{
  "best_metric": 0.9130434782608695,
  "best_model_checkpoint": "vit-base-patch16-224-ve-U13b-80RX3\\checkpoint-875",
  "epoch": 39.61165048543689,
  "eval_steps": 500,
  "global_step": 2040,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.19,
      "learning_rate": 4.647058823529412e-06,
      "loss": 1.3862,
      "step": 10
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.294117647058824e-06,
      "loss": 1.3855,
      "step": 20
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.3941176470588236e-05,
      "loss": 1.3804,
      "step": 30
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.8588235294117647e-05,
      "loss": 1.3673,
      "step": 40
    },
    {
      "epoch": 0.97,
      "learning_rate": 2.3235294117647057e-05,
      "loss": 1.33,
      "step": 50
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.34782608695652173,
      "eval_loss": 1.3133331537246704,
      "eval_runtime": 1.067,
      "eval_samples_per_second": 43.112,
      "eval_steps_per_second": 5.623,
      "step": 51
    },
    {
      "epoch": 1.17,
      "learning_rate": 2.7882352941176473e-05,
      "loss": 1.2898,
      "step": 60
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.252941176470588e-05,
      "loss": 1.213,
      "step": 70
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.7176470588235295e-05,
      "loss": 1.1789,
      "step": 80
    },
    {
      "epoch": 1.75,
      "learning_rate": 4.182352941176471e-05,
      "loss": 1.0432,
      "step": 90
    },
    {
      "epoch": 1.94,
      "learning_rate": 4.6470588235294114e-05,
      "loss": 1.0288,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 1.0045453310012817,
      "eval_runtime": 0.9508,
      "eval_samples_per_second": 48.379,
      "eval_steps_per_second": 6.31,
      "step": 103
    },
    {
      "epoch": 2.14,
      "learning_rate": 4.720433436532508e-05,
      "loss": 0.9473,
      "step": 110
    },
    {
      "epoch": 2.33,
      "learning_rate": 4.695975232198143e-05,
      "loss": 0.9482,
      "step": 120
    },
    {
      "epoch": 2.52,
      "learning_rate": 4.671517027863777e-05,
      "loss": 0.8323,
      "step": 130
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.6470588235294114e-05,
      "loss": 0.7445,
      "step": 140
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.6226006191950463e-05,
      "loss": 0.7322,
      "step": 150
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.7308945059776306,
      "eval_runtime": 0.826,
      "eval_samples_per_second": 55.688,
      "eval_steps_per_second": 7.264,
      "step": 154
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.5981424148606813e-05,
      "loss": 0.6374,
      "step": 160
    },
    {
      "epoch": 3.3,
      "learning_rate": 4.5736842105263157e-05,
      "loss": 0.5266,
      "step": 170
    },
    {
      "epoch": 3.5,
      "learning_rate": 4.5492260061919506e-05,
      "loss": 0.5564,
      "step": 180
    },
    {
      "epoch": 3.69,
      "learning_rate": 4.524767801857585e-05,
      "loss": 0.537,
      "step": 190
    },
    {
      "epoch": 3.88,
      "learning_rate": 4.50030959752322e-05,
      "loss": 0.5476,
      "step": 200
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.6316179633140564,
      "eval_runtime": 0.8212,
      "eval_samples_per_second": 56.017,
      "eval_steps_per_second": 7.307,
      "step": 206
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.475851393188855e-05,
      "loss": 0.405,
      "step": 210
    },
    {
      "epoch": 4.27,
      "learning_rate": 4.451393188854489e-05,
      "loss": 0.3837,
      "step": 220
    },
    {
      "epoch": 4.47,
      "learning_rate": 4.4269349845201236e-05,
      "loss": 0.4488,
      "step": 230
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.4024767801857586e-05,
      "loss": 0.3847,
      "step": 240
    },
    {
      "epoch": 4.85,
      "learning_rate": 4.3780185758513935e-05,
      "loss": 0.2863,
      "step": 250
    },
    {
      "epoch": 4.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.5598096251487732,
      "eval_runtime": 0.8302,
      "eval_samples_per_second": 55.408,
      "eval_steps_per_second": 7.227,
      "step": 257
    },
    {
      "epoch": 5.05,
