{
  "best_metric": 0.18961240351200104,
  "best_model_checkpoint": "mobilevitv2-1.0-imagenet1k-256-finetuned_v2024-7-25-frost/checkpoint-1000",
  "epoch": 30.0,
  "eval_steps": 100,
  "global_step": 1920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15625,
      "grad_norm": 0.2519198954105377,
      "learning_rate": 1.0416666666666668e-05,
      "loss": 0.6958,
      "step": 10
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.21561957895755768,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.6958,
      "step": 20
    },
    {
      "epoch": 0.46875,
      "grad_norm": 0.2186066061258316,
      "learning_rate": 3.125e-05,
      "loss": 0.6948,
      "step": 30
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.22916783392429352,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6934,
      "step": 40
    },
    {
      "epoch": 0.78125,
      "grad_norm": 0.23298271000385284,
      "learning_rate": 5.208333333333334e-05,
      "loss": 0.691,
      "step": 50
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.24950967729091644,
      "learning_rate": 6.25e-05,
      "loss": 0.6895,
      "step": 60
    },
    {
      "epoch": 1.09375,
      "grad_norm": 0.23937764763832092,
      "learning_rate": 7.291666666666667e-05,
      "loss": 0.684,
      "step": 70
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.23057785630226135,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.6794,
      "step": 80
    },
    {
      "epoch": 1.40625,
      "grad_norm": 0.2983661890029907,
      "learning_rate": 9.375e-05,
      "loss": 0.6751,
      "step": 90
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.2652019262313843,
      "learning_rate": 0.00010416666666666667,
      "loss": 0.6687,
      "step": 100
    },
    {
      "epoch": 1.5625,
      "eval_accuracy": 0.7230088495575221,
      "eval_f1": 0.5335320417287631,
      "eval_loss": 0.6623277068138123,
      "eval_precision": 0.40224719101123596,
      "eval_recall": 0.7920353982300885,
      "eval_runtime": 1.2295,
      "eval_samples_per_second": 91.909,
      "eval_steps_per_second": 12.2,
      "step": 100
    },
    {
      "epoch": 1.71875,
      "grad_norm": 0.34594935178756714,
      "learning_rate": 0.00011458333333333333,
      "loss": 0.6586,
      "step": 110
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.26900964975357056,
      "learning_rate": 0.000125,
      "loss": 0.6476,
      "step": 120
    },
    {
      "epoch": 2.03125,
      "grad_norm": 0.30181896686553955,
      "learning_rate": 0.0001354166666666667,
      "loss": 0.6336,
      "step": 130
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.33757150173187256,
      "learning_rate": 0.00014583333333333335,
      "loss": 0.6079,
      "step": 140
    },
    {
      "epoch": 2.34375,
      "grad_norm": 0.54989093542099,
      "learning_rate": 0.00015625,
      "loss": 0.5823,
      "step": 150
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6273936629295349,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.5598,
      "step": 160
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.46115073561668396,
      "learning_rate": 0.00017708333333333335,
      "loss": 0.5239,
      "step": 170
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.47255411744117737,
      "learning_rate": 0.0001875,
      "loss": 0.4972,
      "step": 180
    },
    {
      "epoch": 2.96875,
      "grad_norm": 0.522071361541748,
      "learning_rate": 0.0001979166666666667,
      "loss": 0.4617,
      "step": 190
    },
    {
      "epoch": 3.125,
      "grad_norm": 1.0379081964492798,
      "learning_rate": 0.0001990740740740741,
      "loss": 0.4454,
      "step": 200
    },
    {
      "epoch": 3.125,
      "eval_accuracy": 0.8831858407079646,
      "eval_f1": 0.7490494296577946,
      "eval_loss": 0.41519004106521606,
      "eval_precision": 0.6566666666666666,
      "eval_recall": 0.8716814159292036,
      "eval_runtime": 0.9401,
      "eval_samples_per_second": 120.203,
      "eval_steps_per_second": 15.956,
      "step": 200
    },
    {
      "epoch": 3.28125,
      "grad_norm": 0.545850932598114,
      "learning_rate": 0.0001979166666666667,
      "loss": 0.4247,
      "step": 210
    },
    {
      "epoch": 3.4375,
      "grad_norm": 0.7891977429389954,
      "learning_rate": 0.00019675925925925926,
      "loss": 0.4041,
      "step": 220
    },
    {
      "epoch": 3.59375,
      "grad_norm": 1.542927861213684,
      "learning_rate": 0.00019560185185185186,
      "loss": 0.3845,
      "step": 230
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.44223108887672424,
      "learning_rate": 0.00019444444444444446,
      "loss": 0.3278,
      "step": 240
    },
    {
      "epoch": 3.90625,
      "grad_norm": 0.8207859396934509,
      "learning_rate": 0.00019328703703703706,
