{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 49.01960784313726,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "learning_rate": 0.0003996,
      "loss": 9.2736,
      "step": 10
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0003992,
      "loss": 7.7626,
      "step": 20
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00039880000000000004,
      "loss": 7.5202,
      "step": 30
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.00039840000000000003,
      "loss": 7.371,
      "step": 40
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.000398,
      "loss": 7.2352,
      "step": 50
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0003976,
      "loss": 7.1352,
      "step": 60
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0003972,
      "loss": 7.0625,
      "step": 70
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0003968,
      "loss": 6.911,
      "step": 80
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00039640000000000004,
      "loss": 6.7717,
      "step": 90
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00039600000000000003,
      "loss": 6.6484,
      "step": 100
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0003956,
      "loss": 6.5492,
      "step": 110
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0003952,
      "loss": 6.4417,
      "step": 120
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0003948,
      "loss": 6.3311,
      "step": 130
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0003944,
      "loss": 6.2206,
      "step": 140
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00039400000000000004,
      "loss": 6.1179,
      "step": 150
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0003936,
      "loss": 6.0151,
      "step": 160
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0003932,
      "loss": 5.9035,
      "step": 170
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0003928,
      "loss": 5.8222,
      "step": 180
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0003924,
      "loss": 5.7042,
      "step": 190
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.000392,
      "loss": 5.6265,
      "step": 200
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.00039160000000000003,
      "loss": 5.5338,
      "step": 210
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0003912,
      "loss": 5.4521,
      "step": 220
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0003908,
      "loss": 5.3552,
      "step": 230
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0003904,
      "loss": 5.2771,
      "step": 240
    },
    {
      "epoch": 2.45,
      "learning_rate": 0.00039,
      "loss": 5.1587,
      "step": 250
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0003896,
      "loss": 5.0899,
      "step": 260
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.00038920000000000003,
      "loss": 5.0191,
      "step": 270
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.0003888,
      "loss": 4.9602,
      "step": 280
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.0003884,
      "loss": 4.8366,
      "step": 290
    },
    {
      "epoch": 2.94,
      "learning_rate": 0.000388,
      "loss": 4.7848,
      "step": 300
    },
    {
      "epoch": 3.04,
      "learning_rate": 0.0003876,
      "loss": 4.7199,
      "step": 310
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.00038720000000000003,
      "loss": 4.6306,
      "step": 320
    },
    {
      "epoch": 3.24,