      "learning_rate": 4.353560371517028e-05,
      "loss": 0.3397,
      "step": 260
    },
    {
      "epoch": 5.24,
      "learning_rate": 4.329102167182663e-05,
      "loss": 0.2941,
      "step": 270
    },
    {
      "epoch": 5.44,
      "learning_rate": 4.304643962848297e-05,
      "loss": 0.288,
      "step": 280
    },
    {
      "epoch": 5.63,
      "learning_rate": 4.2801857585139315e-05,
      "loss": 0.3423,
      "step": 290
    },
    {
      "epoch": 5.83,
      "learning_rate": 4.2557275541795665e-05,
      "loss": 0.3149,
      "step": 300
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.5427684783935547,
      "eval_runtime": 0.8152,
      "eval_samples_per_second": 56.427,
      "eval_steps_per_second": 7.36,
      "step": 309
    },
    {
      "epoch": 6.02,
      "learning_rate": 4.2312693498452015e-05,
      "loss": 0.2982,
      "step": 310
    },
    {
      "epoch": 6.21,
      "learning_rate": 4.206811145510836e-05,
      "loss": 0.2823,
      "step": 320
    },
    {
      "epoch": 6.41,
      "learning_rate": 4.182352941176471e-05,
      "loss": 0.2447,
      "step": 330
    },
    {
      "epoch": 6.6,
      "learning_rate": 4.157894736842105e-05,
      "loss": 0.3232,
      "step": 340
    },
    {
      "epoch": 6.8,
      "learning_rate": 4.13343653250774e-05,
      "loss": 0.3179,
      "step": 350
    },
    {
      "epoch": 6.99,
      "learning_rate": 4.1089783281733744e-05,
      "loss": 0.1489,
      "step": 360
    },
    {
      "epoch": 6.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.5149533152580261,
      "eval_runtime": 0.7717,
      "eval_samples_per_second": 59.609,
      "eval_steps_per_second": 7.775,
      "step": 360
    },
    {
      "epoch": 7.18,
      "learning_rate": 4.0845201238390094e-05,
      "loss": 0.1946,
      "step": 370
    },
    {
      "epoch": 7.38,
      "learning_rate": 4.060061919504644e-05,
      "loss": 0.1965,
      "step": 380
    },
    {
      "epoch": 7.57,
      "learning_rate": 4.035603715170279e-05,
      "loss": 0.2141,
      "step": 390
    },
    {
      "epoch": 7.77,
      "learning_rate": 4.011145510835914e-05,
      "loss": 0.3889,
      "step": 400
    },
    {
      "epoch": 7.96,
      "learning_rate": 3.986687306501548e-05,
      "loss": 0.1134,
      "step": 410
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.45851370692253113,
      "eval_runtime": 0.8807,
      "eval_samples_per_second": 52.231,
      "eval_steps_per_second": 6.813,
      "step": 412
    },
    {
      "epoch": 8.16,
      "learning_rate": 3.962229102167183e-05,
      "loss": 0.1778,
      "step": 420
    },
    {
      "epoch": 8.35,
      "learning_rate": 3.937770897832817e-05,
      "loss": 0.1464,
      "step": 430
    },
    {
      "epoch": 8.54,
      "learning_rate": 3.9133126934984516e-05,
      "loss": 0.1724,
      "step": 440
    },
    {
      "epoch": 8.74,
      "learning_rate": 3.8888544891640866e-05,
      "loss": 0.1663,
      "step": 450
    },
    {
      "epoch": 8.93,
      "learning_rate": 3.8643962848297216e-05,
      "loss": 0.1613,
      "step": 460
    },
    {
      "epoch": 8.99,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.6283782720565796,
      "eval_runtime": 0.8873,
      "eval_samples_per_second": 51.841,
      "eval_steps_per_second": 6.762,
      "step": 463
    },
    {
      "epoch": 9.13,
      "learning_rate": 3.839938080495356e-05,
      "loss": 0.1331,
      "step": 470
    },
    {
      "epoch": 9.32,
      "learning_rate": 3.815479876160991e-05,
      "loss": 0.1595,
      "step": 480
    },
    {
      "epoch": 9.51,
      "learning_rate": 3.791021671826625e-05,
      "loss": 0.2071,
      "step": 490
    },
    {
      "epoch": 9.71,
      "learning_rate": 3.76656346749226e-05,
      "loss": 0.11,
      "step": 500
    },
    {
      "epoch": 9.9,
      "learning_rate": 3.7421052631578945e-05,
      "loss": 0.1855,
      "step": 510
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.5985141396522522,
      "eval_runtime": 0.8537,
      "eval_samples_per_second": 53.882,
      "eval_steps_per_second": 7.028,
      "step": 515
    },
    {
      "epoch": 10.1,
      "learning_rate": 3.7176470588235295e-05,
      "loss": 0.0823,
      "step": 520
    },
    {