      "loss": 0.3507,
      "step": 250
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.42037203907966614,
      "learning_rate": 0.00019212962962962963,
      "loss": 0.3481,
      "step": 260
    },
    {
      "epoch": 4.21875,
      "grad_norm": 1.829210877418518,
      "learning_rate": 0.00019097222222222223,
      "loss": 0.3133,
      "step": 270
    },
    {
      "epoch": 4.375,
      "grad_norm": 1.213773250579834,
      "learning_rate": 0.00018981481481481483,
      "loss": 0.2934,
      "step": 280
    },
    {
      "epoch": 4.53125,
      "grad_norm": 0.6154431104660034,
      "learning_rate": 0.00018865740740740743,
      "loss": 0.2923,
      "step": 290
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.3814193606376648,
      "learning_rate": 0.0001875,
      "loss": 0.2835,
      "step": 300
    },
    {
      "epoch": 4.6875,
      "eval_accuracy": 0.9097345132743363,
      "eval_f1": 0.7660550458715596,
      "eval_loss": 0.26609960198402405,
      "eval_precision": 0.7952380952380952,
      "eval_recall": 0.7389380530973452,
      "eval_runtime": 1.4414,
      "eval_samples_per_second": 78.394,
      "eval_steps_per_second": 10.406,
      "step": 300
    },
    {
      "epoch": 4.84375,
      "grad_norm": 0.4515719711780548,
      "learning_rate": 0.0001863425925925926,
      "loss": 0.2672,
      "step": 310
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.077721357345581,
      "learning_rate": 0.0001851851851851852,
      "loss": 0.2914,
      "step": 320
    },
    {
      "epoch": 5.15625,
      "grad_norm": 0.9644371867179871,
      "learning_rate": 0.00018402777777777778,
      "loss": 0.2527,
      "step": 330
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.8245725631713867,
      "learning_rate": 0.00018287037037037038,
      "loss": 0.2477,
      "step": 340
    },
    {
      "epoch": 5.46875,
      "grad_norm": 0.5262947082519531,
      "learning_rate": 0.00018171296296296297,
      "loss": 0.2452,
      "step": 350
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.6464282870292664,
      "learning_rate": 0.00018055555555555557,
      "loss": 0.2308,
      "step": 360
    },
    {
      "epoch": 5.78125,
      "grad_norm": 0.6029626131057739,
      "learning_rate": 0.00017939814814814815,
      "loss": 0.233,
      "step": 370
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.683201789855957,
      "learning_rate": 0.00017824074074074075,
      "loss": 0.2258,
      "step": 380
    },
    {
      "epoch": 6.09375,
      "grad_norm": 0.5622811317443848,
      "learning_rate": 0.00017708333333333335,
      "loss": 0.2342,
      "step": 390
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.5229126214981079,
      "learning_rate": 0.00017592592592592595,
      "loss": 0.2197,
      "step": 400
    },
    {
      "epoch": 6.25,
      "eval_accuracy": 0.9194690265486726,
      "eval_f1": 0.7868852459016394,
      "eval_loss": 0.21510393917560577,
      "eval_precision": 0.835820895522388,
      "eval_recall": 0.7433628318584071,
      "eval_runtime": 0.9737,
      "eval_samples_per_second": 116.048,
      "eval_steps_per_second": 15.405,
      "step": 400
    },
    {
      "epoch": 6.40625,
      "grad_norm": 0.5103662610054016,
      "learning_rate": 0.00017476851851851852,
      "loss": 0.2084,
      "step": 410
    },
    {
      "epoch": 6.5625,
      "grad_norm": 1.2655210494995117,
      "learning_rate": 0.00017361111111111112,
      "loss": 0.2005,
      "step": 420
    },
    {
      "epoch": 6.71875,
      "grad_norm": 0.5232699513435364,
      "learning_rate": 0.00017245370370370372,
      "loss": 0.2296,
      "step": 430
    },
    {
      "epoch": 6.875,
      "grad_norm": 0.8142613172531128,
      "learning_rate": 0.00017129629629629632,
      "loss": 0.187,
      "step": 440
    },
    {
      "epoch": 7.03125,
      "grad_norm": 0.9919219017028809,
      "learning_rate": 0.0001701388888888889,
      "loss": 0.2027,
      "step": 450
    },
    {
      "epoch": 7.1875,
      "grad_norm": 1.2590153217315674,
      "learning_rate": 0.0001689814814814815,
      "loss": 0.1873,
      "step": 460
    },
    {
      "epoch": 7.34375,
      "grad_norm": 0.6513100266456604,
      "learning_rate": 0.0001678240740740741,
      "loss": 0.1939,
      "step": 470
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.0872722864151,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.2156,
      "step": 480
    },
    {
      "epoch": 7.65625,
      "grad_norm": 0.3712750971317291,
      "learning_rate": 0.00016550925925925926,
      "loss": 0.1968,
      "step": 490
    },
    {
      "epoch": 7.8125,
      "grad_norm": 0.8672028183937073,
      "learning_rate": 0.00016435185185185186,
      "loss": 0.1613,
      "step": 500
    },
    {
      "epoch": 7.8125,
