{
  "best_metric": 0.8714733542319749,
  "best_model_checkpoint": "vit-base-patch16-224-pure-ViT/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": 2.127864122390747,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.7058,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.910086750984192,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.6943,
      "step": 20
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.0279719829559326,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.6634,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.858572006225586,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.6503,
      "step": 40
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.1356213092803955,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.6041,
      "step": 50
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.2456114292144775,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.5789,
      "step": 60
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5485836267471313,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.5591,
      "step": 70
    },
    {
      "epoch": 0.4,
      "grad_norm": 2.320420742034912,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.5324,
      "step": 80
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.417330265045166,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.5268,
      "step": 90
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.6199421882629395,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.5218,
      "step": 100
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.233579158782959,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.5062,
      "step": 110
    },
    {
      "epoch": 0.59,
      "grad_norm": 5.436655044555664,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.5447,
      "step": 120
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.305189371109009,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.5187,
      "step": 130
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.4152462482452393,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.5011,
      "step": 140
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.06062912940979,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.5123,
      "step": 150
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.7219951152801514,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.5124,
      "step": 160
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.5283167362213135,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4934,
      "step": 170
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.957411050796509,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.5072,
      "step": 180
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.1534793376922607,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.4785,
      "step": 190
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.588723659515381,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.4676,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8094740508533612,
      "eval_loss": 0.4041886329650879,
      "eval_runtime": 13.1239,
      "eval_samples_per_second": 218.761,
      "eval_steps_per_second": 6.858,
      "step": 202
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.034203290939331,
      "learning_rate": 4.977997799779978e-05,
      "loss": 0.4653,
      "step": 210
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.753670573234558,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.4843,
      "step": 220
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.03153920173645,
      "learning_rate": 4.9229922992299234e-05,
      "loss": 0.4707,
      "step": 230
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.570380926132202,
      "learning_rate": 4.895489548954896e-05,
      "loss": 0.4922,
      "step": 240
    },
    {
      "epoch": 1.24,
      "grad_norm": 1.440894603729248,
      "learning_rate": 4.867986798679868e-05,
      "loss": 0.4689,
      "step": 250
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.8321127891540527,
      "learning_rate": 4.8404840484048406e-05,
      "loss": 0.4417,
      "step": 260
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.853463888168335,
      "learning_rate": 4.812981298129813e-05,
