{
  "best_metric": 0.842911877394636,
  "best_model_checkpoint": "vit-tiny-patch16-224-winkawaks/checkpoint-1818",
  "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": 11.655478477478027,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.8812,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 6.654660224914551,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.7531,
      "step": 20
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.010664939880371,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.7312,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 3.320401191711426,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.6753,
      "step": 40
    },
    {
      "epoch": 0.25,
      "grad_norm": 4.002676010131836,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.6656,
      "step": 50
    },
    {
      "epoch": 0.3,
      "grad_norm": 5.1362690925598145,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.636,
      "step": 60
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.089080333709717,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.6258,
      "step": 70
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.782809257507324,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.6074,
      "step": 80
    },
    {
      "epoch": 0.45,
      "grad_norm": 4.684001922607422,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.6114,
      "step": 90
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.801894187927246,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.5888,
      "step": 100
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.810859680175781,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.5858,
      "step": 110
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.189158916473389,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.5849,
      "step": 120
    },
    {
      "epoch": 0.64,
      "grad_norm": 6.533910274505615,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.5811,
      "step": 130
    },
    {
      "epoch": 0.69,
      "grad_norm": 15.806050300598145,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.6041,
      "step": 140
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.087372064590454,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.5402,
      "step": 150
    },
    {
      "epoch": 0.79,
      "grad_norm": 6.852709770202637,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.5435,
      "step": 160
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.9751667976379395,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.5826,
      "step": 170
    },
    {
      "epoch": 0.89,
      "grad_norm": 8.571554183959961,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.5399,
      "step": 180
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.956425905227661,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.5627,
      "step": 190
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.44449520111084,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.5345,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7770811563915012,
      "eval_loss": 0.4689331650733948,
      "eval_runtime": 5.627,
      "eval_samples_per_second": 510.22,
      "eval_steps_per_second": 15.994,
      "step": 202
    },
    {
      "epoch": 1.04,
      "grad_norm": 5.914970397949219,
      "learning_rate": 4.977997799779978e-05,
      "loss": 0.533,
      "step": 210
    },
    {
      "epoch": 1.09,
      "grad_norm": 7.073808193206787,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.517,
      "step": 220
    },
    {
      "epoch": 1.14,
      "grad_norm": 3.3154642581939697,
      "learning_rate": 4.9229922992299234e-05,
      "loss": 0.5385,
      "step": 230
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.500061273574829,
      "learning_rate": 4.895489548954896e-05,
      "loss": 0.5345,
      "step": 240
    },
    {
      "epoch": 1.24,
      "grad_norm": 8.62484359741211,
      "learning_rate": 4.867986798679868e-05,
      "loss": 0.5368,
      "step": 250
    },
    {
      "epoch": 1.29,
      "grad_norm": 9.027387619018555,
      "learning_rate": 4.8404840484048406e-05,
      "loss": 0.5247,
      "step": 260
    },
    {
      "epoch": 1.34,
      "grad_norm": 7.457286357879639,
      "learning_rate": 4.812981298129813e-05,