      "eval_accuracy": 0.9292035398230089,
      "eval_f1": 0.813953488372093,
      "eval_loss": 0.20068036019802094,
      "eval_precision": 0.8578431372549019,
      "eval_recall": 0.7743362831858407,
      "eval_runtime": 0.9681,
      "eval_samples_per_second": 116.722,
      "eval_steps_per_second": 15.494,
      "step": 500
    },
    {
      "epoch": 7.96875,
      "grad_norm": 1.2281358242034912,
      "learning_rate": 0.00016319444444444446,
      "loss": 0.1864,
      "step": 510
    },
    {
      "epoch": 8.125,
      "grad_norm": 1.1462125778198242,
      "learning_rate": 0.00016203703703703706,
      "loss": 0.1743,
      "step": 520
    },
    {
      "epoch": 8.28125,
      "grad_norm": 0.5552182197570801,
      "learning_rate": 0.00016087962962962963,
      "loss": 0.1963,
      "step": 530
    },
    {
      "epoch": 8.4375,
      "grad_norm": 0.8015744686126709,
      "learning_rate": 0.00015972222222222223,
      "loss": 0.1987,
      "step": 540
    },
    {
      "epoch": 8.59375,
      "grad_norm": 0.8516111969947815,
      "learning_rate": 0.00015856481481481483,
      "loss": 0.1753,
      "step": 550
    },
    {
      "epoch": 8.75,
      "grad_norm": 1.3942711353302002,
      "learning_rate": 0.00015740740740740743,
      "loss": 0.1577,
      "step": 560
    },
    {
      "epoch": 8.90625,
      "grad_norm": 0.812676191329956,
      "learning_rate": 0.00015625,
      "loss": 0.1726,
      "step": 570
    },
    {
      "epoch": 9.0625,
      "grad_norm": 0.567040205001831,
      "learning_rate": 0.0001550925925925926,
      "loss": 0.1665,
      "step": 580
    },
    {
      "epoch": 9.21875,
      "grad_norm": 0.7389497756958008,
      "learning_rate": 0.0001539351851851852,
      "loss": 0.1445,
      "step": 590
    },
    {
      "epoch": 9.375,
      "grad_norm": 0.6939622163772583,
      "learning_rate": 0.00015277777777777777,
      "loss": 0.1655,
      "step": 600
    },
    {
      "epoch": 9.375,
      "eval_accuracy": 0.9309734513274336,
      "eval_f1": 0.8227272727272726,
      "eval_loss": 0.1935483068227768,
      "eval_precision": 0.8457943925233645,
      "eval_recall": 0.8008849557522124,
      "eval_runtime": 1.422,
      "eval_samples_per_second": 79.467,
      "eval_steps_per_second": 10.549,
      "step": 600
    },
    {
      "epoch": 9.53125,
      "grad_norm": 0.6073923110961914,
      "learning_rate": 0.00015162037037037037,
      "loss": 0.159,
      "step": 610
    },
    {
      "epoch": 9.6875,
      "grad_norm": 0.8762220740318298,
      "learning_rate": 0.00015046296296296297,
      "loss": 0.1959,
      "step": 620
    },
    {
      "epoch": 9.84375,
      "grad_norm": 0.7490831017494202,
      "learning_rate": 0.00014930555555555557,
      "loss": 0.1465,
      "step": 630
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.0123506784439087,
      "learning_rate": 0.00014814814814814815,
      "loss": 0.1683,
      "step": 640
    },
    {
      "epoch": 10.15625,
      "grad_norm": 0.5325204133987427,
      "learning_rate": 0.00014699074074074075,
      "loss": 0.1636,
      "step": 650
    },
    {
      "epoch": 10.3125,
      "grad_norm": 0.5814504623413086,
      "learning_rate": 0.00014583333333333335,
      "loss": 0.1729,
      "step": 660
    },
    {
      "epoch": 10.46875,
      "grad_norm": 1.0156935453414917,
      "learning_rate": 0.00014467592592592594,
      "loss": 0.1569,
      "step": 670
    },
    {
      "epoch": 10.625,
      "grad_norm": 1.257921576499939,
      "learning_rate": 0.00014351851851851852,
      "loss": 0.1429,
      "step": 680
    },
    {
      "epoch": 10.78125,
      "grad_norm": 0.929108202457428,
      "learning_rate": 0.00014236111111111112,
      "loss": 0.1554,
      "step": 690
    },
    {
      "epoch": 10.9375,
      "grad_norm": 0.5256894826889038,
      "learning_rate": 0.00014120370370370372,
      "loss": 0.1815,
      "step": 700
    },
    {
      "epoch": 10.9375,
      "eval_accuracy": 0.9265486725663716,
      "eval_f1": 0.8074245939675174,
      "eval_loss": 0.18833249807357788,
      "eval_precision": 0.848780487804878,
      "eval_recall": 0.7699115044247787,
      "eval_runtime": 0.9484,
      "eval_samples_per_second": 119.151,
      "eval_steps_per_second": 15.817,
      "step": 700
    },
    {
      "epoch": 11.09375,
      "grad_norm": 1.1333953142166138,
      "learning_rate": 0.00014004629629629632,
      "loss": 0.1703,
      "step": 710
    },
    {
      "epoch": 11.25,
      "grad_norm": 0.6658828854560852,
      "learning_rate": 0.0001388888888888889,
      "loss": 0.1475,
      "step": 720
    },
    {
      "epoch": 11.40625,
      "grad_norm": 1.04364812374115,
      "learning_rate": 0.0001377314814814815,
      "loss": 0.1598,
      "step": 730
    },
    {