      "loss": 0.4847,
      "step": 270
    },
    {
      "epoch": 1.39,
      "grad_norm": 1.8001580238342285,
      "learning_rate": 4.785478547854786e-05,
      "loss": 0.4815,
      "step": 280
    },
    {
      "epoch": 1.44,
      "grad_norm": 1.7666268348693848,
      "learning_rate": 4.7579757975797585e-05,
      "loss": 0.4424,
      "step": 290
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.373815655708313,
      "learning_rate": 4.730473047304731e-05,
      "loss": 0.4604,
      "step": 300
    },
    {
      "epoch": 1.53,
      "grad_norm": 1.9827806949615479,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.4472,
      "step": 310
    },
    {
      "epoch": 1.58,
      "grad_norm": 4.021182060241699,
      "learning_rate": 4.675467546754676e-05,
      "loss": 0.4614,
      "step": 320
    },
    {
      "epoch": 1.63,
      "grad_norm": 1.6225961446762085,
      "learning_rate": 4.647964796479648e-05,
      "loss": 0.4679,
      "step": 330
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.3169050216674805,
      "learning_rate": 4.62046204620462e-05,
      "loss": 0.4268,
      "step": 340
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.3600728511810303,
      "learning_rate": 4.592959295929593e-05,
      "loss": 0.4601,
      "step": 350
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.267119884490967,
      "learning_rate": 4.5654565456545655e-05,
      "loss": 0.4535,
      "step": 360
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.0034701824188232,
      "learning_rate": 4.537953795379538e-05,
      "loss": 0.4404,
      "step": 370
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.4724442958831787,
      "learning_rate": 4.510451045104511e-05,
      "loss": 0.4393,
      "step": 380
    },
    {
      "epoch": 1.93,
      "grad_norm": 3.4427175521850586,
      "learning_rate": 4.4829482948294834e-05,
      "loss": 0.4411,
      "step": 390
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.7875171899795532,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4605,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8376872169975619,
      "eval_loss": 0.3675467073917389,
      "eval_runtime": 13.0452,
      "eval_samples_per_second": 220.081,
      "eval_steps_per_second": 6.899,
      "step": 404
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.3155332803726196,
      "learning_rate": 4.427942794279428e-05,
      "loss": 0.4514,
      "step": 410
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.455605149269104,
      "learning_rate": 4.4004400440044006e-05,
      "loss": 0.4405,
      "step": 420
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.8905019760131836,
      "learning_rate": 4.372937293729373e-05,
      "loss": 0.418,
      "step": 430
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.4562945365905762,
      "learning_rate": 4.345434543454346e-05,
      "loss": 0.4437,
      "step": 440
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.218812108039856,
      "learning_rate": 4.3179317931793185e-05,
      "loss": 0.422,
      "step": 450
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.373265266418457,
      "learning_rate": 4.2904290429042904e-05,
      "loss": 0.3849,
      "step": 460
    },
    {
      "epoch": 2.33,
      "grad_norm": 1.9199002981185913,
      "learning_rate": 4.262926292629263e-05,
      "loss": 0.4121,
      "step": 470
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.7126922607421875,
      "learning_rate": 4.2354235423542356e-05,
      "loss": 0.4351,
      "step": 480
    },
    {
      "epoch": 2.43,
      "grad_norm": 1.6914509534835815,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4296,
      "step": 490
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.3303816318511963,
      "learning_rate": 4.18041804180418e-05,
      "loss": 0.4435,
      "step": 500
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.5210704803466797,
      "learning_rate": 4.152915291529153e-05,
      "loss": 0.4341,
      "step": 510
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.497227430343628,
      "learning_rate": 4.1254125412541255e-05,
      "loss": 0.4113,
      "step": 520
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.4309931993484497,
      "learning_rate": 4.097909790979098e-05,