      "learning_rate": 0.0003868,
      "loss": 4.5337,
      "step": 330
    },
    {
      "epoch": 3.33,
      "learning_rate": 0.0003864,
      "loss": 4.4796,
      "step": 340
    },
    {
      "epoch": 3.43,
      "learning_rate": 0.000386,
      "loss": 4.3881,
      "step": 350
    },
    {
      "epoch": 3.53,
      "learning_rate": 0.0003856,
      "loss": 4.2989,
      "step": 360
    },
    {
      "epoch": 3.63,
      "learning_rate": 0.0003852,
      "loss": 4.2533,
      "step": 370
    },
    {
      "epoch": 3.73,
      "learning_rate": 0.00038480000000000003,
      "loss": 4.2379,
      "step": 380
    },
    {
      "epoch": 3.82,
      "learning_rate": 0.0003844,
      "loss": 4.142,
      "step": 390
    },
    {
      "epoch": 3.92,
      "learning_rate": 0.000384,
      "loss": 4.0793,
      "step": 400
    },
    {
      "epoch": 4.02,
      "learning_rate": 0.0003836,
      "loss": 4.0005,
      "step": 410
    },
    {
      "epoch": 4.12,
      "learning_rate": 0.0003832,
      "loss": 3.9619,
      "step": 420
    },
    {
      "epoch": 4.22,
      "learning_rate": 0.0003828,
      "loss": 3.8687,
      "step": 430
    },
    {
      "epoch": 4.31,
      "learning_rate": 0.0003824,
      "loss": 3.8486,
      "step": 440
    },
    {
      "epoch": 4.41,
      "learning_rate": 0.000382,
      "loss": 3.7684,
      "step": 450
    },
    {
      "epoch": 4.51,
      "learning_rate": 0.0003816,
      "loss": 3.7013,
      "step": 460
    },
    {
      "epoch": 4.61,
      "learning_rate": 0.0003812,
      "loss": 3.6851,
      "step": 470
    },
    {
      "epoch": 4.71,
      "learning_rate": 0.0003808,
      "loss": 3.6585,
      "step": 480
    },
    {
      "epoch": 4.8,
      "learning_rate": 0.0003804,
      "loss": 3.6172,
      "step": 490
    },
    {
      "epoch": 4.9,
      "learning_rate": 0.00038,
      "loss": 3.5557,
      "step": 500
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0003796,
      "loss": 3.4746,
      "step": 510
    },
    {
      "epoch": 5.1,
      "learning_rate": 0.0003792,
      "loss": 3.4473,
      "step": 520
    },
    {
      "epoch": 5.2,
      "learning_rate": 0.0003788,
      "loss": 3.3828,
      "step": 530
    },
    {
      "epoch": 5.29,
      "learning_rate": 0.0003784,
      "loss": 3.3868,
      "step": 540
    },
    {
      "epoch": 5.39,
      "learning_rate": 0.00037799999999999997,
      "loss": 3.31,
      "step": 550
    },
    {
      "epoch": 5.49,
      "learning_rate": 0.0003776,
      "loss": 3.2628,
      "step": 560
    },
    {
      "epoch": 5.59,
      "learning_rate": 0.0003772,
      "loss": 3.2541,
      "step": 570
    },
    {
      "epoch": 5.69,
      "learning_rate": 0.0003768,
      "loss": 3.223,
      "step": 580
    },
    {
      "epoch": 5.78,
      "learning_rate": 0.0003764,
      "loss": 3.2028,
      "step": 590
    },
    {
      "epoch": 5.88,
      "learning_rate": 0.000376,
      "loss": 3.1659,
      "step": 600
    },
    {
      "epoch": 5.98,
      "learning_rate": 0.0003756,
      "loss": 3.0847,
      "step": 610
    },
    {
      "epoch": 6.08,
      "learning_rate": 0.0003752,
      "loss": 3.0215,
      "step": 620
    },
    {
      "epoch": 6.18,
      "learning_rate": 0.00037480000000000006,
      "loss": 3.0149,
      "step": 630
    },
    {
      "epoch": 6.27,
      "learning_rate": 0.00037440000000000005,
      "loss": 3.0177,
      "step": 640
    },
    {
      "epoch": 6.37,
