{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.6507508336696333,
  "eval_steps": 100,
  "global_step": 7800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.021212395280599594,
      "learning_rate": 0.0001,
      "loss": 8.6987,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.018885262310504913,
      "learning_rate": 0.0001,
      "loss": 8.7228,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.030739275738596916,
      "learning_rate": 0.0001,
      "loss": 8.7384,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.03329615667462349,
      "learning_rate": 0.0001,
      "loss": 8.7324,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.021798215806484222,
      "learning_rate": 0.0001,
      "loss": 8.7257,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.022552112117409706,
      "learning_rate": 0.0001,
      "loss": 8.7345,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.024906286969780922,
      "learning_rate": 0.0001,
      "loss": 8.7341,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.029604945331811905,
      "learning_rate": 0.0001,
      "loss": 8.7372,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.027017420157790184,
      "learning_rate": 0.0001,
      "loss": 8.7178,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.1736399084329605,
      "learning_rate": 0.0001,
      "loss": 8.7289,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.032085757702589035,
      "learning_rate": 0.0001,
      "loss": 8.6985,
      "step": 100
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.0356113463640213,
      "learning_rate": 0.0001,
      "loss": 8.7198,
      "step": 110
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.04567497596144676,
      "learning_rate": 0.0001,
      "loss": 8.7256,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.10645324736833572,
      "learning_rate": 0.0001,
      "loss": 8.7044,
      "step": 130
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.08604108542203903,
      "learning_rate": 0.0001,
      "loss": 8.6868,
      "step": 140
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.21885468065738678,
      "learning_rate": 0.0001,
      "loss": 8.678,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.34108299016952515,
      "learning_rate": 0.0001,
      "loss": 8.6014,
      "step": 160
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.551449179649353,
      "learning_rate": 0.0001,
      "loss": 8.5854,
      "step": 170
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4222724437713623,
      "learning_rate": 0.0001,
      "loss": 8.5096,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4721892178058624,
      "learning_rate": 0.0001,
      "loss": 8.4888,
      "step": 190
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4090125262737274,
      "learning_rate": 0.0001,
      "loss": 8.4725,
      "step": 200
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.3760685920715332,
      "learning_rate": 0.0001,
      "loss": 8.4548,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.44033998250961304,
      "learning_rate": 0.0001,
      "loss": 8.4367,
      "step": 220
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.47666457295417786,
      "learning_rate": 0.0001,
      "loss": 8.4161,
      "step": 230
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4455925524234772,
      "learning_rate": 0.0001,
      "loss": 8.4375,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.46253493428230286,
      "learning_rate": 0.0001,
      "loss": 8.4125,
      "step": 250
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5054658651351929,
      "learning_rate": 0.0001,
      "loss": 8.3878,
      "step": 260
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5697858929634094,
      "learning_rate": 0.0001,
      "loss": 8.4173,
      "step": 270
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5195950269699097,
      "learning_rate": 0.0001,
      "loss": 8.3315,
      "step": 280
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5799770355224609,
      "learning_rate": 0.0001,
      "loss": 8.3224,
      "step": 290
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5585042834281921,
      "learning_rate": 0.0001,
      "loss": 8.3111,
      "step": 300
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6580124497413635,
      "learning_rate": 0.0001,
      "loss": 8.3186,
      "step": 310
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7007073760032654,
      "learning_rate": 0.0001,
      "loss": 8.2862,
      "step": 320
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6406170725822449,
      "learning_rate": 0.0001,
      "loss": 8.3202,
      "step": 330
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6901131272315979,
      "learning_rate": 0.0001,
      "loss": 8.3053,
      "step": 340
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7509775161743164,
      "learning_rate": 0.0001,
      "loss": 8.2628,
      "step": 350
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6910294890403748,
      "learning_rate": 0.0001,
      "loss": 8.2785,
      "step": 360
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.95322185754776,
      "learning_rate": 0.0001,
      "loss": 8.2177,
      "step": 370
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7333555221557617,
      "learning_rate": 0.0001,
      "loss": 8.2437,
      "step": 380
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7888531684875488,
      "learning_rate": 0.0001,
      "loss": 8.2163,
      "step": 390
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.3577535152435303,
      "learning_rate": 0.0001,
      "loss": 8.1334,
      "step": 400
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7684480547904968,
      "learning_rate": 0.0001,
      "loss": 8.2015,
      "step": 410
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.9100524187088013,
      "learning_rate": 0.0001,
      "loss": 8.2179,
      "step": 420
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0707799196243286,
      "learning_rate": 0.0001,
      "loss": 8.2101,
      "step": 430
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8474343419075012,
      "learning_rate": 0.0001,
      "loss": 8.1322,
      "step": 440
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8755463361740112,
      "learning_rate": 0.0001,
      "loss": 8.178,
      "step": 450
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.9474639892578125,
      "learning_rate": 0.0001,
      "loss": 8.1057,
      "step": 460
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.82960045337677,
      "learning_rate": 0.0001,
      "loss": 8.1476,
      "step": 470
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.9738801121711731,
      "learning_rate": 0.0001,
      "loss": 8.1107,
      "step": 480
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9569644927978516,
      "learning_rate": 0.0001,
      "loss": 8.0467,
      "step": 490
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9318249225616455,
      "learning_rate": 0.0001,
      "loss": 8.1014,
      "step": 500
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.0197961330413818,
      "learning_rate": 0.0001,
      "loss": 8.111,
      "step": 510
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0179370641708374,
      "learning_rate": 0.0001,
      "loss": 8.1044,
      "step": 520
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.9549160599708557,
      "learning_rate": 0.0001,
      "loss": 8.1088,
      "step": 530
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.9991666674613953,
      "learning_rate": 0.0001,
      "loss": 8.1583,
      "step": 540
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.237109661102295,
      "learning_rate": 0.0001,
      "loss": 8.048,
      "step": 550
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.0213505029678345,
      "learning_rate": 0.0001,
      "loss": 8.0383,
      "step": 560
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.1742730140686035,
      "learning_rate": 0.0001,
      "loss": 8.0809,
      "step": 570
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.0418646335601807,
      "learning_rate": 0.0001,
      "loss": 8.0511,
      "step": 580
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.026742935180664,
      "learning_rate": 0.0001,
      "loss": 7.9717,
      "step": 590
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2008246183395386,
      "learning_rate": 0.0001,
      "loss": 8.029,
      "step": 600
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0359411239624023,
      "learning_rate": 0.0001,
      "loss": 8.0179,
      "step": 610
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0166168212890625,
      "learning_rate": 0.0001,
      "loss": 8.0339,
      "step": 620
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.0548157691955566,
      "learning_rate": 0.0001,
      "loss": 7.962,
      "step": 630
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0204545259475708,
      "learning_rate": 0.0001,
      "loss": 8.0035,
      "step": 640
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.122182846069336,
      "learning_rate": 0.0001,
      "loss": 8.046,
      "step": 650
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0963354110717773,
      "learning_rate": 0.0001,
      "loss": 8.0241,
      "step": 660
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0327413082122803,
      "learning_rate": 0.0001,
      "loss": 7.9474,
      "step": 670
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.279113531112671,
      "learning_rate": 0.0001,
      "loss": 8.011,
      "step": 680
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.0806117057800293,
      "learning_rate": 0.0001,
      "loss": 7.9235,
      "step": 690
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.0781127214431763,
      "learning_rate": 0.0001,
      "loss": 7.9572,
      "step": 700
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.1420655250549316,
      "learning_rate": 0.0001,
      "loss": 7.9394,
      "step": 710
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.161694049835205,
      "learning_rate": 0.0001,
      "loss": 7.9587,
      "step": 720
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1545701026916504,
      "learning_rate": 0.0001,
      "loss": 7.8929,
      "step": 730
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1458271741867065,
      "learning_rate": 0.0001,
      "loss": 7.8929,
      "step": 740
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.1537671089172363,
      "learning_rate": 0.0001,
      "loss": 7.9182,
      "step": 750
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.217034935951233,
      "learning_rate": 0.0001,
      "loss": 7.9654,
      "step": 760
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1825016736984253,
      "learning_rate": 0.0001,
      "loss": 7.8556,
      "step": 770
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.420203447341919,
      "learning_rate": 0.0001,
      "loss": 7.8865,
      "step": 780
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.126854419708252,
      "learning_rate": 0.0001,
      "loss": 7.9135,
      "step": 790
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.1468126773834229,
      "learning_rate": 0.0001,
      "loss": 7.9101,
      "step": 800
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2076383829116821,
      "learning_rate": 0.0001,
      "loss": 7.8781,
      "step": 810
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2358585596084595,
      "learning_rate": 0.0001,
      "loss": 7.8729,
      "step": 820
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.1873291730880737,
      "learning_rate": 0.0001,
      "loss": 7.8703,
      "step": 830
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1813349723815918,
      "learning_rate": 0.0001,
      "loss": 7.9166,
      "step": 840
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1983938217163086,
      "learning_rate": 0.0001,
      "loss": 7.8662,
      "step": 850
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1260148286819458,
      "learning_rate": 0.0001,
      "loss": 7.8488,
      "step": 860
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1964378356933594,
      "learning_rate": 0.0001,
      "loss": 7.8679,
      "step": 870
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2019832134246826,
      "learning_rate": 0.0001,
      "loss": 7.7503,
      "step": 880
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1462825536727905,
      "learning_rate": 0.0001,
      "loss": 7.8578,
      "step": 890
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.561476230621338,
      "learning_rate": 0.0001,
      "loss": 7.8796,
      "step": 900
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.1881765127182007,
      "learning_rate": 0.0001,
      "loss": 7.8308,
      "step": 910
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2523167133331299,
      "learning_rate": 0.0001,
      "loss": 7.7982,
      "step": 920
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3139631748199463,
      "learning_rate": 0.0001,
      "loss": 7.7945,
      "step": 930
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2442649602890015,
      "learning_rate": 0.0001,
      "loss": 7.8018,
      "step": 940
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2717634439468384,
      "learning_rate": 0.0001,
      "loss": 7.8349,
      "step": 950
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3789329528808594,
      "learning_rate": 0.0001,
      "loss": 7.8695,
      "step": 960
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2785015106201172,
      "learning_rate": 0.0001,
      "loss": 7.7706,
      "step": 970
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3102847337722778,
      "learning_rate": 0.0001,
      "loss": 7.7189,
      "step": 980
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.443925380706787,