      "loss": 0.4353,
      "step": 530
    },
    {
      "epoch": 2.67,
      "grad_norm": 1.4119651317596436,
      "learning_rate": 4.070407040704071e-05,
      "loss": 0.4356,
      "step": 540
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.926058292388916,
      "learning_rate": 4.042904290429043e-05,
      "loss": 0.4124,
      "step": 550
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.4232733249664307,
      "learning_rate": 4.015401540154016e-05,
      "loss": 0.3981,
      "step": 560
    },
    {
      "epoch": 2.82,
      "grad_norm": 2.521343469619751,
      "learning_rate": 3.987898789878988e-05,
      "loss": 0.4113,
      "step": 570
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.658799409866333,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.4261,
      "step": 580
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.6421788930892944,
      "learning_rate": 3.932893289328933e-05,
      "loss": 0.4342,
      "step": 590
    },
    {
      "epoch": 2.97,
      "grad_norm": 1.5527944564819336,
      "learning_rate": 3.905390539053906e-05,
      "loss": 0.4012,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8505747126436781,
      "eval_loss": 0.3486035168170929,
      "eval_runtime": 13.2008,
      "eval_samples_per_second": 217.487,
      "eval_steps_per_second": 6.818,
      "step": 606
    },
    {
      "epoch": 3.02,
      "grad_norm": 1.2584953308105469,
      "learning_rate": 3.877887788778878e-05,
      "loss": 0.4146,
      "step": 610
    },
    {
      "epoch": 3.07,
      "grad_norm": 2.232922315597534,
      "learning_rate": 3.8503850385038503e-05,
      "loss": 0.4259,
      "step": 620
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.4297153949737549,
      "learning_rate": 3.822882288228823e-05,
      "loss": 0.3939,
      "step": 630
    },
    {
      "epoch": 3.17,
      "grad_norm": 1.4966973066329956,
      "learning_rate": 3.7953795379537956e-05,
      "loss": 0.4267,
      "step": 640
    },
    {
      "epoch": 3.22,
      "grad_norm": 2.102534055709839,
      "learning_rate": 3.767876787678768e-05,
      "loss": 0.4007,
      "step": 650
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.261534571647644,
      "learning_rate": 3.74037403740374e-05,
      "loss": 0.3922,
      "step": 660
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.6345826387405396,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.3908,
      "step": 670
    },
    {
      "epoch": 3.37,
      "grad_norm": 2.479656934738159,
      "learning_rate": 3.6853685368536854e-05,
      "loss": 0.4128,
      "step": 680
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.147039532661438,
      "learning_rate": 3.657865786578658e-05,
      "loss": 0.3946,
      "step": 690
    },
    {
      "epoch": 3.47,
      "grad_norm": 1.3348190784454346,
      "learning_rate": 3.6303630363036307e-05,
      "loss": 0.382,
      "step": 700
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.655010938644409,
      "learning_rate": 3.602860286028603e-05,
      "loss": 0.4011,
      "step": 710
    },
    {
      "epoch": 3.56,
      "grad_norm": 1.7104761600494385,
      "learning_rate": 3.575357535753576e-05,
      "loss": 0.362,
      "step": 720
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.740533709526062,
      "learning_rate": 3.5478547854785485e-05,
      "loss": 0.3818,
      "step": 730
    },
    {
      "epoch": 3.66,
      "grad_norm": 1.6973657608032227,
      "learning_rate": 3.5203520352035205e-05,
      "loss": 0.3794,
      "step": 740
    },
    {
      "epoch": 3.71,
      "grad_norm": 1.7627190351486206,
      "learning_rate": 3.492849284928493e-05,
      "loss": 0.4216,
      "step": 750
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.6848688125610352,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.4105,
      "step": 760
    },
    {
      "epoch": 3.81,
      "grad_norm": 1.3963148593902588,
      "learning_rate": 3.4378437843784377e-05,
      "loss": 0.4227,
      "step": 770
    },
    {
      "epoch": 3.86,
      "grad_norm": 1.3992750644683838,
      "learning_rate": 3.41034103410341e-05,
      "loss": 0.3937,
      "step": 780
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.454942226409912,
      "learning_rate": 3.382838283828383e-05,
