{
  "best_metric": 0.8707767328456983,
  "best_model_checkpoint": "hybrid-cnn-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": 3.4049220085144043,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.75,
      "step": 10
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.486193895339966,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.7094,
      "step": 20
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.687983989715576,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.633,
      "step": 30
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.435194969177246,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.5843,
      "step": 40
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.748059272766113,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.5832,
      "step": 50
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.351829767227173,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.5679,
      "step": 60
    },
    {
      "epoch": 0.35,
      "grad_norm": 5.649282455444336,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.5399,
      "step": 70
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.6417155265808105,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.516,
      "step": 80
    },
    {
      "epoch": 0.45,
      "grad_norm": 3.760568857192993,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.5005,
      "step": 90
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.888506889343262,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.5184,
      "step": 100
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.107848644256592,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.5198,
      "step": 110
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.687451124191284,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.5292,
      "step": 120
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.2040014266967773,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.5167,
      "step": 130
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.79260778427124,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.5081,
      "step": 140
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.946630954742432,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.4637,
      "step": 150
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.178718090057373,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.4939,
      "step": 160
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.5444717407226562,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.4808,
      "step": 170
    },
    {
      "epoch": 0.89,
      "grad_norm": 5.634459972381592,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4815,
      "step": 180
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.709349632263184,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.5112,
      "step": 190
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.794649600982666,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.5277,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8209683037269244,
      "eval_loss": 0.39027276635169983,
      "eval_runtime": 34.4714,
      "eval_samples_per_second": 83.286,
      "eval_steps_per_second": 2.611,
      "step": 202
    },
    {
      "epoch": 1.04,
      "grad_norm": 6.730863571166992,
      "learning_rate": 4.977997799779978e-05,
      "loss": 0.4614,
      "step": 210
    },
    {
      "epoch": 1.09,
      "grad_norm": 3.0025227069854736,
      "learning_rate": 4.950495049504951e-05,
      "loss": 0.4952,
      "step": 220
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.116140842437744,
      "learning_rate": 4.9229922992299234e-05,
      "loss": 0.4742,
      "step": 230
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.0951218605041504,
      "learning_rate": 4.895489548954896e-05,
      "loss": 0.47,
      "step": 240
    },
    {
      "epoch": 1.24,
      "grad_norm": 9.745083808898926,
      "learning_rate": 4.867986798679868e-05,
      "loss": 0.4991,
      "step": 250
    },
    {
      "epoch": 1.29,
      "grad_norm": 3.8443119525909424,
      "learning_rate": 4.8404840484048406e-05,
      "loss": 0.5179,
      "step": 260
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.324491262435913,
      "learning_rate": 4.812981298129813e-05,