      "learning_rate": 0.0001,
      "loss": 7.7913,
      "step": 990
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3788172006607056,
      "learning_rate": 0.0001,
      "loss": 7.7747,
      "step": 1000
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3770830631256104,
      "learning_rate": 0.0001,
      "loss": 7.7729,
      "step": 1010
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.492782711982727,
      "learning_rate": 0.0001,
      "loss": 7.773,
      "step": 1020
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.284954309463501,
      "learning_rate": 0.0001,
      "loss": 7.819,
      "step": 1030
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.627119779586792,
      "learning_rate": 0.0001,
      "loss": 7.6785,
      "step": 1040
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3385231494903564,
      "learning_rate": 0.0001,
      "loss": 7.792,
      "step": 1050
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3017216920852661,
      "learning_rate": 0.0001,
      "loss": 7.7978,
      "step": 1060
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3827013969421387,
      "learning_rate": 0.0001,
      "loss": 7.712,
      "step": 1070
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4803731441497803,
      "learning_rate": 0.0001,
      "loss": 7.7092,
      "step": 1080
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.305528163909912,
      "learning_rate": 0.0001,
      "loss": 7.7423,
      "step": 1090
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4556981325149536,
      "learning_rate": 0.0001,
      "loss": 7.7396,
      "step": 1100
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.370935320854187,
      "learning_rate": 0.0001,
      "loss": 7.7866,
      "step": 1110
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5131551027297974,
      "learning_rate": 0.0001,
      "loss": 7.6658,
      "step": 1120
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4023760557174683,
      "learning_rate": 0.0001,
      "loss": 7.7168,
      "step": 1130
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4399409294128418,
      "learning_rate": 0.0001,
      "loss": 7.776,
      "step": 1140
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3380918502807617,
      "learning_rate": 0.0001,
      "loss": 7.6652,
      "step": 1150
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3455390930175781,
      "learning_rate": 0.0001,
      "loss": 7.7044,
      "step": 1160
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.3240538835525513,
      "learning_rate": 0.0001,
      "loss": 7.7477,
      "step": 1170
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.282774567604065,
      "learning_rate": 0.0001,
      "loss": 7.7434,
      "step": 1180
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.3946601152420044,
      "learning_rate": 0.0001,
      "loss": 7.7023,
      "step": 1190
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4039177894592285,
      "learning_rate": 0.0001,
      "loss": 7.6228,
      "step": 1200
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.2924675941467285,
      "learning_rate": 0.0001,
      "loss": 7.6144,
      "step": 1210
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3153537511825562,
      "learning_rate": 0.0001,
      "loss": 7.661,
      "step": 1220
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4871569871902466,
      "learning_rate": 0.0001,
      "loss": 7.6486,
      "step": 1230
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.386214256286621,
      "learning_rate": 0.0001,
      "loss": 7.5779,
      "step": 1240
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.7129216194152832,
      "learning_rate": 0.0001,
      "loss": 7.7321,
      "step": 1250
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4541782140731812,
      "learning_rate": 0.0001,
      "loss": 7.6519,
      "step": 1260
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.459065556526184,
      "learning_rate": 0.0001,
      "loss": 7.68,
      "step": 1270
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4400455951690674,
      "learning_rate": 0.0001,
      "loss": 7.6245,
      "step": 1280
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4551613330841064,
      "learning_rate": 0.0001,
      "loss": 7.6565,
      "step": 1290
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3700193166732788,
      "learning_rate": 0.0001,
      "loss": 7.629,
      "step": 1300
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3817769289016724,
      "learning_rate": 0.0001,
      "loss": 7.6016,
      "step": 1310
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6119599342346191,
      "learning_rate": 0.0001,
      "loss": 7.6298,
      "step": 1320
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4976177215576172,
      "learning_rate": 0.0001,
      "loss": 7.6669,
      "step": 1330
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4302177429199219,
      "learning_rate": 0.0001,
      "loss": 7.6144,
      "step": 1340
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.474238634109497,
      "learning_rate": 0.0001,
      "loss": 7.6288,
      "step": 1350
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4761487245559692,
      "learning_rate": 0.0001,
      "loss": 7.6317,
      "step": 1360
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5451174974441528,
      "learning_rate": 0.0001,
      "loss": 7.5998,
      "step": 1370
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.436260461807251,
      "learning_rate": 0.0001,
      "loss": 7.5431,
      "step": 1380
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5143426656723022,
      "learning_rate": 0.0001,
      "loss": 7.701,
      "step": 1390
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6152522563934326,
      "learning_rate": 0.0001,
      "loss": 7.6366,
      "step": 1400
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4636591672897339,
      "learning_rate": 0.0001,
      "loss": 7.6032,
      "step": 1410
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4276256561279297,
      "learning_rate": 0.0001,
      "loss": 7.621,
      "step": 1420
    },
    {
      "epoch": 0.49,
      "grad_norm": 2.719571590423584,
      "learning_rate": 0.0001,
      "loss": 7.6289,
      "step": 1430
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5515453815460205,
      "learning_rate": 0.0001,
      "loss": 7.5354,
      "step": 1440
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.448424220085144,
      "learning_rate": 0.0001,
      "loss": 7.546,
      "step": 1450
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4829095602035522,
      "learning_rate": 0.0001,
      "loss": 7.6309,
      "step": 1460
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5336196422576904,
      "learning_rate": 0.0001,
      "loss": 7.5966,
      "step": 1470
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4048209190368652,
      "learning_rate": 0.0001,
      "loss": 7.5463,
      "step": 1480
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5213136672973633,
      "learning_rate": 0.0001,
      "loss": 7.5134,
      "step": 1490
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5125114917755127,
      "learning_rate": 0.0001,
      "loss": 7.5627,
      "step": 1500
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3993090391159058,
      "learning_rate": 0.0001,
      "loss": 7.4983,
      "step": 1510
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5257900953292847,
      "learning_rate": 0.0001,
      "loss": 7.5514,
      "step": 1520
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.425870418548584,
      "learning_rate": 0.0001,
      "loss": 7.5658,
      "step": 1530
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5099962949752808,
      "learning_rate": 0.0001,
      "loss": 7.5112,
      "step": 1540
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5617893934249878,
      "learning_rate": 0.0001,
      "loss": 7.4845,
      "step": 1550
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4649708271026611,
      "learning_rate": 0.0001,
      "loss": 7.5263,
      "step": 1560
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5084022283554077,
      "learning_rate": 0.0001,
      "loss": 7.6381,
      "step": 1570
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5481611490249634,
      "learning_rate": 0.0001,
      "loss": 7.5124,
      "step": 1580
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4792182445526123,
      "learning_rate": 0.0001,
      "loss": 7.4572,
      "step": 1590
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4853042364120483,
      "learning_rate": 0.0001,
      "loss": 7.4854,
      "step": 1600
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.365810513496399,
      "learning_rate": 0.0001,
      "loss": 8.7212,
      "step": 1610
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.451943039894104,
      "learning_rate": 0.0001,
      "loss": 8.6955,
      "step": 1620
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3740949630737305,
      "learning_rate": 0.0001,
      "loss": 8.6539,
      "step": 1630
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3139616250991821,
      "learning_rate": 0.0001,
      "loss": 8.6498,
      "step": 1640
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3916865587234497,
      "learning_rate": 0.0001,
      "loss": 8.6318,
      "step": 1650
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4029638767242432,
      "learning_rate": 0.0001,
      "loss": 8.5894,
      "step": 1660
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.49574613571167,
      "learning_rate": 0.0001,
      "loss": 8.568,
      "step": 1670
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4445033073425293,
      "learning_rate": 0.0001,
      "loss": 8.538,
      "step": 1680
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.495067834854126,
      "learning_rate": 0.0001,
      "loss": 8.5365,
      "step": 1690
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4277544021606445,
      "learning_rate": 0.0001,
      "loss": 8.5172,
      "step": 1700
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4678715467453003,
      "learning_rate": 0.0001,
      "loss": 8.4763,
      "step": 1710
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5344581604003906,
      "learning_rate": 0.0001,
      "loss": 8.4817,
      "step": 1720
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4848291873931885,
      "learning_rate": 0.0001,
      "loss": 8.4339,
      "step": 1730
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8756532669067383,
      "learning_rate": 0.0001,
      "loss": 8.4432,
      "step": 1740
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.491782307624817,
      "learning_rate": 0.0001,
      "loss": 8.3983,
      "step": 1750
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4570878744125366,
      "learning_rate": 0.0001,
      "loss": 8.4038,
      "step": 1760
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5003529787063599,
      "learning_rate": 0.0001,
      "loss": 8.3846,
      "step": 1770
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5071924924850464,
      "learning_rate": 0.0001,
      "loss": 8.3769,
      "step": 1780
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.1756210327148438,
      "learning_rate": 0.0001,
      "loss": 8.3569,
      "step": 1790
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5465497970581055,
      "learning_rate": 0.0001,
      "loss": 8.3511,
      "step": 1800
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4990923404693604,
      "learning_rate": 0.0001,
      "loss": 8.3228,
      "step": 1810
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4488916397094727,
      "learning_rate": 0.0001,
      "loss": 8.3425,
      "step": 1820
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4601061344146729,
      "learning_rate": 0.0001,
      "loss": 8.3078,
      "step": 1830
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.4482914209365845,
      "learning_rate": 0.0001,
      "loss": 8.3215,
      "step": 1840
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5367404222488403,
      "learning_rate": 0.0001,
      "loss": 8.279,
      "step": 1850
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5058660507202148,
      "learning_rate": 0.0001,
      "loss": 8.2561,
      "step": 1860
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4864140748977661,
      "learning_rate": 0.0001,
      "loss": 8.2908,
      "step": 1870
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6639902591705322,
      "learning_rate": 0.0001,
      "loss": 8.2606,
      "step": 1880
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6261042356491089,
      "learning_rate": 0.0001,
      "loss": 8.2199,
      "step": 1890
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6041576862335205,
      "learning_rate": 0.0001,
      "loss": 8.2398,
      "step": 1900
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4379785060882568,
      "learning_rate": 0.0001,
      "loss": 8.2312,
      "step": 1910
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5959835052490234,
      "learning_rate": 0.0001,
      "loss": 8.234,
      "step": 1920
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6284434795379639,
      "learning_rate": 0.0001,
      "loss": 8.1663,
      "step": 1930
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.622789978981018,
      "learning_rate": 0.0001,
      "loss": 8.1825,
      "step": 1940
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.634716272354126,
      "learning_rate": 0.0001,
      "loss": 8.1376,
      "step": 1950
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6884840726852417,
      "learning_rate": 0.0001,
      "loss": 8.2065,
      "step": 1960
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5878825187683105,