      "epoch": 10.29,
      "learning_rate": 3.693188854489164e-05,
      "loss": 0.1198,
      "step": 530
    },
    {
      "epoch": 10.49,
      "learning_rate": 3.668730650154799e-05,
      "loss": 0.1222,
      "step": 540
    },
    {
      "epoch": 10.68,
      "learning_rate": 3.644272445820434e-05,
      "loss": 0.1218,
      "step": 550
    },
    {
      "epoch": 10.87,
      "learning_rate": 3.619814241486068e-05,
      "loss": 0.1908,
      "step": 560
    },
    {
      "epoch": 10.99,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.0336034297943115,
      "eval_runtime": 0.844,
      "eval_samples_per_second": 54.503,
      "eval_steps_per_second": 7.109,
      "step": 566
    },
    {
      "epoch": 11.07,
      "learning_rate": 3.5953560371517024e-05,
      "loss": 0.201,
      "step": 570
    },
    {
      "epoch": 11.26,
      "learning_rate": 3.5708978328173374e-05,
      "loss": 0.1526,
      "step": 580
    },
    {
      "epoch": 11.46,
      "learning_rate": 3.5464396284829724e-05,
      "loss": 0.0184,
      "step": 590
    },
    {
      "epoch": 11.65,
      "learning_rate": 3.521981424148607e-05,
      "loss": 0.0832,
      "step": 600
    },
    {
      "epoch": 11.84,
      "learning_rate": 3.497523219814242e-05,
      "loss": 0.2293,
      "step": 610
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.7745507955551147,
      "eval_runtime": 0.8319,
      "eval_samples_per_second": 55.297,
      "eval_steps_per_second": 7.213,
      "step": 618
    },
    {
      "epoch": 12.04,
      "learning_rate": 3.473065015479876e-05,
      "loss": 0.1271,
      "step": 620
    },
    {
      "epoch": 12.23,
      "learning_rate": 3.448606811145511e-05,
      "loss": 0.1235,
      "step": 630
    },
    {
      "epoch": 12.43,
      "learning_rate": 3.424148606811146e-05,
      "loss": 0.1258,
      "step": 640
    },
    {
      "epoch": 12.62,
      "learning_rate": 3.39969040247678e-05,
      "loss": 0.0889,
      "step": 650
    },
    {
      "epoch": 12.82,
      "learning_rate": 3.3752321981424146e-05,
      "loss": 0.1414,
      "step": 660
    },
    {
      "epoch": 12.99,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.6517351865768433,
      "eval_runtime": 0.9004,
      "eval_samples_per_second": 51.091,
      "eval_steps_per_second": 6.664,
      "step": 669
    },
    {
      "epoch": 13.01,
      "learning_rate": 3.3507739938080496e-05,
      "loss": 0.0693,
      "step": 670
    },
    {
      "epoch": 13.2,
      "learning_rate": 3.326315789473684e-05,
      "loss": 0.0435,
      "step": 680
    },
    {
      "epoch": 13.4,
      "learning_rate": 3.301857585139319e-05,
      "loss": 0.2878,
      "step": 690
    },
    {
      "epoch": 13.59,
      "learning_rate": 3.277399380804954e-05,
      "loss": 0.1721,
      "step": 700
    },
    {
      "epoch": 13.79,
      "learning_rate": 3.252941176470588e-05,
      "loss": 0.1421,
      "step": 710
    },
    {
      "epoch": 13.98,
      "learning_rate": 3.2284829721362225e-05,
      "loss": 0.0877,
      "step": 720
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.563931941986084,
      "eval_runtime": 0.8391,
      "eval_samples_per_second": 54.824,
      "eval_steps_per_second": 7.151,
      "step": 721
    },
    {
      "epoch": 14.17,
      "learning_rate": 3.2040247678018575e-05,
      "loss": 0.0886,
      "step": 730
    },
    {
      "epoch": 14.37,
      "learning_rate": 3.1795665634674925e-05,
      "loss": 0.1254,
      "step": 740
    },
    {
      "epoch": 14.56,
      "learning_rate": 3.155108359133127e-05,
      "loss": 0.0781,
      "step": 750
    },
    {
      "epoch": 14.76,
      "learning_rate": 3.130650154798762e-05,
      "loss": 0.1242,
      "step": 760
    },
    {
      "epoch": 14.95,
      "learning_rate": 3.106191950464396e-05,
      "loss": 0.1302,
      "step": 770
    },
    {
      "epoch": 14.99,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.7687188982963562,
      "eval_runtime": 0.8747,
      "eval_samples_per_second": 52.587,
      "eval_steps_per_second": 6.859,
      "step": 772
    },
    {
      "epoch": 15.15,
      "learning_rate": 3.0817337461300304e-05,
      "loss": 0.0822,
      "step": 780
    },
    {
      "epoch": 15.34,
