{
  "best_metric": 0.8732149076976663,
  "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-FER2013/checkpoint-2020",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 2020,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "grad_norm": 0.699901819229126,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.7003,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.40010973811149597,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.6978,
      "step": 20
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3860187530517578,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.6891,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.38579830527305603,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.6837,
      "step": 40
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5694485306739807,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.6643,
      "step": 50
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5950794816017151,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.642,
      "step": 60
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.7590372562408447,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.6213,
      "step": 70
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7452055215835571,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.6079,
      "step": 80
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1108556985855103,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.568,
      "step": 90
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.0815961360931396,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.5555,
      "step": 100
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.206498146057129,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.5587,
      "step": 110
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8447802066802979,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.5502,
      "step": 120
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9129859805107117,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.5442,
      "step": 130
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.3158477544784546,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.509,
      "step": 140
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6786607503890991,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.5129,
      "step": 150
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.0716958045959473,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.5043,
      "step": 160
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8416863679885864,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4766,
      "step": 170
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.109766721725464,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4746,
      "step": 180
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.7481091022491455,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.5162,
      "step": 190
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.037062406539917,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.4811,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.800417972831766,
      "eval_loss": 0.43151769042015076,
      "eval_runtime": 15.3075,
      "eval_samples_per_second": 187.555,
      "eval_steps_per_second": 5.879,
      "step": 202
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.0516308546066284,
      "learning_rate": 4.977997799779978e-05,
      "loss": 0.4625,
      "step": 210
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.0761890411376953,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.4659,
      "step": 220
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.9692820310592651,
      "learning_rate": 4.9229922992299234e-05,
      "loss": 0.4893,
      "step": 230
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.9031820893287659,
      "learning_rate": 4.895489548954896e-05,
      "loss": 0.4801,
      "step": 240
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.241443395614624,
      "learning_rate": 4.867986798679868e-05,
      "loss": 0.465,
      "step": 250
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.1679638624191284,
      "learning_rate": 4.8404840484048406e-05,
      "loss": 0.4396,
      "step": 260
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.4820034503936768,