      "loss": 0.4431,
      "step": 270
    },
    {
      "epoch": 1.39,
      "grad_norm": 4.08018684387207,
      "learning_rate": 4.785478547854786e-05,
      "loss": 0.4497,
      "step": 280
    },
    {
      "epoch": 1.44,
      "grad_norm": 4.045833587646484,
      "learning_rate": 4.7579757975797585e-05,
      "loss": 0.4923,
      "step": 290
    },
    {
      "epoch": 1.49,
      "grad_norm": 5.029973983764648,
      "learning_rate": 4.730473047304731e-05,
      "loss": 0.4931,
      "step": 300
    },
    {
      "epoch": 1.53,
      "grad_norm": 3.3357350826263428,
      "learning_rate": 4.702970297029703e-05,
      "loss": 0.4768,
      "step": 310
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.275207042694092,
      "learning_rate": 4.675467546754676e-05,
      "loss": 0.4646,
      "step": 320
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.9513421058654785,
      "learning_rate": 4.647964796479648e-05,
      "loss": 0.4692,
      "step": 330
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.237048625946045,
      "learning_rate": 4.62046204620462e-05,
      "loss": 0.4692,
      "step": 340
    },
    {
      "epoch": 1.73,
      "grad_norm": 1.937572717666626,
      "learning_rate": 4.592959295929593e-05,
      "loss": 0.4475,
      "step": 350
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.6712441444396973,
      "learning_rate": 4.5654565456545655e-05,
      "loss": 0.465,
      "step": 360
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.3434207439422607,
      "learning_rate": 4.537953795379538e-05,
      "loss": 0.473,
      "step": 370
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.657801389694214,
      "learning_rate": 4.510451045104511e-05,
      "loss": 0.456,
      "step": 380
    },
    {
      "epoch": 1.93,
      "grad_norm": 1.898109793663025,
      "learning_rate": 4.4829482948294834e-05,
      "loss": 0.4769,
      "step": 390
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.6151821613311768,
      "learning_rate": 4.455445544554456e-05,
      "loss": 0.4623,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8415186346220829,
      "eval_loss": 0.3477635383605957,
      "eval_runtime": 34.2991,
      "eval_samples_per_second": 83.705,
      "eval_steps_per_second": 2.624,
      "step": 404
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.7920018434524536,
      "learning_rate": 4.427942794279428e-05,
      "loss": 0.451,
      "step": 410
    },
    {
      "epoch": 2.08,
      "grad_norm": 3.742645502090454,
      "learning_rate": 4.4004400440044006e-05,
      "loss": 0.4107,
      "step": 420
    },
    {
      "epoch": 2.13,
      "grad_norm": 3.6825478076934814,
      "learning_rate": 4.372937293729373e-05,
      "loss": 0.4103,
      "step": 430
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.2625832557678223,
      "learning_rate": 4.345434543454346e-05,
      "loss": 0.4334,
      "step": 440
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.2035701274871826,
      "learning_rate": 4.3179317931793185e-05,
      "loss": 0.4155,
      "step": 450
    },
    {
      "epoch": 2.28,
      "grad_norm": 4.748138904571533,
      "learning_rate": 4.2904290429042904e-05,
      "loss": 0.4452,
      "step": 460
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.5996172428131104,
      "learning_rate": 4.262926292629263e-05,
      "loss": 0.4156,
      "step": 470
    },
    {
      "epoch": 2.38,
      "grad_norm": 1.9976857900619507,
      "learning_rate": 4.2354235423542356e-05,
      "loss": 0.4161,
      "step": 480
    },
    {
      "epoch": 2.43,
      "grad_norm": 4.618232250213623,
      "learning_rate": 4.207920792079208e-05,
      "loss": 0.439,
      "step": 490
    },
    {
      "epoch": 2.48,
      "grad_norm": 4.22369384765625,
      "learning_rate": 4.18041804180418e-05,
      "loss": 0.4711,
      "step": 500
    },
    {
      "epoch": 2.52,
      "grad_norm": 4.014861106872559,
      "learning_rate": 4.152915291529153e-05,
      "loss": 0.4502,
      "step": 510
    },
    {
      "epoch": 2.57,
      "grad_norm": 1.90145742893219,
      "learning_rate": 4.1254125412541255e-05,
      "loss": 0.4216,
      "step": 520
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.8490346670150757,
      "learning_rate": 4.097909790979098e-05,
      "loss": 0.3942,