      "learning_rate": 3.057275541795666e-05,
      "loss": 0.0589,
      "step": 790
    },
    {
      "epoch": 15.53,
      "learning_rate": 3.0328173374613004e-05,
      "loss": 0.0808,
      "step": 800
    },
    {
      "epoch": 15.73,
      "learning_rate": 3.008359133126935e-05,
      "loss": 0.0589,
      "step": 810
    },
    {
      "epoch": 15.92,
      "learning_rate": 2.9839009287925697e-05,
      "loss": 0.047,
      "step": 820
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.6772664189338684,
      "eval_runtime": 0.827,
      "eval_samples_per_second": 55.621,
      "eval_steps_per_second": 7.255,
      "step": 824
    },
    {
      "epoch": 16.12,
      "learning_rate": 2.959442724458204e-05,
      "loss": 0.1881,
      "step": 830
    },
    {
      "epoch": 16.31,
      "learning_rate": 2.9349845201238394e-05,
      "loss": 0.0418,
      "step": 840
    },
    {
      "epoch": 16.5,
      "learning_rate": 2.9105263157894737e-05,
      "loss": 0.0664,
      "step": 850
    },
    {
      "epoch": 16.7,
      "learning_rate": 2.8860681114551083e-05,
      "loss": 0.1497,
      "step": 860
    },
    {
      "epoch": 16.89,
      "learning_rate": 2.861609907120743e-05,
      "loss": 0.1045,
      "step": 870
    },
    {
      "epoch": 16.99,
      "eval_accuracy": 0.9130434782608695,
      "eval_loss": 0.4344385862350464,
      "eval_runtime": 0.8479,
      "eval_samples_per_second": 54.255,
      "eval_steps_per_second": 7.077,
      "step": 875
    },
    {
      "epoch": 17.09,
      "learning_rate": 2.8371517027863776e-05,
      "loss": 0.1355,
      "step": 880
    },
    {
      "epoch": 17.28,
      "learning_rate": 2.8126934984520126e-05,
      "loss": 0.1205,
      "step": 890
    },
    {
      "epoch": 17.48,
      "learning_rate": 2.7882352941176473e-05,
      "loss": 0.0984,
      "step": 900
    },
    {
      "epoch": 17.67,
      "learning_rate": 2.7637770897832816e-05,
      "loss": 0.1224,
      "step": 910
    },
    {
      "epoch": 17.86,
      "learning_rate": 2.7393188854489163e-05,
      "loss": 0.0751,
      "step": 920
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.0160393714904785,
      "eval_runtime": 0.8692,
      "eval_samples_per_second": 52.922,
      "eval_steps_per_second": 6.903,
      "step": 927
    },
    {
      "epoch": 18.06,
      "learning_rate": 2.714860681114551e-05,
      "loss": 0.0833,
      "step": 930
    },
    {
      "epoch": 18.25,
      "learning_rate": 2.690402476780186e-05,
      "loss": 0.0335,
      "step": 940
    },
    {
      "epoch": 18.45,
      "learning_rate": 2.6659442724458205e-05,
      "loss": 0.0702,
      "step": 950
    },
    {
      "epoch": 18.64,
      "learning_rate": 2.6414860681114552e-05,
      "loss": 0.0391,
      "step": 960
    },
    {
      "epoch": 18.83,
      "learning_rate": 2.6170278637770895e-05,
      "loss": 0.1141,
      "step": 970
    },
    {
      "epoch": 18.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.6642520427703857,
      "eval_runtime": 0.9401,
      "eval_samples_per_second": 48.929,
      "eval_steps_per_second": 6.382,
      "step": 978
    },
    {
      "epoch": 19.03,
      "learning_rate": 2.592569659442724e-05,
      "loss": 0.0581,
      "step": 980
    },
    {
      "epoch": 19.22,
      "learning_rate": 2.568111455108359e-05,
      "loss": 0.0739,
      "step": 990
    },
    {
      "epoch": 19.42,
      "learning_rate": 2.5436532507739938e-05,
      "loss": 0.0566,
      "step": 1000
    },
    {
      "epoch": 19.61,
      "learning_rate": 2.5191950464396285e-05,
      "loss": 0.2267,
      "step": 1010
    },
    {
      "epoch": 19.81,
      "learning_rate": 2.494736842105263e-05,
      "loss": 0.0873,
      "step": 1020
    },
    {
      "epoch": 20.0,
      "learning_rate": 2.4702786377708978e-05,
      "loss": 0.1756,
      "step": 1030
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8913043478260869,
      "eval_loss": 0.5581848621368408,
      "eval_runtime": 0.8076,
      "eval_samples_per_second": 56.958,
      "eval_steps_per_second": 7.429,
      "step": 1030
    },
    {
      "epoch": 20.19,
      "learning_rate": 2.4458204334365328e-05,
      "loss": 0.0867,
      "step": 1040
    },
    {