      "learning_rate": 0.00037400000000000004,
      "loss": 2.9474,
      "step": 650
    },
    {
      "epoch": 6.47,
      "learning_rate": 0.00037360000000000003,
      "loss": 2.9245,
      "step": 660
    },
    {
      "epoch": 6.57,
      "learning_rate": 0.0003732,
      "loss": 2.9218,
      "step": 670
    },
    {
      "epoch": 6.67,
      "learning_rate": 0.00037280000000000006,
      "loss": 2.8666,
      "step": 680
    },
    {
      "epoch": 6.76,
      "learning_rate": 0.00037240000000000005,
      "loss": 2.8821,
      "step": 690
    },
    {
      "epoch": 6.86,
      "learning_rate": 0.00037200000000000004,
      "loss": 2.8243,
      "step": 700
    },
    {
      "epoch": 6.96,
      "learning_rate": 0.00037160000000000003,
      "loss": 2.7753,
      "step": 710
    },
    {
      "epoch": 7.06,
      "learning_rate": 0.0003712,
      "loss": 2.7086,
      "step": 720
    },
    {
      "epoch": 7.16,
      "learning_rate": 0.0003708,
      "loss": 2.7104,
      "step": 730
    },
    {
      "epoch": 7.25,
      "learning_rate": 0.00037040000000000006,
      "loss": 2.7103,
      "step": 740
    },
    {
      "epoch": 7.35,
      "learning_rate": 0.00037000000000000005,
      "loss": 2.6707,
      "step": 750
    },
    {
      "epoch": 7.45,
      "learning_rate": 0.00036960000000000004,
      "loss": 2.6413,
      "step": 760
    },
    {
      "epoch": 7.55,
      "learning_rate": 0.00036920000000000003,
      "loss": 2.6359,
      "step": 770
    },
    {
      "epoch": 7.65,
      "learning_rate": 0.0003688,
      "loss": 2.5838,
      "step": 780
    },
    {
      "epoch": 7.75,
      "learning_rate": 0.0003684,
      "loss": 2.6212,
      "step": 790
    },
    {
      "epoch": 7.84,
      "learning_rate": 0.00036800000000000005,
      "loss": 2.5718,
      "step": 800
    },
    {
      "epoch": 7.94,
      "learning_rate": 0.00036760000000000004,
      "loss": 2.5348,
      "step": 810
    },
    {
      "epoch": 8.04,
      "learning_rate": 0.00036720000000000004,
      "loss": 2.4195,
      "step": 820
    },
    {
      "epoch": 8.14,
      "learning_rate": 0.0003668,
      "loss": 2.4938,
      "step": 830
    },
    {
      "epoch": 8.24,
      "learning_rate": 0.0003664,
      "loss": 2.4372,
      "step": 840
    },
    {
      "epoch": 8.33,
      "learning_rate": 0.000366,
      "loss": 2.4567,
      "step": 850
    },
    {
      "epoch": 8.43,
      "learning_rate": 0.00036560000000000005,
      "loss": 2.4108,
      "step": 860
    },
    {
      "epoch": 8.53,
      "learning_rate": 0.00036520000000000004,
      "loss": 2.3993,
      "step": 870
    },
    {
      "epoch": 8.63,
      "learning_rate": 0.00036480000000000003,
      "loss": 2.3739,
      "step": 880
    },
    {
      "epoch": 8.73,
      "learning_rate": 0.0003644,
      "loss": 2.391,
      "step": 890
    },
    {
      "epoch": 8.82,
      "learning_rate": 0.000364,
      "loss": 2.38,
      "step": 900
    },
    {
      "epoch": 8.92,
      "learning_rate": 0.00036360000000000006,
      "loss": 2.3257,
      "step": 910
    },
    {
      "epoch": 9.02,
      "learning_rate": 0.00036320000000000005,
      "loss": 2.1956,
      "step": 920
    },
    {
      "epoch": 9.12,
      "learning_rate": 0.00036280000000000004,
      "loss": 2.2944,
      "step": 930
    },
    {
      "epoch": 9.22,
      "learning_rate": 0.0003624,
      "loss": 2.2163,
      "step": 940
    },
    {
      "epoch": 9.31,