      "step": 530
    },
    {
      "epoch": 2.67,
      "grad_norm": 2.4955575466156006,
      "learning_rate": 4.070407040704071e-05,
      "loss": 0.4037,
      "step": 540
    },
    {
      "epoch": 2.72,
      "grad_norm": 2.651855945587158,
      "learning_rate": 4.042904290429043e-05,
      "loss": 0.4208,
      "step": 550
    },
    {
      "epoch": 2.77,
      "grad_norm": 3.947380781173706,
      "learning_rate": 4.015401540154016e-05,
      "loss": 0.4216,
      "step": 560
    },
    {
      "epoch": 2.82,
      "grad_norm": 3.179884433746338,
      "learning_rate": 3.987898789878988e-05,
      "loss": 0.4367,
      "step": 570
    },
    {
      "epoch": 2.87,
      "grad_norm": 2.080159902572632,
      "learning_rate": 3.9603960396039605e-05,
      "loss": 0.4239,
      "step": 580
    },
    {
      "epoch": 2.92,
      "grad_norm": 2.788001775741577,
      "learning_rate": 3.932893289328933e-05,
      "loss": 0.4063,
      "step": 590
    },
    {
      "epoch": 2.97,
      "grad_norm": 2.00004243850708,
      "learning_rate": 3.905390539053906e-05,
      "loss": 0.4497,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8519679554162313,
      "eval_loss": 0.3333553075790405,
      "eval_runtime": 34.2147,
      "eval_samples_per_second": 83.911,
      "eval_steps_per_second": 2.63,
      "step": 606
    },
    {
      "epoch": 3.02,
      "grad_norm": 2.456493854522705,
      "learning_rate": 3.877887788778878e-05,
      "loss": 0.4136,
      "step": 610
    },
    {
      "epoch": 3.07,
      "grad_norm": 2.922100305557251,
      "learning_rate": 3.8503850385038503e-05,
      "loss": 0.4026,
      "step": 620
    },
    {
      "epoch": 3.12,
      "grad_norm": 2.9372498989105225,
      "learning_rate": 3.822882288228823e-05,
      "loss": 0.4093,
      "step": 630
    },
    {
      "epoch": 3.17,
      "grad_norm": 3.000676393508911,
      "learning_rate": 3.7953795379537956e-05,
      "loss": 0.3901,
      "step": 640
    },
    {
      "epoch": 3.22,
      "grad_norm": 2.4853076934814453,
      "learning_rate": 3.767876787678768e-05,
      "loss": 0.3956,
      "step": 650
    },
    {
      "epoch": 3.27,
      "grad_norm": 2.6341211795806885,
      "learning_rate": 3.74037403740374e-05,
      "loss": 0.4142,
      "step": 660
    },
    {
      "epoch": 3.32,
      "grad_norm": 2.0509798526763916,
      "learning_rate": 3.712871287128713e-05,
      "loss": 0.3922,
      "step": 670
    },
    {
      "epoch": 3.37,
      "grad_norm": 2.622851848602295,
      "learning_rate": 3.6853685368536854e-05,
      "loss": 0.4163,
      "step": 680
    },
    {
      "epoch": 3.42,
      "grad_norm": 2.008538007736206,
      "learning_rate": 3.657865786578658e-05,
      "loss": 0.3921,
      "step": 690
    },
    {
      "epoch": 3.47,
      "grad_norm": 2.5022664070129395,
      "learning_rate": 3.6303630363036307e-05,
      "loss": 0.3626,
      "step": 700
    },
    {
      "epoch": 3.51,
      "grad_norm": 3.2395951747894287,
      "learning_rate": 3.602860286028603e-05,
      "loss": 0.4153,
      "step": 710
    },
    {
      "epoch": 3.56,
      "grad_norm": 2.134999990463257,
      "learning_rate": 3.575357535753576e-05,
      "loss": 0.3853,
      "step": 720
    },
    {
      "epoch": 3.61,
      "grad_norm": 3.0224032402038574,
      "learning_rate": 3.5478547854785485e-05,
      "loss": 0.3913,
      "step": 730
    },
    {
      "epoch": 3.66,
      "grad_norm": 2.18223237991333,
      "learning_rate": 3.5203520352035205e-05,
      "loss": 0.4183,
      "step": 740
    },
    {
      "epoch": 3.71,
      "grad_norm": 2.258599281311035,
      "learning_rate": 3.492849284928493e-05,
      "loss": 0.397,
      "step": 750
    },
    {
      "epoch": 3.76,
      "grad_norm": 2.017249822616577,
      "learning_rate": 3.465346534653465e-05,
      "loss": 0.3807,
      "step": 760
    },
    {
      "epoch": 3.81,
      "grad_norm": 2.045938491821289,
      "learning_rate": 3.4378437843784377e-05,
      "loss": 0.4119,
      "step": 770
    },
    {
      "epoch": 3.86,
      "grad_norm": 5.812851428985596,
      "learning_rate": 3.41034103410341e-05,
      "loss": 0.4088,
      "step": 780
    },
    {
      "epoch": 3.91,
      "grad_norm": 1.954240083694458,
      "learning_rate": 3.382838283828383e-05,
