{
  "best_metric": 0.03466709703207016,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold10/checkpoint-1250",
  "epoch": 4.994708994708994,
  "eval_steps": 50,
  "global_step": 3245,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015392015392015393,
      "grad_norm": 5.839231491088867,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 1.2829,
      "step": 10
    },
    {
      "epoch": 0.030784030784030785,
      "grad_norm": 5.680546283721924,
      "learning_rate": 6.153846153846155e-06,
      "loss": 1.35,
      "step": 20
    },
    {
      "epoch": 0.046176046176046176,
      "grad_norm": 3.828277349472046,
      "learning_rate": 9.230769230769232e-06,
      "loss": 1.1992,
      "step": 30
    },
    {
      "epoch": 0.06156806156806157,
      "grad_norm": 2.330836296081543,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.8478,
      "step": 40
    },
    {
      "epoch": 0.07696007696007696,
      "grad_norm": 1.7859842777252197,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.5221,
      "step": 50
    },
    {
      "epoch": 0.07696007696007696,
      "eval_loss": 0.3092120289802551,
      "eval_runtime": 192.742,
      "eval_samples_per_second": 5.992,
      "eval_steps_per_second": 5.992,
      "step": 50
    },
    {
      "epoch": 0.09235209235209235,
      "grad_norm": 1.266290307044983,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.2921,
      "step": 60
    },
    {
      "epoch": 0.10774410774410774,
      "grad_norm": 2.233220338821411,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.2192,
      "step": 70
    },
    {
      "epoch": 0.12313612313612314,
      "grad_norm": 1.301505446434021,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.1696,
      "step": 80
    },
    {
      "epoch": 0.13852813852813853,
      "grad_norm": 1.4332795143127441,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.1291,
      "step": 90
    },
    {
      "epoch": 0.15392015392015393,
      "grad_norm": 0.8287798762321472,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0971,
      "step": 100
    },
    {
      "epoch": 0.15392015392015393,
      "eval_loss": 0.0860358327627182,
      "eval_runtime": 201.0316,
      "eval_samples_per_second": 5.745,
      "eval_steps_per_second": 5.745,
      "step": 100
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 1.537462830543518,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0958,
      "step": 110
    },
    {
      "epoch": 0.1847041847041847,
      "grad_norm": 0.8775112628936768,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0727,
      "step": 120
    },
    {
      "epoch": 0.2000962000962001,
      "grad_norm": 1.4552613496780396,
      "learning_rate": 4e-05,
      "loss": 0.0744,
      "step": 130
    },
    {
      "epoch": 0.21548821548821548,
      "grad_norm": 2.025639772415161,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0917,
      "step": 140
    },
    {
      "epoch": 0.23088023088023088,
      "grad_norm": 0.8879976868629456,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0873,
      "step": 150
    },
    {
      "epoch": 0.23088023088023088,
      "eval_loss": 0.06073848158121109,
      "eval_runtime": 200.6239,
      "eval_samples_per_second": 5.757,
      "eval_steps_per_second": 5.757,
      "step": 150
    },
    {
      "epoch": 0.24627224627224628,
      "grad_norm": 1.0195509195327759,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0451,
      "step": 160
    },
    {
      "epoch": 0.26166426166426165,
      "grad_norm": 0.8961577415466309,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0674,
      "step": 170
    },
    {
      "epoch": 0.27705627705627706,
      "grad_norm": 0.774815022945404,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0593,
      "step": 180
    },
    {
      "epoch": 0.29244829244829246,
      "grad_norm": 1.6998802423477173,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0692,
      "step": 190
    },
    {
      "epoch": 0.30784030784030786,
      "grad_norm": 0.5050025582313538,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0616,
      "step": 200
    },
    {
      "epoch": 0.30784030784030786,
      "eval_loss": 0.056508466601371765,
      "eval_runtime": 165.9359,
      "eval_samples_per_second": 6.961,
      "eval_steps_per_second": 6.961,
      "step": 200
    },
    {
      "epoch": 0.32323232323232326,
      "grad_norm": 1.1922398805618286,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.0634,
      "step": 210
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 1.0104138851165771,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.0744,
      "step": 220
    },
    {
      "epoch": 0.354016354016354,
      "grad_norm": 1.652618408203125,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.055,
      "step": 230
    },
    {
      "epoch": 0.3694083694083694,
      "grad_norm": 0.5933212041854858,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0647,
      "step": 240
    },
    {
      "epoch": 0.3848003848003848,
      "grad_norm": 0.8468647003173828,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.071,
      "step": 250
    },
    {
      "epoch": 0.3848003848003848,
      "eval_loss": 0.05420336872339249,
      "eval_runtime": 160.844,
      "eval_samples_per_second": 7.181,
      "eval_steps_per_second": 7.181,
      "step": 250
    },
    {
      "epoch": 0.4001924001924002,
      "grad_norm": 0.3505142629146576,
      "learning_rate": 8e-05,
      "loss": 0.0632,
      "step": 260
    },
    {
      "epoch": 0.4155844155844156,
      "grad_norm": 0.5039116144180298,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0569,
      "step": 270
    },
    {
      "epoch": 0.43097643097643096,
      "grad_norm": 0.6876810193061829,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0413,
      "step": 280
    },
    {
      "epoch": 0.44636844636844636,
      "grad_norm": 0.4580015242099762,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0654,
      "step": 290
    },
    {
      "epoch": 0.46176046176046176,
      "grad_norm": 0.676700234413147,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0615,
      "step": 300
    },
    {
      "epoch": 0.46176046176046176,
      "eval_loss": 0.049673303961753845,
      "eval_runtime": 160.3278,
      "eval_samples_per_second": 7.204,
      "eval_steps_per_second": 7.204,
      "step": 300
    },
    {
      "epoch": 0.47715247715247716,
      "grad_norm": 0.5680780410766602,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0497,
      "step": 310
    },
    {
      "epoch": 0.49254449254449256,
      "grad_norm": 1.1781426668167114,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0625,
      "step": 320
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 0.5670598745346069,
      "learning_rate": 9.999927654251793e-05,
      "loss": 0.0691,
      "step": 330
    },
    {
      "epoch": 0.5233285233285233,
      "grad_norm": 0.393940806388855,
      "learning_rate": 9.99934890082745e-05,
      "loss": 0.0504,
      "step": 340
    },
    {
      "epoch": 0.5387205387205387,
      "grad_norm": 0.6476990580558777,
      "learning_rate": 9.99819146097084e-05,
      "loss": 0.0538,
      "step": 350
    },
    {
      "epoch": 0.5387205387205387,
      "eval_loss": 0.046799372881650925,
      "eval_runtime": 160.4795,
      "eval_samples_per_second": 7.197,
      "eval_steps_per_second": 7.197,
      "step": 350
    },
    {
      "epoch": 0.5541125541125541,
      "grad_norm": 0.47279223799705505,
      "learning_rate": 9.996455468658355e-05,
      "loss": 0.0478,
      "step": 360
    },
    {
      "epoch": 0.5695045695045695,
      "grad_norm": 0.5177505016326904,
      "learning_rate": 9.994141124835203e-05,
      "loss": 0.0453,
      "step": 370
    },
    {
      "epoch": 0.5848965848965849,
      "grad_norm": 0.7338818311691284,
      "learning_rate": 9.991248697392142e-05,
      "loss": 0.0367,
      "step": 380
    },
    {
      "epoch": 0.6002886002886003,
      "grad_norm": 1.0812134742736816,
      "learning_rate": 9.987778521134476e-05,
      "loss": 0.0455,
      "step": 390
    },
    {
      "epoch": 0.6156806156806157,
      "grad_norm": 0.37609776854515076,
      "learning_rate": 9.983730997743294e-05,
      "loss": 0.0532,
      "step": 400
    },
    {
      "epoch": 0.6156806156806157,
      "eval_loss": 0.04624510183930397,
      "eval_runtime": 160.4749,
      "eval_samples_per_second": 7.197,
      "eval_steps_per_second": 7.197,
      "step": 400
    },
    {
      "epoch": 0.6310726310726311,
      "grad_norm": 0.7044768333435059,
      "learning_rate": 9.979106595728978e-05,
      "loss": 0.0568,
      "step": 410
    },
    {
      "epoch": 0.6464646464646465,
      "grad_norm": 0.41372236609458923,