      "epoch": 20.39,
      "learning_rate": 2.4213622291021674e-05,
      "loss": 0.0911,
      "step": 1050
    },
    {
      "epoch": 20.58,
      "learning_rate": 2.3969040247678017e-05,
      "loss": 0.1337,
      "step": 1060
    },
    {
      "epoch": 20.78,
      "learning_rate": 2.3724458204334364e-05,
      "loss": 0.1217,
      "step": 1070
    },
    {
      "epoch": 20.97,
      "learning_rate": 2.3479876160990714e-05,
      "loss": 0.1212,
      "step": 1080
    },
    {
      "epoch": 20.99,
      "eval_accuracy": 0.8913043478260869,
      "eval_loss": 0.5640941858291626,
      "eval_runtime": 0.8177,
      "eval_samples_per_second": 56.253,
      "eval_steps_per_second": 7.337,
      "step": 1081
    },
    {
      "epoch": 21.17,
      "learning_rate": 2.3235294117647057e-05,
      "loss": 0.1123,
      "step": 1090
    },
    {
      "epoch": 21.36,
      "learning_rate": 2.2990712074303407e-05,
      "loss": 0.063,
      "step": 1100
    },
    {
      "epoch": 21.55,
      "learning_rate": 2.2746130030959753e-05,
      "loss": 0.0782,
      "step": 1110
    },
    {
      "epoch": 21.75,
      "learning_rate": 2.25015479876161e-05,
      "loss": 0.0588,
      "step": 1120
    },
    {
      "epoch": 21.94,
      "learning_rate": 2.2256965944272446e-05,
      "loss": 0.0903,
      "step": 1130
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.698965847492218,
      "eval_runtime": 0.827,
      "eval_samples_per_second": 55.62,
      "eval_steps_per_second": 7.255,
      "step": 1133
    },
    {
      "epoch": 22.14,
      "learning_rate": 2.2012383900928793e-05,
      "loss": 0.0266,
      "step": 1140
    },
    {
      "epoch": 22.33,
      "learning_rate": 2.176780185758514e-05,
      "loss": 0.0578,
      "step": 1150
    },
    {
      "epoch": 22.52,
      "learning_rate": 2.1523219814241486e-05,
      "loss": 0.0606,
      "step": 1160
    },
    {
      "epoch": 22.72,
      "learning_rate": 2.1278637770897832e-05,
      "loss": 0.0809,
      "step": 1170
    },
    {
      "epoch": 22.91,
      "learning_rate": 2.103405572755418e-05,
      "loss": 0.0693,
      "step": 1180
    },
    {
      "epoch": 22.99,
      "eval_accuracy": 0.8913043478260869,
      "eval_loss": 0.5548133254051208,
      "eval_runtime": 0.8892,
      "eval_samples_per_second": 51.732,
      "eval_steps_per_second": 6.748,
      "step": 1184
    },
    {
      "epoch": 23.11,
      "learning_rate": 2.0789473684210525e-05,
      "loss": 0.0246,
      "step": 1190
    },
    {
      "epoch": 23.3,
      "learning_rate": 2.0544891640866872e-05,
      "loss": 0.0926,
      "step": 1200
    },
    {
      "epoch": 23.5,
      "learning_rate": 2.030030959752322e-05,
      "loss": 0.0344,
      "step": 1210
    },
    {
      "epoch": 23.69,
      "learning_rate": 2.005572755417957e-05,
      "loss": 0.0526,
      "step": 1220
    },
    {
      "epoch": 23.88,
      "learning_rate": 1.9811145510835915e-05,
      "loss": 0.0048,
      "step": 1230
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.6957914233207703,
      "eval_runtime": 0.8992,
      "eval_samples_per_second": 51.155,
      "eval_steps_per_second": 6.672,
      "step": 1236
    },
    {
      "epoch": 24.08,
      "learning_rate": 1.9566563467492258e-05,
      "loss": 0.0412,
      "step": 1240
    },
    {
      "epoch": 24.27,
      "learning_rate": 1.9321981424148608e-05,
      "loss": 0.1009,
      "step": 1250
    },
    {
      "epoch": 24.47,
      "learning_rate": 1.9077399380804954e-05,
      "loss": 0.1074,
      "step": 1260
    },
    {
      "epoch": 24.66,
      "learning_rate": 1.88328173374613e-05,
      "loss": 0.1063,
      "step": 1270
    },
    {
      "epoch": 24.85,
      "learning_rate": 1.8588235294117647e-05,
      "loss": 0.0785,
      "step": 1280
    },
    {
      "epoch": 24.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.7886456251144409,
      "eval_runtime": 0.7742,
      "eval_samples_per_second": 59.417,
      "eval_steps_per_second": 7.75,
      "step": 1287
    },
    {
      "epoch": 25.05,
      "learning_rate": 1.8343653250773994e-05,
      "loss": 0.0349,
      "step": 1290
    },
    {
      "epoch": 25.24,
      "learning_rate": 1.809907120743034e-05,