      "loss": 0.3862,
      "step": 790
    },
    {
      "epoch": 3.96,
      "grad_norm": 3.0154707431793213,
      "learning_rate": 3.3553355335533555e-05,
      "loss": 0.4074,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8460466736328806,
      "eval_loss": 0.33968210220336914,
      "eval_runtime": 34.3325,
      "eval_samples_per_second": 83.623,
      "eval_steps_per_second": 2.621,
      "step": 808
    },
    {
      "epoch": 4.01,
      "grad_norm": 2.900836706161499,
      "learning_rate": 3.327832783278328e-05,
      "loss": 0.3948,
      "step": 810
    },
    {
      "epoch": 4.06,
      "grad_norm": 3.435845375061035,
      "learning_rate": 3.300330033003301e-05,
      "loss": 0.3919,
      "step": 820
    },
    {
      "epoch": 4.11,
      "grad_norm": 2.006223201751709,
      "learning_rate": 3.272827282728273e-05,
      "loss": 0.3892,
      "step": 830
    },
    {
      "epoch": 4.16,
      "grad_norm": 3.857332229614258,
      "learning_rate": 3.2453245324532453e-05,
      "loss": 0.3901,
      "step": 840
    },
    {
      "epoch": 4.21,
      "grad_norm": 3.8714852333068848,
      "learning_rate": 3.217821782178218e-05,
      "loss": 0.3751,
      "step": 850
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.7595456838607788,
      "learning_rate": 3.1903190319031906e-05,
      "loss": 0.3789,
      "step": 860
    },
    {
      "epoch": 4.31,
      "grad_norm": 2.590575695037842,
      "learning_rate": 3.162816281628163e-05,
      "loss": 0.3714,
      "step": 870
    },
    {
      "epoch": 4.36,
      "grad_norm": 2.044952154159546,
      "learning_rate": 3.135313531353136e-05,
      "loss": 0.394,
      "step": 880
    },
    {
      "epoch": 4.41,
      "grad_norm": 2.6872971057891846,
      "learning_rate": 3.1078107810781085e-05,
      "loss": 0.3501,
      "step": 890
    },
    {
      "epoch": 4.46,
      "grad_norm": 4.981396675109863,
      "learning_rate": 3.0803080308030804e-05,
      "loss": 0.3813,
      "step": 900
    },
    {
      "epoch": 4.5,
      "grad_norm": 2.0973191261291504,
      "learning_rate": 3.052805280528053e-05,
      "loss": 0.3627,
      "step": 910
    },
    {
      "epoch": 4.55,
      "grad_norm": 2.8159642219543457,
      "learning_rate": 3.0253025302530253e-05,
      "loss": 0.372,
      "step": 920
    },
    {
      "epoch": 4.6,
      "grad_norm": 4.265486240386963,
      "learning_rate": 2.9977997799779976e-05,
      "loss": 0.3565,
      "step": 930
    },
    {
      "epoch": 4.65,
      "grad_norm": 1.922487735748291,
      "learning_rate": 2.9702970297029702e-05,
      "loss": 0.3528,
      "step": 940
    },
    {
      "epoch": 4.7,
      "grad_norm": 2.1887781620025635,
      "learning_rate": 2.942794279427943e-05,
      "loss": 0.346,
      "step": 950
    },
    {
      "epoch": 4.75,
      "grad_norm": 2.510910749435425,
      "learning_rate": 2.9152915291529155e-05,
      "loss": 0.3465,
      "step": 960
    },
    {
      "epoch": 4.8,
      "grad_norm": 2.1515932083129883,
      "learning_rate": 2.8877887788778878e-05,
      "loss": 0.3558,
      "step": 970
    },
    {
      "epoch": 4.85,
      "grad_norm": 3.1407530307769775,
      "learning_rate": 2.8602860286028604e-05,
      "loss": 0.3909,
      "step": 980
    },
    {
      "epoch": 4.9,
      "grad_norm": 2.861525535583496,
      "learning_rate": 2.832783278327833e-05,
      "loss": 0.3804,
      "step": 990
    },
    {
      "epoch": 4.95,
      "grad_norm": 2.334779739379883,
      "learning_rate": 2.8052805280528056e-05,
      "loss": 0.3362,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.414121150970459,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.3552,
      "step": 1010
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8624172762103797,
      "eval_loss": 0.3226765990257263,
      "eval_runtime": 34.3538,
      "eval_samples_per_second": 83.572,
      "eval_steps_per_second": 2.62,
      "step": 1010
    },
    {
      "epoch": 5.05,
      "grad_norm": 3.0383219718933105,
      "learning_rate": 2.7502750275027505e-05,
      "loss": 0.3375,
      "step": 1020
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.1559934616088867,
      "learning_rate": 2.722772277227723e-05,
      "loss": 0.347,
      "step": 1030
    },
    {
      "epoch": 5.15,
      "grad_norm": 2.3701679706573486,