      "learning_rate": 9.973905850376977e-05,
      "loss": 0.0396,
      "step": 420
    },
    {
      "epoch": 0.6618566618566618,
      "grad_norm": 0.4322606325149536,
      "learning_rate": 9.968129363685833e-05,
      "loss": 0.0419,
      "step": 430
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 1.1119065284729004,
      "learning_rate": 9.961777804297514e-05,
      "loss": 0.0534,
      "step": 440
    },
    {
      "epoch": 0.6926406926406926,
      "grad_norm": 0.2994139790534973,
      "learning_rate": 9.954851907420009e-05,
      "loss": 0.0501,
      "step": 450
    },
    {
      "epoch": 0.6926406926406926,
      "eval_loss": 0.048226650804281235,
      "eval_runtime": 160.5362,
      "eval_samples_per_second": 7.195,
      "eval_steps_per_second": 7.195,
      "step": 450
    },
    {
      "epoch": 0.708032708032708,
      "grad_norm": 0.6291306018829346,
      "learning_rate": 9.947352474742222e-05,
      "loss": 0.0511,
      "step": 460
    },
    {
      "epoch": 0.7234247234247234,
      "grad_norm": 0.5497981905937195,
      "learning_rate": 9.93928037434118e-05,
      "loss": 0.0533,
      "step": 470
    },
    {
      "epoch": 0.7388167388167388,
      "grad_norm": 0.6592866778373718,
      "learning_rate": 9.930636540581556e-05,
      "loss": 0.0464,
      "step": 480
    },
    {
      "epoch": 0.7542087542087542,
      "grad_norm": 0.438894122838974,
      "learning_rate": 9.921421974007506e-05,
      "loss": 0.0469,
      "step": 490
    },
    {
      "epoch": 0.7696007696007696,
      "grad_norm": 0.7949342727661133,
      "learning_rate": 9.911637741226849e-05,
      "loss": 0.0575,
      "step": 500
    },
    {
      "epoch": 0.7696007696007696,
      "eval_loss": 0.042172204703092575,
      "eval_runtime": 160.3062,
      "eval_samples_per_second": 7.205,
      "eval_steps_per_second": 7.205,
      "step": 500
    },
    {
      "epoch": 0.784992784992785,
      "grad_norm": 0.3464134931564331,
      "learning_rate": 9.90128497478762e-05,
      "loss": 0.0484,
      "step": 510
    },
    {
      "epoch": 0.8003848003848004,
      "grad_norm": 0.37780413031578064,
      "learning_rate": 9.890364873046965e-05,
      "loss": 0.0445,
      "step": 520
    },
    {
      "epoch": 0.8157768157768158,
      "grad_norm": 0.3058973550796509,
      "learning_rate": 9.878878700032427e-05,
      "loss": 0.0338,
      "step": 530
    },
    {
      "epoch": 0.8311688311688312,
      "grad_norm": 0.5813219547271729,
      "learning_rate": 9.866827785295638e-05,
      "loss": 0.0481,
      "step": 540
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 0.25342145562171936,
      "learning_rate": 9.854213523758413e-05,
      "loss": 0.0418,
      "step": 550
    },
    {
      "epoch": 0.8465608465608465,
      "eval_loss": 0.043983206152915955,
      "eval_runtime": 160.2238,
      "eval_samples_per_second": 7.209,
      "eval_steps_per_second": 7.209,
      "step": 550
    },
    {
      "epoch": 0.8619528619528619,
      "grad_norm": 0.4131356179714203,
      "learning_rate": 9.841037375551294e-05,
      "loss": 0.0508,
      "step": 560
    },
    {
      "epoch": 0.8773448773448773,
      "grad_norm": 0.8691533803939819,
      "learning_rate": 9.827300865844527e-05,
      "loss": 0.0467,
      "step": 570
    },
    {
      "epoch": 0.8927368927368927,
      "grad_norm": 0.4405064284801483,
      "learning_rate": 9.813005584671522e-05,
      "loss": 0.0365,
      "step": 580
    },
    {
      "epoch": 0.9081289081289081,
      "grad_norm": 0.2739037573337555,
      "learning_rate": 9.79815318674481e-05,
      "loss": 0.0442,
      "step": 590
    },
    {
      "epoch": 0.9235209235209235,
      "grad_norm": 0.328683465719223,
      "learning_rate": 9.782745391264498e-05,
      "loss": 0.048,
      "step": 600
    },
    {
      "epoch": 0.9235209235209235,
      "eval_loss": 0.03977407515048981,
      "eval_runtime": 160.2864,
      "eval_samples_per_second": 7.206,
      "eval_steps_per_second": 7.206,
      "step": 600
    },
    {
      "epoch": 0.9389129389129389,
      "grad_norm": 0.5097044706344604,
      "learning_rate": 9.76678398171927e-05,
      "loss": 0.0469,
      "step": 610
    },
    {
      "epoch": 0.9543049543049543,
      "grad_norm": 0.2837560772895813,
      "learning_rate": 9.750270805679945e-05,
      "loss": 0.0309,
      "step": 620
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.3343970775604248,
      "learning_rate": 9.733207774585618e-05,
      "loss": 0.062,
      "step": 630
    },
    {
      "epoch": 0.9850889850889851,
      "grad_norm": 0.39156869053840637,
      "learning_rate": 9.715596863522398e-05,
      "loss": 0.0386,
      "step": 640
    },
    {
      "epoch": 1.0004810004810005,
      "grad_norm": 0.332334041595459,
      "learning_rate": 9.697440110994801e-05,
      "loss": 0.0559,
      "step": 650
    },
    {
      "epoch": 1.0004810004810005,
      "eval_loss": 0.039704564958810806,
      "eval_runtime": 160.3208,
      "eval_samples_per_second": 7.204,
      "eval_steps_per_second": 7.204,
      "step": 650
    },
    {
      "epoch": 1.0158730158730158,
      "grad_norm": 0.25134462118148804,
      "learning_rate": 9.678739618689775e-05,
      "loss": 0.0366,
      "step": 660
    },
    {
      "epoch": 1.0312650312650313,
      "grad_norm": 0.24178971350193024,
      "learning_rate": 9.659497551233431e-05,
      "loss": 0.0338,
      "step": 670
    },
    {
      "epoch": 1.0466570466570466,
      "grad_norm": 0.19782131910324097,
      "learning_rate": 9.639716135940485e-05,
      "loss": 0.0328,
      "step": 680
    },
    {
      "epoch": 1.0620490620490621,
      "grad_norm": 0.18410953879356384,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.0331,
      "step": 690
    },
    {
      "epoch": 1.0774410774410774,
      "grad_norm": 0.4528217613697052,
      "learning_rate": 9.598544482992517e-05,
      "loss": 0.0358,
      "step": 700
    },
    {
      "epoch": 1.0774410774410774,
      "eval_loss": 0.04308833181858063,
      "eval_runtime": 160.413,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 700
    },
    {
      "epoch": 1.092833092833093,
      "grad_norm": 0.6040289998054504,
      "learning_rate": 9.57715901105348e-05,
      "loss": 0.037,
      "step": 710
    },
    {
      "epoch": 1.1082251082251082,
      "grad_norm": 0.5413057804107666,
      "learning_rate": 9.555243722158158e-05,
      "loss": 0.035,
      "step": 720
    },
    {
      "epoch": 1.1236171236171235,
      "grad_norm": 0.27406299114227295,
      "learning_rate": 9.532801153052958e-05,
      "loss": 0.0408,
      "step": 730
    },
    {
      "epoch": 1.139009139009139,
      "grad_norm": 0.371649831533432,
      "learning_rate": 9.509833901518209e-05,
      "loss": 0.0377,
      "step": 740
    },
    {
      "epoch": 1.1544011544011543,
      "grad_norm": 0.4681012034416199,
      "learning_rate": 9.486344626067475e-05,
      "loss": 0.0277,
      "step": 750
    },
    {
      "epoch": 1.1544011544011543,
      "eval_loss": 0.03915223106741905,
      "eval_runtime": 160.4858,
      "eval_samples_per_second": 7.197,
      "eval_steps_per_second": 7.197,
      "step": 750
    },
    {
      "epoch": 1.1697931697931698,
      "grad_norm": 0.13481155037879944,
      "learning_rate": 9.462336045639815e-05,
      "loss": 0.0355,
      "step": 760
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 0.10173361003398895,
      "learning_rate": 9.437810939285068e-05,
      "loss": 0.0228,
      "step": 770
    },
    {
      "epoch": 1.2005772005772006,
      "grad_norm": 0.24143601953983307,
      "learning_rate": 9.412772145842165e-05,
      "loss": 0.0266,
      "step": 780
    },
    {
      "epoch": 1.215969215969216,
      "grad_norm": 0.43506941199302673,
      "learning_rate": 9.387222563610535e-05,
      "loss": 0.0391,
      "step": 790
    },
    {
      "epoch": 1.2313612313612314,
      "grad_norm": 0.23058776557445526,
      "learning_rate": 9.361165150014606e-05,
      "loss": 0.029,
      "step": 800
    },
    {
      "epoch": 1.2313612313612314,
      "eval_loss": 0.03759562969207764,
      "eval_runtime": 160.4221,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 800
    },
    {
      "epoch": 1.2467532467532467,
      "grad_norm": 0.5037282109260559,
      "learning_rate": 9.334602921261492e-05,
      "loss": 0.0461,
      "step": 810
    },
    {
      "epoch": 1.2621452621452622,
      "grad_norm": 0.2980537414550781,
      "learning_rate": 9.307538951991852e-05,
      "loss": 0.0364,
      "step": 820
    },
    {
      "epoch": 1.2775372775372775,
      "grad_norm": 0.203804150223732,
      "learning_rate": 9.279976374923988e-05,
      "loss": 0.0387,
      "step": 830