      "learning_rate": 0.0001,
      "loss": 8.1239,
      "step": 1970
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5738948583602905,
      "learning_rate": 0.0001,
      "loss": 8.1848,
      "step": 1980
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.68436861038208,
      "learning_rate": 0.0001,
      "loss": 8.1342,
      "step": 1990
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5968626737594604,
      "learning_rate": 0.0001,
      "loss": 8.1406,
      "step": 2000
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5772038698196411,
      "learning_rate": 0.0001,
      "loss": 8.1452,
      "step": 2010
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.751119613647461,
      "learning_rate": 0.0001,
      "loss": 8.0963,
      "step": 2020
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6801263093948364,
      "learning_rate": 0.0001,
      "loss": 8.1235,
      "step": 2030
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6694085597991943,
      "learning_rate": 0.0001,
      "loss": 8.0969,
      "step": 2040
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7246935367584229,
      "learning_rate": 0.0001,
      "loss": 8.1064,
      "step": 2050
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.708856225013733,
      "learning_rate": 0.0001,
      "loss": 8.0842,
      "step": 2060
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.648662805557251,
      "learning_rate": 0.0001,
      "loss": 8.0483,
      "step": 2070
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6835368871688843,
      "learning_rate": 0.0001,
      "loss": 8.0595,
      "step": 2080
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.725957989692688,
      "learning_rate": 0.0001,
      "loss": 8.0901,
      "step": 2090
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6435219049453735,
      "learning_rate": 0.0001,
      "loss": 8.0726,
      "step": 2100
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6143757104873657,
      "learning_rate": 0.0001,
      "loss": 7.9797,
      "step": 2110
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8680998086929321,
      "learning_rate": 0.0001,
      "loss": 8.0061,
      "step": 2120
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7306435108184814,
      "learning_rate": 0.0001,
      "loss": 8.0243,
      "step": 2130
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7048896551132202,
      "learning_rate": 0.0001,
      "loss": 8.0616,
      "step": 2140
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7379491329193115,
      "learning_rate": 0.0001,
      "loss": 8.0672,
      "step": 2150
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7977313995361328,
      "learning_rate": 0.0001,
      "loss": 8.0118,
      "step": 2160
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.70112943649292,
      "learning_rate": 0.0001,
      "loss": 7.9786,
      "step": 2170
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7871379852294922,
      "learning_rate": 0.0001,
      "loss": 8.0074,
      "step": 2180
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7438380718231201,
      "learning_rate": 0.0001,
      "loss": 8.0325,
      "step": 2190
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8116419315338135,
      "learning_rate": 0.0001,
      "loss": 8.0519,
      "step": 2200
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.9109293222427368,
      "learning_rate": 0.0001,
      "loss": 8.0154,
      "step": 2210
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7357542514801025,
      "learning_rate": 0.0001,
      "loss": 8.0222,
      "step": 2220
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.745240330696106,
      "learning_rate": 0.0001,
      "loss": 7.9692,
      "step": 2230
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.757216453552246,
      "learning_rate": 0.0001,
      "loss": 7.8783,
      "step": 2240
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8003623485565186,
      "learning_rate": 0.0001,
      "loss": 7.9668,
      "step": 2250
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7825720310211182,
      "learning_rate": 0.0001,
      "loss": 7.9397,
      "step": 2260
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.9213591814041138,
      "learning_rate": 0.0001,
      "loss": 8.0387,
      "step": 2270
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8614377975463867,
      "learning_rate": 0.0001,
      "loss": 7.9599,
      "step": 2280
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8351399898529053,
      "learning_rate": 0.0001,
      "loss": 7.9109,
      "step": 2290
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7677947282791138,
      "learning_rate": 0.0001,
      "loss": 7.9349,
      "step": 2300
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.916437029838562,
      "learning_rate": 0.0001,
      "loss": 7.9398,
      "step": 2310
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7147809267044067,
      "learning_rate": 0.0001,
      "loss": 7.8875,
      "step": 2320
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7566685676574707,
      "learning_rate": 0.0001,
      "loss": 8.0169,
      "step": 2330
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.0143845081329346,
      "learning_rate": 0.0001,
      "loss": 7.9364,
      "step": 2340
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8107824325561523,
      "learning_rate": 0.0001,
      "loss": 7.9327,
      "step": 2350
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.817542314529419,
      "learning_rate": 0.0001,
      "loss": 7.923,
      "step": 2360
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7315311431884766,
      "learning_rate": 0.0001,
      "loss": 7.8709,
      "step": 2370
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8613227605819702,
      "learning_rate": 0.0001,
      "loss": 7.8522,
      "step": 2380
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8293215036392212,
      "learning_rate": 0.0001,
      "loss": 7.9198,
      "step": 2390
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8429689407348633,
      "learning_rate": 0.0001,
      "loss": 7.8768,
      "step": 2400
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8556689023971558,
      "learning_rate": 0.0001,
      "loss": 7.8559,
      "step": 2410
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.832610845565796,
      "learning_rate": 0.0001,
      "loss": 7.8548,
      "step": 2420
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8542442321777344,
      "learning_rate": 0.0001,
      "loss": 7.8004,
      "step": 2430
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8391692638397217,
      "learning_rate": 0.0001,
      "loss": 7.8565,
      "step": 2440
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.823958396911621,
      "learning_rate": 0.0001,
      "loss": 7.8082,
      "step": 2450
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.9417130947113037,
      "learning_rate": 0.0001,
      "loss": 7.8148,
      "step": 2460
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.882686734199524,
      "learning_rate": 0.0001,
      "loss": 7.8759,
      "step": 2470
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.1273910999298096,
      "learning_rate": 0.0001,
      "loss": 7.7771,
      "step": 2480
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.8162022829055786,
      "learning_rate": 0.0001,
      "loss": 7.8569,
      "step": 2490
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7690000534057617,
      "learning_rate": 0.0001,
      "loss": 7.8044,
      "step": 2500
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.802037000656128,
      "learning_rate": 0.0001,
      "loss": 7.7328,
      "step": 2510
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.9244282245635986,
      "learning_rate": 0.0001,
      "loss": 7.9286,
      "step": 2520
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8962161540985107,
      "learning_rate": 0.0001,
      "loss": 7.8084,
      "step": 2530
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.9069162607192993,
      "learning_rate": 0.0001,
      "loss": 7.7968,
      "step": 2540
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8832062482833862,
      "learning_rate": 0.0001,
      "loss": 7.8471,
      "step": 2550
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8760149478912354,
      "learning_rate": 0.0001,
      "loss": 7.7647,
      "step": 2560
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.9329017400741577,
      "learning_rate": 0.0001,
      "loss": 7.7911,
      "step": 2570
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.9016661643981934,
      "learning_rate": 0.0001,
      "loss": 7.728,
      "step": 2580
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.900335431098938,
      "learning_rate": 0.0001,
      "loss": 7.7218,
      "step": 2590
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.9823989868164062,
      "learning_rate": 0.0001,
      "loss": 7.7499,
      "step": 2600
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8205046653747559,
      "learning_rate": 0.0001,
      "loss": 7.7044,
      "step": 2610
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.9539875984191895,
      "learning_rate": 0.0001,
      "loss": 7.7712,
      "step": 2620
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.910416841506958,
      "learning_rate": 0.0001,
      "loss": 7.7335,
      "step": 2630
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.0304157733917236,
      "learning_rate": 0.0001,
      "loss": 7.7906,
      "step": 2640
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.9031319618225098,
      "learning_rate": 0.0001,
      "loss": 7.7231,
      "step": 2650
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8728690147399902,
      "learning_rate": 0.0001,
      "loss": 7.7539,
      "step": 2660
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.2379097938537598,
      "learning_rate": 0.0001,
      "loss": 7.8118,
      "step": 2670
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.289486885070801,
      "learning_rate": 0.0001,
      "loss": 7.7609,
      "step": 2680
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.0566112995147705,
      "learning_rate": 0.0001,
      "loss": 7.7602,
      "step": 2690
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.9911762475967407,
      "learning_rate": 0.0001,
      "loss": 7.6783,
      "step": 2700
    },
    {
      "epoch": 0.92,
      "grad_norm": 1.8722708225250244,
      "learning_rate": 0.0001,
      "loss": 7.7683,
      "step": 2710
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.0248823165893555,
      "learning_rate": 0.0001,
      "loss": 7.6759,
      "step": 2720
    },
    {
      "epoch": 0.93,
      "grad_norm": 2.038174867630005,
      "learning_rate": 0.0001,
      "loss": 7.6805,
      "step": 2730
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.9096912145614624,
      "learning_rate": 0.0001,
      "loss": 7.622,
      "step": 2740
    },
    {
      "epoch": 0.93,
      "grad_norm": 1.9723740816116333,
      "learning_rate": 0.0001,
      "loss": 7.7454,
      "step": 2750
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.0898845195770264,
      "learning_rate": 0.0001,
      "loss": 7.7406,
      "step": 2760
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.9320049285888672,
      "learning_rate": 0.0001,
      "loss": 7.6579,
      "step": 2770
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.080528497695923,
      "learning_rate": 0.0001,
      "loss": 7.7255,
      "step": 2780
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.2108402252197266,
      "learning_rate": 0.0001,
      "loss": 7.6704,
      "step": 2790
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.9102885723114014,
      "learning_rate": 0.0001,
      "loss": 7.6922,
      "step": 2800
    },
    {
      "epoch": 0.95,
      "grad_norm": 2.090742349624634,
      "learning_rate": 0.0001,
      "loss": 7.6837,
      "step": 2810
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.0163843631744385,
      "learning_rate": 0.0001,
      "loss": 7.6791,
      "step": 2820
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.0053796768188477,
      "learning_rate": 0.0001,
      "loss": 7.7171,
      "step": 2830
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.956453800201416,
      "learning_rate": 0.0001,
      "loss": 7.6368,
      "step": 2840
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.0385525226593018,
      "learning_rate": 0.0001,
      "loss": 7.5701,
      "step": 2850
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.113161325454712,
      "learning_rate": 0.0001,
      "loss": 7.666,
      "step": 2860
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.1949219703674316,
      "learning_rate": 0.0001,
      "loss": 7.7483,
      "step": 2870
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.9858150482177734,
      "learning_rate": 0.0001,
      "loss": 7.5969,
      "step": 2880
    },
    {
      "epoch": 0.98,
      "grad_norm": 2.061603546142578,
      "learning_rate": 0.0001,
      "loss": 7.6143,
      "step": 2890
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.8981903791427612,
      "learning_rate": 0.0001,
      "loss": 7.6184,
      "step": 2900
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.9247734546661377,
      "learning_rate": 0.0001,
      "loss": 7.5995,
      "step": 2910
    },
    {
      "epoch": 0.99,
      "grad_norm": 2.1106338500976562,
      "learning_rate": 0.0001,
      "loss": 7.7138,
      "step": 2920
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.051254987716675,
      "learning_rate": 0.0001,
      "loss": 7.7863,
      "step": 2930
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.010615825653076,
      "learning_rate": 0.0001,
      "loss": 7.6377,
      "step": 2940
