "grad_norm": 0.44820183515548706,
      "learning_rate": 2.4424104720535735e-05,
      "loss": 0.0108,
      "step": 2270
    },
    {
      "epoch": 3.5348837209302326,
      "grad_norm": 0.16855889558792114,
      "learning_rate": 2.3960504853780462e-05,
      "loss": 0.0095,
      "step": 2280
    },
    {
      "epoch": 3.550387596899225,
      "grad_norm": 0.04795743152499199,
      "learning_rate": 2.349995662666547e-05,
      "loss": 0.0071,
      "step": 2290
    },
    {
      "epoch": 3.565891472868217,
      "grad_norm": 0.37006649374961853,
      "learning_rate": 2.3042514012096843e-05,
      "loss": 0.0116,
      "step": 2300
    },
    {
      "epoch": 3.565891472868217,
      "eval_loss": 0.05101173743605614,
      "eval_runtime": 152.5604,
      "eval_samples_per_second": 7.518,
      "eval_steps_per_second": 7.518,
      "step": 2300
    },
    {
      "epoch": 3.5813953488372094,
      "grad_norm": 0.5885526537895203,
      "learning_rate": 2.2588230619025407e-05,
      "loss": 0.0164,
      "step": 2310
    },
    {
      "epoch": 3.5968992248062017,
      "grad_norm": 0.3644309937953949,
      "learning_rate": 2.213715968616425e-05,
      "loss": 0.0095,
      "step": 2320
    },
    {
      "epoch": 3.612403100775194,
      "grad_norm": 0.11602207273244858,
      "learning_rate": 2.16893540757494e-05,
      "loss": 0.0076,
      "step": 2330
    },
    {
      "epoch": 3.6279069767441863,
      "grad_norm": 0.21394725143909454,
      "learning_rate": 2.1244866267344866e-05,
      "loss": 0.0081,
      "step": 2340
    },
    {
      "epoch": 3.6434108527131785,
      "grad_norm": 0.3326941132545471,
      "learning_rate": 2.080374835169235e-05,
      "loss": 0.01,
      "step": 2350
    },
    {
      "epoch": 3.6434108527131785,
      "eval_loss": 0.04809526726603508,
      "eval_runtime": 152.6122,
      "eval_samples_per_second": 7.516,
      "eval_steps_per_second": 7.516,
      "step": 2350
    },
    {
      "epoch": 3.6589147286821704,
      "grad_norm": 0.2461967021226883,
      "learning_rate": 2.0366052024606612e-05,
      "loss": 0.009,
      "step": 2360
    },
    {
      "epoch": 3.6744186046511627,
      "grad_norm": 0.40642136335372925,
      "learning_rate": 1.9931828580917107e-05,
      "loss": 0.0096,
      "step": 2370
    },
    {
      "epoch": 3.689922480620155,
      "grad_norm": 0.46074432134628296,
      "learning_rate": 1.9501128908456523e-05,
      "loss": 0.0103,
      "step": 2380
    },
    {
      "epoch": 3.705426356589147,
      "grad_norm": 0.3104134202003479,
      "learning_rate": 1.9074003482097258e-05,
      "loss": 0.0068,
      "step": 2390
    },
    {
      "epoch": 3.7209302325581395,
      "grad_norm": 0.39811044931411743,
      "learning_rate": 1.8650502357835925e-05,
      "loss": 0.0126,
      "step": 2400
    },
    {
      "epoch": 3.7209302325581395,
      "eval_loss": 0.047415368258953094,
      "eval_runtime": 152.6813,
      "eval_samples_per_second": 7.512,
      "eval_steps_per_second": 7.512,
      "step": 2400
    },
    {
      "epoch": 3.7364341085271318,
      "grad_norm": 0.2357431799173355,
      "learning_rate": 1.82306751669274e-05,
      "loss": 0.0077,
      "step": 2410
    },
    {
      "epoch": 3.751937984496124,
      "grad_norm": 0.45044174790382385,
      "learning_rate": 1.7814571110068135e-05,
      "loss": 0.0143,
      "step": 2420
    },
    {
      "epoch": 3.7674418604651163,
      "grad_norm": 0.2097638100385666,
      "learning_rate": 1.740223895163039e-05,