      "learning_rate": 2.6952695269526958e-05,
      "loss": 0.3319,
      "step": 1040
    },
    {
      "epoch": 5.2,
      "grad_norm": 3.0003325939178467,
      "learning_rate": 2.667766776677668e-05,
      "loss": 0.342,
      "step": 1050
    },
    {
      "epoch": 5.25,
      "grad_norm": 3.154733657836914,
      "learning_rate": 2.64026402640264e-05,
      "loss": 0.3753,
      "step": 1060
    },
    {
      "epoch": 5.3,
      "grad_norm": 2.575256824493408,
      "learning_rate": 2.6127612761276126e-05,
      "loss": 0.3037,
      "step": 1070
    },
    {
      "epoch": 5.35,
      "grad_norm": 2.572767734527588,
      "learning_rate": 2.5852585258525853e-05,
      "loss": 0.3334,
      "step": 1080
    },
    {
      "epoch": 5.4,
      "grad_norm": 2.2883052825927734,
      "learning_rate": 2.557755775577558e-05,
      "loss": 0.3367,
      "step": 1090
    },
    {
      "epoch": 5.45,
      "grad_norm": 2.747040271759033,
      "learning_rate": 2.53025302530253e-05,
      "loss": 0.3135,
      "step": 1100
    },
    {
      "epoch": 5.5,
      "grad_norm": 3.9614763259887695,
      "learning_rate": 2.5027502750275028e-05,
      "loss": 0.3513,
      "step": 1110
    },
    {
      "epoch": 5.54,
      "grad_norm": 3.847574234008789,
      "learning_rate": 2.4752475247524754e-05,
      "loss": 0.3443,
      "step": 1120
    },
    {
      "epoch": 5.59,
      "grad_norm": 2.754849433898926,
      "learning_rate": 2.447744774477448e-05,
      "loss": 0.3472,
      "step": 1130
    },
    {
      "epoch": 5.64,
      "grad_norm": 2.95867657661438,
      "learning_rate": 2.4202420242024203e-05,
      "loss": 0.3424,
      "step": 1140
    },
    {
      "epoch": 5.69,
      "grad_norm": 3.1151840686798096,
      "learning_rate": 2.392739273927393e-05,
      "loss": 0.3431,
      "step": 1150
    },
    {
      "epoch": 5.74,
      "grad_norm": 2.3718581199645996,
      "learning_rate": 2.3652365236523656e-05,
      "loss": 0.3366,
      "step": 1160
    },
    {
      "epoch": 5.79,
      "grad_norm": 2.323246479034424,
      "learning_rate": 2.337733773377338e-05,
      "loss": 0.3162,
      "step": 1170
    },
    {
      "epoch": 5.84,
      "grad_norm": 2.7909843921661377,
      "learning_rate": 2.31023102310231e-05,
      "loss": 0.3521,
      "step": 1180
    },
    {
      "epoch": 5.89,
      "grad_norm": 3.047494888305664,
      "learning_rate": 2.2827282728272828e-05,
      "loss": 0.348,
      "step": 1190
    },
    {
      "epoch": 5.94,
      "grad_norm": 2.6819052696228027,
      "learning_rate": 2.2552255225522554e-05,
      "loss": 0.3573,
      "step": 1200
    },
    {
      "epoch": 5.99,
      "grad_norm": 2.541433334350586,
      "learning_rate": 2.227722772277228e-05,
      "loss": 0.3637,
      "step": 1210
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8617206548241031,
      "eval_loss": 0.3230111300945282,
      "eval_runtime": 34.3385,
      "eval_samples_per_second": 83.609,
      "eval_steps_per_second": 2.621,
      "step": 1212
    },
    {
      "epoch": 6.04,
      "grad_norm": 4.097253322601318,
      "learning_rate": 2.2002200220022003e-05,
      "loss": 0.3407,
      "step": 1220
    },
    {
      "epoch": 6.09,
      "grad_norm": 2.9143059253692627,
      "learning_rate": 2.172717271727173e-05,
      "loss": 0.3196,
      "step": 1230
    },
    {
      "epoch": 6.14,
      "grad_norm": 3.2299258708953857,
      "learning_rate": 2.1452145214521452e-05,
      "loss": 0.3093,
      "step": 1240
    },
    {
      "epoch": 6.19,
      "grad_norm": 3.449723958969116,
      "learning_rate": 2.1177117711771178e-05,
      "loss": 0.3107,
      "step": 1250
    },
    {
      "epoch": 6.24,
      "grad_norm": 4.260531425476074,
      "learning_rate": 2.09020902090209e-05,
      "loss": 0.3447,
      "step": 1260
    },
    {
      "epoch": 6.29,
      "grad_norm": 2.3418753147125244,
      "learning_rate": 2.0627062706270627e-05,
      "loss": 0.2984,
      "step": 1270
    },
    {
      "epoch": 6.34,
      "grad_norm": 2.2192769050598145,
      "learning_rate": 2.0352035203520354e-05,
      "loss": 0.2918,
      "step": 1280
    },
    {
      "epoch": 6.39,
      "grad_norm": 2.858307123184204,
      "learning_rate": 2.007700770077008e-05,
      "loss": 0.3481,
      "step": 1290
    },