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9900637865066528,
      "learning_rate": 0.0001,
      "loss": 7.6389,
      "step": 2950
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.054337501525879,
      "learning_rate": 0.0001,
      "loss": 7.6445,
      "step": 2960
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.9585248231887817,
      "learning_rate": 0.0001,
      "loss": 7.6182,
      "step": 2970
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.0628204345703125,
      "learning_rate": 0.0001,
      "loss": 7.5937,
      "step": 2980
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.0061392784118652,
      "learning_rate": 0.0001,
      "loss": 7.6077,
      "step": 2990
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.9314968585968018,
      "learning_rate": 0.0001,
      "loss": 7.592,
      "step": 3000
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.9211483001708984,
      "learning_rate": 0.0001,
      "loss": 7.5896,
      "step": 3010
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.967714786529541,
      "learning_rate": 0.0001,
      "loss": 7.5541,
      "step": 3020
    },
    {
      "epoch": 1.03,
      "grad_norm": 1.9651085138320923,
      "learning_rate": 0.0001,
      "loss": 7.5424,
      "step": 3030
    },
    {
      "epoch": 1.03,
      "grad_norm": 2.044400453567505,
      "learning_rate": 0.0001,
      "loss": 7.5674,
      "step": 3040
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.03792142868042,
      "learning_rate": 0.0001,
      "loss": 7.5019,
      "step": 3050
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.0937745571136475,
      "learning_rate": 0.0001,
      "loss": 7.6097,
      "step": 3060
    },
    {
      "epoch": 1.04,
      "grad_norm": 2.1240177154541016,
      "learning_rate": 0.0001,
      "loss": 7.5857,
      "step": 3070
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.1132490634918213,
      "learning_rate": 0.0001,
      "loss": 7.5241,
      "step": 3080
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.132042646408081,
      "learning_rate": 0.0001,
      "loss": 7.5651,
      "step": 3090
    },
    {
      "epoch": 1.05,
      "grad_norm": 2.032273292541504,
      "learning_rate": 0.0001,
      "loss": 7.5327,
      "step": 3100
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.105538845062256,
      "learning_rate": 0.0001,
      "loss": 7.5053,
      "step": 3110
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.1169042587280273,
      "learning_rate": 0.0001,
      "loss": 7.5133,
      "step": 3120
    },
    {
      "epoch": 1.06,
      "grad_norm": 2.0327932834625244,
      "learning_rate": 0.0001,
      "loss": 7.5425,
      "step": 3130
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.1249454021453857,
      "learning_rate": 0.0001,
      "loss": 7.5475,
      "step": 3140
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.0636510848999023,
      "learning_rate": 0.0001,
      "loss": 7.5902,
      "step": 3150
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.4280169010162354,
      "learning_rate": 0.0001,
      "loss": 7.5048,
      "step": 3160
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.1936709880828857,
      "learning_rate": 0.0001,
      "loss": 7.5676,
      "step": 3170
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.1290299892425537,
      "learning_rate": 0.0001,
      "loss": 7.45,
      "step": 3180
    },
    {
      "epoch": 1.08,
      "grad_norm": 2.076470136642456,
      "learning_rate": 0.0001,
      "loss": 7.4977,
      "step": 3190
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.9378889799118042,
      "learning_rate": 0.0001,
      "loss": 7.4409,
      "step": 3200
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.13578724861145,
      "learning_rate": 0.0001,
      "loss": 7.5446,
      "step": 3210
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.1130871772766113,
      "learning_rate": 0.0001,
      "loss": 7.5431,
      "step": 3220
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.072411298751831,
      "learning_rate": 0.0001,
      "loss": 7.3864,
      "step": 3230
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.1405575275421143,
      "learning_rate": 0.0001,
      "loss": 7.4161,
      "step": 3240
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.0535531044006348,
      "learning_rate": 0.0001,
      "loss": 7.4866,
      "step": 3250
    },
    {
      "epoch": 1.11,
      "grad_norm": 2.2262203693389893,
      "learning_rate": 0.0001,
      "loss": 7.5469,
      "step": 3260
    },
    {
      "epoch": 1.11,
      "grad_norm": 2.0053820610046387,
      "learning_rate": 0.0001,
      "loss": 7.4489,
      "step": 3270
    },
    {
      "epoch": 1.11,
      "grad_norm": 2.116321325302124,
      "learning_rate": 0.0001,
      "loss": 7.4585,
      "step": 3280
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.023549795150757,
      "learning_rate": 0.0001,
      "loss": 7.544,
      "step": 3290
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.096761465072632,
      "learning_rate": 0.0001,
      "loss": 7.408,
      "step": 3300
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.1344809532165527,
      "learning_rate": 0.0001,
      "loss": 7.4602,
      "step": 3310
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.1644928455352783,
      "learning_rate": 0.0001,
      "loss": 7.4324,
      "step": 3320
    },
    {
      "epoch": 1.13,
      "grad_norm": 2.2048068046569824,
      "learning_rate": 0.0001,
      "loss": 7.5078,
      "step": 3330
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.1286168098449707,
      "learning_rate": 0.0001,
      "loss": 7.4117,
      "step": 3340
    },
    {
      "epoch": 1.14,
      "grad_norm": 1.9755229949951172,
      "learning_rate": 0.0001,
      "loss": 7.4233,
      "step": 3350
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.165160655975342,
      "learning_rate": 0.0001,
      "loss": 7.5074,
      "step": 3360
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.085435152053833,
      "learning_rate": 0.0001,
      "loss": 7.4901,
      "step": 3370
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.0615718364715576,
      "learning_rate": 0.0001,
      "loss": 7.4103,
      "step": 3380
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.113546371459961,
      "learning_rate": 0.0001,
      "loss": 7.4724,
      "step": 3390
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.0932693481445312,
      "learning_rate": 0.0001,
      "loss": 7.4305,
      "step": 3400
    },
    {
      "epoch": 1.16,
      "grad_norm": 2.0436511039733887,
      "learning_rate": 0.0001,
      "loss": 7.4556,
      "step": 3410
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.9966588020324707,
      "learning_rate": 0.0001,
      "loss": 7.3095,
      "step": 3420
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.1284430027008057,
      "learning_rate": 0.0001,
      "loss": 7.3911,
      "step": 3430
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.105534553527832,
      "learning_rate": 0.0001,
      "loss": 7.4409,
      "step": 3440
    },
    {
      "epoch": 1.17,
      "grad_norm": 2.23077654838562,
      "learning_rate": 0.0001,
      "loss": 7.4349,
      "step": 3450
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.1125681400299072,
      "learning_rate": 0.0001,
      "loss": 7.4981,
      "step": 3460
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.067131757736206,
      "learning_rate": 0.0001,
      "loss": 7.4678,
      "step": 3470
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.220170021057129,
      "learning_rate": 0.0001,
      "loss": 7.4274,
      "step": 3480
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.143610954284668,
      "learning_rate": 0.0001,
      "loss": 7.4003,
      "step": 3490
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.099076747894287,
      "learning_rate": 0.0001,
      "loss": 7.383,
      "step": 3500
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.046013355255127,
      "learning_rate": 0.0001,
      "loss": 7.3424,
      "step": 3510
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.070675849914551,
      "learning_rate": 0.0001,
      "loss": 7.4347,
      "step": 3520
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.0689220428466797,
      "learning_rate": 0.0001,
      "loss": 7.402,
      "step": 3530
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.303022623062134,
      "learning_rate": 0.0001,
      "loss": 7.3843,
      "step": 3540
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.153381586074829,
      "learning_rate": 0.0001,
      "loss": 7.3525,
      "step": 3550
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.286656618118286,
      "learning_rate": 0.0001,
      "loss": 7.3751,
      "step": 3560
    },
    {
      "epoch": 1.21,
      "grad_norm": 2.0820603370666504,
      "learning_rate": 0.0001,
      "loss": 7.3258,
      "step": 3570
    },
    {
      "epoch": 1.22,
      "grad_norm": 3.858593702316284,
      "learning_rate": 0.0001,
      "loss": 7.3927,
      "step": 3580
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.2527058124542236,
      "learning_rate": 0.0001,
      "loss": 7.3588,
      "step": 3590
    },
    {
      "epoch": 1.22,
      "grad_norm": 2.079145669937134,
      "learning_rate": 0.0001,
      "loss": 7.374,
      "step": 3600
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.1242687702178955,
      "learning_rate": 0.0001,
      "loss": 7.3516,
      "step": 3610
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.272284984588623,
      "learning_rate": 0.0001,
      "loss": 7.3299,
      "step": 3620
    },
    {
      "epoch": 1.23,
      "grad_norm": 2.10646653175354,
      "learning_rate": 0.0001,
      "loss": 7.4278,
      "step": 3630
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.0857315063476562,
      "learning_rate": 0.0001,
      "loss": 7.3894,
      "step": 3640
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.268275260925293,
      "learning_rate": 0.0001,
      "loss": 7.342,
      "step": 3650
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.0469658374786377,
      "learning_rate": 0.0001,
      "loss": 7.3556,
      "step": 3660
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.1571669578552246,
      "learning_rate": 0.0001,
      "loss": 7.3198,
      "step": 3670
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.3082923889160156,
      "learning_rate": 0.0001,
      "loss": 7.3687,
      "step": 3680
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.8455331325531006,
      "learning_rate": 0.0001,
      "loss": 7.3044,
      "step": 3690
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.215144634246826,
      "learning_rate": 0.0001,
      "loss": 7.2532,
      "step": 3700
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.159938097000122,
      "learning_rate": 0.0001,
      "loss": 7.292,
      "step": 3710
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.175513744354248,
      "learning_rate": 0.0001,
      "loss": 7.4197,
      "step": 3720
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.287431478500366,
      "learning_rate": 0.0001,
      "loss": 7.3119,
      "step": 3730
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.1502156257629395,
      "learning_rate": 0.0001,
      "loss": 7.2915,
      "step": 3740
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.184866189956665,
      "learning_rate": 0.0001,
      "loss": 7.3259,
      "step": 3750
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.73464298248291,
      "learning_rate": 0.0001,
      "loss": 7.299,
      "step": 3760
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.0966055393218994,
      "learning_rate": 0.0001,
      "loss": 7.3202,
      "step": 3770
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.2224791049957275,
      "learning_rate": 0.0001,
      "loss": 7.2796,
      "step": 3780
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.3259451389312744,
      "learning_rate": 0.0001,
      "loss": 7.3665,
      "step": 3790
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.251563310623169,
      "learning_rate": 0.0001,
      "loss": 7.3481,
      "step": 3800
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.183901786804199,
      "learning_rate": 0.0001,
      "loss": 7.2849,
      "step": 3810
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.1441636085510254,
      "learning_rate": 0.0001,
      "loss": 7.3627,
      "step": 3820
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.17110538482666,
      "learning_rate": 0.0001,
      "loss": 7.2228,
      "step": 3830
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.1089303493499756,
      "learning_rate": 0.0001,
      "loss": 7.3896,
      "step": 3840
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.2746925354003906,
      "learning_rate": 0.0001,
      "loss": 7.3009,
      "step": 3850
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.163635730743408,
      "learning_rate": 0.0001,
      "loss": 7.2944,
      "step": 3860
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.1355977058410645,
      "learning_rate": 0.0001,
      "loss": 7.2778,
      "step": 3870
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.229297161102295,
      "learning_rate": 0.0001,
      "loss": 7.3898,
      "step": 3880
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.233773708343506,
      "learning_rate": 0.0001,
      "loss": 7.2333,
      "step": 3890
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.2442920207977295,
      "learning_rate": 0.0001,
      "loss": 7.3783,
      "step": 3900
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.113609552383423,
      "learning_rate": 0.0001,
      "loss": 7.2231,
      "step": 3910
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.049269199371338,