      "loss": 0.009,
      "step": 2430
    },
    {
      "epoch": 3.7829457364341086,
      "grad_norm": 0.04716344177722931,
      "learning_rate": 1.6993727013947336e-05,
      "loss": 0.0106,
      "step": 2440
    },
    {
      "epoch": 3.798449612403101,
      "grad_norm": 0.2624518573284149,
      "learning_rate": 1.6589083171649977e-05,
      "loss": 0.0061,
      "step": 2450
    },
    {
      "epoch": 3.798449612403101,
      "eval_loss": 0.047738756984472275,
      "eval_runtime": 152.5966,
      "eval_samples_per_second": 7.517,
      "eval_steps_per_second": 7.517,
      "step": 2450
    },
    {
      "epoch": 3.813953488372093,
      "grad_norm": 0.08795194327831268,
      "learning_rate": 1.6188354846056698e-05,
      "loss": 0.0071,
      "step": 2460
    },
    {
      "epoch": 3.8294573643410854,
      "grad_norm": 0.09950519353151321,
      "learning_rate": 1.579158899961575e-05,
      "loss": 0.007,
      "step": 2470
    },
    {
      "epoch": 3.8449612403100772,
      "grad_norm": 0.47530969977378845,
      "learning_rate": 1.5398832130401637e-05,
      "loss": 0.0082,
      "step": 2480
    },
    {
      "epoch": 3.8604651162790695,
      "grad_norm": 0.19909849762916565,
      "learning_rate": 1.5010130266665807e-05,
      "loss": 0.0126,
      "step": 2490
    },
    {
      "epoch": 3.875968992248062,
      "grad_norm": 0.5359178185462952,
      "learning_rate": 1.4625528961442591e-05,
      "loss": 0.0088,
      "step": 2500
    },
    {
      "epoch": 3.875968992248062,
      "eval_loss": 0.04871485382318497,
      "eval_runtime": 152.5893,
      "eval_samples_per_second": 7.517,
      "eval_steps_per_second": 7.517,
      "step": 2500
    },
    {
      "epoch": 3.891472868217054,
      "grad_norm": 0.42270898818969727,
      "learning_rate": 1.4245073287210608e-05,
      "loss": 0.0097,
      "step": 2510
    },
    {
      "epoch": 3.9069767441860463,
      "grad_norm": 0.41609057784080505,
      "learning_rate": 1.3868807830610653e-05,
      "loss": 0.0147,
      "step": 2520
    },
    {
      "epoch": 3.9224806201550386,
      "grad_norm": 0.200226828455925,
      "learning_rate": 1.3496776687220514e-05,
      "loss": 0.009,
      "step": 2530
    },
    {
      "epoch": 3.937984496124031,
      "grad_norm": 0.31112906336784363,
      "learning_rate": 1.3129023456387151e-05,
      "loss": 0.0039,
      "step": 2540
    },
    {
      "epoch": 3.953488372093023,
      "grad_norm": 0.30395790934562683,
      "learning_rate": 1.2765591236117324e-05,
      "loss": 0.0074,
      "step": 2550
    },
    {
      "epoch": 3.953488372093023,
      "eval_loss": 0.04882879555225372,
      "eval_runtime": 152.5361,
      "eval_samples_per_second": 7.52,
      "eval_steps_per_second": 7.52,
      "step": 2550
    },
    {
      "epoch": 3.9689922480620154,
      "grad_norm": 0.1842828392982483,
      "learning_rate": 1.240652261802669e-05,
      "loss": 0.0077,
      "step": 2560
    },
    {
      "epoch": 3.9844961240310077,
      "grad_norm": 0.10175956040620804,
      "learning_rate": 1.205185968234847e-05,
      "loss": 0.0122,
      "step": 2570
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.42508187890052795,
      "learning_rate": 1.1701643993001916e-05,
      "loss": 0.009,
      "step": 2580
    },
    {
      "epoch": 4.015503875968992,
      "grad_norm": 0.18442735075950623,
      "learning_rate": 1.1355916592721316e-05,
      "loss": 0.0035,
      "step": 2590
    },
    {
      "epoch": 4.0310077519379846,