      "learning_rate": 0.000362,
      "loss": 2.2641,
      "step": 950
    },
    {
      "epoch": 9.41,
      "learning_rate": 0.0003616,
      "loss": 2.2035,
      "step": 960
    },
    {
      "epoch": 9.51,
      "learning_rate": 0.00036120000000000005,
      "loss": 2.188,
      "step": 970
    },
    {
      "epoch": 9.61,
      "learning_rate": 0.00036080000000000004,
      "loss": 2.197,
      "step": 980
    },
    {
      "epoch": 9.71,
      "learning_rate": 0.00036040000000000003,
      "loss": 2.18,
      "step": 990
    },
    {
      "epoch": 9.8,
      "learning_rate": 0.00036,
      "loss": 2.1994,
      "step": 1000
    },
    {
      "epoch": 9.9,
      "learning_rate": 0.0003596,
      "loss": 2.1538,
      "step": 1010
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.0003592,
      "loss": 2.0495,
      "step": 1020
    },
    {
      "epoch": 10.1,
      "learning_rate": 0.00035880000000000005,
      "loss": 2.0501,
      "step": 1030
    },
    {
      "epoch": 10.2,
      "learning_rate": 0.00035840000000000004,
      "loss": 2.0302,
      "step": 1040
    },
    {
      "epoch": 10.29,
      "learning_rate": 0.00035800000000000003,
      "loss": 2.0996,
      "step": 1050
    },
    {
      "epoch": 10.39,
      "learning_rate": 0.0003576,
      "loss": 2.0369,
      "step": 1060
    },
    {
      "epoch": 10.49,
      "learning_rate": 0.0003572,
      "loss": 1.9996,
      "step": 1070
    },
    {
      "epoch": 10.59,
      "learning_rate": 0.0003568,
      "loss": 2.0254,
      "step": 1080
    },
    {
      "epoch": 10.69,
      "learning_rate": 0.00035640000000000004,
      "loss": 2.0004,
      "step": 1090
    },
    {
      "epoch": 10.78,
      "learning_rate": 0.00035600000000000003,
      "loss": 2.0154,
      "step": 1100
    },
    {
      "epoch": 10.88,
      "learning_rate": 0.0003556,
      "loss": 2.0014,
      "step": 1110
    },
    {
      "epoch": 10.98,
      "learning_rate": 0.00035524,
      "loss": 1.9377,
      "step": 1120
    },
    {
      "epoch": 11.08,
      "learning_rate": 0.00035484000000000004,
      "loss": 1.8296,
      "step": 1130
    },
    {
      "epoch": 11.18,
      "learning_rate": 0.00035444000000000003,
      "loss": 1.9002,
      "step": 1140
    },
    {
      "epoch": 11.27,
      "learning_rate": 0.00035404,
      "loss": 1.9336,
      "step": 1150
    },
    {
      "epoch": 11.37,
      "learning_rate": 0.00035364,
      "loss": 1.9041,
      "step": 1160
    },
    {
      "epoch": 11.47,
      "learning_rate": 0.00035324,
      "loss": 1.8324,
      "step": 1170
    },
    {
      "epoch": 11.57,
      "learning_rate": 0.00035284,
      "loss": 1.8663,
      "step": 1180
    },
    {
      "epoch": 11.67,
      "learning_rate": 0.00035244000000000003,
      "loss": 1.8358,
      "step": 1190
    },
    {
      "epoch": 11.76,
      "learning_rate": 0.00035204,
      "loss": 1.8586,
      "step": 1200
    },
    {
      "epoch": 11.86,
      "learning_rate": 0.00035164,
      "loss": 1.8452,
      "step": 1210
    },
    {
      "epoch": 11.96,
      "learning_rate": 0.00035124,
      "loss": 1.7957,
      "step": 1220
    },
    {
      "epoch": 12.06,
      "learning_rate": 0.00035084,
      "loss": 1.6603,
      "step": 1230
    },
    {
      "epoch": 12.16,
      "learning_rate": 0.00035044000000000004,
      "loss": 1.7903,
      "step": 1240
    },
    {
      "epoch": 12.25,
      "learning_rate": 0.00035004000000000003,
      "loss": 1.7574,