      "learning_rate": 4.812981298129813e-05,
      "loss": 0.4573,
      "step": 270
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.0649698972702026,
      "learning_rate": 4.785478547854786e-05,
      "loss": 0.4616,
      "step": 280
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.2591346502304077,
      "learning_rate": 4.7579757975797585e-05,
      "loss": 0.4661,
      "step": 290
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.9309344291687012,
      "learning_rate": 4.730473047304731e-05,
      "loss": 0.4792,
      "step": 300
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.9727767109870911,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.471,
      "step": 310
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.0227768421173096,
      "learning_rate": 4.675467546754676e-05,
      "loss": 0.4779,
      "step": 320
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.7914155721664429,
      "learning_rate": 4.647964796479648e-05,
      "loss": 0.4552,
      "step": 330
    },
    {
      "epoch": 1.68,
      "grad_norm": 1.1211057901382446,
      "learning_rate": 4.62046204620462e-05,
      "loss": 0.4542,
      "step": 340
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.8994653820991516,
      "learning_rate": 4.592959295929593e-05,
      "loss": 0.469,
      "step": 350
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.6254159212112427,
      "learning_rate": 4.5654565456545655e-05,
      "loss": 0.4209,
      "step": 360
    },
    {
      "epoch": 1.83,
      "grad_norm": 1.175451397895813,
      "learning_rate": 4.537953795379538e-05,
      "loss": 0.4707,
      "step": 370
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.6570440530776978,
      "learning_rate": 4.510451045104511e-05,
      "loss": 0.4587,
      "step": 380
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.0916353464126587,
      "learning_rate": 4.4829482948294834e-05,
      "loss": 0.4507,
      "step": 390
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.0758109092712402,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4287,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8432601880877743,
      "eval_loss": 0.357921838760376,
      "eval_runtime": 15.3987,
      "eval_samples_per_second": 186.444,
      "eval_steps_per_second": 5.845,
      "step": 404
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.7765569686889648,
      "learning_rate": 4.427942794279428e-05,
      "loss": 0.4157,
      "step": 410
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.1473923921585083,
      "learning_rate": 4.4004400440044006e-05,
      "loss": 0.4227,
      "step": 420
    },
    {
      "epoch": 2.13,
      "grad_norm": 7.452829837799072,
      "learning_rate": 4.372937293729373e-05,
      "loss": 0.4033,
      "step": 430
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.391101360321045,
      "learning_rate": 4.345434543454346e-05,
      "loss": 0.4563,
      "step": 440
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.0371557474136353,
      "learning_rate": 4.3179317931793185e-05,
      "loss": 0.4487,
      "step": 450
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.998115062713623,
      "learning_rate": 4.2904290429042904e-05,
      "loss": 0.4178,
      "step": 460
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.1916583776474,
      "learning_rate": 4.262926292629263e-05,
      "loss": 0.4361,
      "step": 470
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.2483221292495728,
      "learning_rate": 4.2354235423542356e-05,
      "loss": 0.4317,
      "step": 480
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.2603436708450317,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4252,
      "step": 490
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.614512324333191,
      "learning_rate": 4.18041804180418e-05,
      "loss": 0.4213,
      "step": 500
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.5845041275024414,
      "learning_rate": 4.152915291529153e-05,
      "loss": 0.4276,
      "step": 510
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.9629144072532654,
      "learning_rate": 4.1254125412541255e-05,
      "loss": 0.4168,
      "step": 520
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.0476891994476318,