      "learning_rate": 0.0001,
      "loss": 7.2754,
      "step": 3920
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.2847328186035156,
      "learning_rate": 0.0001,
      "loss": 7.2955,
      "step": 3930
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.0207245349884033,
      "learning_rate": 0.0001,
      "loss": 7.2091,
      "step": 3940
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.120429277420044,
      "learning_rate": 0.0001,
      "loss": 7.3541,
      "step": 3950
    },
    {
      "epoch": 1.35,
      "grad_norm": 2.1240029335021973,
      "learning_rate": 0.0001,
      "loss": 7.2128,
      "step": 3960
    },
    {
      "epoch": 1.35,
      "grad_norm": 2.112290859222412,
      "learning_rate": 0.0001,
      "loss": 7.3131,
      "step": 3970
    },
    {
      "epoch": 1.35,
      "grad_norm": 2.1971917152404785,
      "learning_rate": 0.0001,
      "loss": 7.1242,
      "step": 3980
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.2146995067596436,
      "learning_rate": 0.0001,
      "loss": 7.307,
      "step": 3990
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.1895055770874023,
      "learning_rate": 0.0001,
      "loss": 7.2525,
      "step": 4000
    },
    {
      "epoch": 1.36,
      "grad_norm": 2.1644904613494873,
      "learning_rate": 0.0001,
      "loss": 7.2756,
      "step": 4010
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.307317018508911,
      "learning_rate": 0.0001,
      "loss": 7.2806,
      "step": 4020
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.1317138671875,
      "learning_rate": 0.0001,
      "loss": 7.1696,
      "step": 4030
    },
    {
      "epoch": 1.37,
      "grad_norm": 2.378333568572998,
      "learning_rate": 0.0001,
      "loss": 7.2553,
      "step": 4040
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.156888246536255,
      "learning_rate": 0.0001,
      "loss": 7.2355,
      "step": 4050
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.226720094680786,
      "learning_rate": 0.0001,
      "loss": 7.24,
      "step": 4060
    },
    {
      "epoch": 1.38,
      "grad_norm": 2.2522284984588623,
      "learning_rate": 0.0001,
      "loss": 7.2193,
      "step": 4070
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.2693936824798584,
      "learning_rate": 0.0001,
      "loss": 7.2314,
      "step": 4080
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.3695926666259766,
      "learning_rate": 0.0001,
      "loss": 7.2259,
      "step": 4090
    },
    {
      "epoch": 1.39,
      "grad_norm": 2.290940761566162,
      "learning_rate": 0.0001,
      "loss": 7.2789,
      "step": 4100
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.2457990646362305,
      "learning_rate": 0.0001,
      "loss": 7.2184,
      "step": 4110
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.299743413925171,
      "learning_rate": 0.0001,
      "loss": 7.2948,
      "step": 4120
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.2357537746429443,
      "learning_rate": 0.0001,
      "loss": 7.2268,
      "step": 4130
    },
    {
      "epoch": 1.41,
      "grad_norm": 2.3687903881073,
      "learning_rate": 0.0001,
      "loss": 7.2935,
      "step": 4140
    },
    {
      "epoch": 1.41,
      "grad_norm": 2.255699396133423,
      "learning_rate": 0.0001,
      "loss": 7.1908,
      "step": 4150
    },
    {
      "epoch": 1.41,
      "grad_norm": 2.3741886615753174,
      "learning_rate": 0.0001,
      "loss": 7.1706,
      "step": 4160
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.2647061347961426,
      "learning_rate": 0.0001,
      "loss": 7.3281,
      "step": 4170
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.179403781890869,
      "learning_rate": 0.0001,
      "loss": 7.2573,
      "step": 4180
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.2076845169067383,
      "learning_rate": 0.0001,
      "loss": 7.2239,
      "step": 4190
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.1713955402374268,
      "learning_rate": 0.0001,
      "loss": 7.1854,
      "step": 4200
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.224914312362671,
      "learning_rate": 0.0001,
      "loss": 7.1896,
      "step": 4210
    },
    {
      "epoch": 1.43,
      "grad_norm": 2.277390241622925,
      "learning_rate": 0.0001,
      "loss": 7.1617,
      "step": 4220
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.262953996658325,
      "learning_rate": 0.0001,
      "loss": 7.1675,
      "step": 4230
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.307990312576294,
      "learning_rate": 0.0001,
      "loss": 7.2488,
      "step": 4240
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.343857526779175,
      "learning_rate": 0.0001,
      "loss": 7.1453,
      "step": 4250
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.225097179412842,
      "learning_rate": 0.0001,
      "loss": 7.1337,
      "step": 4260
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.240400791168213,
      "learning_rate": 0.0001,
      "loss": 7.2477,
      "step": 4270
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.144904613494873,
      "learning_rate": 0.0001,
      "loss": 7.2,
      "step": 4280
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.24607253074646,
      "learning_rate": 0.0001,
      "loss": 7.1434,
      "step": 4290
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.3587682247161865,
      "learning_rate": 0.0001,
      "loss": 7.2904,
      "step": 4300
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.3396973609924316,
      "learning_rate": 0.0001,
      "loss": 7.157,
      "step": 4310
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.254444122314453,
      "learning_rate": 0.0001,
      "loss": 7.203,
      "step": 4320
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.3054215908050537,
      "learning_rate": 0.0001,
      "loss": 7.177,
      "step": 4330
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.1978611946105957,
      "learning_rate": 0.0001,
      "loss": 7.1707,
      "step": 4340
    },
    {
      "epoch": 1.48,
      "grad_norm": 2.1555399894714355,
      "learning_rate": 0.0001,
      "loss": 7.0313,
      "step": 4350
    },
    {
      "epoch": 1.48,
      "grad_norm": 2.261692762374878,
      "learning_rate": 0.0001,
      "loss": 7.1372,
      "step": 4360
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.1888833045959473,
      "learning_rate": 0.0001,
      "loss": 7.1493,
      "step": 4370
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.3072457313537598,
      "learning_rate": 0.0001,
      "loss": 7.0555,
      "step": 4380
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.1015727519989014,
      "learning_rate": 0.0001,
      "loss": 7.2164,
      "step": 4390
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.2788422107696533,
      "learning_rate": 0.0001,
      "loss": 7.1169,
      "step": 4400
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.180575132369995,
      "learning_rate": 0.0001,
      "loss": 7.1008,
      "step": 4410
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.234184503555298,
      "learning_rate": 0.0001,
      "loss": 7.1023,
      "step": 4420
    },
    {
      "epoch": 1.51,
      "grad_norm": 2.4093639850616455,
      "learning_rate": 0.0001,
      "loss": 7.1913,
      "step": 4430
    },
    {
      "epoch": 1.51,
      "grad_norm": 2.1984338760375977,
      "learning_rate": 0.0001,
      "loss": 7.2237,
      "step": 4440
    },
    {
      "epoch": 1.51,
      "grad_norm": 2.2517895698547363,
      "learning_rate": 0.0001,
      "loss": 7.114,
      "step": 4450
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.231074094772339,
      "learning_rate": 0.0001,
      "loss": 7.1897,
      "step": 4460
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.7863545417785645,
      "learning_rate": 0.0001,
      "loss": 7.0944,
      "step": 4470
    },
    {
      "epoch": 1.52,
      "grad_norm": 2.2908711433410645,
      "learning_rate": 0.0001,
      "loss": 7.1134,
      "step": 4480
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.216499090194702,
      "learning_rate": 0.0001,
      "loss": 7.0969,
      "step": 4490
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.1753108501434326,
      "learning_rate": 0.0001,
      "loss": 7.1187,
      "step": 4500
    },
    {
      "epoch": 1.53,
      "grad_norm": 2.310413360595703,
      "learning_rate": 0.0001,
      "loss": 7.183,
      "step": 4510
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.270989179611206,
      "learning_rate": 0.0001,
      "loss": 7.1332,
      "step": 4520
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.1965484619140625,
      "learning_rate": 0.0001,
      "loss": 7.1457,
      "step": 4530
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.1956119537353516,
      "learning_rate": 0.0001,
      "loss": 7.2644,
      "step": 4540
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.54412579536438,
      "learning_rate": 0.0001,
      "loss": 7.1193,
      "step": 4550
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.1950461864471436,
      "learning_rate": 0.0001,
      "loss": 7.1085,
      "step": 4560
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.4147634506225586,
      "learning_rate": 0.0001,
      "loss": 7.1386,
      "step": 4570
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.2770273685455322,
      "learning_rate": 0.0001,
      "loss": 7.1362,
      "step": 4580
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.305471181869507,
      "learning_rate": 0.0001,
      "loss": 7.0386,
      "step": 4590
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.2918567657470703,
      "learning_rate": 0.0001,
      "loss": 7.1585,
      "step": 4600
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.2294692993164062,
      "learning_rate": 0.0001,
      "loss": 7.1507,
      "step": 4610
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.0887460708618164,
      "learning_rate": 0.0001,
      "loss": 7.099,
      "step": 4620
    },
    {
      "epoch": 1.57,
      "grad_norm": 2.202714681625366,
      "learning_rate": 0.0001,
      "loss": 7.193,
      "step": 4630
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.342766761779785,
      "learning_rate": 0.0001,
      "loss": 7.0854,
      "step": 4640
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.3064093589782715,
      "learning_rate": 0.0001,
      "loss": 7.1019,
      "step": 4650
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.1647086143493652,
      "learning_rate": 0.0001,
      "loss": 6.9739,
      "step": 4660
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.220046281814575,
      "learning_rate": 0.0001,
      "loss": 7.1119,
      "step": 4670
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.478766918182373,
      "learning_rate": 0.0001,
      "loss": 7.0424,
      "step": 4680
    },
    {
      "epoch": 1.59,
      "grad_norm": 2.349714994430542,
      "learning_rate": 0.0001,
      "loss": 7.1136,
      "step": 4690
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.365969657897949,
      "learning_rate": 0.0001,
      "loss": 7.1204,
      "step": 4700
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.3459932804107666,
      "learning_rate": 0.0001,
      "loss": 7.0308,
      "step": 4710
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.2051196098327637,
      "learning_rate": 0.0001,
      "loss": 7.1009,
      "step": 4720
    },
    {
      "epoch": 1.61,
      "grad_norm": 2.2125892639160156,
      "learning_rate": 0.0001,
      "loss": 7.012,
      "step": 4730
    },
    {
      "epoch": 1.61,
      "grad_norm": 2.402265787124634,
      "learning_rate": 0.0001,
      "loss": 7.1284,
      "step": 4740
    },
    {
      "epoch": 1.61,
      "grad_norm": 2.1848082542419434,
      "learning_rate": 0.0001,
      "loss": 7.0313,
      "step": 4750
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.2981090545654297,
      "learning_rate": 0.0001,
      "loss": 7.0365,
      "step": 4760
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.3186306953430176,
      "learning_rate": 0.0001,
      "loss": 7.099,
      "step": 4770
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.371321678161621,
      "learning_rate": 0.0001,
      "loss": 7.0968,
      "step": 4780
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.278888702392578,
      "learning_rate": 0.0001,
      "loss": 7.131,
      "step": 4790
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.2284281253814697,
      "learning_rate": 0.0001,
      "loss": 6.8648,
      "step": 4800
    },
    {
      "epoch": 1.63,
      "grad_norm": 2.272829532623291,
      "learning_rate": 0.0001,
      "loss": 7.0416,
      "step": 4810
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.224722146987915,
      "learning_rate": 0.0001,
      "loss": 7.0871,
      "step": 4820
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.2158703804016113,
      "learning_rate": 0.0001,
      "loss": 7.1796,
      "step": 4830
    },
    {
      "epoch": 1.64,
      "grad_norm": 2.2695987224578857,
      "learning_rate": 0.0001,
      "loss": 7.095,
      "step": 4840
    },
    {
      "epoch": 1.65,
      "grad_norm": 2.4090540409088135,
      "learning_rate": 0.0001,
      "loss": 7.0629,
      "step": 4850
    },
    {
      "epoch": 1.65,
      "grad_norm": 2.3584811687469482,
      "learning_rate": 0.0001,
      "loss": 7.1282,
      "step": 4860
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.288571357727051,
      "learning_rate": 0.0001,
      "loss": 7.0856,
      "step": 4870
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.195589065551758,
      "learning_rate": 0.0001,
      "loss": 6.9848,
      "step": 4880
    },
    {
      "epoch": 1.66,
      "grad_norm": 2.2008442878723145,
      "learning_rate": 0.0001,
      "loss": 7.0626,
      "step": 4890