      "step": 1250
    },
    {
      "epoch": 12.35,
      "learning_rate": 0.00034964,
      "loss": 1.7823,
      "step": 1260
    },
    {
      "epoch": 12.45,
      "learning_rate": 0.00034924,
      "loss": 1.7038,
      "step": 1270
    },
    {
      "epoch": 12.55,
      "learning_rate": 0.00034884,
      "loss": 1.7186,
      "step": 1280
    },
    {
      "epoch": 12.65,
      "learning_rate": 0.00034844,
      "loss": 1.6948,
      "step": 1290
    },
    {
      "epoch": 12.75,
      "learning_rate": 0.00034804000000000004,
      "loss": 1.7347,
      "step": 1300
    },
    {
      "epoch": 12.84,
      "learning_rate": 0.00034764,
      "loss": 1.7304,
      "step": 1310
    },
    {
      "epoch": 12.94,
      "learning_rate": 0.00034724,
      "loss": 1.6862,
      "step": 1320
    },
    {
      "epoch": 13.04,
      "learning_rate": 0.00034684,
      "loss": 1.4857,
      "step": 1330
    },
    {
      "epoch": 13.14,
      "learning_rate": 0.00034644,
      "loss": 1.6981,
      "step": 1340
    },
    {
      "epoch": 13.24,
      "learning_rate": 0.00034604,
      "loss": 1.5926,
      "step": 1350
    },
    {
      "epoch": 13.33,
      "learning_rate": 0.00034564000000000003,
      "loss": 1.6545,
      "step": 1360
    },
    {
      "epoch": 13.43,
      "learning_rate": 0.00034524,
      "loss": 1.5918,
      "step": 1370
    },
    {
      "epoch": 13.53,
      "learning_rate": 0.00034484,
      "loss": 1.5743,
      "step": 1380
    },
    {
      "epoch": 13.63,
      "learning_rate": 0.00034444,
      "loss": 1.5867,
      "step": 1390
    },
    {
      "epoch": 13.73,
      "learning_rate": 0.00034404,
      "loss": 1.6042,
      "step": 1400
    },
    {
      "epoch": 13.82,
      "learning_rate": 0.00034364,
      "loss": 1.6308,
      "step": 1410
    },
    {
      "epoch": 13.92,
      "learning_rate": 0.00034324000000000003,
      "loss": 1.5813,
      "step": 1420
    },
    {
      "epoch": 14.02,
      "learning_rate": 0.00034284,
      "loss": 1.364,
      "step": 1430
    },
    {
      "epoch": 14.12,
      "learning_rate": 0.00034244,
      "loss": 1.5849,
      "step": 1440
    },
    {
      "epoch": 14.22,
      "learning_rate": 0.00034204,
      "loss": 1.4957,
      "step": 1450
    },
    {
      "epoch": 14.31,
      "learning_rate": 0.00034164,
      "loss": 1.5312,
      "step": 1460
    },
    {
      "epoch": 14.41,
      "learning_rate": 0.00034124,
      "loss": 1.4995,
      "step": 1470
    },
    {
      "epoch": 14.51,
      "learning_rate": 0.00034084,
      "loss": 1.4501,
      "step": 1480
    },
    {
      "epoch": 14.61,
      "learning_rate": 0.00034044,
      "loss": 1.4721,
      "step": 1490
    },
    {
      "epoch": 14.71,
      "learning_rate": 0.00034004,
      "loss": 1.4792,
      "step": 1500
    },
    {
      "epoch": 14.8,
      "learning_rate": 0.00033964,
      "loss": 1.513,
      "step": 1510
    },
    {
      "epoch": 14.9,
      "learning_rate": 0.00033924,
      "loss": 1.5028,
      "step": 1520
    },
    {
      "epoch": 15.0,
      "learning_rate": 0.00033884000000000003,
      "loss": 1.3511,
      "step": 1530
    },
    {
      "epoch": 15.1,
      "learning_rate": 0.00033844,
      "loss": 1.4152,
      "step": 1540
    },
    {
      "epoch": 15.2,
      "learning_rate": 0.00033804,
      "loss": 1.4216,
      "step": 1550
    },
    {
      "epoch": 15.29,
      "learning_rate": 0.00033764,
      "loss": 1.4177,
      "step": 1560