      "loss": 0.1957,
      "step": 1300
    },
    {
      "epoch": 25.44,
      "learning_rate": 1.7854489164086687e-05,
      "loss": 0.0344,
      "step": 1310
    },
    {
      "epoch": 25.63,
      "learning_rate": 1.7609907120743034e-05,
      "loss": 0.1233,
      "step": 1320
    },
    {
      "epoch": 25.83,
      "learning_rate": 1.736532507739938e-05,
      "loss": 0.0373,
      "step": 1330
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.6345128417015076,
      "eval_runtime": 0.8942,
      "eval_samples_per_second": 51.442,
      "eval_steps_per_second": 6.71,
      "step": 1339
    },
    {
      "epoch": 26.02,
      "learning_rate": 1.712074303405573e-05,
      "loss": 0.0051,
      "step": 1340
    },
    {
      "epoch": 26.21,
      "learning_rate": 1.6876160990712073e-05,
      "loss": 0.1126,
      "step": 1350
    },
    {
      "epoch": 26.41,
      "learning_rate": 1.663157894736842e-05,
      "loss": 0.0579,
      "step": 1360
    },
    {
      "epoch": 26.6,
      "learning_rate": 1.638699690402477e-05,
      "loss": 0.0228,
      "step": 1370
    },
    {
      "epoch": 26.8,
      "learning_rate": 1.6142414860681113e-05,
      "loss": 0.0734,
      "step": 1380
    },
    {
      "epoch": 26.99,
      "learning_rate": 1.5897832817337463e-05,
      "loss": 0.0763,
      "step": 1390
    },
    {
      "epoch": 26.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.6829540729522705,
      "eval_runtime": 0.8532,
      "eval_samples_per_second": 53.914,
      "eval_steps_per_second": 7.032,
      "step": 1390
    },
    {
      "epoch": 27.18,
      "learning_rate": 1.565325077399381e-05,
      "loss": 0.0457,
      "step": 1400
    },
    {
      "epoch": 27.38,
      "learning_rate": 1.5408668730650152e-05,
      "loss": 0.1175,
      "step": 1410
    },
    {
      "epoch": 27.57,
      "learning_rate": 1.5164086687306502e-05,
      "loss": 0.0137,
      "step": 1420
    },
    {
      "epoch": 27.77,
      "learning_rate": 1.4919504643962849e-05,
      "loss": 0.0147,
      "step": 1430
    },
    {
      "epoch": 27.96,
      "learning_rate": 1.4674922600619197e-05,
      "loss": 0.0621,
      "step": 1440
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.7293574213981628,
      "eval_runtime": 0.9042,
      "eval_samples_per_second": 50.873,
      "eval_steps_per_second": 6.636,
      "step": 1442
    },
    {
      "epoch": 28.16,
      "learning_rate": 1.4430340557275542e-05,
      "loss": 0.071,
      "step": 1450
    },
    {
      "epoch": 28.35,
      "learning_rate": 1.4185758513931888e-05,
      "loss": 0.0818,
      "step": 1460
    },
    {
      "epoch": 28.54,
      "learning_rate": 1.3941176470588236e-05,
      "loss": 0.009,
      "step": 1470
    },
    {
      "epoch": 28.74,
      "learning_rate": 1.3696594427244581e-05,
      "loss": 0.0417,
      "step": 1480
    },
    {
      "epoch": 28.93,
      "learning_rate": 1.345201238390093e-05,
      "loss": 0.0367,
      "step": 1490
    },
    {
      "epoch": 28.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.6636242866516113,
      "eval_runtime": 0.8167,
      "eval_samples_per_second": 56.324,
      "eval_steps_per_second": 7.347,
      "step": 1493
    },
    {
      "epoch": 29.13,
      "learning_rate": 1.3207430340557276e-05,
      "loss": 0.0583,
      "step": 1500
    },
    {
      "epoch": 29.32,
      "learning_rate": 1.296284829721362e-05,
      "loss": 0.0745,
      "step": 1510
    },
    {
      "epoch": 29.51,
      "learning_rate": 1.2718266253869969e-05,
      "loss": 0.0502,
      "step": 1520
    },
    {
      "epoch": 29.71,
      "learning_rate": 1.2473684210526316e-05,
      "loss": 0.0385,
      "step": 1530
    },
    {
      "epoch": 29.9,
      "learning_rate": 1.2229102167182664e-05,
      "loss": 0.0124,
      "step": 1540
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.8030955195426941,
      "eval_runtime": 0.7682,
      "eval_samples_per_second": 59.881,
      "eval_steps_per_second": 7.811,
      "step": 1545
    },
    {
      "epoch": 30.1,
      "learning_rate": 1.1984520123839009e-05,
      "loss": 0.0158,
      "step": 1550
    },
    {
      "epoch": 30.29,
      "learning_rate": 1.1739938080495357e-05,
      "loss": 0.0406,