      "loss": 0.3917,
      "step": 790
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.409183382987976,
      "learning_rate": 3.3553355335533555e-05,
      "loss": 0.3727,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8481365377917102,
      "eval_loss": 0.34128066897392273,
      "eval_runtime": 13.0732,
      "eval_samples_per_second": 219.609,
      "eval_steps_per_second": 6.884,
      "step": 808
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.3102561235427856,
      "learning_rate": 3.327832783278328e-05,
      "loss": 0.3735,
      "step": 810
    },
    {
      "epoch": 4.06,
      "grad_norm": 1.6926517486572266,
      "learning_rate": 3.300330033003301e-05,
      "loss": 0.3925,
      "step": 820
    },
    {
      "epoch": 4.11,
      "grad_norm": 2.1403160095214844,
      "learning_rate": 3.272827282728273e-05,
      "loss": 0.38,
      "step": 830
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.4817917346954346,
      "learning_rate": 3.2453245324532453e-05,
      "loss": 0.3588,
      "step": 840
    },
    {
      "epoch": 4.21,
      "grad_norm": 2.6763010025024414,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.3728,
      "step": 850
    },
    {
      "epoch": 4.26,
      "grad_norm": 2.9526965618133545,
      "learning_rate": 3.1903190319031906e-05,
      "loss": 0.3934,
      "step": 860
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.8583509922027588,
      "learning_rate": 3.162816281628163e-05,
      "loss": 0.3603,
      "step": 870
    },
    {
      "epoch": 4.36,
      "grad_norm": 2.0919086933135986,
      "learning_rate": 3.135313531353136e-05,
      "loss": 0.3527,
      "step": 880
    },
    {
      "epoch": 4.41,
      "grad_norm": 2.18446946144104,
      "learning_rate": 3.1078107810781085e-05,
      "loss": 0.3735,
      "step": 890
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.4681159257888794,
      "learning_rate": 3.0803080308030804e-05,
      "loss": 0.371,
      "step": 900
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.7022130489349365,
      "learning_rate": 3.052805280528053e-05,
      "loss": 0.3642,
      "step": 910
    },
    {
      "epoch": 4.55,
      "grad_norm": 2.031651258468628,
      "learning_rate": 3.0253025302530253e-05,
      "loss": 0.3561,
      "step": 920
    },
    {
      "epoch": 4.6,
      "grad_norm": 2.2234113216400146,
      "learning_rate": 2.9977997799779976e-05,
      "loss": 0.3568,
      "step": 930
    },
    {
      "epoch": 4.65,
      "grad_norm": 1.4382578134536743,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.3685,
      "step": 940
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.374062180519104,
      "learning_rate": 2.942794279427943e-05,
      "loss": 0.3478,
      "step": 950
    },
    {
      "epoch": 4.75,
      "grad_norm": 1.6396955251693726,
      "learning_rate": 2.9152915291529155e-05,
      "loss": 0.3549,
      "step": 960
    },
    {
      "epoch": 4.8,
      "grad_norm": 2.2647740840911865,
      "learning_rate": 2.8877887788778878e-05,
      "loss": 0.3598,
      "step": 970
    },
    {
      "epoch": 4.85,
      "grad_norm": 1.7425304651260376,
      "learning_rate": 2.8602860286028604e-05,
      "loss": 0.3898,
      "step": 980
    },
    {
      "epoch": 4.9,
      "grad_norm": 1.8157033920288086,
      "learning_rate": 2.832783278327833e-05,
      "loss": 0.3747,
      "step": 990
    },
    {
      "epoch": 4.95,
      "grad_norm": 1.7114821672439575,
      "learning_rate": 2.8052805280528056e-05,
      "loss": 0.3655,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.7515342235565186,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.3482,
      "step": 1010
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8613723441309649,
      "eval_loss": 0.33393073081970215,
      "eval_runtime": 13.1135,
      "eval_samples_per_second": 218.934,
      "eval_steps_per_second": 6.863,
      "step": 1010
    },
    {
      "epoch": 5.05,
      "grad_norm": 2.1413614749908447,
      "learning_rate": 2.7502750275027505e-05,
      "loss": 0.3581,
      "step": 1020
    },
    {
      "epoch": 5.1,
      "grad_norm": 1.9296377897262573,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.3561,
      "step": 1030
    },
    {