      "loss": 0.5044,
      "step": 270
    },
    {
      "epoch": 1.39,
      "grad_norm": 7.2032790184021,
      "learning_rate": 4.785478547854786e-05,
      "loss": 0.5156,
      "step": 280
    },
    {
      "epoch": 1.44,
      "grad_norm": 5.909176826477051,
      "learning_rate": 4.7579757975797585e-05,
      "loss": 0.5191,
      "step": 290
    },
    {
      "epoch": 1.49,
      "grad_norm": 4.56859827041626,
      "learning_rate": 4.730473047304731e-05,
      "loss": 0.4891,
      "step": 300
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.181271553039551,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.5016,
      "step": 310
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.948479175567627,
      "learning_rate": 4.675467546754676e-05,
      "loss": 0.5008,
      "step": 320
    },
    {
      "epoch": 1.63,
      "grad_norm": 7.155640125274658,
      "learning_rate": 4.647964796479648e-05,
      "loss": 0.5451,
      "step": 330
    },
    {
      "epoch": 1.68,
      "grad_norm": 3.5051815509796143,
      "learning_rate": 4.62046204620462e-05,
      "loss": 0.4998,
      "step": 340
    },
    {
      "epoch": 1.73,
      "grad_norm": 12.015359878540039,
      "learning_rate": 4.592959295929593e-05,
      "loss": 0.5302,
      "step": 350
    },
    {
      "epoch": 1.78,
      "grad_norm": 5.719058990478516,
      "learning_rate": 4.5654565456545655e-05,
      "loss": 0.5052,
      "step": 360
    },
    {
      "epoch": 1.83,
      "grad_norm": 6.292553424835205,
      "learning_rate": 4.537953795379538e-05,
      "loss": 0.5192,
      "step": 370
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.8755393028259277,
      "learning_rate": 4.510451045104511e-05,
      "loss": 0.4966,
      "step": 380
    },
    {
      "epoch": 1.93,
      "grad_norm": 3.096843719482422,
      "learning_rate": 4.4829482948294834e-05,
      "loss": 0.5032,
      "step": 390
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.9766151905059814,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4936,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7485196795541623,
      "eval_loss": 0.5022404789924622,
      "eval_runtime": 5.6861,
      "eval_samples_per_second": 504.916,
      "eval_steps_per_second": 15.828,
      "step": 404
    },
    {
      "epoch": 2.03,
      "grad_norm": 7.3236870765686035,
      "learning_rate": 4.427942794279428e-05,
      "loss": 0.4828,
      "step": 410
    },
    {
      "epoch": 2.08,
      "grad_norm": 4.203598976135254,
      "learning_rate": 4.4004400440044006e-05,
      "loss": 0.5108,
      "step": 420
    },
    {
      "epoch": 2.13,
      "grad_norm": 4.851642608642578,
      "learning_rate": 4.372937293729373e-05,
      "loss": 0.474,
      "step": 430
    },
    {
      "epoch": 2.18,
      "grad_norm": 8.77710247039795,
      "learning_rate": 4.345434543454346e-05,
      "loss": 0.4902,
      "step": 440
    },
    {
      "epoch": 2.23,
      "grad_norm": 6.436154365539551,
      "learning_rate": 4.3179317931793185e-05,
      "loss": 0.4838,
      "step": 450
    },
    {
      "epoch": 2.28,
      "grad_norm": 4.809104919433594,
      "learning_rate": 4.2904290429042904e-05,
      "loss": 0.5064,
      "step": 460
    },
    {
      "epoch": 2.33,
      "grad_norm": 4.125964641571045,
      "learning_rate": 4.262926292629263e-05,
      "loss": 0.4819,
      "step": 470
    },
    {
      "epoch": 2.38,
      "grad_norm": 6.8852763175964355,
      "learning_rate": 4.2354235423542356e-05,
      "loss": 0.5011,
      "step": 480
    },
    {
      "epoch": 2.43,
      "grad_norm": 3.3697924613952637,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4676,
      "step": 490
    },
    {
      "epoch": 2.48,
      "grad_norm": 5.354856967926025,
      "learning_rate": 4.18041804180418e-05,
      "loss": 0.507,
      "step": 500
    },
    {
      "epoch": 2.52,
      "grad_norm": 7.638093948364258,
      "learning_rate": 4.152915291529153e-05,
      "loss": 0.5046,
      "step": 510
    },
    {
      "epoch": 2.57,
      "grad_norm": 4.444268226623535,
      "learning_rate": 4.1254125412541255e-05,
      "loss": 0.4635,
      "step": 520
    },
    {
      "epoch": 2.62,
      "grad_norm": 8.037334442138672,
      "learning_rate": 4.097909790979098e-05,