    {
      "epoch": 6.44,
      "grad_norm": 2.215548038482666,
      "learning_rate": 1.9801980198019803e-05,
      "loss": 0.3107,
      "step": 1300
    },
    {
      "epoch": 6.49,
      "grad_norm": 3.317269802093506,
      "learning_rate": 1.952695269526953e-05,
      "loss": 0.3133,
      "step": 1310
    },
    {
      "epoch": 6.53,
      "grad_norm": 3.2768638134002686,
      "learning_rate": 1.9251925192519252e-05,
      "loss": 0.2935,
      "step": 1320
    },
    {
      "epoch": 6.58,
      "grad_norm": 3.264039993286133,
      "learning_rate": 1.8976897689768978e-05,
      "loss": 0.3061,
      "step": 1330
    },
    {
      "epoch": 6.63,
      "grad_norm": 4.440883159637451,
      "learning_rate": 1.87018701870187e-05,
      "loss": 0.335,
      "step": 1340
    },
    {
      "epoch": 6.68,
      "grad_norm": 2.5896713733673096,
      "learning_rate": 1.8426842684268427e-05,
      "loss": 0.3186,
      "step": 1350
    },
    {
      "epoch": 6.73,
      "grad_norm": 3.132344961166382,
      "learning_rate": 1.8151815181518153e-05,
      "loss": 0.3179,
      "step": 1360
    },
    {
      "epoch": 6.78,
      "grad_norm": 2.9426112174987793,
      "learning_rate": 1.787678767876788e-05,
      "loss": 0.3051,
      "step": 1370
    },
    {
      "epoch": 6.83,
      "grad_norm": 3.3997907638549805,
      "learning_rate": 1.7601760176017602e-05,
      "loss": 0.3032,
      "step": 1380
    },
    {
      "epoch": 6.88,
      "grad_norm": 3.3179874420166016,
      "learning_rate": 1.7326732673267325e-05,
      "loss": 0.3073,
      "step": 1390
    },
    {
      "epoch": 6.93,
      "grad_norm": 3.1379783153533936,
      "learning_rate": 1.705170517051705e-05,
      "loss": 0.3032,
      "step": 1400
    },
    {
      "epoch": 6.98,
      "grad_norm": 2.739069700241089,
      "learning_rate": 1.6776677667766778e-05,
      "loss": 0.3316,
      "step": 1410
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8672936259143156,
      "eval_loss": 0.31887274980545044,
      "eval_runtime": 34.4865,
      "eval_samples_per_second": 83.25,
      "eval_steps_per_second": 2.61,
      "step": 1414
    },
    {
      "epoch": 7.03,
      "grad_norm": 2.8164913654327393,
      "learning_rate": 1.6501650165016504e-05,
      "loss": 0.2943,
      "step": 1420
    },
    {
      "epoch": 7.08,
      "grad_norm": 2.99489688873291,
      "learning_rate": 1.6226622662266227e-05,
      "loss": 0.286,
      "step": 1430
    },
    {
      "epoch": 7.13,
      "grad_norm": 2.5896992683410645,
      "learning_rate": 1.5951595159515953e-05,
      "loss": 0.2976,
      "step": 1440
    },
    {
      "epoch": 7.18,
      "grad_norm": 2.5396080017089844,
      "learning_rate": 1.567656765676568e-05,
      "loss": 0.2813,
      "step": 1450
    },
    {
      "epoch": 7.23,
      "grad_norm": 3.6244957447052,
      "learning_rate": 1.5401540154015402e-05,
      "loss": 0.2735,
      "step": 1460
    },
    {
      "epoch": 7.28,
      "grad_norm": 3.4044344425201416,
      "learning_rate": 1.5126512651265127e-05,
      "loss": 0.288,
      "step": 1470
    },
    {
      "epoch": 7.33,
      "grad_norm": 3.5277624130249023,
      "learning_rate": 1.4851485148514851e-05,
      "loss": 0.3093,
      "step": 1480
    },
    {
      "epoch": 7.38,
      "grad_norm": 3.3465511798858643,
      "learning_rate": 1.4576457645764577e-05,
      "loss": 0.2734,
      "step": 1490
    },
    {
      "epoch": 7.43,
      "grad_norm": 3.2350246906280518,
      "learning_rate": 1.4301430143014302e-05,
      "loss": 0.2807,
      "step": 1500
    },
    {
      "epoch": 7.48,
      "grad_norm": 2.3579447269439697,
      "learning_rate": 1.4026402640264028e-05,
      "loss": 0.2724,
      "step": 1510
    },
    {
      "epoch": 7.52,
      "grad_norm": 3.0606653690338135,
      "learning_rate": 1.3751375137513753e-05,
      "loss": 0.2754,
      "step": 1520
    },
    {
      "epoch": 7.57,
      "grad_norm": 3.078380584716797,
      "learning_rate": 1.3476347634763479e-05,
      "loss": 0.2796,
      "step": 1530
    },
    {
      "epoch": 7.62,
      "grad_norm": 3.4750142097473145,
      "learning_rate": 1.32013201320132e-05,
      "loss": 0.2913,
      "step": 1540
    },
    {
      "epoch": 7.67,
      "grad_norm": 2.9180691242218018,