      "learning_rate": 4.097909790979098e-05,
      "loss": 0.4166,
      "step": 530
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.2000118494033813,
      "learning_rate": 4.070407040704071e-05,
      "loss": 0.4238,
      "step": 540
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.608628511428833,
      "learning_rate": 4.042904290429043e-05,
      "loss": 0.3974,
      "step": 550
    },
    {
      "epoch": 2.77,
      "grad_norm": 2.0195207595825195,
      "learning_rate": 4.015401540154016e-05,
      "loss": 0.4087,
      "step": 560
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.2935571670532227,
      "learning_rate": 3.987898789878988e-05,
      "loss": 0.396,
      "step": 570
    },
    {
      "epoch": 2.87,
      "grad_norm": 1.2056093215942383,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.4191,
      "step": 580
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.9370687007904053,
      "learning_rate": 3.932893289328933e-05,
      "loss": 0.4323,
      "step": 590
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.9331934452056885,
      "learning_rate": 3.905390539053906e-05,
      "loss": 0.4184,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8467432950191571,
      "eval_loss": 0.35166114568710327,
      "eval_runtime": 15.3286,
      "eval_samples_per_second": 187.297,
      "eval_steps_per_second": 5.871,
      "step": 606
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.35804283618927,
      "learning_rate": 3.877887788778878e-05,
      "loss": 0.4034,
      "step": 610
    },
    {
      "epoch": 3.07,
      "grad_norm": 1.5728434324264526,
      "learning_rate": 3.8503850385038503e-05,
      "loss": 0.3917,
      "step": 620
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.0493104457855225,
      "learning_rate": 3.822882288228823e-05,
      "loss": 0.416,
      "step": 630
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.2043938636779785,
      "learning_rate": 3.7953795379537956e-05,
      "loss": 0.3913,
      "step": 640
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.6263302564620972,
      "learning_rate": 3.767876787678768e-05,
      "loss": 0.3815,
      "step": 650
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.8161449432373047,
      "learning_rate": 3.74037403740374e-05,
      "loss": 0.3802,
      "step": 660
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.2450302839279175,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.387,
      "step": 670
    },
    {
      "epoch": 3.37,
      "grad_norm": 1.1230440139770508,
      "learning_rate": 3.6853685368536854e-05,
      "loss": 0.4023,
      "step": 680
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.3349334001541138,
      "learning_rate": 3.657865786578658e-05,
      "loss": 0.4092,
      "step": 690
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.1348090171813965,
      "learning_rate": 3.6303630363036307e-05,
      "loss": 0.3724,
      "step": 700
    },
    {
      "epoch": 3.51,
      "grad_norm": 1.1959893703460693,
      "learning_rate": 3.602860286028603e-05,
      "loss": 0.415,
      "step": 710
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.0938156843185425,
      "learning_rate": 3.575357535753576e-05,
      "loss": 0.3931,
      "step": 720
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.199876308441162,
      "learning_rate": 3.5478547854785485e-05,
      "loss": 0.3817,
      "step": 730
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.3452043533325195,
      "learning_rate": 3.5203520352035205e-05,
      "loss": 0.3897,
      "step": 740
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.3791470527648926,
      "learning_rate": 3.492849284928493e-05,
      "loss": 0.3828,
      "step": 750
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.7758994102478027,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.382,
      "step": 760
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.0924941301345825,
      "learning_rate": 3.4378437843784377e-05,
      "loss": 0.4048,
      "step": 770
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.1729379892349243,
      "learning_rate": 3.41034103410341e-05,
      "loss": 0.3716,
      "step": 780
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.1679847240447998,