    },
    {
      "epoch": 1.67,
      "grad_norm": 2.3728606700897217,
      "learning_rate": 0.0001,
      "loss": 7.09,
      "step": 4900
    },
    {
      "epoch": 1.67,
      "grad_norm": 2.2774598598480225,
      "learning_rate": 0.0001,
      "loss": 7.1203,
      "step": 4910
    },
    {
      "epoch": 1.67,
      "grad_norm": 2.3884873390197754,
      "learning_rate": 0.0001,
      "loss": 7.0542,
      "step": 4920
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.33353590965271,
      "learning_rate": 0.0001,
      "loss": 7.0604,
      "step": 4930
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.2535629272460938,
      "learning_rate": 0.0001,
      "loss": 7.04,
      "step": 4940
    },
    {
      "epoch": 1.68,
      "grad_norm": 2.2882211208343506,
      "learning_rate": 0.0001,
      "loss": 7.0488,
      "step": 4950
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.2261734008789062,
      "learning_rate": 0.0001,
      "loss": 7.0044,
      "step": 4960
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.2294704914093018,
      "learning_rate": 0.0001,
      "loss": 7.0224,
      "step": 4970
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.355883836746216,
      "learning_rate": 0.0001,
      "loss": 6.996,
      "step": 4980
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.255924940109253,
      "learning_rate": 0.0001,
      "loss": 7.0028,
      "step": 4990
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.121638298034668,
      "learning_rate": 0.0001,
      "loss": 6.9282,
      "step": 5000
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.522393226623535,
      "learning_rate": 0.0001,
      "loss": 7.0418,
      "step": 5010
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.3869247436523438,
      "learning_rate": 0.0001,
      "loss": 7.1096,
      "step": 5020
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.2230443954467773,
      "learning_rate": 0.0001,
      "loss": 7.0023,
      "step": 5030
    },
    {
      "epoch": 1.71,
      "grad_norm": 2.2751498222351074,
      "learning_rate": 0.0001,
      "loss": 6.9971,
      "step": 5040
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.3354780673980713,
      "learning_rate": 0.0001,
      "loss": 6.9604,
      "step": 5050
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.2902698516845703,
      "learning_rate": 0.0001,
      "loss": 6.9917,
      "step": 5060
    },
    {
      "epoch": 1.72,
      "grad_norm": 2.3068253993988037,
      "learning_rate": 0.0001,
      "loss": 7.0955,
      "step": 5070
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.175537109375,
      "learning_rate": 0.0001,
      "loss": 7.0525,
      "step": 5080
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.30596923828125,
      "learning_rate": 0.0001,
      "loss": 6.9501,
      "step": 5090
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.3087093830108643,
      "learning_rate": 0.0001,
      "loss": 7.068,
      "step": 5100
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.311290979385376,
      "learning_rate": 0.0001,
      "loss": 7.0114,
      "step": 5110
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.3126213550567627,
      "learning_rate": 0.0001,
      "loss": 6.8703,
      "step": 5120
    },
    {
      "epoch": 1.74,
      "grad_norm": 2.339921236038208,
      "learning_rate": 0.0001,
      "loss": 6.9398,
      "step": 5130
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.385369062423706,
      "learning_rate": 0.0001,
      "loss": 7.092,
      "step": 5140
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.2324063777923584,
      "learning_rate": 0.0001,
      "loss": 7.0395,
      "step": 5150
    },
    {
      "epoch": 1.75,
      "grad_norm": 2.1126644611358643,
      "learning_rate": 0.0001,
      "loss": 6.9577,
      "step": 5160
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.3116745948791504,
      "learning_rate": 0.0001,
      "loss": 7.007,
      "step": 5170
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.302995443344116,
      "learning_rate": 0.0001,
      "loss": 7.0157,
      "step": 5180
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.2893717288970947,
      "learning_rate": 0.0001,
      "loss": 7.0359,
      "step": 5190
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.49497652053833,
      "learning_rate": 0.0001,
      "loss": 6.8841,
      "step": 5200
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.4017276763916016,
      "learning_rate": 0.0001,
      "loss": 6.9358,
      "step": 5210
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.2474894523620605,
      "learning_rate": 0.0001,
      "loss": 6.9088,
      "step": 5220
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.430718183517456,
      "learning_rate": 0.0001,
      "loss": 6.9641,
      "step": 5230
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.2758383750915527,
      "learning_rate": 0.0001,
      "loss": 6.8806,
      "step": 5240
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.4353394508361816,
      "learning_rate": 0.0001,
      "loss": 6.9995,
      "step": 5250
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.167844533920288,
      "learning_rate": 0.0001,
      "loss": 6.9315,
      "step": 5260
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.2327005863189697,
      "learning_rate": 0.0001,
      "loss": 6.9489,
      "step": 5270
    },
    {
      "epoch": 1.79,
      "grad_norm": 2.4071407318115234,
      "learning_rate": 0.0001,
      "loss": 6.9209,
      "step": 5280
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.3251023292541504,
      "learning_rate": 0.0001,
      "loss": 7.0063,
      "step": 5290
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.3144476413726807,
      "learning_rate": 0.0001,
      "loss": 6.9339,
      "step": 5300
    },
    {
      "epoch": 1.8,
      "grad_norm": 2.239985942840576,
      "learning_rate": 0.0001,
      "loss": 6.8448,
      "step": 5310
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.3071162700653076,
      "learning_rate": 0.0001,
      "loss": 6.9222,
      "step": 5320
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.3900599479675293,
      "learning_rate": 0.0001,
      "loss": 6.9269,
      "step": 5330
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.181514024734497,
      "learning_rate": 0.0001,
      "loss": 6.8864,
      "step": 5340
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.2172696590423584,
      "learning_rate": 0.0001,
      "loss": 6.9524,
      "step": 5350
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.4472038745880127,
      "learning_rate": 0.0001,
      "loss": 7.0239,
      "step": 5360
    },
    {
      "epoch": 1.82,
      "grad_norm": 2.205364942550659,
      "learning_rate": 0.0001,
      "loss": 6.8968,
      "step": 5370
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.275344133377075,
      "learning_rate": 0.0001,
      "loss": 6.9347,
      "step": 5380
    },
    {
      "epoch": 1.83,
      "grad_norm": 2.2779319286346436,
      "learning_rate": 0.0001,
      "loss": 6.9653,
      "step": 5390
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.321763038635254,
      "learning_rate": 0.0001,
      "loss": 6.9756,
      "step": 5400
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.4092564582824707,
      "learning_rate": 0.0001,
      "loss": 6.9021,
      "step": 5410
    },
    {
      "epoch": 1.84,
      "grad_norm": 2.5679686069488525,
      "learning_rate": 0.0001,
      "loss": 6.9372,
      "step": 5420
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.332031488418579,
      "learning_rate": 0.0001,
      "loss": 6.9713,
      "step": 5430
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.4025161266326904,
      "learning_rate": 0.0001,
      "loss": 6.8268,
      "step": 5440
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.3501081466674805,
      "learning_rate": 0.0001,
      "loss": 6.9569,
      "step": 5450
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.3788626194000244,
      "learning_rate": 0.0001,
      "loss": 6.8993,
      "step": 5460
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.3662962913513184,
      "learning_rate": 0.0001,
      "loss": 6.8359,
      "step": 5470
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.401604175567627,
      "learning_rate": 0.0001,
      "loss": 6.9646,
      "step": 5480
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.348583698272705,
      "learning_rate": 0.0001,
      "loss": 6.913,
      "step": 5490
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.356675386428833,
      "learning_rate": 0.0001,
      "loss": 6.9592,
      "step": 5500
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.2990479469299316,
      "learning_rate": 0.0001,
      "loss": 6.9444,
      "step": 5510
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.4327516555786133,
      "learning_rate": 0.0001,
      "loss": 6.9434,
      "step": 5520
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.2638111114501953,
      "learning_rate": 0.0001,
      "loss": 6.9039,
      "step": 5530
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.3576090335845947,
      "learning_rate": 0.0001,
      "loss": 7.0159,
      "step": 5540
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.226438522338867,
      "learning_rate": 0.0001,
      "loss": 6.9251,
      "step": 5550
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.3056225776672363,
      "learning_rate": 0.0001,
      "loss": 6.915,
      "step": 5560
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.361037254333496,
      "learning_rate": 0.0001,
      "loss": 6.9419,
      "step": 5570
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.2873964309692383,
      "learning_rate": 0.0001,
      "loss": 6.9656,
      "step": 5580
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.1307671070098877,
      "learning_rate": 0.0001,
      "loss": 6.9237,
      "step": 5590
    },
    {
      "epoch": 1.9,
      "grad_norm": 2.331777572631836,
      "learning_rate": 0.0001,
      "loss": 6.8829,
      "step": 5600
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.4056396484375,
      "learning_rate": 0.0001,
      "loss": 6.8417,
      "step": 5610
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.3956289291381836,
      "learning_rate": 0.0001,
      "loss": 6.848,
      "step": 5620
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.404996156692505,
      "learning_rate": 0.0001,
      "loss": 6.8649,
      "step": 5630
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.32171893119812,
      "learning_rate": 0.0001,
      "loss": 6.977,
      "step": 5640
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.45902681350708,
      "learning_rate": 0.0001,
      "loss": 6.905,
      "step": 5650
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.273726224899292,
      "learning_rate": 0.0001,
      "loss": 6.8997,
      "step": 5660
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.427234172821045,
      "learning_rate": 0.0001,
      "loss": 6.8337,
      "step": 5670
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.4655344486236572,
      "learning_rate": 0.0001,
      "loss": 6.8642,
      "step": 5680
    },
    {
      "epoch": 1.93,
      "grad_norm": 2.356198787689209,
      "learning_rate": 0.0001,
      "loss": 6.9213,
      "step": 5690
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.2886674404144287,
      "learning_rate": 0.0001,
      "loss": 6.9349,
      "step": 5700
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.36144757270813,
      "learning_rate": 0.0001,
      "loss": 6.9109,
      "step": 5710
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.34747576713562,
      "learning_rate": 0.0001,
      "loss": 6.875,
      "step": 5720
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.3149468898773193,
      "learning_rate": 0.0001,
      "loss": 6.8872,
      "step": 5730
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.38944673538208,
      "learning_rate": 0.0001,
      "loss": 6.8233,
      "step": 5740
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.441582441329956,
      "learning_rate": 0.0001,
      "loss": 6.9502,
      "step": 5750
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.380875587463379,
      "learning_rate": 0.0001,
      "loss": 6.8632,
      "step": 5760
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.284801721572876,
      "learning_rate": 0.0001,
      "loss": 6.7719,
      "step": 5770
    },
    {
      "epoch": 1.96,
      "grad_norm": 2.3175783157348633,
      "learning_rate": 0.0001,
      "loss": 6.8299,
      "step": 5780
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.3551747798919678,
      "learning_rate": 0.0001,
      "loss": 6.8656,
      "step": 5790
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.3243958950042725,
      "learning_rate": 0.0001,
      "loss": 6.9141,
      "step": 5800
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.3328282833099365,
      "learning_rate": 0.0001,
      "loss": 6.8811,
      "step": 5810
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.4374606609344482,
      "learning_rate": 0.0001,
      "loss": 6.9106,
      "step": 5820
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.3840887546539307,
      "learning_rate": 0.0001,
      "loss": 6.865,
      "step": 5830
    },
    {
      "epoch": 1.98,
      "grad_norm": 2.421757221221924,
      "learning_rate": 0.0001,
      "loss": 6.8836,
      "step": 5840
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.400735855102539,
      "learning_rate": 0.0001,
      "loss": 6.8079,
      "step": 5850
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.3208372592926025,
      "learning_rate": 0.0001,
      "loss": 6.8372,
      "step": 5860
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.2900898456573486,