      "learning_rate": 1.2926292629262926e-05,
      "loss": 0.2556,
      "step": 1550
    },
    {
      "epoch": 7.72,
      "grad_norm": 3.396652936935425,
      "learning_rate": 1.265126512651265e-05,
      "loss": 0.2898,
      "step": 1560
    },
    {
      "epoch": 7.77,
      "grad_norm": 3.599966287612915,
      "learning_rate": 1.2376237623762377e-05,
      "loss": 0.2612,
      "step": 1570
    },
    {
      "epoch": 7.82,
      "grad_norm": 3.002462387084961,
      "learning_rate": 1.2101210121012102e-05,
      "loss": 0.2932,
      "step": 1580
    },
    {
      "epoch": 7.87,
      "grad_norm": 3.0669169425964355,
      "learning_rate": 1.1826182618261828e-05,
      "loss": 0.3011,
      "step": 1590
    },
    {
      "epoch": 7.92,
      "grad_norm": 2.8159451484680176,
      "learning_rate": 1.155115511551155e-05,
      "loss": 0.259,
      "step": 1600
    },
    {
      "epoch": 7.97,
      "grad_norm": 2.9513814449310303,
      "learning_rate": 1.1276127612761277e-05,
      "loss": 0.31,
      "step": 1610
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8491814698711251,
      "eval_loss": 0.380447119474411,
      "eval_runtime": 34.2815,
      "eval_samples_per_second": 83.748,
      "eval_steps_per_second": 2.625,
      "step": 1616
    },
    {
      "epoch": 8.02,
      "grad_norm": 3.6924493312835693,
      "learning_rate": 1.1001100110011001e-05,
      "loss": 0.2748,
      "step": 1620
    },
    {
      "epoch": 8.07,
      "grad_norm": 2.8759872913360596,
      "learning_rate": 1.0726072607260726e-05,
      "loss": 0.2577,
      "step": 1630
    },
    {
      "epoch": 8.12,
      "grad_norm": 4.450769424438477,
      "learning_rate": 1.045104510451045e-05,
      "loss": 0.2568,
      "step": 1640
    },
    {
      "epoch": 8.17,
      "grad_norm": 3.2912204265594482,
      "learning_rate": 1.0176017601760177e-05,
      "loss": 0.2741,
      "step": 1650
    },
    {
      "epoch": 8.22,
      "grad_norm": 3.2611083984375,
      "learning_rate": 9.900990099009901e-06,
      "loss": 0.2713,
      "step": 1660
    },
    {
      "epoch": 8.27,
      "grad_norm": 3.5394034385681152,
      "learning_rate": 9.625962596259626e-06,
      "loss": 0.2614,
      "step": 1670
    },
    {
      "epoch": 8.32,
      "grad_norm": 2.817716121673584,
      "learning_rate": 9.35093509350935e-06,
      "loss": 0.2278,
      "step": 1680
    },
    {
      "epoch": 8.37,
      "grad_norm": 2.6742143630981445,
      "learning_rate": 9.075907590759077e-06,
      "loss": 0.2568,
      "step": 1690
    },
    {
      "epoch": 8.42,
      "grad_norm": 3.2295913696289062,
      "learning_rate": 8.800880088008801e-06,
      "loss": 0.2782,
      "step": 1700
    },
    {
      "epoch": 8.47,
      "grad_norm": 2.844656467437744,
      "learning_rate": 8.525852585258526e-06,
      "loss": 0.245,
      "step": 1710
    },
    {
      "epoch": 8.51,
      "grad_norm": 3.8193488121032715,
      "learning_rate": 8.250825082508252e-06,
      "loss": 0.2474,
      "step": 1720
    },
    {
      "epoch": 8.56,
      "grad_norm": 3.309065103530884,
      "learning_rate": 7.975797579757976e-06,
      "loss": 0.2521,
      "step": 1730
    },
    {
      "epoch": 8.61,
      "grad_norm": 3.4286129474639893,
      "learning_rate": 7.700770077007701e-06,
      "loss": 0.2595,
      "step": 1740
    },
    {
      "epoch": 8.66,
      "grad_norm": 2.9357540607452393,
      "learning_rate": 7.4257425742574256e-06,
      "loss": 0.2674,
      "step": 1750
    },
    {
      "epoch": 8.71,
      "grad_norm": 2.765472650527954,
      "learning_rate": 7.150715071507151e-06,
      "loss": 0.2803,
      "step": 1760
    },
    {
      "epoch": 8.76,
      "grad_norm": 2.7337937355041504,
      "learning_rate": 6.875687568756876e-06,
      "loss": 0.2561,
      "step": 1770
    },
    {
      "epoch": 8.81,
      "grad_norm": 3.0260534286499023,
      "learning_rate": 6.6006600660066e-06,
      "loss": 0.2595,
      "step": 1780
    },
    {
      "epoch": 8.86,
      "grad_norm": 3.255251884460449,
      "learning_rate": 6.325632563256325e-06,
      "loss": 0.2571,
      "step": 1790
    },
    {
      "epoch": 8.91,
      "grad_norm": 3.837414503097534,
      "learning_rate": 6.050605060506051e-06,
      "loss": 0.2672,
      "step": 1800
    },
    {
      "epoch": 8.96,