      "epoch": 11.5625,
      "grad_norm": 0.8811527490615845,
      "learning_rate": 0.0001365740740740741,
      "loss": 0.1678,
      "step": 740
    },
    {
      "epoch": 11.71875,
      "grad_norm": 0.8651083111763,
      "learning_rate": 0.0001354166666666667,
      "loss": 0.1681,
      "step": 750
    },
    {
      "epoch": 11.875,
      "grad_norm": 0.833223283290863,
      "learning_rate": 0.00013425925925925926,
      "loss": 0.1616,
      "step": 760
    },
    {
      "epoch": 12.03125,
      "grad_norm": 0.5667290687561035,
      "learning_rate": 0.00013310185185185186,
      "loss": 0.1185,
      "step": 770
    },
    {
      "epoch": 12.1875,
      "grad_norm": 1.3427128791809082,
      "learning_rate": 0.00013194444444444446,
      "loss": 0.1442,
      "step": 780
    },
    {
      "epoch": 12.34375,
      "grad_norm": 0.859018087387085,
      "learning_rate": 0.00013078703703703706,
      "loss": 0.1552,
      "step": 790
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.6311579942703247,
      "learning_rate": 0.00012962962962962963,
      "loss": 0.1316,
      "step": 800
    },
    {
      "epoch": 12.5,
      "eval_accuracy": 0.9327433628318584,
      "eval_f1": 0.8272727272727272,
      "eval_loss": 0.18246687948703766,
      "eval_precision": 0.8504672897196262,
      "eval_recall": 0.8053097345132744,
      "eval_runtime": 0.9594,
      "eval_samples_per_second": 117.786,
      "eval_steps_per_second": 15.635,
      "step": 800
    },
    {
      "epoch": 12.65625,
      "grad_norm": 0.8464061617851257,
      "learning_rate": 0.00012847222222222223,
      "loss": 0.1344,
      "step": 810
    },
    {
      "epoch": 12.8125,
      "grad_norm": 0.6711329221725464,
      "learning_rate": 0.00012731481481481483,
      "loss": 0.1602,
      "step": 820
    },
    {
      "epoch": 12.96875,
      "grad_norm": 1.0340158939361572,
      "learning_rate": 0.00012615740740740743,
      "loss": 0.1483,
      "step": 830
    },
    {
      "epoch": 13.125,
      "grad_norm": 0.711726725101471,
      "learning_rate": 0.000125,
      "loss": 0.1507,
      "step": 840
    },
    {
      "epoch": 13.28125,
      "grad_norm": 0.8784794211387634,
      "learning_rate": 0.00012384259259259258,
      "loss": 0.1515,
      "step": 850
    },
    {
      "epoch": 13.4375,
      "grad_norm": 0.9908888339996338,
      "learning_rate": 0.0001226851851851852,
      "loss": 0.1583,
      "step": 860
    },
    {
      "epoch": 13.59375,
      "grad_norm": 0.5473937392234802,
      "learning_rate": 0.00012152777777777777,
      "loss": 0.1433,
      "step": 870
    },
    {
      "epoch": 13.75,
      "grad_norm": 1.6888905763626099,
      "learning_rate": 0.00012037037037037037,
      "loss": 0.1371,
      "step": 880
    },
    {
      "epoch": 13.90625,
      "grad_norm": 1.0640438795089722,
      "learning_rate": 0.00011921296296296296,
      "loss": 0.1376,
      "step": 890
    },
    {
      "epoch": 14.0625,
      "grad_norm": 1.9941257238388062,
      "learning_rate": 0.00011805555555555556,
      "loss": 0.1612,
      "step": 900
    },
    {
      "epoch": 14.0625,
      "eval_accuracy": 0.9256637168141593,
      "eval_f1": 0.8099547511312217,
      "eval_loss": 0.18371373414993286,
      "eval_precision": 0.8287037037037037,
      "eval_recall": 0.7920353982300885,
      "eval_runtime": 1.4255,
      "eval_samples_per_second": 79.27,
      "eval_steps_per_second": 10.523,
      "step": 900
    },
    {
      "epoch": 14.21875,
      "grad_norm": 0.650867760181427,
      "learning_rate": 0.00011689814814814815,
      "loss": 0.1468,
      "step": 910
    },
    {
      "epoch": 14.375,
      "grad_norm": 0.7459059357643127,
      "learning_rate": 0.00011574074074074075,
      "loss": 0.1468,
      "step": 920
    },
    {
      "epoch": 14.53125,
      "grad_norm": 0.7468872666358948,
      "learning_rate": 0.00011458333333333333,
      "loss": 0.1169,
      "step": 930
    },
    {
      "epoch": 14.6875,
      "grad_norm": 0.6512945890426636,
      "learning_rate": 0.00011342592592592593,
      "loss": 0.1373,
      "step": 940
    },
    {
      "epoch": 14.84375,
      "grad_norm": 0.710382878780365,
      "learning_rate": 0.00011226851851851852,
      "loss": 0.1223,
      "step": 950
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.2112369537353516,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.1257,
      "step": 960
    },
    {
      "epoch": 15.15625,
      "grad_norm": 4.069777965545654,
      "learning_rate": 0.0001099537037037037,
      "loss": 0.1369,
      "step": 970
    },
    {
      "epoch": 15.3125,
      "grad_norm": 0.9751072525978088,
      "learning_rate": 0.0001087962962962963,
      "loss": 0.1316,
      "step": 980
    },
    {
      "epoch": 15.46875,