      "learning_rate": 3.382838283828383e-05,
      "loss": 0.3726,
      "step": 790
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.166401743888855,
      "learning_rate": 3.3553355335533555e-05,
      "loss": 0.3931,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.855451062347614,
      "eval_loss": 0.33083683252334595,
      "eval_runtime": 15.5188,
      "eval_samples_per_second": 185.001,
      "eval_steps_per_second": 5.799,
      "step": 808
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.225820541381836,
      "learning_rate": 3.327832783278328e-05,
      "loss": 0.3976,
      "step": 810
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.0681790113449097,
      "learning_rate": 3.300330033003301e-05,
      "loss": 0.3641,
      "step": 820
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.3162754774093628,
      "learning_rate": 3.272827282728273e-05,
      "loss": 0.3601,
      "step": 830
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.5509825944900513,
      "learning_rate": 3.2453245324532453e-05,
      "loss": 0.3777,
      "step": 840
    },
    {
      "epoch": 4.21,
      "grad_norm": 1.4965535402297974,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.3662,
      "step": 850
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.846308946609497,
      "learning_rate": 3.1903190319031906e-05,
      "loss": 0.3755,
      "step": 860
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.008911371231079,
      "learning_rate": 3.162816281628163e-05,
      "loss": 0.3283,
      "step": 870
    },
    {
      "epoch": 4.36,
      "grad_norm": 1.7026519775390625,
      "learning_rate": 3.135313531353136e-05,
      "loss": 0.3782,
      "step": 880
    },
    {
      "epoch": 4.41,
      "grad_norm": 1.3682829141616821,
      "learning_rate": 3.1078107810781085e-05,
      "loss": 0.3943,
      "step": 890
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.2787247896194458,
      "learning_rate": 3.0803080308030804e-05,
      "loss": 0.3651,
      "step": 900
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.2230846881866455,
      "learning_rate": 3.052805280528053e-05,
      "loss": 0.3489,
      "step": 910
    },
    {
      "epoch": 4.55,
      "grad_norm": 1.4894435405731201,
      "learning_rate": 3.0253025302530253e-05,
      "loss": 0.372,
      "step": 920
    },
    {
      "epoch": 4.6,
      "grad_norm": 1.6204265356063843,
      "learning_rate": 2.9977997799779976e-05,
      "loss": 0.3637,
      "step": 930
    },
    {
      "epoch": 4.65,
      "grad_norm": 1.4866546392440796,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.3657,
      "step": 940
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.8092739582061768,
      "learning_rate": 2.942794279427943e-05,
      "loss": 0.375,
      "step": 950
    },
    {
      "epoch": 4.75,
      "grad_norm": 1.3936327695846558,
      "learning_rate": 2.9152915291529155e-05,
      "loss": 0.3737,
      "step": 960
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.3176568746566772,
      "learning_rate": 2.8877887788778878e-05,
      "loss": 0.3565,
      "step": 970
    },
    {
      "epoch": 4.85,
      "grad_norm": 1.4575417041778564,
      "learning_rate": 2.8602860286028604e-05,
      "loss": 0.3611,
      "step": 980
    },
    {
      "epoch": 4.9,
      "grad_norm": 1.5857540369033813,
      "learning_rate": 2.832783278327833e-05,
      "loss": 0.3656,
      "step": 990
    },
    {
      "epoch": 4.95,
      "grad_norm": 1.0636128187179565,
      "learning_rate": 2.8052805280528056e-05,
      "loss": 0.3396,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4022269248962402,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.3667,
      "step": 1010
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8610240334378265,
      "eval_loss": 0.3203551471233368,
      "eval_runtime": 15.3956,
      "eval_samples_per_second": 186.482,
      "eval_steps_per_second": 5.846,
      "step": 1010
    },
    {
      "epoch": 5.05,
      "grad_norm": 1.3242789506912231,
      "learning_rate": 2.7502750275027505e-05,
      "loss": 0.3563,
      "step": 1020
    },
    {
      "epoch": 5.1,
      "grad_norm": 1.8484835624694824,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.3295,
      "step": 1030
    },
    {
      "epoch": 5.15,