      "epoch": 5.15,
      "grad_norm": 2.4745469093322754,
      "learning_rate": 2.6952695269526958e-05,
      "loss": 0.3617,
      "step": 1040
    },
    {
      "epoch": 5.2,
      "grad_norm": 2.292569160461426,
      "learning_rate": 2.667766776677668e-05,
      "loss": 0.353,
      "step": 1050
    },
    {
      "epoch": 5.25,
      "grad_norm": 2.5686182975769043,
      "learning_rate": 2.64026402640264e-05,
      "loss": 0.3437,
      "step": 1060
    },
    {
      "epoch": 5.3,
      "grad_norm": 1.8417789936065674,
      "learning_rate": 2.6127612761276126e-05,
      "loss": 0.3262,
      "step": 1070
    },
    {
      "epoch": 5.35,
      "grad_norm": 2.6065661907196045,
      "learning_rate": 2.5852585258525853e-05,
      "loss": 0.3617,
      "step": 1080
    },
    {
      "epoch": 5.4,
      "grad_norm": 2.017815113067627,
      "learning_rate": 2.557755775577558e-05,
      "loss": 0.34,
      "step": 1090
    },
    {
      "epoch": 5.45,
      "grad_norm": 1.7035880088806152,
      "learning_rate": 2.53025302530253e-05,
      "loss": 0.3282,
      "step": 1100
    },
    {
      "epoch": 5.5,
      "grad_norm": 2.314371347427368,
      "learning_rate": 2.5027502750275028e-05,
      "loss": 0.3385,
      "step": 1110
    },
    {
      "epoch": 5.54,
      "grad_norm": 1.7718466520309448,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.3415,
      "step": 1120
    },
    {
      "epoch": 5.59,
      "grad_norm": 1.4826040267944336,
      "learning_rate": 2.447744774477448e-05,
      "loss": 0.3368,
      "step": 1130
    },
    {
      "epoch": 5.64,
      "grad_norm": 1.6457180976867676,
      "learning_rate": 2.4202420242024203e-05,
      "loss": 0.3774,
      "step": 1140
    },
    {
      "epoch": 5.69,
      "grad_norm": 2.1672353744506836,
      "learning_rate": 2.392739273927393e-05,
      "loss": 0.3556,
      "step": 1150
    },
    {
      "epoch": 5.74,
      "grad_norm": 1.5077142715454102,
      "learning_rate": 2.3652365236523656e-05,
      "loss": 0.34,
      "step": 1160
    },
    {
      "epoch": 5.79,
      "grad_norm": 1.6296086311340332,
      "learning_rate": 2.337733773377338e-05,
      "loss": 0.3325,
      "step": 1170
    },
    {
      "epoch": 5.84,
      "grad_norm": 2.324458599090576,
      "learning_rate": 2.31023102310231e-05,
      "loss": 0.3276,
      "step": 1180
    },
    {
      "epoch": 5.89,
      "grad_norm": 2.119600534439087,
      "learning_rate": 2.2827282728272828e-05,
      "loss": 0.3472,
      "step": 1190
    },
    {
      "epoch": 5.94,
      "grad_norm": 1.9447872638702393,
      "learning_rate": 2.2552255225522554e-05,
      "loss": 0.3702,
      "step": 1200
    },
    {
      "epoch": 5.99,
      "grad_norm": 1.564266562461853,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.354,
      "step": 1210
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8561476837338906,
      "eval_loss": 0.343646377325058,
      "eval_runtime": 13.2461,
      "eval_samples_per_second": 216.743,
      "eval_steps_per_second": 6.794,
      "step": 1212
    },
    {
      "epoch": 6.04,
      "grad_norm": 1.6272279024124146,
      "learning_rate": 2.2002200220022003e-05,
      "loss": 0.3351,
      "step": 1220
    },
    {
      "epoch": 6.09,
      "grad_norm": 1.7797123193740845,
      "learning_rate": 2.172717271727173e-05,
      "loss": 0.3326,
      "step": 1230
    },
    {
      "epoch": 6.14,
      "grad_norm": 1.8236846923828125,
      "learning_rate": 2.1452145214521452e-05,
      "loss": 0.3223,
      "step": 1240
    },
    {
      "epoch": 6.19,
      "grad_norm": 1.5644466876983643,
      "learning_rate": 2.1177117711771178e-05,
      "loss": 0.3195,
      "step": 1250
    },
    {
      "epoch": 6.24,
      "grad_norm": 1.7544844150543213,
      "learning_rate": 2.09020902090209e-05,
      "loss": 0.3086,
      "step": 1260
    },
    {
      "epoch": 6.29,
      "grad_norm": 1.6309467554092407,
      "learning_rate": 2.0627062706270627e-05,
      "loss": 0.3423,
      "step": 1270
    },
    {
      "epoch": 6.34,
      "grad_norm": 2.1340713500976562,
      "learning_rate": 2.0352035203520354e-05,
      "loss": 0.3361,
      "step": 1280
    },
    {
      "epoch": 6.39,
      "grad_norm": 1.5964475870132446,
      "learning_rate": 2.007700770077008e-05,