      "grad_norm": 0.49943211674690247,
      "learning_rate": 0.00010763888888888889,
      "loss": 0.1088,
      "step": 990
    },
    {
      "epoch": 15.625,
      "grad_norm": 0.7845533490180969,
      "learning_rate": 0.00010648148148148149,
      "loss": 0.118,
      "step": 1000
    },
    {
      "epoch": 15.625,
      "eval_accuracy": 0.9309734513274336,
      "eval_f1": 0.8227272727272726,
      "eval_loss": 0.18961240351200104,
      "eval_precision": 0.8457943925233645,
      "eval_recall": 0.8008849557522124,
      "eval_runtime": 0.9541,
      "eval_samples_per_second": 118.435,
      "eval_steps_per_second": 15.721,
      "step": 1000
    },
    {
      "epoch": 15.78125,
      "grad_norm": 0.5193383693695068,
      "learning_rate": 0.00010532407407407407,
      "loss": 0.1233,
      "step": 1010
    },
    {
      "epoch": 15.9375,
      "grad_norm": 0.5976629257202148,
      "learning_rate": 0.00010416666666666667,
      "loss": 0.1351,
      "step": 1020
    },
    {
      "epoch": 16.09375,
      "grad_norm": 1.0629384517669678,
      "learning_rate": 0.00010300925925925926,
      "loss": 0.1597,
      "step": 1030
    },
    {
      "epoch": 16.25,
      "grad_norm": 0.8576996326446533,
      "learning_rate": 0.00010185185185185186,
      "loss": 0.1268,
      "step": 1040
    },
    {
      "epoch": 16.40625,
      "grad_norm": 0.7236841917037964,
      "learning_rate": 0.00010069444444444445,
      "loss": 0.1411,
      "step": 1050
    },
    {
      "epoch": 16.5625,
      "grad_norm": 1.1142785549163818,
      "learning_rate": 9.953703703703704e-05,
      "loss": 0.1297,
      "step": 1060
    },
    {
      "epoch": 16.71875,
      "grad_norm": 0.8304411768913269,
      "learning_rate": 9.837962962962963e-05,
      "loss": 0.1231,
      "step": 1070
    },
    {
      "epoch": 16.875,
      "grad_norm": 0.8226402997970581,
      "learning_rate": 9.722222222222223e-05,
      "loss": 0.1399,
      "step": 1080
    },
    {
      "epoch": 17.03125,
      "grad_norm": 0.6692397594451904,
      "learning_rate": 9.606481481481482e-05,
      "loss": 0.1833,
      "step": 1090
    },
    {
      "epoch": 17.1875,
      "grad_norm": 0.6689762473106384,
      "learning_rate": 9.490740740740742e-05,
      "loss": 0.1178,
      "step": 1100
    },
    {
      "epoch": 17.1875,
      "eval_accuracy": 0.9238938053097345,
      "eval_f1": 0.8027522935779817,
      "eval_loss": 0.19371576607227325,
      "eval_precision": 0.8333333333333334,
      "eval_recall": 0.7743362831858407,
      "eval_runtime": 0.9499,
      "eval_samples_per_second": 118.958,
      "eval_steps_per_second": 15.791,
      "step": 1100
    },
    {
      "epoch": 17.34375,
      "grad_norm": 0.6079881191253662,
      "learning_rate": 9.375e-05,
      "loss": 0.1259,
      "step": 1110
    },
    {
      "epoch": 17.5,
      "grad_norm": 0.37670084834098816,
      "learning_rate": 9.25925925925926e-05,
      "loss": 0.1101,
      "step": 1120
    },
    {
      "epoch": 17.65625,
      "grad_norm": 0.7734571695327759,
      "learning_rate": 9.143518518518519e-05,
      "loss": 0.1279,
      "step": 1130
    },
    {
      "epoch": 17.8125,
      "grad_norm": 1.0208630561828613,
      "learning_rate": 9.027777777777779e-05,
      "loss": 0.1258,
      "step": 1140
    },
    {
      "epoch": 17.96875,
      "grad_norm": 0.5698951482772827,
      "learning_rate": 8.912037037037037e-05,
      "loss": 0.1111,
      "step": 1150
    },
    {
      "epoch": 18.125,
      "grad_norm": 1.77188241481781,
      "learning_rate": 8.796296296296297e-05,
      "loss": 0.1254,
      "step": 1160
    },
    {
      "epoch": 18.28125,
      "grad_norm": 0.8389852643013,
      "learning_rate": 8.680555555555556e-05,
      "loss": 0.1042,
      "step": 1170
    },
    {
      "epoch": 18.4375,
      "grad_norm": 0.6655524969100952,
      "learning_rate": 8.564814814814816e-05,
      "loss": 0.1272,
      "step": 1180
    },
    {
      "epoch": 18.59375,
      "grad_norm": 0.4668845236301422,
      "learning_rate": 8.449074074074074e-05,
      "loss": 0.1096,
      "step": 1190
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.8379706740379333,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.1248,
      "step": 1200
    },
    {
      "epoch": 18.75,
      "eval_accuracy": 0.9300884955752212,
      "eval_f1": 0.8192219679633868,
      "eval_loss": 0.19132623076438904,
      "eval_precision": 0.8483412322274881,
      "eval_recall": 0.7920353982300885,
      "eval_runtime": 1.2549,
      "eval_samples_per_second": 90.047,
      "eval_steps_per_second": 11.953,
      "step": 1200
    },
    {
      "epoch": 18.90625,
      "grad_norm": 0.9271652698516846,
      "learning_rate": 8.217592592592593e-05,
      "loss": 0.1126,
      "step": 1210
    },
    {
      "epoch": 19.0625,