    },
    {
      "epoch": 1.2929292929292928,
      "grad_norm": 0.3414406478404999,
      "learning_rate": 9.251918380491238e-05,
      "loss": 0.0282,
      "step": 840
    },
    {
      "epoch": 1.3083213083213083,
      "grad_norm": 0.8123178482055664,
      "learning_rate": 9.223368216472668e-05,
      "loss": 0.0283,
      "step": 850
    },
    {
      "epoch": 1.3083213083213083,
      "eval_loss": 0.03833651915192604,
      "eval_runtime": 160.4154,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 850
    },
    {
      "epoch": 1.3237133237133238,
      "grad_norm": 0.40936821699142456,
      "learning_rate": 9.194329187617127e-05,
      "loss": 0.0343,
      "step": 860
    },
    {
      "epoch": 1.3391053391053391,
      "grad_norm": 0.3399330973625183,
      "learning_rate": 9.164804655260735e-05,
      "loss": 0.036,
      "step": 870
    },
    {
      "epoch": 1.3544973544973544,
      "grad_norm": 0.3808479607105255,
      "learning_rate": 9.134798036937778e-05,
      "loss": 0.0428,
      "step": 880
    },
    {
      "epoch": 1.36988936988937,
      "grad_norm": 0.18996691703796387,
      "learning_rate": 9.104312805985134e-05,
      "loss": 0.0337,
      "step": 890
    },
    {
      "epoch": 1.3852813852813852,
      "grad_norm": 0.30460020899772644,
      "learning_rate": 9.073352491140221e-05,
      "loss": 0.035,
      "step": 900
    },
    {
      "epoch": 1.3852813852813852,
      "eval_loss": 0.03713306412100792,
      "eval_runtime": 160.2728,
      "eval_samples_per_second": 7.206,
      "eval_steps_per_second": 7.206,
      "step": 900
    },
    {
      "epoch": 1.4006734006734007,
      "grad_norm": 0.5870276689529419,
      "learning_rate": 9.041920676132543e-05,
      "loss": 0.0424,
      "step": 910
    },
    {
      "epoch": 1.416065416065416,
      "grad_norm": 0.5088344216346741,
      "learning_rate": 9.010020999268854e-05,
      "loss": 0.0342,
      "step": 920
    },
    {
      "epoch": 1.4314574314574315,
      "grad_norm": 0.22709380090236664,
      "learning_rate": 8.977657153012027e-05,
      "loss": 0.0363,
      "step": 930
    },
    {
      "epoch": 1.4468494468494468,
      "grad_norm": 0.32548844814300537,
      "learning_rate": 8.944832883553633e-05,
      "loss": 0.0401,
      "step": 940
    },
    {
      "epoch": 1.4622414622414621,
      "grad_norm": 0.30676841735839844,
      "learning_rate": 8.91155199038032e-05,
      "loss": 0.0367,
      "step": 950
    },
    {
      "epoch": 1.4622414622414621,
      "eval_loss": 0.03731577843427658,
      "eval_runtime": 160.4408,
      "eval_samples_per_second": 7.199,
      "eval_steps_per_second": 7.199,
      "step": 950
    },
    {
      "epoch": 1.4776334776334776,
      "grad_norm": 0.3863791227340698,
      "learning_rate": 8.877818325834006e-05,
      "loss": 0.0343,
      "step": 960
    },
    {
      "epoch": 1.4930254930254931,
      "grad_norm": 0.6713721752166748,
      "learning_rate": 8.843635794665961e-05,
      "loss": 0.0392,
      "step": 970
    },
    {
      "epoch": 1.5084175084175084,
      "grad_norm": 0.2896818518638611,
      "learning_rate": 8.809008353584829e-05,
      "loss": 0.0364,
      "step": 980
    },
    {
      "epoch": 1.5238095238095237,
      "grad_norm": 0.2636934518814087,
      "learning_rate": 8.773940010798627e-05,
      "loss": 0.0279,
      "step": 990
    },
    {
      "epoch": 1.5392015392015392,
      "grad_norm": 0.48908543586730957,
      "learning_rate": 8.738434825550785e-05,
      "loss": 0.0272,
      "step": 1000
    },
    {
      "epoch": 1.5392015392015392,
      "eval_loss": 0.042762041091918945,
      "eval_runtime": 160.4777,
      "eval_samples_per_second": 7.197,
      "eval_steps_per_second": 7.197,
      "step": 1000
    },
    {
      "epoch": 1.5545935545935547,
      "grad_norm": 0.3237270712852478,
      "learning_rate": 8.702496907650274e-05,
      "loss": 0.0337,
      "step": 1010
    },
    {
      "epoch": 1.56998556998557,
      "grad_norm": 0.25954726338386536,
      "learning_rate": 8.666130416995897e-05,
      "loss": 0.026,
      "step": 1020
    },
    {
      "epoch": 1.5853775853775853,
      "grad_norm": 0.523502767086029,
      "learning_rate": 8.629339563094758e-05,
      "loss": 0.0229,
      "step": 1030
    },
    {
      "epoch": 1.6007696007696008,
      "grad_norm": 0.6461522579193115,
      "learning_rate": 8.592128604575014e-05,
      "loss": 0.0384,
      "step": 1040
    },
    {
      "epoch": 1.6161616161616161,
      "grad_norm": 0.2046208530664444,
      "learning_rate": 8.554501848692921e-05,
      "loss": 0.0435,
      "step": 1050
    },
    {
      "epoch": 1.6161616161616161,
      "eval_loss": 0.036668434739112854,
      "eval_runtime": 160.4211,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 1050
    },
    {
      "epoch": 1.6315536315536314,
      "grad_norm": 0.311105877161026,
      "learning_rate": 8.51646365083426e-05,
      "loss": 0.0393,
      "step": 1060
    },
    {
      "epoch": 1.646945646945647,
      "grad_norm": 0.42447835206985474,
      "learning_rate": 8.478018414010191e-05,
      "loss": 0.0451,
      "step": 1070
    },
    {
      "epoch": 1.6623376623376624,
      "grad_norm": 0.3853769600391388,
      "learning_rate": 8.439170588347598e-05,
      "loss": 0.0365,
      "step": 1080
    },
    {
      "epoch": 1.6777296777296777,
      "grad_norm": 0.18537352979183197,
      "learning_rate": 8.399924670573969e-05,
      "loss": 0.0249,
      "step": 1090
    },
    {
      "epoch": 1.693121693121693,
      "grad_norm": 0.52433180809021,
      "learning_rate": 8.360285203496894e-05,
      "loss": 0.0379,
      "step": 1100
    },
    {
      "epoch": 1.693121693121693,
      "eval_loss": 0.036822497844696045,
      "eval_runtime": 160.5884,
      "eval_samples_per_second": 7.192,
      "eval_steps_per_second": 7.192,
      "step": 1100
    },
    {
      "epoch": 1.7085137085137085,
      "grad_norm": 0.4827461540699005,
      "learning_rate": 8.320256775478228e-05,
      "loss": 0.0403,
      "step": 1110
    },
    {
      "epoch": 1.723905723905724,
      "grad_norm": 0.3847942352294922,
      "learning_rate": 8.279844019902968e-05,
      "loss": 0.0265,
      "step": 1120
    },
    {
      "epoch": 1.7392977392977393,
      "grad_norm": 0.5013036131858826,
      "learning_rate": 8.239051614642934e-05,
      "loss": 0.0227,
      "step": 1130
    },
    {
      "epoch": 1.7546897546897546,
      "grad_norm": 0.2711322605609894,
      "learning_rate": 8.197884281515292e-05,
      "loss": 0.0299,
      "step": 1140
    },
    {
      "epoch": 1.7700817700817701,
      "grad_norm": 0.3275453448295593,
      "learning_rate": 8.156346785735997e-05,
      "loss": 0.0296,
      "step": 1150
    },
    {
      "epoch": 1.7700817700817701,
      "eval_loss": 0.03776984289288521,
      "eval_runtime": 160.4557,
      "eval_samples_per_second": 7.198,
      "eval_steps_per_second": 7.198,
      "step": 1150
    },
    {
      "epoch": 1.7854737854737854,
      "grad_norm": 0.40839052200317383,
      "learning_rate": 8.114443935368198e-05,
      "loss": 0.0346,
      "step": 1160
    },
    {
      "epoch": 1.8008658008658007,
      "grad_norm": 0.22129075229167938,
      "learning_rate": 8.072180580765709e-05,
      "loss": 0.029,
      "step": 1170
    },
    {
      "epoch": 1.8162578162578162,
      "grad_norm": 0.2696245014667511,
      "learning_rate": 8.02956161401155e-05,
      "loss": 0.0317,
      "step": 1180
    },
    {
      "epoch": 1.8316498316498318,
      "grad_norm": 0.3445149064064026,
      "learning_rate": 7.98659196835169e-05,
      "loss": 0.0397,
      "step": 1190
    },
    {
      "epoch": 1.847041847041847,
      "grad_norm": 0.45406243205070496,
      "learning_rate": 7.943276617624014e-05,
      "loss": 0.0423,
      "step": 1200
    },
    {
      "epoch": 1.847041847041847,
      "eval_loss": 0.037733353674411774,
      "eval_runtime": 160.3445,
      "eval_samples_per_second": 7.203,
      "eval_steps_per_second": 7.203,
      "step": 1200
    },
    {
      "epoch": 1.8624338624338623,
      "grad_norm": 0.37593597173690796,
      "learning_rate": 7.899620575682578e-05,
      "loss": 0.026,
      "step": 1210
    },
    {
      "epoch": 1.8778258778258778,
      "grad_norm": 0.1280842423439026,
      "learning_rate": 7.85562889581725e-05,
      "loss": 0.0318,
      "step": 1220
    },
    {
      "epoch": 1.8932178932178934,
      "grad_norm": 0.2298801690340042,
      "learning_rate": 7.811306670168788e-05,
      "loss": 0.0265,
      "step": 1230
    },
    {
      "epoch": 1.9086099086099086,
      "grad_norm": 0.392292857170105,
      "learning_rate": 7.766659029139393e-05,
      "loss": 0.0399,
      "step": 1240