      "grad_norm": 0.3040939271450043,
      "learning_rate": 1.1014717998246099e-05,
      "loss": 0.0076,
      "step": 2600
    },
    {
      "epoch": 4.0310077519379846,
      "eval_loss": 0.04985535144805908,
      "eval_runtime": 152.6125,
      "eval_samples_per_second": 7.516,
      "eval_steps_per_second": 7.516,
      "step": 2600
    },
    {
      "epoch": 4.046511627906977,
      "grad_norm": 0.06122979894280434,
      "learning_rate": 1.0678088195572517e-05,
      "loss": 0.0033,
      "step": 2610
    },
    {
      "epoch": 4.062015503875969,
      "grad_norm": 0.12987752258777618,
      "learning_rate": 1.0346066635267676e-05,
      "loss": 0.0025,
      "step": 2620
    },
    {
      "epoch": 4.077519379844961,
      "grad_norm": 0.12539908289909363,
      "learning_rate": 1.001869222784611e-05,
      "loss": 0.0041,
      "step": 2630
    },
    {
      "epoch": 4.093023255813954,
      "grad_norm": 0.2488107532262802,
      "learning_rate": 9.696003339209819e-06,
      "loss": 0.0059,
      "step": 2640
    },
    {
      "epoch": 4.108527131782946,
      "grad_norm": 0.16544918715953827,
      "learning_rate": 9.378037786152055e-06,
      "loss": 0.0051,
      "step": 2650
    },
    {
      "epoch": 4.108527131782946,
      "eval_loss": 0.05235826596617699,
      "eval_runtime": 152.6712,
      "eval_samples_per_second": 7.513,
      "eval_steps_per_second": 7.513,
      "step": 2650
    },
    {
      "epoch": 4.124031007751938,
      "grad_norm": 0.253421813249588,
      "learning_rate": 9.064832831925441e-06,
      "loss": 0.0037,
      "step": 2660
    },
    {
      "epoch": 4.1395348837209305,
      "grad_norm": 0.055169735103845596,
      "learning_rate": 8.756425181875028e-06,
      "loss": 0.002,
      "step": 2670
    },
    {
      "epoch": 4.155038759689923,
      "grad_norm": 0.11390373855829239,
      "learning_rate": 8.452850979136617e-06,
      "loss": 0.0024,
      "step": 2680
    },
    {
      "epoch": 4.170542635658915,
      "grad_norm": 0.02611367031931877,
      "learning_rate": 8.15414580040117e-06,
      "loss": 0.0017,
      "step": 2690
    },
    {
      "epoch": 4.186046511627907,
      "grad_norm": 0.13717994093894958,
      "learning_rate": 7.860344651745366e-06,
      "loss": 0.0038,
      "step": 2700
    },
    {
      "epoch": 4.186046511627907,
      "eval_loss": 0.05555489659309387,
      "eval_runtime": 152.6773,
      "eval_samples_per_second": 7.513,
      "eval_steps_per_second": 7.513,
      "step": 2700
    },
    {
      "epoch": 4.2015503875969,
      "grad_norm": 0.1446123719215393,
      "learning_rate": 7.571481964529226e-06,
      "loss": 0.003,
      "step": 2710
    },
    {
      "epoch": 4.217054263565892,
      "grad_norm": 0.11501077562570572,
      "learning_rate": 7.287591591360926e-06,
      "loss": 0.0014,
      "step": 2720
    },
    {
      "epoch": 4.232558139534884,
      "grad_norm": 0.05887759476900101,
      "learning_rate": 7.008706802129606e-06,
      "loss": 0.0055,
      "step": 2730
    },
    {
      "epoch": 4.248062015503876,
      "grad_norm": 0.01087168138474226,
      "learning_rate": 6.734860280106292e-06,
      "loss": 0.0019,
      "step": 2740
    },
    {
      "epoch": 4.263565891472869,
      "grad_norm": 0.13784486055374146,
      "learning_rate": 6.466084118113735e-06,
      "loss": 0.0031,
      "step": 2750
    },
    {
      "epoch": 4.263565891472869,
      "eval_loss": 0.05840018764138222,
      "eval_runtime": 152.725,