    },
    {
      "epoch": 15.39,
      "learning_rate": 0.00033724,
      "loss": 1.4074,
      "step": 1570
    },
    {
      "epoch": 15.49,
      "learning_rate": 0.00033684,
      "loss": 1.3532,
      "step": 1580
    },
    {
      "epoch": 15.59,
      "learning_rate": 0.00033644,
      "loss": 1.3525,
      "step": 1590
    },
    {
      "epoch": 15.69,
      "learning_rate": 0.00033604,
      "loss": 1.3755,
      "step": 1600
    },
    {
      "epoch": 15.78,
      "learning_rate": 0.00033564,
      "loss": 1.3989,
      "step": 1610
    },
    {
      "epoch": 15.88,
      "learning_rate": 0.00033524,
      "loss": 1.4012,
      "step": 1620
    },
    {
      "epoch": 15.98,
      "learning_rate": 0.00033484,
      "loss": 1.3386,
      "step": 1630
    },
    {
      "epoch": 16.08,
      "learning_rate": 0.00033444,
      "loss": 1.2476,
      "step": 1640
    },
    {
      "epoch": 16.18,
      "learning_rate": 0.00033404,
      "loss": 1.3435,
      "step": 1650
    },
    {
      "epoch": 16.27,
      "learning_rate": 0.00033364,
      "loss": 1.3131,
      "step": 1660
    },
    {
      "epoch": 16.37,
      "learning_rate": 0.00033324,
      "loss": 1.3099,
      "step": 1670
    },
    {
      "epoch": 16.47,
      "learning_rate": 0.00033284,
      "loss": 1.2693,
      "step": 1680
    },
    {
      "epoch": 16.57,
      "learning_rate": 0.00033244,
      "loss": 1.2649,
      "step": 1690
    },
    {
      "epoch": 16.67,
      "learning_rate": 0.00033203999999999997,
      "loss": 1.2777,
      "step": 1700
    },
    {
      "epoch": 16.76,
      "learning_rate": 0.00033164,
      "loss": 1.2889,
      "step": 1710
    },
    {
      "epoch": 16.86,
      "learning_rate": 0.00033124,
      "loss": 1.2729,
      "step": 1720
    },
    {
      "epoch": 16.96,
      "learning_rate": 0.00033084,
      "loss": 1.2529,
      "step": 1730
    },
    {
      "epoch": 17.06,
      "learning_rate": 0.00033044,
      "loss": 1.1566,
      "step": 1740
    },
    {
      "epoch": 17.16,
      "learning_rate": 0.00033004,
      "loss": 1.2668,
      "step": 1750
    },
    {
      "epoch": 17.25,
      "learning_rate": 0.00032964,
      "loss": 1.2052,
      "step": 1760
    },
    {
      "epoch": 17.35,
      "learning_rate": 0.00032924,
      "loss": 1.2159,
      "step": 1770
    },
    {
      "epoch": 17.45,
      "learning_rate": 0.00032884000000000006,
      "loss": 1.2003,
      "step": 1780
    },
    {
      "epoch": 17.55,
      "learning_rate": 0.00032844000000000005,
      "loss": 1.1836,
      "step": 1790
    },
    {
      "epoch": 17.65,
      "learning_rate": 0.00032804000000000004,
      "loss": 1.1857,
      "step": 1800
    },
    {
      "epoch": 17.75,
      "learning_rate": 0.00032764000000000003,
      "loss": 1.2012,
      "step": 1810
    },
    {
      "epoch": 17.84,
      "learning_rate": 0.00032724,
      "loss": 1.1907,
      "step": 1820
    },
    {
      "epoch": 17.94,
      "learning_rate": 0.00032684,
      "loss": 1.1723,
      "step": 1830
    },
    {
      "epoch": 18.04,
      "learning_rate": 0.00032644000000000005,
      "loss": 1.0299,
      "step": 1840
    },
    {
      "epoch": 18.14,
      "learning_rate": 0.00032604000000000004,
      "loss": 1.1996,
      "step": 1850
    },
    {
      "epoch": 18.24,
      "learning_rate": 0.00032564000000000003,
      "loss": 1.0919,
      "step": 1860
    },
    {
      "epoch": 18.33,
      "learning_rate": 0.00032524,