      "grad_norm": 2.7164084911346436,
      "learning_rate": 5.775577557755775e-06,
      "loss": 0.2324,
      "step": 1810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8662486938349008,
      "eval_loss": 0.3381957411766052,
      "eval_runtime": 34.3859,
      "eval_samples_per_second": 83.493,
      "eval_steps_per_second": 2.617,
      "step": 1818
    },
    {
      "epoch": 9.01,
      "grad_norm": 3.442255735397339,
      "learning_rate": 5.500550055005501e-06,
      "loss": 0.2454,
      "step": 1820
    },
    {
      "epoch": 9.06,
      "grad_norm": 3.476314067840576,
      "learning_rate": 5.225522552255225e-06,
      "loss": 0.2488,
      "step": 1830
    },
    {
      "epoch": 9.11,
      "grad_norm": 2.7462522983551025,
      "learning_rate": 4.950495049504951e-06,
      "loss": 0.2386,
      "step": 1840
    },
    {
      "epoch": 9.16,
      "grad_norm": 4.024281978607178,
      "learning_rate": 4.675467546754675e-06,
      "loss": 0.2434,
      "step": 1850
    },
    {
      "epoch": 9.21,
      "grad_norm": 3.229743003845215,
      "learning_rate": 4.400440044004401e-06,
      "loss": 0.2382,
      "step": 1860
    },
    {
      "epoch": 9.26,
      "grad_norm": 3.476935386657715,
      "learning_rate": 4.125412541254126e-06,
      "loss": 0.2706,
      "step": 1870
    },
    {
      "epoch": 9.31,
      "grad_norm": 4.351961135864258,
      "learning_rate": 3.8503850385038505e-06,
      "loss": 0.2529,
      "step": 1880
    },
    {
      "epoch": 9.36,
      "grad_norm": 3.750527858734131,
      "learning_rate": 3.5753575357535755e-06,
      "loss": 0.2189,
      "step": 1890
    },
    {
      "epoch": 9.41,
      "grad_norm": 4.025125026702881,
      "learning_rate": 3.3003300330033e-06,
      "loss": 0.2535,
      "step": 1900
    },
    {
      "epoch": 9.46,
      "grad_norm": 3.364833354949951,
      "learning_rate": 3.0253025302530254e-06,
      "loss": 0.2475,
      "step": 1910
    },
    {
      "epoch": 9.5,
      "grad_norm": 2.76141619682312,
      "learning_rate": 2.7502750275027504e-06,
      "loss": 0.2248,
      "step": 1920
    },
    {
      "epoch": 9.55,
      "grad_norm": 3.184504270553589,
      "learning_rate": 2.4752475247524753e-06,
      "loss": 0.2374,
      "step": 1930
    },
    {
      "epoch": 9.6,
      "grad_norm": 3.67197585105896,
      "learning_rate": 2.2002200220022003e-06,
      "loss": 0.2336,
      "step": 1940
    },
    {
      "epoch": 9.65,
      "grad_norm": 3.4729063510894775,
      "learning_rate": 1.9251925192519253e-06,
      "loss": 0.2345,
      "step": 1950
    },
    {
      "epoch": 9.7,
      "grad_norm": 2.817584991455078,
      "learning_rate": 1.65016501650165e-06,
      "loss": 0.2057,
      "step": 1960
    },
    {
      "epoch": 9.75,
      "grad_norm": 3.606154441833496,
      "learning_rate": 1.3751375137513752e-06,
      "loss": 0.2526,
      "step": 1970
    },
    {
      "epoch": 9.8,
      "grad_norm": 3.766678810119629,
      "learning_rate": 1.1001100110011001e-06,
      "loss": 0.2465,
      "step": 1980
    },
    {
      "epoch": 9.85,
      "grad_norm": 2.8478949069976807,
      "learning_rate": 8.25082508250825e-07,
      "loss": 0.2305,
      "step": 1990
    },
    {
      "epoch": 9.9,
      "grad_norm": 3.010897159576416,
      "learning_rate": 5.500550055005501e-07,
      "loss": 0.2269,
      "step": 2000
    },
    {
      "epoch": 9.95,
      "grad_norm": 2.9162042140960693,
      "learning_rate": 2.7502750275027504e-07,
      "loss": 0.2408,
      "step": 2010
    },
    {
      "epoch": 10.0,
      "grad_norm": 5.191891670227051,
      "learning_rate": 0.0,
      "loss": 0.234,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8707767328456983,
      "eval_loss": 0.3384123742580414,
      "eval_runtime": 34.2477,
      "eval_samples_per_second": 83.83,
      "eval_steps_per_second": 2.628,
      "step": 2020
    },
    {
      "epoch": 10.0,
      "step": 2020,
      "total_flos": 6.733106817319895e+19,
      "train_loss": 0.3648681934517209,
      "train_runtime": 8202.686,
      "train_samples_per_second": 31.498,
      "train_steps_per_second": 0.246
    }
  ],
  "logging_steps": 10,
  "max_steps": 2020,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 6.733106817319895e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}