      "eval_samples_per_second": 7.51,
      "eval_steps_per_second": 7.51,
      "step": 2750
    },
    {
      "epoch": 4.27906976744186,
      "grad_norm": 0.25398412346839905,
      "learning_rate": 6.202409814765325e-06,
      "loss": 0.0051,
      "step": 2760
    },
    {
      "epoch": 4.294573643410852,
      "grad_norm": 0.03926202654838562,
      "learning_rate": 5.9438682707736725e-06,
      "loss": 0.0033,
      "step": 2770
    },
    {
      "epoch": 4.310077519379845,
      "grad_norm": 0.1600087583065033,
      "learning_rate": 5.690489785329301e-06,
      "loss": 0.0031,
      "step": 2780
    },
    {
      "epoch": 4.325581395348837,
      "grad_norm": 0.01682668924331665,
      "learning_rate": 5.442304052549752e-06,
      "loss": 0.0029,
      "step": 2790
    },
    {
      "epoch": 4.341085271317829,
      "grad_norm": 0.13927067816257477,
      "learning_rate": 5.199340157999733e-06,
      "loss": 0.0028,
      "step": 2800
    },
    {
      "epoch": 4.341085271317829,
      "eval_loss": 0.06022586300969124,
      "eval_runtime": 152.7632,
      "eval_samples_per_second": 7.508,
      "eval_steps_per_second": 7.508,
      "step": 2800
    },
    {
      "epoch": 4.3565891472868215,
      "grad_norm": 0.02230786345899105,
      "learning_rate": 4.961626575282396e-06,
      "loss": 0.0019,
      "step": 2810
    },
    {
      "epoch": 4.372093023255814,
      "grad_norm": 0.08988174796104431,
      "learning_rate": 4.7291911627025235e-06,
      "loss": 0.002,
      "step": 2820
    },
    {
      "epoch": 4.387596899224806,
      "grad_norm": 0.07500022649765015,
      "learning_rate": 4.502061160001725e-06,
      "loss": 0.0021,
      "step": 2830
    },
    {
      "epoch": 4.403100775193798,
      "grad_norm": 0.1328381896018982,
      "learning_rate": 4.280263185166078e-06,
      "loss": 0.0037,
      "step": 2840
    },
    {
      "epoch": 4.4186046511627906,
      "grad_norm": 0.2652447819709778,
      "learning_rate": 4.063823231306757e-06,
      "loss": 0.0037,
      "step": 2850
    },
    {
      "epoch": 4.4186046511627906,
      "eval_loss": 0.061209216713905334,
      "eval_runtime": 152.4706,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 2850
    },
    {
      "epoch": 4.434108527131783,
      "grad_norm": 0.07953423261642456,
      "learning_rate": 3.8527666636137885e-06,
      "loss": 0.0018,
      "step": 2860
    },
    {
      "epoch": 4.449612403100775,
      "grad_norm": 0.18091465532779694,
      "learning_rate": 3.6471182163834825e-06,
      "loss": 0.0032,
      "step": 2870
    },
    {
      "epoch": 4.465116279069767,
      "grad_norm": 0.09644496440887451,
      "learning_rate": 3.4469019901197054e-06,
      "loss": 0.0018,
      "step": 2880
    },
    {
      "epoch": 4.48062015503876,
      "grad_norm": 0.44461509585380554,
      "learning_rate": 3.252141448709495e-06,
      "loss": 0.0032,
      "step": 2890
    },
    {
      "epoch": 4.496124031007752,
      "grad_norm": 0.21934878826141357,
      "learning_rate": 3.0628594166732693e-06,
      "loss": 0.0037,
      "step": 2900
    },
    {
      "epoch": 4.496124031007752,
      "eval_loss": 0.06198548525571823,
      "eval_runtime": 152.3492,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 2900
    },
    {
      "epoch": 4.511627906976744,
      "grad_norm": 0.04599190875887871,
      "learning_rate": 2.8790780764899384e-06,
      "loss": 0.0029,
      "step": 2910
    },
    {
      "epoch": 4.5271317829457365,
      "grad_norm": 0.25089895725250244,