      "learning_rate": 0.0001,
      "loss": 6.7931,
      "step": 5870
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.4279627799987793,
      "learning_rate": 0.0001,
      "loss": 6.9407,
      "step": 5880
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.536397695541382,
      "learning_rate": 0.0001,
      "loss": 6.7606,
      "step": 5890
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.3431828022003174,
      "learning_rate": 0.0001,
      "loss": 6.7804,
      "step": 5900
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.306082248687744,
      "learning_rate": 0.0001,
      "loss": 6.7596,
      "step": 5910
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.324713945388794,
      "learning_rate": 0.0001,
      "loss": 6.7527,
      "step": 5920
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.4252986907958984,
      "learning_rate": 0.0001,
      "loss": 6.8721,
      "step": 5930
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.2642593383789062,
      "learning_rate": 0.0001,
      "loss": 6.7373,
      "step": 5940
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.3272111415863037,
      "learning_rate": 0.0001,
      "loss": 6.7637,
      "step": 5950
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.428049325942993,
      "learning_rate": 0.0001,
      "loss": 6.77,
      "step": 5960
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.4869771003723145,
      "learning_rate": 0.0001,
      "loss": 6.858,
      "step": 5970
    },
    {
      "epoch": 2.03,
      "grad_norm": 2.3572263717651367,
      "learning_rate": 0.0001,
      "loss": 6.8538,
      "step": 5980
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.6547634601593018,
      "learning_rate": 0.0001,
      "loss": 6.786,
      "step": 5990
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.362126111984253,
      "learning_rate": 0.0001,
      "loss": 6.7479,
      "step": 6000
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.273364305496216,
      "learning_rate": 0.0001,
      "loss": 6.8455,
      "step": 6010
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.3831865787506104,
      "learning_rate": 0.0001,
      "loss": 6.7612,
      "step": 6020
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.3428845405578613,
      "learning_rate": 0.0001,
      "loss": 6.8112,
      "step": 6030
    },
    {
      "epoch": 2.05,
      "grad_norm": 2.3353376388549805,
      "learning_rate": 0.0001,
      "loss": 6.8605,
      "step": 6040
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.5539698600769043,
      "learning_rate": 0.0001,
      "loss": 6.7803,
      "step": 6050
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.363743305206299,
      "learning_rate": 0.0001,
      "loss": 6.7285,
      "step": 6060
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.30212664604187,
      "learning_rate": 0.0001,
      "loss": 6.7403,
      "step": 6070
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.4871535301208496,
      "learning_rate": 0.0001,
      "loss": 6.8449,
      "step": 6080
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.4670472145080566,
      "learning_rate": 0.0001,
      "loss": 6.6744,
      "step": 6090
    },
    {
      "epoch": 2.07,
      "grad_norm": 2.3175411224365234,
      "learning_rate": 0.0001,
      "loss": 6.7953,
      "step": 6100
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.3637678623199463,
      "learning_rate": 0.0001,
      "loss": 6.7753,
      "step": 6110
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.4457314014434814,
      "learning_rate": 0.0001,
      "loss": 6.8608,
      "step": 6120
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.5131587982177734,
      "learning_rate": 0.0001,
      "loss": 6.823,
      "step": 6130
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.2857093811035156,
      "learning_rate": 0.0001,
      "loss": 6.667,
      "step": 6140
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.4054253101348877,
      "learning_rate": 0.0001,
      "loss": 6.6906,
      "step": 6150
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.414931535720825,
      "learning_rate": 0.0001,
      "loss": 6.7225,
      "step": 6160
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.3389484882354736,
      "learning_rate": 0.0001,
      "loss": 6.7414,
      "step": 6170
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.2892231941223145,
      "learning_rate": 0.0001,
      "loss": 6.8815,
      "step": 6180
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.4546141624450684,
      "learning_rate": 0.0001,
      "loss": 6.8125,
      "step": 6190
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.5399043560028076,
      "learning_rate": 0.0001,
      "loss": 6.7728,
      "step": 6200
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.4791290760040283,
      "learning_rate": 0.0001,
      "loss": 6.8109,
      "step": 6210
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.2605090141296387,
      "learning_rate": 0.0001,
      "loss": 6.6687,
      "step": 6220
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.2917070388793945,
      "learning_rate": 0.0001,
      "loss": 6.7358,
      "step": 6230
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.355299711227417,
      "learning_rate": 0.0001,
      "loss": 6.8014,
      "step": 6240
    },
    {
      "epoch": 2.12,
      "grad_norm": 2.325913667678833,
      "learning_rate": 0.0001,
      "loss": 6.6981,
      "step": 6250
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.4129645824432373,
      "learning_rate": 0.0001,
      "loss": 6.7411,
      "step": 6260
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.5154759883880615,
      "learning_rate": 0.0001,
      "loss": 6.6815,
      "step": 6270
    },
    {
      "epoch": 2.13,
      "grad_norm": 2.3484439849853516,
      "learning_rate": 0.0001,
      "loss": 6.7147,
      "step": 6280
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.524163007736206,
      "learning_rate": 0.0001,
      "loss": 6.7066,
      "step": 6290
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.343323230743408,
      "learning_rate": 0.0001,
      "loss": 6.805,
      "step": 6300
    },
    {
      "epoch": 2.14,
      "grad_norm": 2.652834415435791,
      "learning_rate": 0.0001,
      "loss": 6.6583,
      "step": 6310
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.4346814155578613,
      "learning_rate": 0.0001,
      "loss": 6.8433,
      "step": 6320
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.6968090534210205,
      "learning_rate": 0.0001,
      "loss": 6.6649,
      "step": 6330
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.486971855163574,
      "learning_rate": 0.0001,
      "loss": 6.7352,
      "step": 6340
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.4433326721191406,
      "learning_rate": 0.0001,
      "loss": 6.7313,
      "step": 6350
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.3915398120880127,
      "learning_rate": 0.0001,
      "loss": 6.6529,
      "step": 6360
    },
    {
      "epoch": 2.16,
      "grad_norm": 2.344801902770996,
      "learning_rate": 0.0001,
      "loss": 6.6492,
      "step": 6370
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.4664182662963867,
      "learning_rate": 0.0001,
      "loss": 6.7902,
      "step": 6380
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.3616039752960205,
      "learning_rate": 0.0001,
      "loss": 6.7068,
      "step": 6390
    },
    {
      "epoch": 2.17,
      "grad_norm": 2.502997875213623,
      "learning_rate": 0.0001,
      "loss": 6.8245,
      "step": 6400
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.3835978507995605,
      "learning_rate": 0.0001,
      "loss": 6.6895,
      "step": 6410
    },
    {
      "epoch": 2.18,
      "grad_norm": 2.433032512664795,
      "learning_rate": 0.0001,
      "loss": 6.7404,
      "step": 6420
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.5085232257843018,
      "learning_rate": 0.0001,
      "loss": 6.8267,
      "step": 6430
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.3694140911102295,
      "learning_rate": 0.0001,
      "loss": 6.707,
      "step": 6440
    },
    {
      "epoch": 2.19,
      "grad_norm": 2.4146511554718018,
      "learning_rate": 0.0001,
      "loss": 6.7517,
      "step": 6450
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.4081177711486816,
      "learning_rate": 0.0001,
      "loss": 6.6947,
      "step": 6460
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.3583593368530273,
      "learning_rate": 0.0001,
      "loss": 6.6277,
      "step": 6470
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.445613384246826,
      "learning_rate": 0.0001,
      "loss": 6.5899,
      "step": 6480
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.3775651454925537,
      "learning_rate": 0.0001,
      "loss": 6.6387,
      "step": 6490
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.419865608215332,
      "learning_rate": 0.0001,
      "loss": 6.7063,
      "step": 6500
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.3780555725097656,
      "learning_rate": 0.0001,
      "loss": 6.7098,
      "step": 6510
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.3151533603668213,
      "learning_rate": 0.0001,
      "loss": 6.7985,
      "step": 6520
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.419513463973999,
      "learning_rate": 0.0001,
      "loss": 6.7104,
      "step": 6530
    },
    {
      "epoch": 2.22,
      "grad_norm": 2.4487075805664062,
      "learning_rate": 0.0001,
      "loss": 6.7433,
      "step": 6540
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.4610846042633057,
      "learning_rate": 0.0001,
      "loss": 6.7578,
      "step": 6550
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.4538321495056152,
      "learning_rate": 0.0001,
      "loss": 6.7022,
      "step": 6560
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.5543904304504395,
      "learning_rate": 0.0001,
      "loss": 6.5997,
      "step": 6570
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.461637020111084,
      "learning_rate": 0.0001,
      "loss": 6.647,
      "step": 6580
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.477705240249634,
      "learning_rate": 0.0001,
      "loss": 6.6654,
      "step": 6590
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.4382028579711914,
      "learning_rate": 0.0001,
      "loss": 6.6494,
      "step": 6600
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.578843832015991,
      "learning_rate": 0.0001,
      "loss": 6.679,
      "step": 6610
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.321014165878296,
      "learning_rate": 0.0001,
      "loss": 6.6854,
      "step": 6620
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.3366098403930664,
      "learning_rate": 0.0001,
      "loss": 6.7736,
      "step": 6630
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.338735580444336,
      "learning_rate": 0.0001,
      "loss": 6.6428,
      "step": 6640
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.5356178283691406,
      "learning_rate": 0.0001,
      "loss": 6.631,
      "step": 6650
    },
    {
      "epoch": 2.26,
      "grad_norm": 2.184365749359131,
      "learning_rate": 0.0001,
      "loss": 6.6344,
      "step": 6660
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.3265435695648193,
      "learning_rate": 0.0001,
      "loss": 6.5805,
      "step": 6670
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.421135425567627,
      "learning_rate": 0.0001,
      "loss": 6.6785,
      "step": 6680
    },
    {
      "epoch": 2.27,
      "grad_norm": 2.321599006652832,
      "learning_rate": 0.0001,
      "loss": 6.704,
      "step": 6690
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.3654885292053223,
      "learning_rate": 0.0001,
      "loss": 6.7158,
      "step": 6700
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.5177831649780273,
      "learning_rate": 0.0001,
      "loss": 6.6503,
      "step": 6710
    },
    {
      "epoch": 2.28,
      "grad_norm": 2.313040018081665,
      "learning_rate": 0.0001,
      "loss": 6.5879,
      "step": 6720
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.3558781147003174,
      "learning_rate": 0.0001,
      "loss": 6.7163,
      "step": 6730
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.357043504714966,
      "learning_rate": 0.0001,
      "loss": 6.7328,
      "step": 6740
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.4913806915283203,
      "learning_rate": 0.0001,
      "loss": 6.6572,
      "step": 6750
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.4111156463623047,
      "learning_rate": 0.0001,
      "loss": 6.6991,
      "step": 6760
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.5237784385681152,
      "learning_rate": 0.0001,
      "loss": 6.7251,
      "step": 6770
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.480854034423828,
      "learning_rate": 0.0001,
      "loss": 6.5924,
      "step": 6780
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.342013359069824,
      "learning_rate": 0.0001,
      "loss": 6.6509,
      "step": 6790
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.5419070720672607,
      "learning_rate": 0.0001,
      "loss": 6.7271,
      "step": 6800
    },
    {
      "epoch": 2.31,
      "grad_norm": 2.566915512084961,
      "learning_rate": 0.0001,
      "loss": 6.6383,
      "step": 6810
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.4860012531280518,
      "learning_rate": 0.0001,
      "loss": 6.6661,
      "step": 6820
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.4008114337921143,
      "learning_rate": 0.0001,
      "loss": 6.6478,
      "step": 6830
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.404358386993408,
      "learning_rate": 0.0001,
      "loss": 6.5861,
      "step": 6840