      "loss": 0.3254,
      "step": 1290
    },
    {
      "epoch": 6.44,
      "grad_norm": 1.533130168914795,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.3294,
      "step": 1300
    },
    {
      "epoch": 6.49,
      "grad_norm": 1.8504540920257568,
      "learning_rate": 1.952695269526953e-05,
      "loss": 0.3146,
      "step": 1310
    },
    {
      "epoch": 6.53,
      "grad_norm": 2.3233487606048584,
      "learning_rate": 1.9251925192519252e-05,
      "loss": 0.3308,
      "step": 1320
    },
    {
      "epoch": 6.58,
      "grad_norm": 1.8644599914550781,
      "learning_rate": 1.8976897689768978e-05,
      "loss": 0.3213,
      "step": 1330
    },
    {
      "epoch": 6.63,
      "grad_norm": 1.5879623889923096,
      "learning_rate": 1.87018701870187e-05,
      "loss": 0.3121,
      "step": 1340
    },
    {
      "epoch": 6.68,
      "grad_norm": 1.4578536748886108,
      "learning_rate": 1.8426842684268427e-05,
      "loss": 0.3052,
      "step": 1350
    },
    {
      "epoch": 6.73,
      "grad_norm": 1.7176955938339233,
      "learning_rate": 1.8151815181518153e-05,
      "loss": 0.3236,
      "step": 1360
    },
    {
      "epoch": 6.78,
      "grad_norm": 2.3816351890563965,
      "learning_rate": 1.787678767876788e-05,
      "loss": 0.3168,
      "step": 1370
    },
    {
      "epoch": 6.83,
      "grad_norm": 2.024005889892578,
      "learning_rate": 1.7601760176017602e-05,
      "loss": 0.3594,
      "step": 1380
    },
    {
      "epoch": 6.88,
      "grad_norm": 1.6122448444366455,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.3304,
      "step": 1390
    },
    {
      "epoch": 6.93,
      "grad_norm": 2.809938430786133,
      "learning_rate": 1.705170517051705e-05,
      "loss": 0.3376,
      "step": 1400
    },
    {
      "epoch": 6.98,
      "grad_norm": 2.1670022010803223,
      "learning_rate": 1.6776677667766778e-05,
      "loss": 0.3212,
      "step": 1410
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8533611981887844,
      "eval_loss": 0.3415059745311737,
      "eval_runtime": 13.0333,
      "eval_samples_per_second": 220.283,
      "eval_steps_per_second": 6.905,
      "step": 1414
    },
    {
      "epoch": 7.03,
      "grad_norm": 2.0041117668151855,
      "learning_rate": 1.6501650165016504e-05,
      "loss": 0.3216,
      "step": 1420
    },
    {
      "epoch": 7.08,
      "grad_norm": 1.5932248830795288,
      "learning_rate": 1.6226622662266227e-05,
      "loss": 0.2902,
      "step": 1430
    },
    {
      "epoch": 7.13,
      "grad_norm": 1.8543118238449097,
      "learning_rate": 1.5951595159515953e-05,
      "loss": 0.3383,
      "step": 1440
    },
    {
      "epoch": 7.18,
      "grad_norm": 2.0562777519226074,
      "learning_rate": 1.567656765676568e-05,
      "loss": 0.3155,
      "step": 1450
    },
    {
      "epoch": 7.23,
      "grad_norm": 2.037494421005249,
      "learning_rate": 1.5401540154015402e-05,
      "loss": 0.3104,
      "step": 1460
    },
    {
      "epoch": 7.28,
      "grad_norm": 1.9015072584152222,
      "learning_rate": 1.5126512651265127e-05,
      "loss": 0.3259,
      "step": 1470
    },
    {
      "epoch": 7.33,
      "grad_norm": 1.8102400302886963,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.3087,
      "step": 1480
    },
    {
      "epoch": 7.38,
      "grad_norm": 2.07037353515625,
      "learning_rate": 1.4576457645764577e-05,
      "loss": 0.301,
      "step": 1490
    },
    {
      "epoch": 7.43,
      "grad_norm": 2.2085139751434326,
      "learning_rate": 1.4301430143014302e-05,
      "loss": 0.3115,
      "step": 1500
    },
    {
      "epoch": 7.48,
      "grad_norm": 2.438488721847534,
      "learning_rate": 1.4026402640264028e-05,
      "loss": 0.3406,
      "step": 1510
    },
    {
      "epoch": 7.52,
      "grad_norm": 1.43858003616333,
      "learning_rate": 1.3751375137513753e-05,
      "loss": 0.2928,
      "step": 1520
    },
    {
      "epoch": 7.57,
      "grad_norm": 1.8977525234222412,
      "learning_rate": 1.3476347634763479e-05,
      "loss": 0.3201,
      "step": 1530
    },
    {
      "epoch": 7.62,
      "grad_norm": 1.6580339670181274,
      "learning_rate": 1.32013201320132e-05,
      "loss": 0.2892,
      "step": 1540
    },
    {