      "loss": 0.4796,
      "step": 530
    },
    {
      "epoch": 2.67,
      "grad_norm": 6.52307653427124,
      "learning_rate": 4.070407040704071e-05,
      "loss": 0.4668,
      "step": 540
    },
    {
      "epoch": 2.72,
      "grad_norm": 7.490472316741943,
      "learning_rate": 4.042904290429043e-05,
      "loss": 0.4747,
      "step": 550
    },
    {
      "epoch": 2.77,
      "grad_norm": 3.393075942993164,
      "learning_rate": 4.015401540154016e-05,
      "loss": 0.4667,
      "step": 560
    },
    {
      "epoch": 2.82,
      "grad_norm": 4.454715251922607,
      "learning_rate": 3.987898789878988e-05,
      "loss": 0.4795,
      "step": 570
    },
    {
      "epoch": 2.87,
      "grad_norm": 5.316873550415039,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.499,
      "step": 580
    },
    {
      "epoch": 2.92,
      "grad_norm": 2.8893849849700928,
      "learning_rate": 3.932893289328933e-05,
      "loss": 0.4913,
      "step": 590
    },
    {
      "epoch": 2.97,
      "grad_norm": 2.6538729667663574,
      "learning_rate": 3.905390539053906e-05,
      "loss": 0.4911,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.82793451758969,
      "eval_loss": 0.3886525630950928,
      "eval_runtime": 5.5731,
      "eval_samples_per_second": 515.151,
      "eval_steps_per_second": 16.149,
      "step": 606
    },
    {
      "epoch": 3.02,
      "grad_norm": 9.823522567749023,
      "learning_rate": 3.877887788778878e-05,
      "loss": 0.4731,
      "step": 610
    },
    {
      "epoch": 3.07,
      "grad_norm": 5.899537086486816,
      "learning_rate": 3.8503850385038503e-05,
      "loss": 0.4621,
      "step": 620
    },
    {
      "epoch": 3.12,
      "grad_norm": 4.964333534240723,
      "learning_rate": 3.822882288228823e-05,
      "loss": 0.4762,
      "step": 630
    },
    {
      "epoch": 3.17,
      "grad_norm": 4.469970703125,
      "learning_rate": 3.7953795379537956e-05,
      "loss": 0.4774,
      "step": 640
    },
    {
      "epoch": 3.22,
      "grad_norm": 3.393183469772339,
      "learning_rate": 3.767876787678768e-05,
      "loss": 0.4501,
      "step": 650
    },
    {
      "epoch": 3.27,
      "grad_norm": 11.101873397827148,
      "learning_rate": 3.74037403740374e-05,
      "loss": 0.4596,
      "step": 660
    },
    {
      "epoch": 3.32,
      "grad_norm": 4.9833831787109375,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.4516,
      "step": 670
    },
    {
      "epoch": 3.37,
      "grad_norm": 5.423743724822998,
      "learning_rate": 3.6853685368536854e-05,
      "loss": 0.4887,
      "step": 680
    },
    {
      "epoch": 3.42,
      "grad_norm": 4.94615364074707,
      "learning_rate": 3.657865786578658e-05,
      "loss": 0.4538,
      "step": 690
    },
    {
      "epoch": 3.47,
      "grad_norm": 5.113562107086182,
      "learning_rate": 3.6303630363036307e-05,
      "loss": 0.4645,
      "step": 700
    },
    {
      "epoch": 3.51,
      "grad_norm": 7.3234357833862305,
      "learning_rate": 3.602860286028603e-05,
      "loss": 0.4645,
      "step": 710
    },
    {
      "epoch": 3.56,
      "grad_norm": 4.696035385131836,
      "learning_rate": 3.575357535753576e-05,
      "loss": 0.4584,
      "step": 720
    },
    {
      "epoch": 3.61,
      "grad_norm": 3.724952459335327,
      "learning_rate": 3.5478547854785485e-05,
      "loss": 0.427,
      "step": 730
    },
    {
      "epoch": 3.66,
      "grad_norm": 3.8439409732818604,
      "learning_rate": 3.5203520352035205e-05,
      "loss": 0.4554,
      "step": 740
    },
    {
      "epoch": 3.71,
      "grad_norm": 3.4024925231933594,
      "learning_rate": 3.492849284928493e-05,
      "loss": 0.4413,
      "step": 750
    },
    {
      "epoch": 3.76,
      "grad_norm": 8.046459197998047,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.4638,
      "step": 760
    },
    {
      "epoch": 3.81,
      "grad_norm": 5.1486358642578125,
      "learning_rate": 3.4378437843784377e-05,
      "loss": 0.4852,
      "step": 770
    },
    {
      "epoch": 3.86,
      "grad_norm": 3.408259868621826,
      "learning_rate": 3.41034103410341e-05,
      "loss": 0.4347,
      "step": 780
    },
    {
      "epoch": 3.91,
      "grad_norm": 5.650546073913574,
      "learning_rate": 3.382838283828383e-05,