      "grad_norm": 1.2119253873825073,
      "learning_rate": 2.6952695269526958e-05,
      "loss": 0.3789,
      "step": 1040
    },
    {
      "epoch": 5.2,
      "grad_norm": 1.6589637994766235,
      "learning_rate": 2.667766776677668e-05,
      "loss": 0.3379,
      "step": 1050
    },
    {
      "epoch": 5.25,
      "grad_norm": 1.5714747905731201,
      "learning_rate": 2.64026402640264e-05,
      "loss": 0.3675,
      "step": 1060
    },
    {
      "epoch": 5.3,
      "grad_norm": 1.3527103662490845,
      "learning_rate": 2.6127612761276126e-05,
      "loss": 0.3573,
      "step": 1070
    },
    {
      "epoch": 5.35,
      "grad_norm": 1.055513620376587,
      "learning_rate": 2.5852585258525853e-05,
      "loss": 0.3478,
      "step": 1080
    },
    {
      "epoch": 5.4,
      "grad_norm": 1.6614227294921875,
      "learning_rate": 2.557755775577558e-05,
      "loss": 0.3692,
      "step": 1090
    },
    {
      "epoch": 5.45,
      "grad_norm": 1.5745677947998047,
      "learning_rate": 2.53025302530253e-05,
      "loss": 0.331,
      "step": 1100
    },
    {
      "epoch": 5.5,
      "grad_norm": 1.4897499084472656,
      "learning_rate": 2.5027502750275028e-05,
      "loss": 0.352,
      "step": 1110
    },
    {
      "epoch": 5.54,
      "grad_norm": 1.4070786237716675,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.345,
      "step": 1120
    },
    {
      "epoch": 5.59,
      "grad_norm": 2.2010486125946045,
      "learning_rate": 2.447744774477448e-05,
      "loss": 0.3213,
      "step": 1130
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.4076497554779053,
      "learning_rate": 2.4202420242024203e-05,
      "loss": 0.3414,
      "step": 1140
    },
    {
      "epoch": 5.69,
      "grad_norm": 1.168713092803955,
      "learning_rate": 2.392739273927393e-05,
      "loss": 0.3521,
      "step": 1150
    },
    {
      "epoch": 5.74,
      "grad_norm": 1.8131704330444336,
      "learning_rate": 2.3652365236523656e-05,
      "loss": 0.3576,
      "step": 1160
    },
    {
      "epoch": 5.79,
      "grad_norm": 1.2432703971862793,
      "learning_rate": 2.337733773377338e-05,
      "loss": 0.3416,
      "step": 1170
    },
    {
      "epoch": 5.84,
      "grad_norm": 2.248685598373413,
      "learning_rate": 2.31023102310231e-05,
      "loss": 0.3396,
      "step": 1180
    },
    {
      "epoch": 5.89,
      "grad_norm": 1.8308826684951782,
      "learning_rate": 2.2827282728272828e-05,
      "loss": 0.3466,
      "step": 1190
    },
    {
      "epoch": 5.94,
      "grad_norm": 2.5257883071899414,
      "learning_rate": 2.2552255225522554e-05,
      "loss": 0.3186,
      "step": 1200
    },
    {
      "epoch": 5.99,
      "grad_norm": 1.7196460962295532,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.3545,
      "step": 1210
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8659003831417624,
      "eval_loss": 0.31437239050865173,
      "eval_runtime": 15.7136,
      "eval_samples_per_second": 182.707,
      "eval_steps_per_second": 5.728,
      "step": 1212
    },
    {
      "epoch": 6.04,
      "grad_norm": 1.643741250038147,
      "learning_rate": 2.2002200220022003e-05,
      "loss": 0.3356,
      "step": 1220
    },
    {
      "epoch": 6.09,
      "grad_norm": 1.5010825395584106,
      "learning_rate": 2.172717271727173e-05,
      "loss": 0.3288,
      "step": 1230
    },
    {
      "epoch": 6.14,
      "grad_norm": 1.2441235780715942,
      "learning_rate": 2.1452145214521452e-05,
      "loss": 0.3395,
      "step": 1240
    },
    {
      "epoch": 6.19,
      "grad_norm": 1.351552128791809,
      "learning_rate": 2.1177117711771178e-05,
      "loss": 0.3011,
      "step": 1250
    },
    {
      "epoch": 6.24,
      "grad_norm": 1.3512260913848877,
      "learning_rate": 2.09020902090209e-05,
      "loss": 0.3416,
      "step": 1260
    },
    {
      "epoch": 6.29,
      "grad_norm": 1.5813019275665283,
      "learning_rate": 2.0627062706270627e-05,
      "loss": 0.3523,
      "step": 1270
    },
    {
      "epoch": 6.34,
      "grad_norm": 1.5249056816101074,
      "learning_rate": 2.0352035203520354e-05,
      "loss": 0.3288,
      "step": 1280
    },
    {
      "epoch": 6.39,
      "grad_norm": 1.9175376892089844,
      "learning_rate": 2.007700770077008e-05,
      "loss": 0.3527,