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.487964391708374,
      "learning_rate": 0.0001,
      "loss": 6.6282,
      "step": 6850
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.448789358139038,
      "learning_rate": 0.0001,
      "loss": 6.5526,
      "step": 6860
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.5977635383605957,
      "learning_rate": 0.0001,
      "loss": 6.7969,
      "step": 6870
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.435678720474243,
      "learning_rate": 0.0001,
      "loss": 6.6237,
      "step": 6880
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.430389881134033,
      "learning_rate": 0.0001,
      "loss": 6.6098,
      "step": 6890
    },
    {
      "epoch": 2.34,
      "grad_norm": 2.3108503818511963,
      "learning_rate": 0.0001,
      "loss": 6.5015,
      "step": 6900
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.4835143089294434,
      "learning_rate": 0.0001,
      "loss": 6.5734,
      "step": 6910
    },
    {
      "epoch": 2.35,
      "grad_norm": 2.467932939529419,
      "learning_rate": 0.0001,
      "loss": 6.5835,
      "step": 6920
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.4854071140289307,
      "learning_rate": 0.0001,
      "loss": 6.6903,
      "step": 6930
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.3827455043792725,
      "learning_rate": 0.0001,
      "loss": 6.6991,
      "step": 6940
    },
    {
      "epoch": 2.36,
      "grad_norm": 2.40759539604187,
      "learning_rate": 0.0001,
      "loss": 6.6489,
      "step": 6950
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.494851589202881,
      "learning_rate": 0.0001,
      "loss": 6.6051,
      "step": 6960
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.38189435005188,
      "learning_rate": 0.0001,
      "loss": 6.6455,
      "step": 6970
    },
    {
      "epoch": 2.37,
      "grad_norm": 2.2972683906555176,
      "learning_rate": 0.0001,
      "loss": 6.6477,
      "step": 6980
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.3755240440368652,
      "learning_rate": 0.0001,
      "loss": 6.6287,
      "step": 6990
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.4599435329437256,
      "learning_rate": 0.0001,
      "loss": 6.497,
      "step": 7000
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.5824034214019775,
      "learning_rate": 0.0001,
      "loss": 6.5575,
      "step": 7010
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.6116127967834473,
      "learning_rate": 0.0001,
      "loss": 6.6656,
      "step": 7020
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.4924373626708984,
      "learning_rate": 0.0001,
      "loss": 6.5141,
      "step": 7030
    },
    {
      "epoch": 2.39,
      "grad_norm": 2.6754868030548096,
      "learning_rate": 0.0001,
      "loss": 6.6132,
      "step": 7040
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.6676344871520996,
      "learning_rate": 0.0001,
      "loss": 6.6724,
      "step": 7050
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.3747360706329346,
      "learning_rate": 0.0001,
      "loss": 6.6109,
      "step": 7060
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.4569485187530518,
      "learning_rate": 0.0001,
      "loss": 6.5757,
      "step": 7070
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.354471445083618,
      "learning_rate": 0.0001,
      "loss": 6.6857,
      "step": 7080
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.4351491928100586,
      "learning_rate": 0.0001,
      "loss": 6.7106,
      "step": 7090
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.420145034790039,
      "learning_rate": 0.0001,
      "loss": 6.593,
      "step": 7100
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.480693817138672,
      "learning_rate": 0.0001,
      "loss": 6.7345,
      "step": 7110
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.398705005645752,
      "learning_rate": 0.0001,
      "loss": 6.573,
      "step": 7120
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.4723408222198486,
      "learning_rate": 0.0001,
      "loss": 6.6368,
      "step": 7130
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.353672981262207,
      "learning_rate": 0.0001,
      "loss": 6.5385,
      "step": 7140
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.4100985527038574,
      "learning_rate": 0.0001,
      "loss": 6.5854,
      "step": 7150
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.4232449531555176,
      "learning_rate": 0.0001,
      "loss": 6.4987,
      "step": 7160
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.713700532913208,
      "learning_rate": 0.0001,
      "loss": 6.5794,
      "step": 7170
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.6835007667541504,
      "learning_rate": 0.0001,
      "loss": 6.6516,
      "step": 7180
    },
    {
      "epoch": 2.44,
      "grad_norm": 2.540566921234131,
      "learning_rate": 0.0001,
      "loss": 6.5251,
      "step": 7190
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.3966240882873535,
      "learning_rate": 0.0001,
      "loss": 6.6168,
      "step": 7200
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.3429226875305176,
      "learning_rate": 0.0001,
      "loss": 6.7033,
      "step": 7210
    },
    {
      "epoch": 2.45,
      "grad_norm": 2.2682909965515137,
      "learning_rate": 0.0001,
      "loss": 6.5747,
      "step": 7220
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.472114324569702,
      "learning_rate": 0.0001,
      "loss": 6.6046,
      "step": 7230
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.375086784362793,
      "learning_rate": 0.0001,
      "loss": 6.5942,
      "step": 7240
    },
    {
      "epoch": 2.46,
      "grad_norm": 2.3834826946258545,
      "learning_rate": 0.0001,
      "loss": 6.6195,
      "step": 7250
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.443159818649292,
      "learning_rate": 0.0001,
      "loss": 6.5226,
      "step": 7260
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.3855631351470947,
      "learning_rate": 0.0001,
      "loss": 6.5294,
      "step": 7270
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.5067007541656494,
      "learning_rate": 0.0001,
      "loss": 6.5628,
      "step": 7280
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.2735891342163086,
      "learning_rate": 0.0001,
      "loss": 6.5801,
      "step": 7290
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.338221549987793,
      "learning_rate": 0.0001,
      "loss": 6.5488,
      "step": 7300
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.345191240310669,
      "learning_rate": 0.0001,
      "loss": 6.3966,
      "step": 7310
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.5452232360839844,
      "learning_rate": 0.0001,
      "loss": 6.582,
      "step": 7320
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.5252342224121094,
      "learning_rate": 0.0001,
      "loss": 6.4424,
      "step": 7330
    },
    {
      "epoch": 2.49,
      "grad_norm": 2.3293297290802,
      "learning_rate": 0.0001,
      "loss": 6.4929,
      "step": 7340
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.3897252082824707,
      "learning_rate": 0.0001,
      "loss": 6.5216,
      "step": 7350
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.3954155445098877,
      "learning_rate": 0.0001,
      "loss": 6.6009,
      "step": 7360
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.4169862270355225,
      "learning_rate": 0.0001,
      "loss": 6.5498,
      "step": 7370
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.3694748878479004,
      "learning_rate": 0.0001,
      "loss": 6.6092,
      "step": 7380
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.1525871753692627,
      "learning_rate": 0.0001,
      "loss": 6.4587,
      "step": 7390
    },
    {
      "epoch": 2.51,
      "grad_norm": 2.3603813648223877,
      "learning_rate": 0.0001,
      "loss": 6.5468,
      "step": 7400
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.374941110610962,
      "learning_rate": 0.0001,
      "loss": 6.5933,
      "step": 7410
    },
    {
      "epoch": 2.52,
      "grad_norm": 2.58593487739563,
      "learning_rate": 0.0001,
      "loss": 6.5777,
      "step": 7420
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.3613600730895996,
      "learning_rate": 0.0001,
      "loss": 6.4953,
      "step": 7430
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.400717258453369,
      "learning_rate": 0.0001,
      "loss": 6.5551,
      "step": 7440
    },
    {
      "epoch": 2.53,
      "grad_norm": 2.4552764892578125,
      "learning_rate": 0.0001,
      "loss": 6.595,
      "step": 7450
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.431891679763794,
      "learning_rate": 0.0001,
      "loss": 6.5206,
      "step": 7460
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.622408628463745,
      "learning_rate": 0.0001,
      "loss": 6.6461,
      "step": 7470
    },
    {
      "epoch": 2.54,
      "grad_norm": 2.530582904815674,
      "learning_rate": 0.0001,
      "loss": 6.5773,
      "step": 7480
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.5170624256134033,
      "learning_rate": 0.0001,
      "loss": 6.5591,
      "step": 7490
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.525491714477539,
      "learning_rate": 0.0001,
      "loss": 6.6803,
      "step": 7500
    },
    {
      "epoch": 2.55,
      "grad_norm": 2.6200342178344727,
      "learning_rate": 0.0001,
      "loss": 6.4981,
      "step": 7510
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.6727588176727295,
      "learning_rate": 0.0001,
      "loss": 6.565,
      "step": 7520
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.4518978595733643,
      "learning_rate": 0.0001,
      "loss": 6.5723,
      "step": 7530
    },
    {
      "epoch": 2.56,
      "grad_norm": 2.5073177814483643,
      "learning_rate": 0.0001,
      "loss": 6.5436,
      "step": 7540
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.4163527488708496,
      "learning_rate": 0.0001,
      "loss": 6.5592,
      "step": 7550
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.3171191215515137,
      "learning_rate": 0.0001,
      "loss": 6.4982,
      "step": 7560
    },
    {
      "epoch": 2.57,
      "grad_norm": 2.4813661575317383,
      "learning_rate": 0.0001,
      "loss": 6.5612,
      "step": 7570
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.391533613204956,
      "learning_rate": 0.0001,
      "loss": 6.5979,
      "step": 7580
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.3852477073669434,
      "learning_rate": 0.0001,
      "loss": 6.5185,
      "step": 7590
    },
    {
      "epoch": 2.58,
      "grad_norm": 2.3847193717956543,
      "learning_rate": 0.0001,
      "loss": 6.5222,
      "step": 7600
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.582272529602051,
      "learning_rate": 0.0001,
      "loss": 6.6064,
      "step": 7610
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.2116689682006836,
      "learning_rate": 0.0001,
      "loss": 6.4506,
      "step": 7620
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.3123679161071777,
      "learning_rate": 0.0001,
      "loss": 6.4978,
      "step": 7630
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.4657137393951416,
      "learning_rate": 0.0001,
      "loss": 6.5312,
      "step": 7640
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.466048002243042,
      "learning_rate": 0.0001,
      "loss": 6.5716,
      "step": 7650
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.464794874191284,
      "learning_rate": 0.0001,
      "loss": 6.5371,
      "step": 7660
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.514782428741455,
      "learning_rate": 0.0001,
      "loss": 6.485,
      "step": 7670
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.3275089263916016,
      "learning_rate": 0.0001,
      "loss": 6.4874,
      "step": 7680
    },
    {
      "epoch": 2.61,
      "grad_norm": 2.54858136177063,
      "learning_rate": 0.0001,
      "loss": 6.4682,
      "step": 7690
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.384810447692871,
      "learning_rate": 0.0001,
      "loss": 6.6035,
      "step": 7700
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.4897470474243164,
      "learning_rate": 0.0001,
      "loss": 6.5514,
      "step": 7710
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.4654860496520996,
      "learning_rate": 0.0001,
      "loss": 6.4483,
      "step": 7720
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.2626514434814453,
      "learning_rate": 0.0001,
      "loss": 6.4642,
      "step": 7730
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.4894731044769287,
      "learning_rate": 0.0001,
      "loss": 6.5365,
      "step": 7740
    },
    {
      "epoch": 2.63,
      "grad_norm": 2.3856000900268555,
      "learning_rate": 0.0001,
      "loss": 6.4537,
      "step": 7750
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.338667392730713,
      "learning_rate": 0.0001,
      "loss": 6.5448,
      "step": 7760
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.4658851623535156,
      "learning_rate": 0.0001,
      "loss": 6.638,
      "step": 7770
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.5155911445617676,
      "learning_rate": 0.0001,
      "loss": 6.4092,
      "step": 7780
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.4454259872436523,
      "learning_rate": 0.0001,
      "loss": 6.4091,
      "step": 7790
    },
    {
      "epoch": 2.65,
      "grad_norm": 2.4755043983459473,
      "learning_rate": 0.0001,
      "loss": 6.4448,
      "step": 7800
    }
  ],
  "logging_steps": 10,
  "max_steps": 17652,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 200,
  "total_flos": 1.3151692470722863e+19,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}