      "loss": 0.4294,
      "step": 790
    },
    {
      "epoch": 3.96,
      "grad_norm": 3.8805394172668457,
      "learning_rate": 3.3553355335533555e-05,
      "loss": 0.4191,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8098223615464994,
      "eval_loss": 0.4121237099170685,
      "eval_runtime": 5.4396,
      "eval_samples_per_second": 527.797,
      "eval_steps_per_second": 16.545,
      "step": 808
    },
    {
      "epoch": 4.01,
      "grad_norm": 4.736340522766113,
      "learning_rate": 3.327832783278328e-05,
      "loss": 0.4824,
      "step": 810
    },
    {
      "epoch": 4.06,
      "grad_norm": 4.468814373016357,
      "learning_rate": 3.300330033003301e-05,
      "loss": 0.468,
      "step": 820
    },
    {
      "epoch": 4.11,
      "grad_norm": 3.311174154281616,
      "learning_rate": 3.272827282728273e-05,
      "loss": 0.442,
      "step": 830
    },
    {
      "epoch": 4.16,
      "grad_norm": 6.016156196594238,
      "learning_rate": 3.2453245324532453e-05,
      "loss": 0.4748,
      "step": 840
    },
    {
      "epoch": 4.21,
      "grad_norm": 6.195199966430664,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.4488,
      "step": 850
    },
    {
      "epoch": 4.26,
      "grad_norm": 3.6519343852996826,
      "learning_rate": 3.1903190319031906e-05,
      "loss": 0.4285,
      "step": 860
    },
    {
      "epoch": 4.31,
      "grad_norm": 5.461367130279541,
      "learning_rate": 3.162816281628163e-05,
      "loss": 0.4124,
      "step": 870
    },
    {
      "epoch": 4.36,
      "grad_norm": 3.7743630409240723,
      "learning_rate": 3.135313531353136e-05,
      "loss": 0.4361,
      "step": 880
    },
    {
      "epoch": 4.41,
      "grad_norm": 3.741748571395874,
      "learning_rate": 3.1078107810781085e-05,
      "loss": 0.4196,
      "step": 890
    },
    {
      "epoch": 4.46,
      "grad_norm": 5.458765983581543,
      "learning_rate": 3.0803080308030804e-05,
      "loss": 0.4814,
      "step": 900
    },
    {
      "epoch": 4.5,
      "grad_norm": 5.118380546569824,
      "learning_rate": 3.052805280528053e-05,
      "loss": 0.4441,
      "step": 910
    },
    {
      "epoch": 4.55,
      "grad_norm": 4.428480625152588,
      "learning_rate": 3.0253025302530253e-05,
      "loss": 0.4217,
      "step": 920
    },
    {
      "epoch": 4.6,
      "grad_norm": 3.2278711795806885,
      "learning_rate": 2.9977997799779976e-05,
      "loss": 0.4259,
      "step": 930
    },
    {
      "epoch": 4.65,
      "grad_norm": 3.5108141899108887,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.4507,
      "step": 940
    },
    {
      "epoch": 4.7,
      "grad_norm": 3.02825665473938,
      "learning_rate": 2.942794279427943e-05,
      "loss": 0.4095,
      "step": 950
    },
    {
      "epoch": 4.75,
      "grad_norm": 3.8650879859924316,
      "learning_rate": 2.9152915291529155e-05,
      "loss": 0.4251,
      "step": 960
    },
    {
      "epoch": 4.8,
      "grad_norm": 5.335313320159912,
      "learning_rate": 2.8877887788778878e-05,
      "loss": 0.428,
      "step": 970
    },
    {
      "epoch": 4.85,
      "grad_norm": 5.564798831939697,
      "learning_rate": 2.8602860286028604e-05,
      "loss": 0.4713,
      "step": 980
    },
    {
      "epoch": 4.9,
      "grad_norm": 4.036501407623291,
      "learning_rate": 2.832783278327833e-05,
      "loss": 0.4152,
      "step": 990
    },
    {
      "epoch": 4.95,
      "grad_norm": 4.295027732849121,
      "learning_rate": 2.8052805280528056e-05,
      "loss": 0.4209,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.8911187648773193,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.4408,
      "step": 1010
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8254963427377221,
      "eval_loss": 0.38967224955558777,
      "eval_runtime": 5.6692,
      "eval_samples_per_second": 506.425,
      "eval_steps_per_second": 15.875,
      "step": 1010
    },
    {
      "epoch": 5.05,
      "grad_norm": 7.270760536193848,
      "learning_rate": 2.7502750275027505e-05,
      "loss": 0.4415,
      "step": 1020
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.75962495803833,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.4282,
      "step": 1030
    },
    {
      "epoch": 5.15,