      "step": 1290
    },
    {
      "epoch": 6.44,
      "grad_norm": 1.5546538829803467,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.3076,
      "step": 1300
    },
    {
      "epoch": 6.49,
      "grad_norm": 1.5470608472824097,
      "learning_rate": 1.952695269526953e-05,
      "loss": 0.3303,
      "step": 1310
    },
    {
      "epoch": 6.53,
      "grad_norm": 1.3036069869995117,
      "learning_rate": 1.9251925192519252e-05,
      "loss": 0.3527,
      "step": 1320
    },
    {
      "epoch": 6.58,
      "grad_norm": 1.524294137954712,
      "learning_rate": 1.8976897689768978e-05,
      "loss": 0.3243,
      "step": 1330
    },
    {
      "epoch": 6.63,
      "grad_norm": 1.5366973876953125,
      "learning_rate": 1.87018701870187e-05,
      "loss": 0.3299,
      "step": 1340
    },
    {
      "epoch": 6.68,
      "grad_norm": 1.2275673151016235,
      "learning_rate": 1.8426842684268427e-05,
      "loss": 0.2928,
      "step": 1350
    },
    {
      "epoch": 6.73,
      "grad_norm": 1.4205721616744995,
      "learning_rate": 1.8151815181518153e-05,
      "loss": 0.3215,
      "step": 1360
    },
    {
      "epoch": 6.78,
      "grad_norm": 1.4984817504882812,
      "learning_rate": 1.787678767876788e-05,
      "loss": 0.3008,
      "step": 1370
    },
    {
      "epoch": 6.83,
      "grad_norm": 1.6792049407958984,
      "learning_rate": 1.7601760176017602e-05,
      "loss": 0.3403,
      "step": 1380
    },
    {
      "epoch": 6.88,
      "grad_norm": 1.5373748540878296,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.3353,
      "step": 1390
    },
    {
      "epoch": 6.93,
      "grad_norm": 2.501495361328125,
      "learning_rate": 1.705170517051705e-05,
      "loss": 0.3219,
      "step": 1400
    },
    {
      "epoch": 6.98,
      "grad_norm": 1.3856743574142456,
      "learning_rate": 1.6776677667766778e-05,
      "loss": 0.3137,
      "step": 1410
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.864158829676071,
      "eval_loss": 0.3307534158229828,
      "eval_runtime": 15.9606,
      "eval_samples_per_second": 179.88,
      "eval_steps_per_second": 5.639,
      "step": 1414
    },
    {
      "epoch": 7.03,
      "grad_norm": 1.5321968793869019,
      "learning_rate": 1.6501650165016504e-05,
      "loss": 0.308,
      "step": 1420
    },
    {
      "epoch": 7.08,
      "grad_norm": 1.4582158327102661,
      "learning_rate": 1.6226622662266227e-05,
      "loss": 0.3168,
      "step": 1430
    },
    {
      "epoch": 7.13,
      "grad_norm": 1.4425067901611328,
      "learning_rate": 1.5951595159515953e-05,
      "loss": 0.3213,
      "step": 1440
    },
    {
      "epoch": 7.18,
      "grad_norm": 1.933104395866394,
      "learning_rate": 1.567656765676568e-05,
      "loss": 0.2978,
      "step": 1450
    },
    {
      "epoch": 7.23,
      "grad_norm": 1.5549027919769287,
      "learning_rate": 1.5401540154015402e-05,
      "loss": 0.3281,
      "step": 1460
    },
    {
      "epoch": 7.28,
      "grad_norm": 1.4376338720321655,
      "learning_rate": 1.5126512651265127e-05,
      "loss": 0.2896,
      "step": 1470
    },
    {
      "epoch": 7.33,
      "grad_norm": 1.414781093597412,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.3135,
      "step": 1480
    },
    {
      "epoch": 7.38,
      "grad_norm": 1.3260300159454346,
      "learning_rate": 1.4576457645764577e-05,
      "loss": 0.2991,
      "step": 1490
    },
    {
      "epoch": 7.43,
      "grad_norm": 1.8632981777191162,
      "learning_rate": 1.4301430143014302e-05,
      "loss": 0.3135,
      "step": 1500
    },
    {
      "epoch": 7.48,
      "grad_norm": 1.6400821208953857,
      "learning_rate": 1.4026402640264028e-05,
      "loss": 0.2882,
      "step": 1510
    },
    {
      "epoch": 7.52,
      "grad_norm": 1.8777371644973755,
      "learning_rate": 1.3751375137513753e-05,
      "loss": 0.3539,
      "step": 1520
    },
    {
      "epoch": 7.57,
      "grad_norm": 1.431818962097168,
      "learning_rate": 1.3476347634763479e-05,
      "loss": 0.3065,
      "step": 1530
    },
    {
      "epoch": 7.62,
      "grad_norm": 2.002371311187744,
      "learning_rate": 1.32013201320132e-05,
      "loss": 0.3362,
      "step": 1540
    },
    {
      "epoch": 7.67,
      "grad_norm": 1.3172950744628906,