    },
    {
      "epoch": 1.924001924001924,
      "grad_norm": 0.7621856331825256,
      "learning_rate": 7.72169114079887e-05,
      "loss": 0.0389,
      "step": 1250
    },
    {
      "epoch": 1.924001924001924,
      "eval_loss": 0.03466709703207016,
      "eval_runtime": 160.3061,
      "eval_samples_per_second": 7.205,
      "eval_steps_per_second": 7.205,
      "step": 1250
    },
    {
      "epoch": 1.9393939393939394,
      "grad_norm": 0.1976393312215805,
      "learning_rate": 7.676408210286407e-05,
      "loss": 0.025,
      "step": 1260
    },
    {
      "epoch": 1.9547859547859547,
      "grad_norm": 0.20227721333503723,
      "learning_rate": 7.630815479208066e-05,
      "loss": 0.0315,
      "step": 1270
    },
    {
      "epoch": 1.97017797017797,
      "grad_norm": 0.22260995209217072,
      "learning_rate": 7.584918225030053e-05,
      "loss": 0.029,
      "step": 1280
    },
    {
      "epoch": 1.9855699855699855,
      "grad_norm": 0.17172077298164368,
      "learning_rate": 7.538721760467845e-05,
      "loss": 0.0401,
      "step": 1290
    },
    {
      "epoch": 2.000962000962001,
      "grad_norm": 0.2766900658607483,
      "learning_rate": 7.49223143287122e-05,
      "loss": 0.0349,
      "step": 1300
    },
    {
      "epoch": 2.000962000962001,
      "eval_loss": 0.03779321163892746,
      "eval_runtime": 160.4154,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 1300
    },
    {
      "epoch": 2.0163540163540166,
      "grad_norm": 0.2699854373931885,
      "learning_rate": 7.445452623605307e-05,
      "loss": 0.0297,
      "step": 1310
    },
    {
      "epoch": 2.0317460317460316,
      "grad_norm": 0.1711016446352005,
      "learning_rate": 7.398390747427662e-05,
      "loss": 0.022,
      "step": 1320
    },
    {
      "epoch": 2.047138047138047,
      "grad_norm": 0.24682311713695526,
      "learning_rate": 7.351051251861501e-05,
      "loss": 0.0201,
      "step": 1330
    },
    {
      "epoch": 2.0625300625300627,
      "grad_norm": 0.14136731624603271,
      "learning_rate": 7.303439616565146e-05,
      "loss": 0.0171,
      "step": 1340
    },
    {
      "epoch": 2.0779220779220777,
      "grad_norm": 0.3170979917049408,
      "learning_rate": 7.255561352697732e-05,
      "loss": 0.0191,
      "step": 1350
    },
    {
      "epoch": 2.0779220779220777,
      "eval_loss": 0.0376242995262146,
      "eval_runtime": 160.4017,
      "eval_samples_per_second": 7.201,
      "eval_steps_per_second": 7.201,
      "step": 1350
    },
    {
      "epoch": 2.0933140933140932,
      "grad_norm": 0.3596189618110657,
      "learning_rate": 7.207422002281281e-05,
      "loss": 0.0161,
      "step": 1360
    },
    {
      "epoch": 2.1087061087061088,
      "grad_norm": 0.2732468843460083,
      "learning_rate": 7.159027137559197e-05,
      "loss": 0.019,
      "step": 1370
    },
    {
      "epoch": 2.1240981240981243,
      "grad_norm": 0.27598610520362854,
      "learning_rate": 7.110382360351267e-05,
      "loss": 0.0213,
      "step": 1380
    },
    {
      "epoch": 2.1394901394901393,
      "grad_norm": 0.24308344721794128,
      "learning_rate": 7.061493301405244e-05,
      "loss": 0.0151,
      "step": 1390
    },
    {
      "epoch": 2.154882154882155,
      "grad_norm": 0.40665531158447266,
      "learning_rate": 7.012365619745064e-05,
      "loss": 0.0252,
      "step": 1400
    },
    {
      "epoch": 2.154882154882155,
      "eval_loss": 0.037112053483724594,
      "eval_runtime": 160.559,
      "eval_samples_per_second": 7.194,
      "eval_steps_per_second": 7.194,
      "step": 1400
    },
    {
      "epoch": 2.1702741702741704,
      "grad_norm": 0.23677869141101837,
      "learning_rate": 6.963005002015808e-05,
      "loss": 0.0197,
      "step": 1410
    },
    {
      "epoch": 2.185666185666186,
      "grad_norm": 0.07495041936635971,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0147,
      "step": 1420
    },
    {
      "epoch": 2.201058201058201,
      "grad_norm": 0.27357932925224304,
      "learning_rate": 6.863607839083507e-05,
      "loss": 0.03,
      "step": 1430
    },
    {
      "epoch": 2.2164502164502164,
      "grad_norm": 0.19724859297275543,
      "learning_rate": 6.813582799336624e-05,
      "loss": 0.0163,
      "step": 1440
    },
    {
      "epoch": 2.231842231842232,
      "grad_norm": 0.21828219294548035,
      "learning_rate": 6.763347833101192e-05,
      "loss": 0.016,
      "step": 1450
    },
    {
      "epoch": 2.231842231842232,
      "eval_loss": 0.03809448331594467,
      "eval_runtime": 160.626,
      "eval_samples_per_second": 7.191,
      "eval_steps_per_second": 7.191,
      "step": 1450
    },
    {
      "epoch": 2.247234247234247,
      "grad_norm": 0.3522093594074249,
      "learning_rate": 6.712908755193094e-05,
      "loss": 0.0191,
      "step": 1460
    },
    {
      "epoch": 2.2626262626262625,
      "grad_norm": 0.3476664125919342,
      "learning_rate": 6.662271404054621e-05,
      "loss": 0.0229,
      "step": 1470
    },
    {
      "epoch": 2.278018278018278,
      "grad_norm": 0.5309289693832397,
      "learning_rate": 6.611441641078657e-05,
      "loss": 0.0252,
      "step": 1480
    },
    {
      "epoch": 2.2934102934102936,
      "grad_norm": 0.33545804023742676,
      "learning_rate": 6.560425349930208e-05,
      "loss": 0.0221,
      "step": 1490
    },
    {
      "epoch": 2.3088023088023086,
      "grad_norm": 0.22372335195541382,
      "learning_rate": 6.509228435865357e-05,
      "loss": 0.0211,
      "step": 1500
    },
    {
      "epoch": 2.3088023088023086,
      "eval_loss": 0.036216478794813156,
      "eval_runtime": 160.4231,
      "eval_samples_per_second": 7.2,
      "eval_steps_per_second": 7.2,
      "step": 1500
    },
    {
      "epoch": 2.324194324194324,
      "grad_norm": 0.20684416592121124,
      "learning_rate": 6.45785682504772e-05,
      "loss": 0.0227,
      "step": 1510
    },
    {
      "epoch": 2.3395863395863397,
      "grad_norm": 0.2306625097990036,
      "learning_rate": 6.406316463862467e-05,
      "loss": 0.0191,
      "step": 1520
    },
    {
      "epoch": 2.354978354978355,
      "grad_norm": 0.16803477704524994,
      "learning_rate": 6.35461331822803e-05,
      "loss": 0.0266,
      "step": 1530
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 0.24731677770614624,
      "learning_rate": 6.302753372905515e-05,
      "loss": 0.0307,
      "step": 1540
    },
    {
      "epoch": 2.3857623857623858,
      "grad_norm": 0.37987345457077026,
      "learning_rate": 6.250742630805971e-05,
      "loss": 0.0223,
      "step": 1550
    },
    {
      "epoch": 2.3857623857623858,
      "eval_loss": 0.03545283153653145,
      "eval_runtime": 160.3714,
      "eval_samples_per_second": 7.202,
      "eval_steps_per_second": 7.202,
      "step": 1550
    },
    {
      "epoch": 2.4011544011544013,
      "grad_norm": 0.1934383660554886,
      "learning_rate": 6.198587112295526e-05,
      "loss": 0.0223,
      "step": 1560
    },
    {
      "epoch": 2.4165464165464163,
      "grad_norm": 0.09964929521083832,
      "learning_rate": 6.14629285449852e-05,
      "loss": 0.0182,
      "step": 1570
    },
    {
      "epoch": 2.431938431938432,
      "grad_norm": 0.19195552170276642,
      "learning_rate": 6.093865910598688e-05,
      "loss": 0.0145,
      "step": 1580
    },
    {
      "epoch": 2.4473304473304474,
      "grad_norm": 0.0327489972114563,
      "learning_rate": 6.0413123491385025e-05,
      "loss": 0.0205,
      "step": 1590
    },
    {
      "epoch": 2.462722462722463,
      "grad_norm": 0.3766339421272278,
      "learning_rate": 5.988638253316713e-05,
      "loss": 0.0227,
      "step": 1600
    },
    {
      "epoch": 2.462722462722463,
      "eval_loss": 0.03853141516447067,
      "eval_runtime": 160.3204,
      "eval_samples_per_second": 7.204,
      "eval_steps_per_second": 7.204,
      "step": 1600
    },
    {
      "epoch": 2.478114478114478,
      "grad_norm": 0.2539356052875519,
      "learning_rate": 5.935849720284206e-05,
      "loss": 0.0182,
      "step": 1610
    },
    {
      "epoch": 2.4935064935064934,
      "grad_norm": 0.31089529395103455,
      "learning_rate": 5.8829528604382436e-05,
      "loss": 0.0201,
      "step": 1620
    },
    {
      "epoch": 2.508898508898509,
      "grad_norm": 0.5152587294578552,
      "learning_rate": 5.829953796715176e-05,
      "loss": 0.0292,
      "step": 1630
    },
    {
      "epoch": 2.5242905242905245,
      "grad_norm": 0.28368091583251953,
      "learning_rate": 5.7768586638816924e-05,
      "loss": 0.0242,
      "step": 1640
    },
    {
      "epoch": 2.5396825396825395,
      "grad_norm": 0.3478910028934479,
      "learning_rate": 5.723673607824705e-05,
      "loss": 0.0268,