      "grad_norm": 3.6654067039489746,
      "learning_rate": 2.6952695269526958e-05,
      "loss": 0.4265,
      "step": 1040
    },
    {
      "epoch": 5.2,
      "grad_norm": 5.461548805236816,
      "learning_rate": 2.667766776677668e-05,
      "loss": 0.4188,
      "step": 1050
    },
    {
      "epoch": 5.25,
      "grad_norm": 3.629088878631592,
      "learning_rate": 2.64026402640264e-05,
      "loss": 0.438,
      "step": 1060
    },
    {
      "epoch": 5.3,
      "grad_norm": 5.7036614418029785,
      "learning_rate": 2.6127612761276126e-05,
      "loss": 0.4073,
      "step": 1070
    },
    {
      "epoch": 5.35,
      "grad_norm": 7.140379905700684,
      "learning_rate": 2.5852585258525853e-05,
      "loss": 0.4287,
      "step": 1080
    },
    {
      "epoch": 5.4,
      "grad_norm": 3.2449722290039062,
      "learning_rate": 2.557755775577558e-05,
      "loss": 0.425,
      "step": 1090
    },
    {
      "epoch": 5.45,
      "grad_norm": 4.488211631774902,
      "learning_rate": 2.53025302530253e-05,
      "loss": 0.4203,
      "step": 1100
    },
    {
      "epoch": 5.5,
      "grad_norm": 3.7270421981811523,
      "learning_rate": 2.5027502750275028e-05,
      "loss": 0.4174,
      "step": 1110
    },
    {
      "epoch": 5.54,
      "grad_norm": 4.254336357116699,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.4533,
      "step": 1120
    },
    {
      "epoch": 5.59,
      "grad_norm": 3.1208856105804443,
      "learning_rate": 2.447744774477448e-05,
      "loss": 0.4221,
      "step": 1130
    },
    {
      "epoch": 5.64,
      "grad_norm": 6.634130001068115,
      "learning_rate": 2.4202420242024203e-05,
      "loss": 0.4377,
      "step": 1140
    },
    {
      "epoch": 5.69,
      "grad_norm": 5.457045078277588,
      "learning_rate": 2.392739273927393e-05,
      "loss": 0.4459,
      "step": 1150
    },
    {
      "epoch": 5.74,
      "grad_norm": 3.537189483642578,
      "learning_rate": 2.3652365236523656e-05,
      "loss": 0.4423,
      "step": 1160
    },
    {
      "epoch": 5.79,
      "grad_norm": 5.883282661437988,
      "learning_rate": 2.337733773377338e-05,
      "loss": 0.4121,
      "step": 1170
    },
    {
      "epoch": 5.84,
      "grad_norm": 4.0562744140625,
      "learning_rate": 2.31023102310231e-05,
      "loss": 0.4114,
      "step": 1180
    },
    {
      "epoch": 5.89,
      "grad_norm": 5.0152974128723145,
      "learning_rate": 2.2827282728272828e-05,
      "loss": 0.4287,
      "step": 1190
    },
    {
      "epoch": 5.94,
      "grad_norm": 4.672534465789795,
      "learning_rate": 2.2552255225522554e-05,
      "loss": 0.4124,
      "step": 1200
    },
    {
      "epoch": 5.99,
      "grad_norm": 5.874300479888916,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.4134,
      "step": 1210
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8331591779867642,
      "eval_loss": 0.3714359402656555,
      "eval_runtime": 5.6334,
      "eval_samples_per_second": 509.636,
      "eval_steps_per_second": 15.976,
      "step": 1212
    },
    {
      "epoch": 6.04,
      "grad_norm": 3.3733067512512207,
      "learning_rate": 2.2002200220022003e-05,
      "loss": 0.404,
      "step": 1220
    },
    {
      "epoch": 6.09,
      "grad_norm": 3.479614734649658,
      "learning_rate": 2.172717271727173e-05,
      "loss": 0.4102,
      "step": 1230
    },
    {
      "epoch": 6.14,
      "grad_norm": 6.636502742767334,
      "learning_rate": 2.1452145214521452e-05,
      "loss": 0.4261,
      "step": 1240
    },
    {
      "epoch": 6.19,
      "grad_norm": 3.6467645168304443,
      "learning_rate": 2.1177117711771178e-05,
      "loss": 0.4061,
      "step": 1250
    },
    {
      "epoch": 6.24,
      "grad_norm": 3.435811996459961,
      "learning_rate": 2.09020902090209e-05,
      "loss": 0.4198,
      "step": 1260
    },
    {
      "epoch": 6.29,
      "grad_norm": 5.085671424865723,
      "learning_rate": 2.0627062706270627e-05,
      "loss": 0.3963,
      "step": 1270
    },
    {
      "epoch": 6.34,
      "grad_norm": 9.153255462646484,
      "learning_rate": 2.0352035203520354e-05,
      "loss": 0.4239,
      "step": 1280
    },
    {
      "epoch": 6.39,
      "grad_norm": 5.1853251457214355,
      "learning_rate": 2.007700770077008e-05,
      "loss": 0.3908,