      "step": 1560
    },
    {
      "epoch": 30.49,
      "learning_rate": 1.1495356037151703e-05,
      "loss": 0.1925,
      "step": 1570
    },
    {
      "epoch": 30.68,
      "learning_rate": 1.125077399380805e-05,
      "loss": 0.0512,
      "step": 1580
    },
    {
      "epoch": 30.87,
      "learning_rate": 1.1006191950464396e-05,
      "loss": 0.0759,
      "step": 1590
    },
    {
      "epoch": 30.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.7075828313827515,
      "eval_runtime": 0.8152,
      "eval_samples_per_second": 56.429,
      "eval_steps_per_second": 7.36,
      "step": 1596
    },
    {
      "epoch": 31.07,
      "learning_rate": 1.0761609907120743e-05,
      "loss": 0.0585,
      "step": 1600
    },
    {
      "epoch": 31.26,
      "learning_rate": 1.051702786377709e-05,
      "loss": 0.0113,
      "step": 1610
    },
    {
      "epoch": 31.46,
      "learning_rate": 1.0272445820433436e-05,
      "loss": 0.0053,
      "step": 1620
    },
    {
      "epoch": 31.65,
      "learning_rate": 1.0027863777089784e-05,
      "loss": 0.0227,
      "step": 1630
    },
    {
      "epoch": 31.84,
      "learning_rate": 9.783281733746129e-06,
      "loss": 0.0786,
      "step": 1640
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.8023592829704285,
      "eval_runtime": 0.8222,
      "eval_samples_per_second": 55.947,
      "eval_steps_per_second": 7.297,
      "step": 1648
    },
    {
      "epoch": 32.04,
      "learning_rate": 9.538699690402477e-06,
      "loss": 0.004,
      "step": 1650
    },
    {
      "epoch": 32.23,
      "learning_rate": 9.294117647058824e-06,
      "loss": 0.0131,
      "step": 1660
    },
    {
      "epoch": 32.43,
      "learning_rate": 9.04953560371517e-06,
      "loss": 0.0616,
      "step": 1670
    },
    {
      "epoch": 32.62,
      "learning_rate": 8.804953560371517e-06,
      "loss": 0.0058,
      "step": 1680
    },
    {
      "epoch": 32.82,
      "learning_rate": 8.560371517027865e-06,
      "loss": 0.0487,
      "step": 1690
    },
    {
      "epoch": 32.99,
      "eval_accuracy": 0.8695652173913043,
      "eval_loss": 0.7926502227783203,
      "eval_runtime": 0.8442,
      "eval_samples_per_second": 54.49,
      "eval_steps_per_second": 7.107,
      "step": 1699
    },
    {
      "epoch": 33.01,
      "learning_rate": 8.31578947368421e-06,
      "loss": 0.1104,
      "step": 1700
    },
    {
      "epoch": 33.2,
      "learning_rate": 8.071207430340556e-06,
      "loss": 0.0528,
      "step": 1710
    },
    {
      "epoch": 33.4,
      "learning_rate": 7.826625386996905e-06,
      "loss": 0.056,
      "step": 1720
    },
    {
      "epoch": 33.59,
      "learning_rate": 7.582043343653251e-06,
      "loss": 0.0205,
      "step": 1730
    },
    {
      "epoch": 33.79,
      "learning_rate": 7.3374613003095984e-06,
      "loss": 0.0431,
      "step": 1740
    },
    {
      "epoch": 33.98,
      "learning_rate": 7.092879256965944e-06,
      "loss": 0.0664,
      "step": 1750
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.9607385396957397,
      "eval_runtime": 0.8292,
      "eval_samples_per_second": 55.474,
      "eval_steps_per_second": 7.236,
      "step": 1751
    },
    {
      "epoch": 34.17,
      "learning_rate": 6.848297213622291e-06,
      "loss": 0.0596,
      "step": 1760
    },
    {
      "epoch": 34.37,
      "learning_rate": 6.603715170278638e-06,
      "loss": 0.0592,
      "step": 1770
    },
    {
      "epoch": 34.56,
      "learning_rate": 6.3591331269349845e-06,
      "loss": 0.0469,
      "step": 1780
    },
    {
      "epoch": 34.76,
      "learning_rate": 6.114551083591332e-06,
      "loss": 0.0376,
      "step": 1790
    },
    {
      "epoch": 34.95,
      "learning_rate": 5.869969040247678e-06,
      "loss": 0.0054,
      "step": 1800
    },
    {
      "epoch": 34.99,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.9701578617095947,
      "eval_runtime": 0.8192,
      "eval_samples_per_second": 56.152,
      "eval_steps_per_second": 7.324,
      "step": 1802
    },
    {
      "epoch": 35.15,
      "learning_rate": 5.625386996904025e-06,
      "loss": 0.0512,
      "step": 1810
    },
    {
      "epoch": 35.34,
      "learning_rate": 5.3808049535603715e-06,