      "learning_rate": 1.2926292629262926e-05,
      "loss": 0.3062,
      "step": 1550
    },
    {
      "epoch": 7.72,
      "grad_norm": 2.155853271484375,
      "learning_rate": 1.265126512651265e-05,
      "loss": 0.3151,
      "step": 1560
    },
    {
      "epoch": 7.77,
      "grad_norm": 1.8984310626983643,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.3488,
      "step": 1570
    },
    {
      "epoch": 7.82,
      "grad_norm": 1.3027973175048828,
      "learning_rate": 1.2101210121012102e-05,
      "loss": 0.3197,
      "step": 1580
    },
    {
      "epoch": 7.87,
      "grad_norm": 1.9755516052246094,
      "learning_rate": 1.1826182618261828e-05,
      "loss": 0.3287,
      "step": 1590
    },
    {
      "epoch": 7.92,
      "grad_norm": 1.6631439924240112,
      "learning_rate": 1.155115511551155e-05,
      "loss": 0.303,
      "step": 1600
    },
    {
      "epoch": 7.97,
      "grad_norm": 1.8642737865447998,
      "learning_rate": 1.1276127612761277e-05,
      "loss": 0.3178,
      "step": 1610
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8645071403692093,
      "eval_loss": 0.32295528054237366,
      "eval_runtime": 15.3242,
      "eval_samples_per_second": 187.351,
      "eval_steps_per_second": 5.873,
      "step": 1616
    },
    {
      "epoch": 8.02,
      "grad_norm": 1.4315767288208008,
      "learning_rate": 1.1001100110011001e-05,
      "loss": 0.3078,
      "step": 1620
    },
    {
      "epoch": 8.07,
      "grad_norm": 1.6558310985565186,
      "learning_rate": 1.0726072607260726e-05,
      "loss": 0.2853,
      "step": 1630
    },
    {
      "epoch": 8.12,
      "grad_norm": 1.906076431274414,
      "learning_rate": 1.045104510451045e-05,
      "loss": 0.2814,
      "step": 1640
    },
    {
      "epoch": 8.17,
      "grad_norm": 2.086242914199829,
      "learning_rate": 1.0176017601760177e-05,
      "loss": 0.2847,
      "step": 1650
    },
    {
      "epoch": 8.22,
      "grad_norm": 1.793742299079895,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.2945,
      "step": 1660
    },
    {
      "epoch": 8.27,
      "grad_norm": 1.7942755222320557,
      "learning_rate": 9.625962596259626e-06,
      "loss": 0.296,
      "step": 1670
    },
    {
      "epoch": 8.32,
      "grad_norm": 1.7433489561080933,
      "learning_rate": 9.35093509350935e-06,
      "loss": 0.2885,
      "step": 1680
    },
    {
      "epoch": 8.37,
      "grad_norm": 1.6308104991912842,
      "learning_rate": 9.075907590759077e-06,
      "loss": 0.3134,
      "step": 1690
    },
    {
      "epoch": 8.42,
      "grad_norm": 1.6005682945251465,
      "learning_rate": 8.800880088008801e-06,
      "loss": 0.3206,
      "step": 1700
    },
    {
      "epoch": 8.47,
      "grad_norm": 1.563693642616272,
      "learning_rate": 8.525852585258526e-06,
      "loss": 0.3112,
      "step": 1710
    },
    {
      "epoch": 8.51,
      "grad_norm": 1.796925663948059,
      "learning_rate": 8.250825082508252e-06,
      "loss": 0.3076,
      "step": 1720
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.5874933004379272,
      "learning_rate": 7.975797579757976e-06,
      "loss": 0.3197,
      "step": 1730
    },
    {
      "epoch": 8.61,
      "grad_norm": 1.5310964584350586,
      "learning_rate": 7.700770077007701e-06,
      "loss": 0.3026,
      "step": 1740
    },
    {
      "epoch": 8.66,
      "grad_norm": 1.9441354274749756,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.3051,
      "step": 1750
    },
    {
      "epoch": 8.71,
      "grad_norm": 1.8038005828857422,
      "learning_rate": 7.150715071507151e-06,
      "loss": 0.3107,
      "step": 1760
    },
    {
      "epoch": 8.76,
      "grad_norm": 4.733364582061768,
      "learning_rate": 6.875687568756876e-06,
      "loss": 0.2939,
      "step": 1770
    },
    {
      "epoch": 8.81,
      "grad_norm": 1.5854134559631348,
      "learning_rate": 6.6006600660066e-06,
      "loss": 0.3049,
      "step": 1780
    },
    {
      "epoch": 8.86,
      "grad_norm": 1.7545024156570435,
      "learning_rate": 6.325632563256325e-06,
      "loss": 0.2936,
      "step": 1790
    },
    {
      "epoch": 8.91,
      "grad_norm": 1.6307064294815063,
      "learning_rate": 6.050605060506051e-06,
      "loss": 0.2855,
      "step": 1800
    },
    {
      "epoch": 8.96,