      "step": 1290
    },
    {
      "epoch": 6.44,
      "grad_norm": 7.626086235046387,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.4132,
      "step": 1300
    },
    {
      "epoch": 6.49,
      "grad_norm": 5.670557022094727,
      "learning_rate": 1.952695269526953e-05,
      "loss": 0.4228,
      "step": 1310
    },
    {
      "epoch": 6.53,
      "grad_norm": 5.062951564788818,
      "learning_rate": 1.9251925192519252e-05,
      "loss": 0.4385,
      "step": 1320
    },
    {
      "epoch": 6.58,
      "grad_norm": 3.8000316619873047,
      "learning_rate": 1.8976897689768978e-05,
      "loss": 0.414,
      "step": 1330
    },
    {
      "epoch": 6.63,
      "grad_norm": 3.444866895675659,
      "learning_rate": 1.87018701870187e-05,
      "loss": 0.4246,
      "step": 1340
    },
    {
      "epoch": 6.68,
      "grad_norm": 3.809499740600586,
      "learning_rate": 1.8426842684268427e-05,
      "loss": 0.4198,
      "step": 1350
    },
    {
      "epoch": 6.73,
      "grad_norm": 4.515871047973633,
      "learning_rate": 1.8151815181518153e-05,
      "loss": 0.4051,
      "step": 1360
    },
    {
      "epoch": 6.78,
      "grad_norm": 4.38671350479126,
      "learning_rate": 1.787678767876788e-05,
      "loss": 0.4153,
      "step": 1370
    },
    {
      "epoch": 6.83,
      "grad_norm": 3.238205909729004,
      "learning_rate": 1.7601760176017602e-05,
      "loss": 0.3947,
      "step": 1380
    },
    {
      "epoch": 6.88,
      "grad_norm": 3.455275058746338,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.4018,
      "step": 1390
    },
    {
      "epoch": 6.93,
      "grad_norm": 3.587259531021118,
      "learning_rate": 1.705170517051705e-05,
      "loss": 0.3983,
      "step": 1400
    },
    {
      "epoch": 6.98,
      "grad_norm": 4.972535610198975,
      "learning_rate": 1.6776677667766778e-05,
      "loss": 0.4117,
      "step": 1410
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8376872169975619,
      "eval_loss": 0.36851075291633606,
      "eval_runtime": 5.6001,
      "eval_samples_per_second": 512.666,
      "eval_steps_per_second": 16.071,
      "step": 1414
    },
    {
      "epoch": 7.03,
      "grad_norm": 4.680665493011475,
      "learning_rate": 1.6501650165016504e-05,
      "loss": 0.3791,
      "step": 1420
    },
    {
      "epoch": 7.08,
      "grad_norm": 4.995221138000488,
      "learning_rate": 1.6226622662266227e-05,
      "loss": 0.4119,
      "step": 1430
    },
    {
      "epoch": 7.13,
      "grad_norm": 3.969187021255493,
      "learning_rate": 1.5951595159515953e-05,
      "loss": 0.3864,
      "step": 1440
    },
    {
      "epoch": 7.18,
      "grad_norm": 3.4930386543273926,
      "learning_rate": 1.567656765676568e-05,
      "loss": 0.3914,
      "step": 1450
    },
    {
      "epoch": 7.23,
      "grad_norm": 4.228696823120117,
      "learning_rate": 1.5401540154015402e-05,
      "loss": 0.4072,
      "step": 1460
    },
    {
      "epoch": 7.28,
      "grad_norm": 3.5527050495147705,
      "learning_rate": 1.5126512651265127e-05,
      "loss": 0.3864,
      "step": 1470
    },
    {
      "epoch": 7.33,
      "grad_norm": 3.4274041652679443,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.3957,
      "step": 1480
    },
    {
      "epoch": 7.38,
      "grad_norm": 4.775713920593262,
      "learning_rate": 1.4576457645764577e-05,
      "loss": 0.4034,
      "step": 1490
    },
    {
      "epoch": 7.43,
      "grad_norm": 4.741795539855957,
      "learning_rate": 1.4301430143014302e-05,
      "loss": 0.4042,
      "step": 1500
    },
    {
      "epoch": 7.48,
      "grad_norm": 6.42737340927124,
      "learning_rate": 1.4026402640264028e-05,
      "loss": 0.4197,
      "step": 1510
    },
    {
      "epoch": 7.52,
      "grad_norm": 5.1894121170043945,
      "learning_rate": 1.3751375137513753e-05,
      "loss": 0.3956,
      "step": 1520
    },
    {
      "epoch": 7.57,
      "grad_norm": 5.648491859436035,
      "learning_rate": 1.3476347634763479e-05,
      "loss": 0.4005,
      "step": 1530
    },
    {
      "epoch": 7.62,
      "grad_norm": 4.315338611602783,
      "learning_rate": 1.32013201320132e-05,
      "loss": 0.3784,
      "step": 1540
    },
    {
      "epoch": 7.67,
      "grad_norm": 4.007127285003662,