      "step": 1650
    },
    {
      "epoch": 2.5396825396825395,
      "eval_loss": 0.03544799983501434,
      "eval_runtime": 160.3923,
      "eval_samples_per_second": 7.201,
      "eval_steps_per_second": 7.201,
      "step": 1650
    },
    {
      "epoch": 2.555074555074555,
      "grad_norm": 0.1506284922361374,
      "learning_rate": 5.670404784839953e-05,
      "loss": 0.0312,
      "step": 1660
    },
    {
      "epoch": 2.5704665704665706,
      "grad_norm": 0.18113912642002106,
      "learning_rate": 5.6170583609193984e-05,
      "loss": 0.0241,
      "step": 1670
    },
    {
      "epoch": 2.5858585858585856,
      "grad_norm": 0.3110048770904541,
      "learning_rate": 5.56364051103749e-05,
      "loss": 0.0247,
      "step": 1680
    },
    {
      "epoch": 2.601250601250601,
      "grad_norm": 0.24057482182979584,
      "learning_rate": 5.5101574184364056e-05,
      "loss": 0.0158,
      "step": 1690
    },
    {
      "epoch": 2.6166426166426167,
      "grad_norm": 0.46047157049179077,
      "learning_rate": 5.4566152739103116e-05,
      "loss": 0.0267,
      "step": 1700
    },
    {
      "epoch": 2.6166426166426167,
      "eval_loss": 0.03490716964006424,
      "eval_runtime": 160.4306,
      "eval_samples_per_second": 7.199,
      "eval_steps_per_second": 7.199,
      "step": 1700
    },
    {
      "epoch": 2.632034632034632,
      "grad_norm": 0.11632063239812851,
      "learning_rate": 5.403020275088786e-05,
      "loss": 0.0153,
      "step": 1710
    },
    {
      "epoch": 2.6474266474266477,
      "grad_norm": 0.24772502481937408,
      "learning_rate": 5.349378625719408e-05,
      "loss": 0.0182,
      "step": 1720
    },
    {
      "epoch": 2.6628186628186628,
      "grad_norm": 0.3546585142612457,
      "learning_rate": 5.2956965349496743e-05,
      "loss": 0.0245,
      "step": 1730
    },
    {
      "epoch": 2.6782106782106783,
      "grad_norm": 0.1844964474439621,
      "learning_rate": 5.241980216608261e-05,
      "loss": 0.0197,
      "step": 1740
    },
    {
      "epoch": 2.6936026936026938,
      "grad_norm": 0.31028082966804504,
      "learning_rate": 5.188235888485773e-05,
      "loss": 0.0158,
      "step": 1750
    },
    {
      "epoch": 2.6936026936026938,
      "eval_loss": 0.0352238304913044,
      "eval_runtime": 160.4591,
      "eval_samples_per_second": 7.198,
      "eval_steps_per_second": 7.198,
      "step": 1750
    },
    {
      "epoch": 2.708994708994709,
      "grad_norm": 0.28572559356689453,
      "learning_rate": 5.13446977161501e-05,
      "loss": 0.0176,
      "step": 1760
    },
    {
      "epoch": 2.7243867243867244,
      "grad_norm": 0.3480417728424072,
      "learning_rate": 5.0806880895508766e-05,
      "loss": 0.0244,
      "step": 1770
    },
    {
      "epoch": 2.73977873977874,
      "grad_norm": 0.30054420232772827,
      "learning_rate": 5.026897067649976e-05,
      "loss": 0.0252,
      "step": 1780
    },
    {
      "epoch": 2.755170755170755,
      "grad_norm": 0.29019275307655334,
      "learning_rate": 4.973102932350026e-05,
      "loss": 0.0154,
      "step": 1790
    },
    {
      "epoch": 2.7705627705627704,
      "grad_norm": 0.3516891300678253,
      "learning_rate": 4.9193119104491245e-05,
      "loss": 0.0186,
      "step": 1800
    },
    {
      "epoch": 2.7705627705627704,
      "eval_loss": 0.03843548893928528,
      "eval_runtime": 160.6061,
      "eval_samples_per_second": 7.192,
      "eval_steps_per_second": 7.192,
      "step": 1800
    },
    {
      "epoch": 2.785954785954786,
      "grad_norm": 0.48940154910087585,
      "learning_rate": 4.86553022838499e-05,
      "loss": 0.0214,
      "step": 1810
    },
    {
      "epoch": 2.8013468013468015,
      "grad_norm": 0.21716728806495667,
      "learning_rate": 4.8117641115142285e-05,
      "loss": 0.0226,
      "step": 1820
    },
    {
      "epoch": 2.816738816738817,
      "grad_norm": 0.1930004209280014,
      "learning_rate": 4.758019783391741e-05,
      "loss": 0.0238,
      "step": 1830
    },
    {
      "epoch": 2.832130832130832,
      "grad_norm": 0.4503800868988037,
      "learning_rate": 4.704303465050327e-05,
      "loss": 0.0182,
      "step": 1840
    },
    {
      "epoch": 2.8475228475228476,
      "grad_norm": 0.28708982467651367,
      "learning_rate": 4.6506213742805924e-05,
      "loss": 0.0155,
      "step": 1850
    },
    {
      "epoch": 2.8475228475228476,
      "eval_loss": 0.04005544260144234,
      "eval_runtime": 160.4557,
      "eval_samples_per_second": 7.198,
      "eval_steps_per_second": 7.198,
      "step": 1850
    },
    {
      "epoch": 2.862914862914863,
      "grad_norm": 0.3030581772327423,
      "learning_rate": 4.5969797249112146e-05,
      "loss": 0.0163,
      "step": 1860
    },
    {
      "epoch": 2.878306878306878,
      "grad_norm": 0.3970169723033905,
      "learning_rate": 4.54338472608969e-05,
      "loss": 0.0283,
      "step": 1870
    },
    {
      "epoch": 2.8936988936988937,
      "grad_norm": 0.30313968658447266,
      "learning_rate": 4.489842581563597e-05,
      "loss": 0.017,
      "step": 1880
    },
    {
      "epoch": 2.909090909090909,
      "grad_norm": 0.07091408222913742,
      "learning_rate": 4.436359488962511e-05,
      "loss": 0.0162,
      "step": 1890
    },
    {
      "epoch": 2.9244829244829242,
      "grad_norm": 0.12560871243476868,
      "learning_rate": 4.382941639080602e-05,
      "loss": 0.0158,
      "step": 1900
    },
    {
      "epoch": 2.9244829244829242,
      "eval_loss": 0.03647943586111069,
      "eval_runtime": 160.33,
      "eval_samples_per_second": 7.204,
      "eval_steps_per_second": 7.204,
      "step": 1900
    },
    {
      "epoch": 2.9398749398749398,
      "grad_norm": 0.27986544370651245,
      "learning_rate": 4.3295952151600475e-05,
      "loss": 0.0199,
      "step": 1910
    },
    {
      "epoch": 2.9552669552669553,
      "grad_norm": 0.31035858392715454,
      "learning_rate": 4.276326392175296e-05,
      "loss": 0.0193,
      "step": 1920
    },
    {
      "epoch": 2.9706589706589708,
      "grad_norm": 0.15130609273910522,
      "learning_rate": 4.2231413361183095e-05,
      "loss": 0.026,
      "step": 1930
    },
    {
      "epoch": 2.9860509860509863,
      "grad_norm": 0.2772333025932312,
      "learning_rate": 4.1700462032848244e-05,
      "loss": 0.0216,
      "step": 1940
    },
    {
      "epoch": 3.0014430014430014,
      "grad_norm": 0.3355155289173126,
      "learning_rate": 4.117047139561758e-05,
      "loss": 0.0185,
      "step": 1950
    },
    {
      "epoch": 3.0014430014430014,
      "eval_loss": 0.03621777892112732,
      "eval_runtime": 160.2832,
      "eval_samples_per_second": 7.206,
      "eval_steps_per_second": 7.206,
      "step": 1950
    },
    {
      "epoch": 3.016835016835017,
      "grad_norm": 0.10336437821388245,
      "learning_rate": 4.0641502797157966e-05,
      "loss": 0.0123,
      "step": 1960
    },
    {
      "epoch": 3.0322270322270324,
      "grad_norm": 0.0928972139954567,
      "learning_rate": 4.011361746683288e-05,
      "loss": 0.01,
      "step": 1970
    },
    {
      "epoch": 3.0476190476190474,
      "grad_norm": 0.15533702075481415,
      "learning_rate": 3.958687650861499e-05,
      "loss": 0.0078,
      "step": 1980
    },
    {
      "epoch": 3.063011063011063,
      "grad_norm": 0.0917314663529396,
      "learning_rate": 3.906134089401313e-05,
      "loss": 0.0094,
      "step": 1990
    },
    {
      "epoch": 3.0784030784030785,
      "grad_norm": 0.14643442630767822,
      "learning_rate": 3.853707145501482e-05,
      "loss": 0.0103,
      "step": 2000
    },
    {
      "epoch": 3.0784030784030785,
      "eval_loss": 0.040098611265420914,
      "eval_runtime": 160.3706,
      "eval_samples_per_second": 7.202,
      "eval_steps_per_second": 7.202,
      "step": 2000
    },
    {
      "epoch": 3.093795093795094,
      "grad_norm": 0.21054565906524658,
      "learning_rate": 3.801412887704475e-05,
      "loss": 0.0098,
      "step": 2010
    },
    {
      "epoch": 3.109187109187109,
      "grad_norm": 0.15137901902198792,
      "learning_rate": 3.749257369194029e-05,
      "loss": 0.0092,
      "step": 2020
    },
    {
      "epoch": 3.1245791245791246,
      "grad_norm": 0.4027303159236908,
      "learning_rate": 3.697246627094487e-05,
      "loss": 0.0131,
      "step": 2030
    },
    {
      "epoch": 3.13997113997114,
      "grad_norm": 0.2530450224876404,
      "learning_rate": 3.6453866817719726e-05,
      "loss": 0.01,
      "step": 2040
    },
    {
      "epoch": 3.1553631553631556,
      "grad_norm": 0.16488143801689148,
      "learning_rate": 3.593683536137533e-05,
      "loss": 0.0111,
      "step": 2050
    },
    {
      "epoch": 3.1553631553631556,