      "grad_norm": 1.1356163024902344,
      "learning_rate": 8.113425925925926e-05,
      "loss": 0.129,
      "step": 1220
    },
    {
      "epoch": 19.21875,
      "grad_norm": 0.4993898570537567,
      "learning_rate": 7.997685185185186e-05,
      "loss": 0.1385,
      "step": 1230
    },
    {
      "epoch": 19.375,
      "grad_norm": 1.2999491691589355,
      "learning_rate": 7.881944444444444e-05,
      "loss": 0.1242,
      "step": 1240
    },
    {
      "epoch": 19.53125,
      "grad_norm": 1.1871651411056519,
      "learning_rate": 7.766203703703704e-05,
      "loss": 0.113,
      "step": 1250
    },
    {
      "epoch": 19.6875,
      "grad_norm": 1.5129660367965698,
      "learning_rate": 7.650462962962963e-05,
      "loss": 0.1024,
      "step": 1260
    },
    {
      "epoch": 19.84375,
      "grad_norm": 0.7286781072616577,
      "learning_rate": 7.534722222222223e-05,
      "loss": 0.0994,
      "step": 1270
    },
    {
      "epoch": 20.0,
      "grad_norm": 1.0448476076126099,
      "learning_rate": 7.418981481481481e-05,
      "loss": 0.1115,
      "step": 1280
    },
    {
      "epoch": 20.15625,
      "grad_norm": 0.6149379014968872,
      "learning_rate": 7.303240740740741e-05,
      "loss": 0.1243,
      "step": 1290
    },
    {
      "epoch": 20.3125,
      "grad_norm": 0.7020682692527771,
      "learning_rate": 7.1875e-05,
      "loss": 0.1169,
      "step": 1300
    },
    {
      "epoch": 20.3125,
      "eval_accuracy": 0.9300884955752212,
      "eval_f1": 0.8167053364269141,
      "eval_loss": 0.19162432849407196,
      "eval_precision": 0.8585365853658536,
      "eval_recall": 0.7787610619469026,
      "eval_runtime": 0.958,
      "eval_samples_per_second": 117.951,
      "eval_steps_per_second": 15.657,
      "step": 1300
    },
    {
      "epoch": 20.46875,
      "grad_norm": 1.8554919958114624,
      "learning_rate": 7.07175925925926e-05,
      "loss": 0.1175,
      "step": 1310
    },
    {
      "epoch": 20.625,
      "grad_norm": 1.1227444410324097,
      "learning_rate": 6.956018518518518e-05,
      "loss": 0.1004,
      "step": 1320
    },
    {
      "epoch": 20.78125,
      "grad_norm": 1.376546025276184,
      "learning_rate": 6.840277777777778e-05,
      "loss": 0.1099,
      "step": 1330
    },
    {
      "epoch": 20.9375,
      "grad_norm": 0.86075758934021,
      "learning_rate": 6.724537037037037e-05,
      "loss": 0.133,
      "step": 1340
    },
    {
      "epoch": 21.09375,
      "grad_norm": 1.093257188796997,
      "learning_rate": 6.608796296296297e-05,
      "loss": 0.1143,
      "step": 1350
    },
    {
      "epoch": 21.25,
      "grad_norm": 0.5665271282196045,
      "learning_rate": 6.493055555555556e-05,
      "loss": 0.1065,
      "step": 1360
    },
    {
      "epoch": 21.40625,
      "grad_norm": 0.607912003993988,
      "learning_rate": 6.377314814814816e-05,
      "loss": 0.1221,
      "step": 1370
    },
    {
      "epoch": 21.5625,
      "grad_norm": 0.4708748161792755,
      "learning_rate": 6.261574074074074e-05,
      "loss": 0.0941,
      "step": 1380
    },
    {
      "epoch": 21.71875,
      "grad_norm": 0.8719390630722046,
      "learning_rate": 6.145833333333334e-05,
      "loss": 0.112,
      "step": 1390
    },
    {
      "epoch": 21.875,
      "grad_norm": 0.45299583673477173,
      "learning_rate": 6.0300925925925934e-05,
      "loss": 0.1094,
      "step": 1400
    },
    {
      "epoch": 21.875,
      "eval_accuracy": 0.9292035398230089,
      "eval_f1": 0.8181818181818182,
      "eval_loss": 0.19246041774749756,
      "eval_precision": 0.8411214953271028,
      "eval_recall": 0.7964601769911505,
      "eval_runtime": 1.2007,
      "eval_samples_per_second": 94.111,
      "eval_steps_per_second": 12.493,
      "step": 1400
    },
    {
      "epoch": 22.03125,
      "grad_norm": 0.7893074154853821,
      "learning_rate": 5.9143518518518527e-05,
      "loss": 0.1142,
      "step": 1410
    },
    {
      "epoch": 22.1875,
      "grad_norm": 1.0832182168960571,
      "learning_rate": 5.798611111111112e-05,
      "loss": 0.096,
      "step": 1420
    },
    {
      "epoch": 22.34375,
      "grad_norm": 0.5880953669548035,
      "learning_rate": 5.682870370370371e-05,
      "loss": 0.1142,
      "step": 1430
    },
    {
      "epoch": 22.5,
      "grad_norm": 0.6071570515632629,
      "learning_rate": 5.567129629629629e-05,
      "loss": 0.1146,
      "step": 1440
    },
    {
      "epoch": 22.65625,
      "grad_norm": 1.7968199253082275,
      "learning_rate": 5.4513888888888884e-05,
      "loss": 0.1141,
      "step": 1450
    },
    {
      "epoch": 22.8125,
      "grad_norm": 0.6409327983856201,
      "learning_rate": 5.335648148148148e-05,
      "loss": 0.0907,
      "step": 1460
    },
    {
      "epoch": 22.96875,