      "loss": 0.0149,
      "step": 1820
    },
    {
      "epoch": 35.53,
      "learning_rate": 5.136222910216718e-06,
      "loss": 0.0083,
      "step": 1830
    },
    {
      "epoch": 35.73,
      "learning_rate": 4.8916408668730645e-06,
      "loss": 0.0085,
      "step": 1840
    },
    {
      "epoch": 35.92,
      "learning_rate": 4.647058823529412e-06,
      "loss": 0.0277,
      "step": 1850
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.8351179957389832,
      "eval_runtime": 0.8282,
      "eval_samples_per_second": 55.543,
      "eval_steps_per_second": 7.245,
      "step": 1854
    },
    {
      "epoch": 36.12,
      "learning_rate": 4.402476780185758e-06,
      "loss": 0.0365,
      "step": 1860
    },
    {
      "epoch": 36.31,
      "learning_rate": 4.157894736842105e-06,
      "loss": 0.0119,
      "step": 1870
    },
    {
      "epoch": 36.5,
      "learning_rate": 3.913312693498452e-06,
      "loss": 0.0608,
      "step": 1880
    },
    {
      "epoch": 36.7,
      "learning_rate": 3.6687306501547992e-06,
      "loss": 0.135,
      "step": 1890
    },
    {
      "epoch": 36.89,
      "learning_rate": 3.4241486068111453e-06,
      "loss": 0.0025,
      "step": 1900
    },
    {
      "epoch": 36.99,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.9318453073501587,
      "eval_runtime": 1.1833,
      "eval_samples_per_second": 38.875,
      "eval_steps_per_second": 5.071,
      "step": 1905
    },
    {
      "epoch": 37.09,
      "learning_rate": 3.1795665634674923e-06,
      "loss": 0.027,
      "step": 1910
    },
    {
      "epoch": 37.28,
      "learning_rate": 2.934984520123839e-06,
      "loss": 0.0794,
      "step": 1920
    },
    {
      "epoch": 37.48,
      "learning_rate": 2.6904024767801857e-06,
      "loss": 0.0461,
      "step": 1930
    },
    {
      "epoch": 37.67,
      "learning_rate": 2.4458204334365322e-06,
      "loss": 0.0464,
      "step": 1940
    },
    {
      "epoch": 37.86,
      "learning_rate": 2.201238390092879e-06,
      "loss": 0.0188,
      "step": 1950
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.899507462978363,
      "eval_runtime": 0.8587,
      "eval_samples_per_second": 53.568,
      "eval_steps_per_second": 6.987,
      "step": 1957
    },
    {
      "epoch": 38.06,
      "learning_rate": 1.956656346749226e-06,
      "loss": 0.0413,
      "step": 1960
    },
    {
      "epoch": 38.25,
      "learning_rate": 1.7120743034055727e-06,
      "loss": 0.0481,
      "step": 1970
    },
    {
      "epoch": 38.45,
      "learning_rate": 1.4674922600619196e-06,
      "loss": 0.0494,
      "step": 1980
    },
    {
      "epoch": 38.64,
      "learning_rate": 1.2229102167182661e-06,
      "loss": 0.0201,
      "step": 1990
    },
    {
      "epoch": 38.83,
      "learning_rate": 9.78328173374613e-07,
      "loss": 0.0385,
      "step": 2000
    },
    {
      "epoch": 38.99,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.8927894830703735,
      "eval_runtime": 0.8062,
      "eval_samples_per_second": 57.057,
      "eval_steps_per_second": 7.442,
      "step": 2008
    },
    {
      "epoch": 39.03,
      "learning_rate": 7.337461300309598e-07,
      "loss": 0.0231,
      "step": 2010
    },
    {
      "epoch": 39.22,
      "learning_rate": 4.891640866873065e-07,
      "loss": 0.0227,
      "step": 2020
    },
    {
      "epoch": 39.42,
      "learning_rate": 2.4458204334365327e-07,
      "loss": 0.1008,
      "step": 2030
    },
    {
      "epoch": 39.61,
      "learning_rate": 0.0,
      "loss": 0.0474,
      "step": 2040
    },
    {
      "epoch": 39.61,
      "eval_accuracy": 0.8478260869565217,
      "eval_loss": 0.8863323926925659,
      "eval_runtime": 1.0372,
      "eval_samples_per_second": 44.349,
      "eval_steps_per_second": 5.785,
      "step": 2040
    },
    {
      "epoch": 39.61,
      "step": 2040,
      "total_flos": 2.5142726714989363e+18,
      "train_loss": 0.19007185283390915,
      "train_runtime": 695.8956,
      "train_samples_per_second": 47.076,
      "train_steps_per_second": 2.931
    }
  ],
  "logging_steps": 10,
  "max_steps": 2040,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 2.5142726714989363e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}