      "learning_rate": 2.700818965997315e-06,
      "loss": 0.004,
      "step": 2920
    },
    {
      "epoch": 4.542635658914729,
      "grad_norm": 0.008679673075675964,
      "learning_rate": 2.528102975867991e-06,
      "loss": 0.0007,
      "step": 2930
    },
    {
      "epoch": 4.558139534883721,
      "grad_norm": 0.06974011659622192,
      "learning_rate": 2.3609503471611284e-06,
      "loss": 0.0014,
      "step": 2940
    },
    {
      "epoch": 4.573643410852713,
      "grad_norm": 0.015127426944673061,
      "learning_rate": 2.1993806689503738e-06,
      "loss": 0.0013,
      "step": 2950
    },
    {
      "epoch": 4.573643410852713,
      "eval_loss": 0.06263046711683273,
      "eval_runtime": 152.5658,
      "eval_samples_per_second": 7.518,
      "eval_steps_per_second": 7.518,
      "step": 2950
    },
    {
      "epoch": 4.589147286821706,
      "grad_norm": 0.0138722388073802,
      "learning_rate": 2.0434128760281167e-06,
      "loss": 0.001,
      "step": 2960
    },
    {
      "epoch": 4.604651162790698,
      "grad_norm": 0.027368668466806412,
      "learning_rate": 1.893065246686504e-06,
      "loss": 0.0005,
      "step": 2970
    },
    {
      "epoch": 4.62015503875969,
      "grad_norm": 0.05920499190688133,
      "learning_rate": 1.74835540057533e-06,
      "loss": 0.004,
      "step": 2980
    },
    {
      "epoch": 4.635658914728682,
      "grad_norm": 0.10905849933624268,
      "learning_rate": 1.6093002966371617e-06,
      "loss": 0.0013,
      "step": 2990
    },
    {
      "epoch": 4.651162790697675,
      "grad_norm": 0.09890134632587433,
      "learning_rate": 1.4759162311198783e-06,
      "loss": 0.0013,
      "step": 3000
    },
    {
      "epoch": 4.651162790697675,
      "eval_loss": 0.06311814486980438,
      "eval_runtime": 152.5486,
      "eval_samples_per_second": 7.519,
      "eval_steps_per_second": 7.519,
      "step": 3000
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.0647616907954216,
      "learning_rate": 1.348218835666859e-06,
      "loss": 0.003,
      "step": 3010
    },
    {
      "epoch": 4.682170542635659,
      "grad_norm": 0.022763436660170555,
      "learning_rate": 1.2262230754850445e-06,
      "loss": 0.0029,
      "step": 3020
    },
    {
      "epoch": 4.6976744186046515,
      "grad_norm": 0.07503198832273483,
      "learning_rate": 1.109943247591172e-06,
      "loss": 0.0037,
      "step": 3030
    },
    {
      "epoch": 4.713178294573644,
      "grad_norm": 0.16900363564491272,
      "learning_rate": 9.993929791362323e-07,
      "loss": 0.003,
      "step": 3040
    },
    {
      "epoch": 4.728682170542635,
      "grad_norm": 0.054068733006715775,
      "learning_rate": 8.945852258084863e-07,
      "loss": 0.0023,
      "step": 3050
    },
    {
      "epoch": 4.728682170542635,
      "eval_loss": 0.06340909004211426,
      "eval_runtime": 152.4344,
      "eval_samples_per_second": 7.525,
      "eval_steps_per_second": 7.525,
      "step": 3050
    },
    {
      "epoch": 4.7441860465116275,
      "grad_norm": 0.3682853579521179,
      "learning_rate": 7.955322703151358e-07,
      "loss": 0.0051,
      "step": 3060
    },
    {
      "epoch": 4.75968992248062,
      "grad_norm": 0.08928383886814117,
      "learning_rate": 7.022457209428901e-07,
      "loss": 0.0033,
      "step": 3070
    },
    {
      "epoch": 4.775193798449612,
      "grad_norm": 0.5383804440498352,
      "learning_rate": 6.147365101975666e-07,
      "loss": 0.0029,
      "step": 3080
    },