      "learning_rate": 1.2926292629262926e-05,
      "loss": 0.3948,
      "step": 1550
    },
    {
      "epoch": 7.72,
      "grad_norm": 4.685734748840332,
      "learning_rate": 1.265126512651265e-05,
      "loss": 0.4165,
      "step": 1560
    },
    {
      "epoch": 7.77,
      "grad_norm": 5.51500940322876,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.3872,
      "step": 1570
    },
    {
      "epoch": 7.82,
      "grad_norm": 3.208813428878784,
      "learning_rate": 1.2101210121012102e-05,
      "loss": 0.3886,
      "step": 1580
    },
    {
      "epoch": 7.87,
      "grad_norm": 4.27180814743042,
      "learning_rate": 1.1826182618261828e-05,
      "loss": 0.3837,
      "step": 1590
    },
    {
      "epoch": 7.92,
      "grad_norm": 9.008807182312012,
      "learning_rate": 1.155115511551155e-05,
      "loss": 0.4174,
      "step": 1600
    },
    {
      "epoch": 7.97,
      "grad_norm": 3.7245564460754395,
      "learning_rate": 1.1276127612761277e-05,
      "loss": 0.3991,
      "step": 1610
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8411703239289446,
      "eval_loss": 0.3601776659488678,
      "eval_runtime": 5.6002,
      "eval_samples_per_second": 512.663,
      "eval_steps_per_second": 16.071,
      "step": 1616
    },
    {
      "epoch": 8.02,
      "grad_norm": 4.604074954986572,
      "learning_rate": 1.1001100110011001e-05,
      "loss": 0.4121,
      "step": 1620
    },
    {
      "epoch": 8.07,
      "grad_norm": 6.1971611976623535,
      "learning_rate": 1.0726072607260726e-05,
      "loss": 0.3925,
      "step": 1630
    },
    {
      "epoch": 8.12,
      "grad_norm": 3.589857816696167,
      "learning_rate": 1.045104510451045e-05,
      "loss": 0.4339,
      "step": 1640
    },
    {
      "epoch": 8.17,
      "grad_norm": 3.007169246673584,
      "learning_rate": 1.0176017601760177e-05,
      "loss": 0.3841,
      "step": 1650
    },
    {
      "epoch": 8.22,
      "grad_norm": 8.007019996643066,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.4005,
      "step": 1660
    },
    {
      "epoch": 8.27,
      "grad_norm": 4.514450550079346,
      "learning_rate": 9.625962596259626e-06,
      "loss": 0.4072,
      "step": 1670
    },
    {
      "epoch": 8.32,
      "grad_norm": 6.048130989074707,
      "learning_rate": 9.35093509350935e-06,
      "loss": 0.3756,
      "step": 1680
    },
    {
      "epoch": 8.37,
      "grad_norm": 6.211109638214111,
      "learning_rate": 9.075907590759077e-06,
      "loss": 0.3953,
      "step": 1690
    },
    {
      "epoch": 8.42,
      "grad_norm": 3.7817318439483643,
      "learning_rate": 8.800880088008801e-06,
      "loss": 0.3685,
      "step": 1700
    },
    {
      "epoch": 8.47,
      "grad_norm": 5.379549980163574,
      "learning_rate": 8.525852585258526e-06,
      "loss": 0.3713,
      "step": 1710
    },
    {
      "epoch": 8.51,
      "grad_norm": 4.123336315155029,
      "learning_rate": 8.250825082508252e-06,
      "loss": 0.3629,
      "step": 1720
    },
    {
      "epoch": 8.56,
      "grad_norm": 4.12966251373291,
      "learning_rate": 7.975797579757976e-06,
      "loss": 0.3884,
      "step": 1730
    },
    {
      "epoch": 8.61,
      "grad_norm": 4.9768967628479,
      "learning_rate": 7.700770077007701e-06,
      "loss": 0.4092,
      "step": 1740
    },
    {
      "epoch": 8.66,
      "grad_norm": 5.166417598724365,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.3891,
      "step": 1750
    },
    {
      "epoch": 8.71,
      "grad_norm": 6.434201240539551,
      "learning_rate": 7.150715071507151e-06,
      "loss": 0.3592,
      "step": 1760
    },
    {
      "epoch": 8.76,
      "grad_norm": 7.199638843536377,
      "learning_rate": 6.875687568756876e-06,
      "loss": 0.3909,
      "step": 1770
    },
    {
      "epoch": 8.81,
      "grad_norm": 7.480076313018799,
      "learning_rate": 6.6006600660066e-06,
      "loss": 0.3895,
      "step": 1780
    },
    {
      "epoch": 8.86,
      "grad_norm": 5.124411582946777,
      "learning_rate": 6.325632563256325e-06,
      "loss": 0.391,
      "step": 1790
    },
    {
      "epoch": 8.91,
      "grad_norm": 9.751836776733398,
      "learning_rate": 6.050605060506051e-06,
      "loss": 0.3908,
      "step": 1800
    },
    {
      "epoch": 8.96,