      "grad_norm": 1.28959321975708,
      "learning_rate": 5.219907407407407e-05,
      "loss": 0.0945,
      "step": 1470
    },
    {
      "epoch": 23.125,
      "grad_norm": 1.0384379625320435,
      "learning_rate": 5.115740740740741e-05,
      "loss": 0.1062,
      "step": 1480
    },
    {
      "epoch": 23.28125,
      "grad_norm": 0.8010191917419434,
      "learning_rate": 5e-05,
      "loss": 0.1029,
      "step": 1490
    },
    {
      "epoch": 23.4375,
      "grad_norm": 1.078620195388794,
      "learning_rate": 4.8842592592592595e-05,
      "loss": 0.1108,
      "step": 1500
    },
    {
      "epoch": 23.4375,
      "eval_accuracy": 0.9345132743362832,
      "eval_f1": 0.8333333333333334,
      "eval_loss": 0.19605357944965363,
      "eval_precision": 0.8486238532110092,
      "eval_recall": 0.8185840707964602,
      "eval_runtime": 0.9739,
      "eval_samples_per_second": 116.033,
      "eval_steps_per_second": 15.403,
      "step": 1500
    },
    {
      "epoch": 23.59375,
      "grad_norm": 0.9104486703872681,
      "learning_rate": 4.768518518518519e-05,
      "loss": 0.1039,
      "step": 1510
    },
    {
      "epoch": 23.75,
      "grad_norm": 1.1187772750854492,
      "learning_rate": 4.652777777777778e-05,
      "loss": 0.1179,
      "step": 1520
    },
    {
      "epoch": 23.90625,
      "grad_norm": 0.6038283109664917,
      "learning_rate": 4.5370370370370374e-05,
      "loss": 0.0981,
      "step": 1530
    },
    {
      "epoch": 24.0625,
      "grad_norm": 1.483780860900879,
      "learning_rate": 4.4212962962962966e-05,
      "loss": 0.1287,
      "step": 1540
    },
    {
      "epoch": 24.21875,
      "grad_norm": 0.9008955359458923,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.1038,
      "step": 1550
    },
    {
      "epoch": 24.375,
      "grad_norm": 1.3843752145767212,
      "learning_rate": 4.1898148148148145e-05,
      "loss": 0.1177,
      "step": 1560
    },
    {
      "epoch": 24.53125,
      "grad_norm": 1.3227291107177734,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.0992,
      "step": 1570
    },
    {
      "epoch": 24.6875,
      "grad_norm": 0.4530428349971771,
      "learning_rate": 3.958333333333333e-05,
      "loss": 0.1064,
      "step": 1580
    },
    {
      "epoch": 24.84375,
      "grad_norm": 1.476251244544983,
      "learning_rate": 3.8425925925925924e-05,
      "loss": 0.1008,
      "step": 1590
    },
    {
      "epoch": 25.0,
      "grad_norm": 1.4163908958435059,
      "learning_rate": 3.726851851851852e-05,
      "loss": 0.1089,
      "step": 1600
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.9283185840707965,
      "eval_f1": 0.8171557562076749,
      "eval_loss": 0.1992984265089035,
      "eval_precision": 0.8341013824884793,
      "eval_recall": 0.8008849557522124,
      "eval_runtime": 0.9786,
      "eval_samples_per_second": 115.465,
      "eval_steps_per_second": 15.327,
      "step": 1600
    },
    {
      "epoch": 25.15625,
      "grad_norm": 0.805210292339325,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.0922,
      "step": 1610
    },
    {
      "epoch": 25.3125,
      "grad_norm": 0.7946519255638123,
      "learning_rate": 3.49537037037037e-05,
      "loss": 0.1033,
      "step": 1620
    },
    {
      "epoch": 25.46875,
      "grad_norm": 0.7051573991775513,
      "learning_rate": 3.3796296296296295e-05,
      "loss": 0.1031,
      "step": 1630
    },
    {
      "epoch": 25.625,
      "grad_norm": 0.6867948174476624,
      "learning_rate": 3.263888888888889e-05,
      "loss": 0.1203,
      "step": 1640
    },
    {
      "epoch": 25.78125,
      "grad_norm": 0.9575832486152649,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.101,
      "step": 1650
    },
    {
      "epoch": 25.9375,
      "grad_norm": 1.125503420829773,
      "learning_rate": 3.0324074074074077e-05,
      "loss": 0.0868,
      "step": 1660
    },
    {
      "epoch": 26.09375,
      "grad_norm": 0.694492757320404,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.0916,
      "step": 1670
    },
    {
      "epoch": 26.25,
      "grad_norm": 0.606955885887146,
      "learning_rate": 2.8009259259259263e-05,
      "loss": 0.0978,
      "step": 1680
    },
    {
      "epoch": 26.40625,
      "grad_norm": 0.855603814125061,
      "learning_rate": 2.6851851851851855e-05,
      "loss": 0.0988,
      "step": 1690
    },
    {
      "epoch": 26.5625,
      "grad_norm": 0.6119447946548462,
      "learning_rate": 2.5694444444444445e-05,
      "loss": 0.0919,
      "step": 1700
    },
    {
      "epoch": 26.5625,
      "eval_accuracy": 0.9318584070796461,
      "eval_f1": 0.8261851015801355,
      "eval_loss": 0.19360247254371643,
      "eval_precision": 0.8433179723502304,
      "eval_recall": 0.8097345132743363,
      "eval_runtime": 1.5343,