      "eval_loss": 0.04021536186337471,
      "eval_runtime": 160.0952,
      "eval_samples_per_second": 7.214,
      "eval_steps_per_second": 7.214,
      "step": 2050
    },
    {
      "epoch": 3.1707551707551707,
      "grad_norm": 0.30931276082992554,
      "learning_rate": 3.542143174952282e-05,
      "loss": 0.0068,
      "step": 2060
    },
    {
      "epoch": 3.186147186147186,
      "grad_norm": 0.9578791260719299,
      "learning_rate": 3.490771564134643e-05,
      "loss": 0.0087,
      "step": 2070
    },
    {
      "epoch": 3.2015392015392017,
      "grad_norm": 0.24608023464679718,
      "learning_rate": 3.4395746500697925e-05,
      "loss": 0.0079,
      "step": 2080
    },
    {
      "epoch": 3.2169312169312168,
      "grad_norm": 0.2659907341003418,
      "learning_rate": 3.388558358921343e-05,
      "loss": 0.0095,
      "step": 2090
    },
    {
      "epoch": 3.2323232323232323,
      "grad_norm": 0.14139080047607422,
      "learning_rate": 3.337728595945378e-05,
      "loss": 0.0105,
      "step": 2100
    },
    {
      "epoch": 3.2323232323232323,
      "eval_loss": 0.04483305290341377,
      "eval_runtime": 159.6267,
      "eval_samples_per_second": 7.236,
      "eval_steps_per_second": 7.236,
      "step": 2100
    },
    {
      "epoch": 3.2477152477152478,
      "grad_norm": 0.18512935936450958,
      "learning_rate": 3.287091244806907e-05,
      "loss": 0.0083,
      "step": 2110
    },
    {
      "epoch": 3.2631072631072633,
      "grad_norm": 0.47649720311164856,
      "learning_rate": 3.2366521668988106e-05,
      "loss": 0.0161,
      "step": 2120
    },
    {
      "epoch": 3.2784992784992784,
      "grad_norm": 0.12076949328184128,
      "learning_rate": 3.186417200663379e-05,
      "loss": 0.0094,
      "step": 2130
    },
    {
      "epoch": 3.293891293891294,
      "grad_norm": 0.5417758822441101,
      "learning_rate": 3.136392160916495e-05,
      "loss": 0.0078,
      "step": 2140
    },
    {
      "epoch": 3.3092833092833094,
      "grad_norm": 0.24188639223575592,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0077,
      "step": 2150
    },
    {
      "epoch": 3.3092833092833094,
      "eval_loss": 0.043526049703359604,
      "eval_runtime": 159.2497,
      "eval_samples_per_second": 7.253,
      "eval_steps_per_second": 7.253,
      "step": 2150
    },
    {
      "epoch": 3.324675324675325,
      "grad_norm": 0.07239481806755066,
      "learning_rate": 3.036994997984194e-05,
      "loss": 0.0072,
      "step": 2160
    },
    {
      "epoch": 3.34006734006734,
      "grad_norm": 0.05550282448530197,
      "learning_rate": 2.9876343802549357e-05,
      "loss": 0.0065,
      "step": 2170
    },
    {
      "epoch": 3.3554593554593555,
      "grad_norm": 0.30957522988319397,
      "learning_rate": 2.938506698594755e-05,
      "loss": 0.0073,
      "step": 2180
    },
    {
      "epoch": 3.370851370851371,
      "grad_norm": 0.12012802809476852,
      "learning_rate": 2.8896176396487334e-05,
      "loss": 0.0071,
      "step": 2190
    },
    {
      "epoch": 3.386243386243386,
      "grad_norm": 0.30071431398391724,
      "learning_rate": 2.8409728624408067e-05,
      "loss": 0.0078,
      "step": 2200
    },
    {
      "epoch": 3.386243386243386,
      "eval_loss": 0.047629620879888535,
      "eval_runtime": 158.7585,
      "eval_samples_per_second": 7.275,
      "eval_steps_per_second": 7.275,
      "step": 2200
    },
    {
      "epoch": 3.4016354016354016,
      "grad_norm": 0.12309174984693527,
      "learning_rate": 2.7925779977187215e-05,
      "loss": 0.0092,
      "step": 2210
    },
    {
      "epoch": 3.417027417027417,
      "grad_norm": 0.13252991437911987,
      "learning_rate": 2.7444386473022686e-05,
      "loss": 0.0082,
      "step": 2220
    },
    {
      "epoch": 3.4324194324194326,
      "grad_norm": 0.0676349624991417,
      "learning_rate": 2.696560383434854e-05,
      "loss": 0.0091,
      "step": 2230
    },
    {
      "epoch": 3.4478114478114477,
      "grad_norm": 0.10048947483301163,
      "learning_rate": 2.6489487481384988e-05,
      "loss": 0.0077,
      "step": 2240
    },
    {
      "epoch": 3.463203463203463,
      "grad_norm": 0.4449813961982727,
      "learning_rate": 2.601609252572339e-05,
      "loss": 0.0072,
      "step": 2250
    },
    {
      "epoch": 3.463203463203463,
      "eval_loss": 0.04566098749637604,
      "eval_runtime": 158.4312,
      "eval_samples_per_second": 7.29,
      "eval_steps_per_second": 7.29,
      "step": 2250
    },
    {
      "epoch": 3.4785954785954787,
      "grad_norm": 0.045422930270433426,
      "learning_rate": 2.554547376394692e-05,
      "loss": 0.0084,
      "step": 2260
    },
    {
      "epoch": 3.493987493987494,
      "grad_norm": 0.14082124829292297,
      "learning_rate": 2.5077685671287808e-05,
      "loss": 0.0115,
      "step": 2270
    },
    {
      "epoch": 3.5093795093795093,
      "grad_norm": 0.698976993560791,
      "learning_rate": 2.4612782395321593e-05,
      "loss": 0.0125,
      "step": 2280
    },
    {
      "epoch": 3.5247715247715248,
      "grad_norm": 0.28054505586624146,
      "learning_rate": 2.4150817749699496e-05,
      "loss": 0.0056,
      "step": 2290
    },
    {
      "epoch": 3.5401635401635403,
      "grad_norm": 0.6282978057861328,
      "learning_rate": 2.3691845207919357e-05,
      "loss": 0.0118,
      "step": 2300
    },
    {
      "epoch": 3.5401635401635403,
      "eval_loss": 0.04520741105079651,
      "eval_runtime": 158.2029,
      "eval_samples_per_second": 7.301,
      "eval_steps_per_second": 7.301,
      "step": 2300
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 0.07654228806495667,
      "learning_rate": 2.3235917897135934e-05,
      "loss": 0.0063,
      "step": 2310
    },
    {
      "epoch": 3.570947570947571,
      "grad_norm": 0.2524222731590271,
      "learning_rate": 2.2783088592011303e-05,
      "loss": 0.0079,
      "step": 2320
    },
    {
      "epoch": 3.5863395863395864,
      "grad_norm": 0.4098080098628998,
      "learning_rate": 2.233340970860608e-05,
      "loss": 0.0094,
      "step": 2330
    },
    {
      "epoch": 3.601731601731602,
      "grad_norm": 0.3212938606739044,
      "learning_rate": 2.1886933298312125e-05,
      "loss": 0.0096,
      "step": 2340
    },
    {
      "epoch": 3.617123617123617,
      "grad_norm": 0.1865200251340866,
      "learning_rate": 2.1443711041827503e-05,
      "loss": 0.0107,
      "step": 2350
    },
    {
      "epoch": 3.617123617123617,
      "eval_loss": 0.04479704424738884,
      "eval_runtime": 158.2191,
      "eval_samples_per_second": 7.3,
      "eval_steps_per_second": 7.3,
      "step": 2350
    },
    {
      "epoch": 3.6325156325156325,
      "grad_norm": 0.07322122901678085,
      "learning_rate": 2.1003794243174252e-05,
      "loss": 0.0049,
      "step": 2360
    },
    {
      "epoch": 3.647907647907648,
      "grad_norm": 0.03634365275502205,
      "learning_rate": 2.0567233823759886e-05,
      "loss": 0.0058,
      "step": 2370
    },
    {
      "epoch": 3.6632996632996635,
      "grad_norm": 0.2598310112953186,
      "learning_rate": 2.0134080316483113e-05,
      "loss": 0.0062,
      "step": 2380
    },
    {
      "epoch": 3.6786916786916786,
      "grad_norm": 0.44899070262908936,
      "learning_rate": 1.970438385988452e-05,
      "loss": 0.0055,
      "step": 2390
    },
    {
      "epoch": 3.694083694083694,
      "grad_norm": 1.2249691486358643,
      "learning_rate": 1.927819419234292e-05,
      "loss": 0.01,
      "step": 2400
    },
    {
      "epoch": 3.694083694083694,
      "eval_loss": 0.04775693640112877,
      "eval_runtime": 158.2614,
      "eval_samples_per_second": 7.298,
      "eval_steps_per_second": 7.298,
      "step": 2400
    },
    {
      "epoch": 3.7094757094757096,
      "grad_norm": 0.03975760191679001,
      "learning_rate": 1.885556064631801e-05,
      "loss": 0.012,
      "step": 2410
    },
    {
      "epoch": 3.7248677248677247,
      "grad_norm": 0.06989089399576187,
      "learning_rate": 1.843653214264003e-05,
      "loss": 0.0087,
      "step": 2420
    },
    {
      "epoch": 3.74025974025974,
      "grad_norm": 0.049630213528871536,
      "learning_rate": 1.8021157184847087e-05,
      "loss": 0.0113,
      "step": 2430
    },
    {
      "epoch": 3.7556517556517557,
      "grad_norm": 0.13058657944202423,
      "learning_rate": 1.7609483853570668e-05,
      "loss": 0.008,
      "step": 2440
    },
    {
      "epoch": 3.771043771043771,
      "grad_norm": 0.413333922624588,
      "learning_rate": 1.720155980097034e-05,
      "loss": 0.0092,
      "step": 2450
    },
    {
      "epoch": 3.771043771043771,
      "eval_loss": 0.04710584878921509,
      "eval_runtime": 158.1627,