      "grad_norm": 4.501916408538818,
      "learning_rate": 5.775577557755775e-06,
      "loss": 0.3936,
      "step": 1810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.842911877394636,
      "eval_loss": 0.3542243540287018,
      "eval_runtime": 5.623,
      "eval_samples_per_second": 510.58,
      "eval_steps_per_second": 16.006,
      "step": 1818
    },
    {
      "epoch": 9.01,
      "grad_norm": 4.195034027099609,
      "learning_rate": 5.500550055005501e-06,
      "loss": 0.4147,
      "step": 1820
    },
    {
      "epoch": 9.06,
      "grad_norm": 4.904818058013916,
      "learning_rate": 5.225522552255225e-06,
      "loss": 0.3822,
      "step": 1830
    },
    {
      "epoch": 9.11,
      "grad_norm": 4.105266094207764,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.401,
      "step": 1840
    },
    {
      "epoch": 9.16,
      "grad_norm": 4.364089488983154,
      "learning_rate": 4.675467546754675e-06,
      "loss": 0.398,
      "step": 1850
    },
    {
      "epoch": 9.21,
      "grad_norm": 5.043737411499023,
      "learning_rate": 4.400440044004401e-06,
      "loss": 0.4218,
      "step": 1860
    },
    {
      "epoch": 9.26,
      "grad_norm": 4.617894172668457,
      "learning_rate": 4.125412541254126e-06,
      "loss": 0.4002,
      "step": 1870
    },
    {
      "epoch": 9.31,
      "grad_norm": 3.750427007675171,
      "learning_rate": 3.8503850385038505e-06,
      "loss": 0.4005,
      "step": 1880
    },
    {
      "epoch": 9.36,
      "grad_norm": 4.708156585693359,
      "learning_rate": 3.5753575357535755e-06,
      "loss": 0.4067,
      "step": 1890
    },
    {
      "epoch": 9.41,
      "grad_norm": 4.125495910644531,
      "learning_rate": 3.3003300330033e-06,
      "loss": 0.3745,
      "step": 1900
    },
    {
      "epoch": 9.46,
      "grad_norm": 3.2616422176361084,
      "learning_rate": 3.0253025302530254e-06,
      "loss": 0.3914,
      "step": 1910
    },
    {
      "epoch": 9.5,
      "grad_norm": 3.8317136764526367,
      "learning_rate": 2.7502750275027504e-06,
      "loss": 0.383,
      "step": 1920
    },
    {
      "epoch": 9.55,
      "grad_norm": 4.044767379760742,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.3713,
      "step": 1930
    },
    {
      "epoch": 9.6,
      "grad_norm": 5.957603931427002,
      "learning_rate": 2.2002200220022003e-06,
      "loss": 0.3524,
      "step": 1940
    },
    {
      "epoch": 9.65,
      "grad_norm": 3.8878390789031982,
      "learning_rate": 1.9251925192519253e-06,
      "loss": 0.3876,
      "step": 1950
    },
    {
      "epoch": 9.7,
      "grad_norm": 4.049957752227783,
      "learning_rate": 1.65016501650165e-06,
      "loss": 0.3726,
      "step": 1960
    },
    {
      "epoch": 9.75,
      "grad_norm": 3.9568910598754883,
      "learning_rate": 1.3751375137513752e-06,
      "loss": 0.397,
      "step": 1970
    },
    {
      "epoch": 9.8,
      "grad_norm": 5.049424171447754,
      "learning_rate": 1.1001100110011001e-06,
      "loss": 0.3771,
      "step": 1980
    },
    {
      "epoch": 9.85,
      "grad_norm": 3.9506354331970215,
      "learning_rate": 8.25082508250825e-07,
      "loss": 0.3816,
      "step": 1990
    },
    {
      "epoch": 9.9,
      "grad_norm": 4.543933391571045,
      "learning_rate": 5.500550055005501e-07,
      "loss": 0.3961,
      "step": 2000
    },
    {
      "epoch": 9.95,
      "grad_norm": 4.267475605010986,
      "learning_rate": 2.7502750275027504e-07,
      "loss": 0.3612,
      "step": 2010
    },
    {
      "epoch": 10.0,
      "grad_norm": 7.206649303436279,
      "learning_rate": 0.0,
      "loss": 0.3422,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8397770811563915,
      "eval_loss": 0.35397040843963623,
      "eval_runtime": 5.6519,
      "eval_samples_per_second": 507.973,
      "eval_steps_per_second": 15.924,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "step": 2020,
      "total_flos": 1.2892209231838003e+18,
      "train_loss": 0.45288851414576614,
      "train_runtime": 810.9556,
      "train_samples_per_second": 318.599,
      "train_steps_per_second": 2.491
    }
  ],
  "logging_steps": 10,
  "max_steps": 2020,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.2892209231838003e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}