      "eval_samples_per_second": 73.651,
      "eval_steps_per_second": 9.777,
      "step": 1700
    },
    {
      "epoch": 26.71875,
      "grad_norm": 0.9873837828636169,
      "learning_rate": 2.4537037037037038e-05,
      "loss": 0.0829,
      "step": 1710
    },
    {
      "epoch": 26.875,
      "grad_norm": 0.9287075996398926,
      "learning_rate": 2.337962962962963e-05,
      "loss": 0.1128,
      "step": 1720
    },
    {
      "epoch": 27.03125,
      "grad_norm": 0.8201906681060791,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1084,
      "step": 1730
    },
    {
      "epoch": 27.1875,
      "grad_norm": 1.1874263286590576,
      "learning_rate": 2.1064814814814816e-05,
      "loss": 0.1065,
      "step": 1740
    },
    {
      "epoch": 27.34375,
      "grad_norm": 1.0616997480392456,
      "learning_rate": 1.990740740740741e-05,
      "loss": 0.1014,
      "step": 1750
    },
    {
      "epoch": 27.5,
      "grad_norm": 0.8941544890403748,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.1008,
      "step": 1760
    },
    {
      "epoch": 27.65625,
      "grad_norm": 0.9521628022193909,
      "learning_rate": 1.7592592592592595e-05,
      "loss": 0.0908,
      "step": 1770
    },
    {
      "epoch": 27.8125,
      "grad_norm": 0.79527348279953,
      "learning_rate": 1.6435185185185187e-05,
      "loss": 0.1137,
      "step": 1780
    },
    {
      "epoch": 27.96875,
      "grad_norm": 0.8606336116790771,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.0929,
      "step": 1790
    },
    {
      "epoch": 28.125,
      "grad_norm": 1.2438716888427734,
      "learning_rate": 1.412037037037037e-05,
      "loss": 0.0969,
      "step": 1800
    },
    {
      "epoch": 28.125,
      "eval_accuracy": 0.9309734513274336,
      "eval_f1": 0.8227272727272726,
      "eval_loss": 0.19780634343624115,
      "eval_precision": 0.8457943925233645,
      "eval_recall": 0.8008849557522124,
      "eval_runtime": 0.9685,
      "eval_samples_per_second": 116.67,
      "eval_steps_per_second": 15.487,
      "step": 1800
    },
    {
      "epoch": 28.28125,
      "grad_norm": 0.9101031422615051,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.0977,
      "step": 1810
    },
    {
      "epoch": 28.4375,
      "grad_norm": 0.6852589249610901,
      "learning_rate": 1.1805555555555555e-05,
      "loss": 0.1018,
      "step": 1820
    },
    {
      "epoch": 28.59375,
      "grad_norm": 0.8925357460975647,
      "learning_rate": 1.0648148148148148e-05,
      "loss": 0.0985,
      "step": 1830
    },
    {
      "epoch": 28.75,
      "grad_norm": 0.8219801783561707,
      "learning_rate": 9.490740740740741e-06,
      "loss": 0.1014,
      "step": 1840
    },
    {
      "epoch": 28.90625,
      "grad_norm": 0.6065762639045715,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1161,
      "step": 1850
    },
    {
      "epoch": 29.0625,
      "grad_norm": 0.7718455791473389,
      "learning_rate": 7.1759259259259266e-06,
      "loss": 0.1101,
      "step": 1860
    },
    {
      "epoch": 29.21875,
      "grad_norm": 1.0950359106063843,
      "learning_rate": 6.0185185185185185e-06,
      "loss": 0.1042,
      "step": 1870
    },
    {
      "epoch": 29.375,
      "grad_norm": 0.7298617362976074,
      "learning_rate": 4.861111111111111e-06,
      "loss": 0.0823,
      "step": 1880
    },
    {
      "epoch": 29.53125,
      "grad_norm": 0.782146692276001,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1051,
      "step": 1890
    },
    {
      "epoch": 29.6875,
      "grad_norm": 0.7693170309066772,
      "learning_rate": 2.546296296296296e-06,
      "loss": 0.1093,
      "step": 1900
    },
    {
      "epoch": 29.6875,
      "eval_accuracy": 0.9283185840707965,
      "eval_f1": 0.8171557562076749,
      "eval_loss": 0.19546246528625488,
      "eval_precision": 0.8341013824884793,
      "eval_recall": 0.8008849557522124,
      "eval_runtime": 1.3969,
      "eval_samples_per_second": 80.891,
      "eval_steps_per_second": 10.738,
      "step": 1900
    },
    {
      "epoch": 29.84375,
      "grad_norm": 1.2105658054351807,
      "learning_rate": 1.388888888888889e-06,
      "loss": 0.0836,
      "step": 1910
    },
    {
      "epoch": 30.0,
      "grad_norm": 2.1491291522979736,
      "learning_rate": 2.3148148148148148e-07,
      "loss": 0.1122,
      "step": 1920
    },
    {
      "epoch": 30.0,
      "step": 1920,
      "total_flos": 1.9916656541540352e+17,
      "train_loss": 0.1991663834390541,
      "train_runtime": 388.9258,
      "train_samples_per_second": 78.061,
      "train_steps_per_second": 4.937
    }
  ],
  "logging_steps": 10,
  "max_steps": 1920,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9916656541540352e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}