      "grad_norm": 1.6208785772323608,
      "learning_rate": 5.775577557755775e-06,
      "loss": 0.2998,
      "step": 1810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8707767328456983,
      "eval_loss": 0.32057538628578186,
      "eval_runtime": 15.4893,
      "eval_samples_per_second": 185.354,
      "eval_steps_per_second": 5.81,
      "step": 1818
    },
    {
      "epoch": 9.01,
      "grad_norm": 1.7324903011322021,
      "learning_rate": 5.500550055005501e-06,
      "loss": 0.2916,
      "step": 1820
    },
    {
      "epoch": 9.06,
      "grad_norm": 1.4792495965957642,
      "learning_rate": 5.225522552255225e-06,
      "loss": 0.2771,
      "step": 1830
    },
    {
      "epoch": 9.11,
      "grad_norm": 2.207402229309082,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.2859,
      "step": 1840
    },
    {
      "epoch": 9.16,
      "grad_norm": 1.9077599048614502,
      "learning_rate": 4.675467546754675e-06,
      "loss": 0.2833,
      "step": 1850
    },
    {
      "epoch": 9.21,
      "grad_norm": 1.7705243825912476,
      "learning_rate": 4.400440044004401e-06,
      "loss": 0.3007,
      "step": 1860
    },
    {
      "epoch": 9.26,
      "grad_norm": 1.74075448513031,
      "learning_rate": 4.125412541254126e-06,
      "loss": 0.2912,
      "step": 1870
    },
    {
      "epoch": 9.31,
      "grad_norm": 1.8058630228042603,
      "learning_rate": 3.8503850385038505e-06,
      "loss": 0.297,
      "step": 1880
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.7161786556243896,
      "learning_rate": 3.5753575357535755e-06,
      "loss": 0.2789,
      "step": 1890
    },
    {
      "epoch": 9.41,
      "grad_norm": 2.8628289699554443,
      "learning_rate": 3.3003300330033e-06,
      "loss": 0.3018,
      "step": 1900
    },
    {
      "epoch": 9.46,
      "grad_norm": 1.9037891626358032,
      "learning_rate": 3.0253025302530254e-06,
      "loss": 0.3043,
      "step": 1910
    },
    {
      "epoch": 9.5,
      "grad_norm": 1.5443955659866333,
      "learning_rate": 2.7502750275027504e-06,
      "loss": 0.2813,
      "step": 1920
    },
    {
      "epoch": 9.55,
      "grad_norm": 1.6395975351333618,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.2942,
      "step": 1930
    },
    {
      "epoch": 9.6,
      "grad_norm": 1.4861618280410767,
      "learning_rate": 2.2002200220022003e-06,
      "loss": 0.2769,
      "step": 1940
    },
    {
      "epoch": 9.65,
      "grad_norm": 1.9940297603607178,
      "learning_rate": 1.9251925192519253e-06,
      "loss": 0.2833,
      "step": 1950
    },
    {
      "epoch": 9.7,
      "grad_norm": 1.800776720046997,
      "learning_rate": 1.65016501650165e-06,
      "loss": 0.2948,
      "step": 1960
    },
    {
      "epoch": 9.75,
      "grad_norm": 1.4160261154174805,
      "learning_rate": 1.3751375137513752e-06,
      "loss": 0.2932,
      "step": 1970
    },
    {
      "epoch": 9.8,
      "grad_norm": 1.8922903537750244,
      "learning_rate": 1.1001100110011001e-06,
      "loss": 0.3189,
      "step": 1980
    },
    {
      "epoch": 9.85,
      "grad_norm": 1.754544734954834,
      "learning_rate": 8.25082508250825e-07,
      "loss": 0.2907,
      "step": 1990
    },
    {
      "epoch": 9.9,
      "grad_norm": 1.8808295726776123,
      "learning_rate": 5.500550055005501e-07,
      "loss": 0.2763,
      "step": 2000
    },
    {
      "epoch": 9.95,
      "grad_norm": 2.0124707221984863,
      "learning_rate": 2.7502750275027504e-07,
      "loss": 0.2611,
      "step": 2010
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.8302078247070312,
      "learning_rate": 0.0,
      "loss": 0.2773,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8732149076976663,
      "eval_loss": 0.3263641893863678,
      "eval_runtime": 15.9034,
      "eval_samples_per_second": 180.527,
      "eval_steps_per_second": 5.659,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "step": 2020,
      "total_flos": 2.0021605356722135e+19,
      "train_loss": 0.37925267927717454,
      "train_runtime": 3266.1275,
      "train_samples_per_second": 79.106,
      "train_steps_per_second": 0.618
    }
  ],
  "logging_steps": 10,
  "max_steps": 2020,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.0021605356722135e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}