      "eval_samples_per_second": 7.303,
      "eval_steps_per_second": 7.303,
      "step": 2450
    },
    {
      "epoch": 3.7864357864357867,
      "grad_norm": 0.03138028830289841,
      "learning_rate": 1.6797432245217736e-05,
      "loss": 0.0096,
      "step": 2460
    },
    {
      "epoch": 3.801827801827802,
      "grad_norm": 0.040889088064432144,
      "learning_rate": 1.6397147965031072e-05,
      "loss": 0.0076,
      "step": 2470
    },
    {
      "epoch": 3.8172198172198173,
      "grad_norm": 0.654095470905304,
      "learning_rate": 1.600075329426033e-05,
      "loss": 0.011,
      "step": 2480
    },
    {
      "epoch": 3.832611832611833,
      "grad_norm": 0.20909631252288818,
      "learning_rate": 1.560829411652403e-05,
      "loss": 0.0096,
      "step": 2490
    },
    {
      "epoch": 3.848003848003848,
      "grad_norm": 0.34065234661102295,
      "learning_rate": 1.521981585989809e-05,
      "loss": 0.0166,
      "step": 2500
    },
    {
      "epoch": 3.848003848003848,
      "eval_loss": 0.04371841251850128,
      "eval_runtime": 158.1627,
      "eval_samples_per_second": 7.303,
      "eval_steps_per_second": 7.303,
      "step": 2500
    },
    {
      "epoch": 3.8633958633958634,
      "grad_norm": 0.1074584573507309,
      "learning_rate": 1.4835363491657412e-05,
      "loss": 0.011,
      "step": 2510
    },
    {
      "epoch": 3.878787878787879,
      "grad_norm": 0.14425139129161835,
      "learning_rate": 1.4454981513070791e-05,
      "loss": 0.0121,
      "step": 2520
    },
    {
      "epoch": 3.894179894179894,
      "grad_norm": 0.2657093107700348,
      "learning_rate": 1.4078713954249873e-05,
      "loss": 0.0061,
      "step": 2530
    },
    {
      "epoch": 3.9095719095719095,
      "grad_norm": 0.09532363712787628,
      "learning_rate": 1.3706604369052434e-05,
      "loss": 0.0059,
      "step": 2540
    },
    {
      "epoch": 3.924963924963925,
      "grad_norm": 0.22441978752613068,
      "learning_rate": 1.3338695830041047e-05,
      "loss": 0.0048,
      "step": 2550
    },
    {
      "epoch": 3.924963924963925,
      "eval_loss": 0.044441528618335724,
      "eval_runtime": 158.141,
      "eval_samples_per_second": 7.304,
      "eval_steps_per_second": 7.304,
      "step": 2550
    },
    {
      "epoch": 3.9403559403559405,
      "grad_norm": 0.21590059995651245,
      "learning_rate": 1.2975030923497262e-05,
      "loss": 0.0052,
      "step": 2560
    },
    {
      "epoch": 3.955747955747956,
      "grad_norm": 0.41011568903923035,
      "learning_rate": 1.2615651744492151e-05,
      "loss": 0.0098,
      "step": 2570
    },
    {
      "epoch": 3.971139971139971,
      "grad_norm": 0.14635717868804932,
      "learning_rate": 1.2260599892013719e-05,
      "loss": 0.0059,
      "step": 2580
    },
    {
      "epoch": 3.9865319865319866,
      "grad_norm": 0.14410582184791565,
      "learning_rate": 1.1909916464151715e-05,
      "loss": 0.0084,
      "step": 2590
    },
    {
      "epoch": 4.001924001924002,
      "grad_norm": 0.11865393072366714,
      "learning_rate": 1.1563642053340406e-05,
      "loss": 0.0057,
      "step": 2600
    },
    {
      "epoch": 4.001924001924002,
      "eval_loss": 0.0453677773475647,
      "eval_runtime": 157.9245,
      "eval_samples_per_second": 7.314,
      "eval_steps_per_second": 7.314,
      "step": 2600
    },
    {
      "epoch": 4.017316017316017,
      "grad_norm": 0.06959637254476547,
      "learning_rate": 1.1221816741659964e-05,
      "loss": 0.0042,
      "step": 2610
    },
    {
      "epoch": 4.032708032708033,
      "grad_norm": 0.13380874693393707,
      "learning_rate": 1.0884480096196803e-05,
      "loss": 0.0037,
      "step": 2620
    },
    {
      "epoch": 4.048100048100048,
      "grad_norm": 0.06643780320882797,
      "learning_rate": 1.0551671164463666e-05,
      "loss": 0.0043,
      "step": 2630
    },
    {
      "epoch": 4.063492063492063,
      "grad_norm": 0.14242886006832123,
      "learning_rate": 1.0223428469879736e-05,
      "loss": 0.0051,
      "step": 2640
    },
    {
      "epoch": 4.078884078884079,
      "grad_norm": 0.16189664602279663,
      "learning_rate": 9.899790007311455e-06,
      "loss": 0.0033,
      "step": 2650
    },
    {
      "epoch": 4.078884078884079,
      "eval_loss": 0.048355016857385635,
      "eval_runtime": 157.8991,
      "eval_samples_per_second": 7.315,
      "eval_steps_per_second": 7.315,
      "step": 2650
    },
    {
      "epoch": 4.094276094276094,
      "grad_norm": 0.022516440600156784,
      "learning_rate": 9.580793238674573e-06,
      "loss": 0.0028,
      "step": 2660
    },
    {
      "epoch": 4.109668109668109,
      "grad_norm": 0.24439583718776703,
      "learning_rate": 9.266475088597781e-06,
      "loss": 0.0019,
      "step": 2670
    },
    {
      "epoch": 4.125060125060125,
      "grad_norm": 0.30291637778282166,
      "learning_rate": 8.956871940148681e-06,
      "loss": 0.0057,
      "step": 2680
    },
    {
      "epoch": 4.14045214045214,
      "grad_norm": 0.07655764371156693,
      "learning_rate": 8.65201963062225e-06,
      "loss": 0.0044,
      "step": 2690
    },
    {
      "epoch": 4.1558441558441555,
      "grad_norm": 0.13116689026355743,
      "learning_rate": 8.351953447392674e-06,
      "loss": 0.0032,
      "step": 2700
    },
    {
      "epoch": 4.1558441558441555,
      "eval_loss": 0.05001786723732948,
      "eval_runtime": 157.9193,
      "eval_samples_per_second": 7.314,
      "eval_steps_per_second": 7.314,
      "step": 2700
    },
    {
      "epoch": 4.171236171236171,
      "grad_norm": 0.23487193882465363,
      "learning_rate": 8.056708123828738e-06,
      "loss": 0.0025,
      "step": 2710
    },
    {
      "epoch": 4.1866281866281865,
      "grad_norm": 0.0992528423666954,
      "learning_rate": 7.76631783527334e-06,
      "loss": 0.0021,
      "step": 2720
    },
    {
      "epoch": 4.202020202020202,
      "grad_norm": 0.30396416783332825,
      "learning_rate": 7.4808161950876145e-06,
      "loss": 0.0046,
      "step": 2730
    },
    {
      "epoch": 4.2174122174122175,
      "grad_norm": 0.0412018857896328,
      "learning_rate": 7.200236250760111e-06,
      "loss": 0.0026,
      "step": 2740
    },
    {
      "epoch": 4.232804232804233,
      "grad_norm": 0.009596334770321846,
      "learning_rate": 6.924610480081478e-06,
      "loss": 0.005,
      "step": 2750
    },
    {
      "epoch": 4.232804232804233,
      "eval_loss": 0.0526924729347229,
      "eval_runtime": 157.9437,
      "eval_samples_per_second": 7.313,
      "eval_steps_per_second": 7.313,
      "step": 2750
    },
    {
      "epoch": 4.2481962481962485,
      "grad_norm": 0.038329195231199265,
      "learning_rate": 6.6539707873850774e-06,
      "loss": 0.0025,
      "step": 2760
    },
    {
      "epoch": 4.263588263588264,
      "grad_norm": 0.0019896586891263723,
      "learning_rate": 6.3883484998539525e-06,
      "loss": 0.0029,
      "step": 2770
    },
    {
      "epoch": 4.278980278980279,
      "grad_norm": 0.06064137443900108,
      "learning_rate": 6.127774363894667e-06,
      "loss": 0.0017,
      "step": 2780
    },
    {
      "epoch": 4.294372294372295,
      "grad_norm": 0.07957368344068527,
      "learning_rate": 5.872278541578352e-06,
      "loss": 0.0051,
      "step": 2790
    },
    {
      "epoch": 4.30976430976431,
      "grad_norm": 0.007363211829215288,
      "learning_rate": 5.6218906071493275e-06,
      "loss": 0.004,
      "step": 2800
    },
    {
      "epoch": 4.30976430976431,
      "eval_loss": 0.05456327274441719,
      "eval_runtime": 157.911,
      "eval_samples_per_second": 7.314,
      "eval_steps_per_second": 7.314,
      "step": 2800
    },
    {
      "epoch": 4.325156325156325,
      "grad_norm": 0.2222929447889328,
      "learning_rate": 5.376639543601858e-06,
      "loss": 0.0028,
      "step": 2810
    },
    {
      "epoch": 4.340548340548341,
      "grad_norm": 0.0447678342461586,
      "learning_rate": 5.136553739325256e-06,
      "loss": 0.0017,
      "step": 2820
    },
    {
      "epoch": 4.355940355940356,
      "grad_norm": 0.04685351252555847,
      "learning_rate": 4.901660984817908e-06,
      "loss": 0.0027,
      "step": 2830
    },
    {
      "epoch": 4.371332371332372,
      "grad_norm": 0.9541874527931213,
      "learning_rate": 4.671988469470434e-06,
      "loss": 0.0036,
      "step": 2840
    },
    {
      "epoch": 4.386724386724387,
      "grad_norm": 0.08022912591695786,
      "learning_rate": 4.447562778418435e-06,
      "loss": 0.0034,
      "step": 2850
    },
    {
      "epoch": 4.386724386724387,
      "eval_loss": 0.055369455367326736,
      "eval_runtime": 157.9015,
      "eval_samples_per_second": 7.315,
      "eval_steps_per_second": 7.315,