      "epoch": 7.67,
      "grad_norm": 2.5153727531433105,
      "learning_rate": 1.2926292629262926e-05,
      "loss": 0.3087,
      "step": 1550
    },
    {
      "epoch": 7.72,
      "grad_norm": 2.4366941452026367,
      "learning_rate": 1.265126512651265e-05,
      "loss": 0.28,
      "step": 1560
    },
    {
      "epoch": 7.77,
      "grad_norm": 1.8391032218933105,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.3104,
      "step": 1570
    },
    {
      "epoch": 7.82,
      "grad_norm": 3.4520180225372314,
      "learning_rate": 1.2101210121012102e-05,
      "loss": 0.3113,
      "step": 1580
    },
    {
      "epoch": 7.87,
      "grad_norm": 2.3579323291778564,
      "learning_rate": 1.1826182618261828e-05,
      "loss": 0.3308,
      "step": 1590
    },
    {
      "epoch": 7.92,
      "grad_norm": 2.1137523651123047,
      "learning_rate": 1.155115511551155e-05,
      "loss": 0.3024,
      "step": 1600
    },
    {
      "epoch": 7.97,
      "grad_norm": 2.0364956855773926,
      "learning_rate": 1.1276127612761277e-05,
      "loss": 0.3263,
      "step": 1610
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.864158829676071,
      "eval_loss": 0.3281286954879761,
      "eval_runtime": 13.04,
      "eval_samples_per_second": 220.169,
      "eval_steps_per_second": 6.902,
      "step": 1616
    },
    {
      "epoch": 8.02,
      "grad_norm": 1.6843018531799316,
      "learning_rate": 1.1001100110011001e-05,
      "loss": 0.3053,
      "step": 1620
    },
    {
      "epoch": 8.07,
      "grad_norm": 1.6904422044754028,
      "learning_rate": 1.0726072607260726e-05,
      "loss": 0.2766,
      "step": 1630
    },
    {
      "epoch": 8.12,
      "grad_norm": 2.0085134506225586,
      "learning_rate": 1.045104510451045e-05,
      "loss": 0.2933,
      "step": 1640
    },
    {
      "epoch": 8.17,
      "grad_norm": 2.3377184867858887,
      "learning_rate": 1.0176017601760177e-05,
      "loss": 0.3185,
      "step": 1650
    },
    {
      "epoch": 8.22,
      "grad_norm": 1.9562610387802124,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.3282,
      "step": 1660
    },
    {
      "epoch": 8.27,
      "grad_norm": 1.9323431253433228,
      "learning_rate": 9.625962596259626e-06,
      "loss": 0.3042,
      "step": 1670
    },
    {
      "epoch": 8.32,
      "grad_norm": 1.8757867813110352,
      "learning_rate": 9.35093509350935e-06,
      "loss": 0.2838,
      "step": 1680
    },
    {
      "epoch": 8.37,
      "grad_norm": 1.5240260362625122,
      "learning_rate": 9.075907590759077e-06,
      "loss": 0.2996,
      "step": 1690
    },
    {
      "epoch": 8.42,
      "grad_norm": 1.236165165901184,
      "learning_rate": 8.800880088008801e-06,
      "loss": 0.2833,
      "step": 1700
    },
    {
      "epoch": 8.47,
      "grad_norm": 1.9543601274490356,
      "learning_rate": 8.525852585258526e-06,
      "loss": 0.2706,
      "step": 1710
    },
    {
      "epoch": 8.51,
      "grad_norm": 2.0328145027160645,
      "learning_rate": 8.250825082508252e-06,
      "loss": 0.2952,
      "step": 1720
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.5521138906478882,
      "learning_rate": 7.975797579757976e-06,
      "loss": 0.2984,
      "step": 1730
    },
    {
      "epoch": 8.61,
      "grad_norm": 1.9917885065078735,
      "learning_rate": 7.700770077007701e-06,
      "loss": 0.2924,
      "step": 1740
    },
    {
      "epoch": 8.66,
      "grad_norm": 1.6940460205078125,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.3079,
      "step": 1750
    },
    {
      "epoch": 8.71,
      "grad_norm": 1.891791582107544,
      "learning_rate": 7.150715071507151e-06,
      "loss": 0.3039,
      "step": 1760
    },
    {
      "epoch": 8.76,
      "grad_norm": 1.9249194860458374,
      "learning_rate": 6.875687568756876e-06,
      "loss": 0.2744,
      "step": 1770
    },
    {
      "epoch": 8.81,
      "grad_norm": 1.7590512037277222,
      "learning_rate": 6.6006600660066e-06,
      "loss": 0.3029,
      "step": 1780
    },
    {
      "epoch": 8.86,
      "grad_norm": 1.5572028160095215,
      "learning_rate": 6.325632563256325e-06,
      "loss": 0.2814,
      "step": 1790
    },
    {
      "epoch": 8.91,
      "grad_norm": 1.9127514362335205,
      "learning_rate": 6.050605060506051e-06,