    {
      "epoch": 4.790697674418604,
      "grad_norm": 0.15904894471168518,
      "learning_rate": 5.33014893522854e-07,
      "loss": 0.003,
      "step": 3090
    },
    {
      "epoch": 4.8062015503875966,
      "grad_norm": 0.237746924161911,
      "learning_rate": 4.57090448098485e-07,
      "loss": 0.0042,
      "step": 3100
    },
    {
      "epoch": 4.8062015503875966,
      "eval_loss": 0.0634797140955925,
      "eval_runtime": 152.4113,
      "eval_samples_per_second": 7.526,
      "eval_steps_per_second": 7.526,
      "step": 3100
    },
    {
      "epoch": 4.821705426356589,
      "grad_norm": 0.24263429641723633,
      "learning_rate": 3.8697207171781714e-07,
      "loss": 0.0043,
      "step": 3110
    },
    {
      "epoch": 4.837209302325581,
      "grad_norm": 0.004827297758311033,
      "learning_rate": 3.2266798174512837e-07,
      "loss": 0.0043,
      "step": 3120
    },
    {
      "epoch": 4.852713178294573,
      "grad_norm": 0.017983857542276382,
      "learning_rate": 2.6418571415255387e-07,
      "loss": 0.0069,
      "step": 3130
    },
    {
      "epoch": 4.868217054263566,
      "grad_norm": 0.022801002487540245,
      "learning_rate": 2.1153212263695378e-07,
      "loss": 0.0026,
      "step": 3140
    },
    {
      "epoch": 4.883720930232558,
      "grad_norm": 0.25147223472595215,
      "learning_rate": 1.6471337781669982e-07,
      "loss": 0.0053,
      "step": 3150
    },
    {
      "epoch": 4.883720930232558,
      "eval_loss": 0.06353544443845749,
      "eval_runtime": 152.3358,
      "eval_samples_per_second": 7.529,
      "eval_steps_per_second": 7.529,
      "step": 3150
    },
    {
      "epoch": 4.89922480620155,
      "grad_norm": 0.06839290261268616,
      "learning_rate": 1.237349665085097e-07,
      "loss": 0.0026,
      "step": 3160
    },
    {
      "epoch": 4.9147286821705425,
      "grad_norm": 0.05276154354214668,
      "learning_rate": 8.86016910844667e-08,
      "loss": 0.0027,
      "step": 3170
    },
    {
      "epoch": 4.930232558139535,
      "grad_norm": 0.4559900164604187,
      "learning_rate": 5.9317668909192323e-08,
      "loss": 0.004,
      "step": 3180
    },
    {
      "epoch": 4.945736434108527,
      "grad_norm": 0.08621206879615784,
      "learning_rate": 3.588633185730994e-08,
      "loss": 0.005,
      "step": 3190
    },
    {
      "epoch": 4.961240310077519,
      "grad_norm": 0.10626496374607086,
      "learning_rate": 1.8310425911294283e-08,
      "loss": 0.0041,
      "step": 3200
    },
    {
      "epoch": 4.961240310077519,
      "eval_loss": 0.06344824284315109,
      "eval_runtime": 152.4596,
      "eval_samples_per_second": 7.523,
      "eval_steps_per_second": 7.523,
      "step": 3200
    },
    {
      "epoch": 4.976744186046512,
      "grad_norm": 0.29534295201301575,
      "learning_rate": 6.592010839612251e-09,
      "loss": 0.0043,
      "step": 3210
    },
    {
      "epoch": 4.992248062015504,
      "grad_norm": 0.03377154842019081,
      "learning_rate": 7.324599553770739e-10,
      "loss": 0.0015,
      "step": 3220
    },
    {
      "epoch": 5.0,
      "step": 3225,
      "total_flos": 8.096268812599296e+17,
      "train_loss": 0.04132688383911123,
      "train_runtime": 34528.1447,
      "train_samples_per_second": 1.494,
      "train_steps_per_second": 0.093
    }
  ],
  "logging_steps": 10,
  "max_steps": 3225,
  "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": 8.096268812599296e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}