      "step": 2850
    },
    {
      "epoch": 4.402116402116402,
      "grad_norm": 0.1095672994852066,
      "learning_rate": 4.228409889465218e-06,
      "loss": 0.0018,
      "step": 2860
    },
    {
      "epoch": 4.417508417508418,
      "grad_norm": 0.7465848922729492,
      "learning_rate": 4.014555170074824e-06,
      "loss": 0.0041,
      "step": 2870
    },
    {
      "epoch": 4.432900432900433,
      "grad_norm": 0.11146574467420578,
      "learning_rate": 3.8060233744356633e-06,
      "loss": 0.0018,
      "step": 2880
    },
    {
      "epoch": 4.448292448292448,
      "grad_norm": 0.06695666164159775,
      "learning_rate": 3.602838640595152e-06,
      "loss": 0.0026,
      "step": 2890
    },
    {
      "epoch": 4.463684463684464,
      "grad_norm": 0.15812787413597107,
      "learning_rate": 3.4050244876656866e-06,
      "loss": 0.0023,
      "step": 2900
    },
    {
      "epoch": 4.463684463684464,
      "eval_loss": 0.05597861111164093,
      "eval_runtime": 158.0063,
      "eval_samples_per_second": 7.31,
      "eval_steps_per_second": 7.31,
      "step": 2900
    },
    {
      "epoch": 4.479076479076479,
      "grad_norm": 0.028353378176689148,
      "learning_rate": 3.2126038131022552e-06,
      "loss": 0.001,
      "step": 2910
    },
    {
      "epoch": 4.494468494468494,
      "grad_norm": 0.13058075308799744,
      "learning_rate": 3.025598890051995e-06,
      "loss": 0.0016,
      "step": 2920
    },
    {
      "epoch": 4.50986050986051,
      "grad_norm": 0.0282717477530241,
      "learning_rate": 2.8440313647760288e-06,
      "loss": 0.0058,
      "step": 2930
    },
    {
      "epoch": 4.525252525252525,
      "grad_norm": 0.2148657888174057,
      "learning_rate": 2.6679222541438397e-06,
      "loss": 0.0041,
      "step": 2940
    },
    {
      "epoch": 4.540644540644541,
      "grad_norm": 0.07000631839036942,
      "learning_rate": 2.4972919432005582e-06,
      "loss": 0.0027,
      "step": 2950
    },
    {
      "epoch": 4.540644540644541,
      "eval_loss": 0.05637305602431297,
      "eval_runtime": 157.8692,
      "eval_samples_per_second": 7.316,
      "eval_steps_per_second": 7.316,
      "step": 2950
    },
    {
      "epoch": 4.556036556036556,
      "grad_norm": 0.058950576931238174,
      "learning_rate": 2.3321601828073135e-06,
      "loss": 0.0054,
      "step": 2960
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.12322745472192764,
      "learning_rate": 2.1725460873550317e-06,
      "loss": 0.0022,
      "step": 2970
    },
    {
      "epoch": 4.586820586820587,
      "grad_norm": 0.0949539840221405,
      "learning_rate": 2.0184681325519096e-06,
      "loss": 0.0043,
      "step": 2980
    },
    {
      "epoch": 4.602212602212602,
      "grad_norm": 0.01033385656774044,
      "learning_rate": 1.869944153284786e-06,
      "loss": 0.0039,
      "step": 2990
    },
    {
      "epoch": 4.617604617604617,
      "grad_norm": 0.09847161173820496,
      "learning_rate": 1.7269913415547456e-06,
      "loss": 0.0025,
      "step": 3000
    },
    {
      "epoch": 4.617604617604617,
      "eval_loss": 0.0562971830368042,
      "eval_runtime": 157.7475,
      "eval_samples_per_second": 7.322,
      "eval_steps_per_second": 7.322,
      "step": 3000
    },
    {
      "epoch": 4.632996632996633,
      "grad_norm": 0.07785726338624954,
      "learning_rate": 1.589626244487069e-06,
      "loss": 0.0029,
      "step": 3010
    },
    {
      "epoch": 4.648388648388648,
      "grad_norm": 0.0484982393682003,
      "learning_rate": 1.4578647624158726e-06,
      "loss": 0.0041,
      "step": 3020
    },
    {
      "epoch": 4.663780663780663,
      "grad_norm": 0.4885905086994171,
      "learning_rate": 1.3317221470436293e-06,
      "loss": 0.0034,
      "step": 3030
    },
    {
      "epoch": 4.679172679172679,
      "grad_norm": 0.2064741998910904,
      "learning_rate": 1.2112129996757315e-06,
      "loss": 0.0036,
      "step": 3040
    },
    {
      "epoch": 4.694564694564694,
      "grad_norm": 0.06070768088102341,
      "learning_rate": 1.0963512695303546e-06,
      "loss": 0.0054,
      "step": 3050
    },
    {
      "epoch": 4.694564694564694,
      "eval_loss": 0.05679197236895561,
      "eval_runtime": 157.7617,
      "eval_samples_per_second": 7.321,
      "eval_steps_per_second": 7.321,
      "step": 3050
    },
    {
      "epoch": 4.70995670995671,
      "grad_norm": 0.015436189249157906,
      "learning_rate": 9.871502521237975e-07,
      "loss": 0.003,
      "step": 3060
    },
    {
      "epoch": 4.725348725348725,
      "grad_norm": 0.08534955233335495,
      "learning_rate": 8.83622587731514e-07,
      "loss": 0.0026,
      "step": 3070
    },
    {
      "epoch": 4.7407407407407405,
      "grad_norm": 0.18290217220783234,
      "learning_rate": 7.857802599249564e-07,
      "loss": 0.0019,
      "step": 3080
    },
    {
      "epoch": 4.756132756132756,
      "grad_norm": 0.012095090933144093,
      "learning_rate": 6.936345941844335e-07,
      "loss": 0.0039,
      "step": 3090
    },
    {
      "epoch": 4.7715247715247715,
      "grad_norm": 0.032935187220573425,
      "learning_rate": 6.071962565881939e-07,
      "loss": 0.0016,
      "step": 3100
    },
    {
      "epoch": 4.7715247715247715,
      "eval_loss": 0.05687952786684036,
      "eval_runtime": 157.7599,
      "eval_samples_per_second": 7.321,
      "eval_steps_per_second": 7.321,
      "step": 3100
    },
    {
      "epoch": 4.786916786916787,
      "grad_norm": 0.03579120710492134,
      "learning_rate": 5.264752525777961e-07,
      "loss": 0.0014,
      "step": 3110
    },
    {
      "epoch": 4.8023088023088025,
      "grad_norm": 0.34775373339653015,
      "learning_rate": 4.5148092579991863e-07,
      "loss": 0.0023,
      "step": 3120
    },
    {
      "epoch": 4.817700817700818,
      "grad_norm": 0.2920624911785126,
      "learning_rate": 3.822219570248531e-07,
      "loss": 0.0033,
      "step": 3130
    },
    {
      "epoch": 4.833092833092833,
      "grad_norm": 0.03081696853041649,
      "learning_rate": 3.187063631416742e-07,
      "loss": 0.0048,
      "step": 3140
    },
    {
      "epoch": 4.848484848484849,
      "grad_norm": 0.07112334668636322,
      "learning_rate": 2.6094149623024365e-07,
      "loss": 0.0024,
      "step": 3150
    },
    {
      "epoch": 4.848484848484849,
      "eval_loss": 0.056653253734111786,
      "eval_runtime": 157.8802,
      "eval_samples_per_second": 7.316,
      "eval_steps_per_second": 7.316,
      "step": 3150
    },
    {
      "epoch": 4.863876863876864,
      "grad_norm": 0.1798284500837326,
      "learning_rate": 2.0893404271022376e-07,
      "loss": 0.002,
      "step": 3160
    },
    {
      "epoch": 4.87926887926888,
      "grad_norm": 0.08457917720079422,
      "learning_rate": 1.626900225670691e-07,
      "loss": 0.0021,
      "step": 3170
    },
    {
      "epoch": 4.894660894660895,
      "grad_norm": 0.057848088443279266,
      "learning_rate": 1.2221478865524493e-07,
      "loss": 0.0038,
      "step": 3180
    },
    {
      "epoch": 4.91005291005291,
      "grad_norm": 0.12058410048484802,
      "learning_rate": 8.751302607857836e-08,
      "loss": 0.0022,
      "step": 3190
    },
    {
      "epoch": 4.925444925444926,
      "grad_norm": 0.028752174228429794,
      "learning_rate": 5.8588751647975506e-08,
      "loss": 0.0018,
      "step": 3200
    },
    {
      "epoch": 4.925444925444926,
      "eval_loss": 0.056797631084918976,
      "eval_runtime": 157.844,
      "eval_samples_per_second": 7.317,
      "eval_steps_per_second": 7.317,
      "step": 3200
    },
    {
      "epoch": 4.940836940836941,
      "grad_norm": 0.13063248991966248,
      "learning_rate": 3.544531341646007e-08,
      "loss": 0.0062,
      "step": 3210
    },
    {
      "epoch": 4.956228956228956,
      "grad_norm": 0.27323055267333984,
      "learning_rate": 1.808539029161116e-08,
      "loss": 0.0031,
      "step": 3220
    },
    {
      "epoch": 4.971620971620972,
      "grad_norm": 0.03140595927834511,
      "learning_rate": 6.510991725500182e-09,
      "loss": 0.0027,
      "step": 3230
    },
    {
      "epoch": 4.987012987012987,
      "grad_norm": 0.02295578084886074,
      "learning_rate": 7.234574820769169e-10,
      "loss": 0.0016,
      "step": 3240
    },
    {
      "epoch": 4.994708994708994,
      "step": 3245,
      "total_flos": 8.15387262412456e+17,
      "train_loss": 0.041617053084775736,
      "train_runtime": 36330.2026,
      "train_samples_per_second": 1.431,
      "train_steps_per_second": 0.089
    }
  ],
  "logging_steps": 10,
  "max_steps": 3245,
  "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": 8.15387262412456e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}