      "loss": 0.2968,
      "step": 1800
    },
    {
      "epoch": 8.96,
      "grad_norm": 1.6464173793792725,
      "learning_rate": 5.775577557755775e-06,
      "loss": 0.285,
      "step": 1810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8672936259143156,
      "eval_loss": 0.3263307809829712,
      "eval_runtime": 13.0994,
      "eval_samples_per_second": 219.17,
      "eval_steps_per_second": 6.871,
      "step": 1818
    },
    {
      "epoch": 9.01,
      "grad_norm": 2.2403345108032227,
      "learning_rate": 5.500550055005501e-06,
      "loss": 0.3191,
      "step": 1820
    },
    {
      "epoch": 9.06,
      "grad_norm": 2.0323376655578613,
      "learning_rate": 5.225522552255225e-06,
      "loss": 0.2864,
      "step": 1830
    },
    {
      "epoch": 9.11,
      "grad_norm": 2.5605928897857666,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.2973,
      "step": 1840
    },
    {
      "epoch": 9.16,
      "grad_norm": 2.1377811431884766,
      "learning_rate": 4.675467546754675e-06,
      "loss": 0.2711,
      "step": 1850
    },
    {
      "epoch": 9.21,
      "grad_norm": 1.5810461044311523,
      "learning_rate": 4.400440044004401e-06,
      "loss": 0.2816,
      "step": 1860
    },
    {
      "epoch": 9.26,
      "grad_norm": 2.0196871757507324,
      "learning_rate": 4.125412541254126e-06,
      "loss": 0.2867,
      "step": 1870
    },
    {
      "epoch": 9.31,
      "grad_norm": 1.4029825925827026,
      "learning_rate": 3.8503850385038505e-06,
      "loss": 0.2873,
      "step": 1880
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.715185284614563,
      "learning_rate": 3.5753575357535755e-06,
      "loss": 0.2944,
      "step": 1890
    },
    {
      "epoch": 9.41,
      "grad_norm": 1.9352003335952759,
      "learning_rate": 3.3003300330033e-06,
      "loss": 0.2926,
      "step": 1900
    },
    {
      "epoch": 9.46,
      "grad_norm": 2.096611738204956,
      "learning_rate": 3.0253025302530254e-06,
      "loss": 0.2593,
      "step": 1910
    },
    {
      "epoch": 9.5,
      "grad_norm": 1.943051815032959,
      "learning_rate": 2.7502750275027504e-06,
      "loss": 0.2894,
      "step": 1920
    },
    {
      "epoch": 9.55,
      "grad_norm": 1.3575071096420288,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.2958,
      "step": 1930
    },
    {
      "epoch": 9.6,
      "grad_norm": 1.547815203666687,
      "learning_rate": 2.2002200220022003e-06,
      "loss": 0.2707,
      "step": 1940
    },
    {
      "epoch": 9.65,
      "grad_norm": 2.169074773788452,
      "learning_rate": 1.9251925192519253e-06,
      "loss": 0.269,
      "step": 1950
    },
    {
      "epoch": 9.7,
      "grad_norm": 1.756319522857666,
      "learning_rate": 1.65016501650165e-06,
      "loss": 0.2827,
      "step": 1960
    },
    {
      "epoch": 9.75,
      "grad_norm": 1.895622730255127,
      "learning_rate": 1.3751375137513752e-06,
      "loss": 0.2791,
      "step": 1970
    },
    {
      "epoch": 9.8,
      "grad_norm": 1.7341846227645874,
      "learning_rate": 1.1001100110011001e-06,
      "loss": 0.2808,
      "step": 1980
    },
    {
      "epoch": 9.85,
      "grad_norm": 1.894599199295044,
      "learning_rate": 8.25082508250825e-07,
      "loss": 0.2961,
      "step": 1990
    },
    {
      "epoch": 9.9,
      "grad_norm": 2.3876936435699463,
      "learning_rate": 5.500550055005501e-07,
      "loss": 0.2771,
      "step": 2000
    },
    {
      "epoch": 9.95,
      "grad_norm": 1.6493264436721802,
      "learning_rate": 2.7502750275027504e-07,
      "loss": 0.2674,
      "step": 2010
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.8131953477859497,
      "learning_rate": 0.0,
      "loss": 0.2779,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8714733542319749,
      "eval_loss": 0.3270263373851776,
      "eval_runtime": 13.1626,
      "eval_samples_per_second": 218.117,
      "eval_steps_per_second": 6.838,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "step": 2020,
      "total_flos": 2.0021605356722135e+19,
      "train_loss": 0.37628526250914773,
      "train_runtime": 2991.3974,
      "train_samples_per_second": 86.371,
      "train_steps_per_second": 0.675
    }
  ],
  "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
}