      "loss": 1.1291,
      "step": 1870
    },
    {
      "epoch": 18.43,
      "learning_rate": 0.00032484,
      "loss": 1.1343,
      "step": 1880
    },
    {
      "epoch": 18.53,
      "learning_rate": 0.00032444000000000006,
      "loss": 1.0965,
      "step": 1890
    },
    {
      "epoch": 18.63,
      "learning_rate": 0.00032404000000000005,
      "loss": 1.1149,
      "step": 1900
    },
    {
      "epoch": 18.73,
      "learning_rate": 0.00032364000000000004,
      "loss": 1.132,
      "step": 1910
    },
    {
      "epoch": 18.82,
      "learning_rate": 0.00032324000000000003,
      "loss": 1.1218,
      "step": 1920
    },
    {
      "epoch": 18.92,
      "learning_rate": 0.00032284,
      "loss": 1.0928,
      "step": 1930
    },
    {
      "epoch": 19.02,
      "learning_rate": 0.00032244,
      "loss": 0.9377,
      "step": 1940
    },
    {
      "epoch": 19.12,
      "learning_rate": 0.00032204000000000005,
      "loss": 1.1132,
      "step": 1950
    },
    {
      "epoch": 19.22,
      "learning_rate": 0.00032164000000000004,
      "loss": 1.0275,
      "step": 1960
    },
    {
      "epoch": 19.31,
      "learning_rate": 0.00032124000000000003,
      "loss": 1.0312,
      "step": 1970
    },
    {
      "epoch": 19.41,
      "learning_rate": 0.00032084,
      "loss": 1.0248,
      "step": 1980
    },
    {
      "epoch": 19.51,
      "learning_rate": 0.00032044,
      "loss": 1.0063,
      "step": 1990
    },
    {
      "epoch": 19.61,
      "learning_rate": 0.00032004,
      "loss": 1.0438,
      "step": 2000
    },
    {
      "epoch": 19.71,
      "learning_rate": 0.00031964000000000005,
      "loss": 1.0608,
      "step": 2010
    },
    {
      "epoch": 19.8,
      "learning_rate": 0.00031924000000000004,
      "loss": 1.0503,
      "step": 2020
    },
    {
      "epoch": 19.9,
      "learning_rate": 0.00031884000000000003,
      "loss": 1.0422,
      "step": 2030
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.00031844,
      "loss": 0.9331,
      "step": 2040
    },
    {
      "epoch": 20.1,
      "learning_rate": 0.00031804,
      "loss": 0.9744,
      "step": 2050
    },
    {
      "epoch": 20.2,
      "learning_rate": 0.00031768000000000003,
      "loss": 0.9977,
      "step": 2060
    },
    {
      "epoch": 20.29,
      "learning_rate": 0.00031728,
      "loss": 0.9579,
      "step": 2070
    },
    {
      "epoch": 20.39,
      "learning_rate": 0.00031688,
      "loss": 0.908,
      "step": 2080
    },
    {
      "epoch": 20.49,
      "learning_rate": 0.00031648,
      "loss": 0.9323,
      "step": 2090
    },
    {
      "epoch": 20.59,
      "learning_rate": 0.00031608,
      "loss": 0.9597,
      "step": 2100
    },
    {
      "epoch": 20.69,
      "learning_rate": 0.00031568000000000004,
      "loss": 0.975,
      "step": 2110
    },
    {
      "epoch": 20.78,
      "learning_rate": 0.00031528000000000003,
      "loss": 0.9676,
      "step": 2120
    },
    {
      "epoch": 20.88,
      "learning_rate": 0.00031488,
      "loss": 0.979,
      "step": 2130
    },
    {
      "epoch": 20.98,
      "learning_rate": 0.00031448,
      "loss": 0.9263,
      "step": 2140
    },
    {
      "epoch": 21.08,
      "learning_rate": 0.00031408,
      "loss": 0.8463,
      "step": 2150
    },
    {
      "epoch": 21.18,
      "learning_rate": 0.00031368,
      "loss": 0.9642,
      "step": 2160
    },
    {
      "epoch": 21.27,
      "learning_rate": 0.00031328000000000004,
      "loss": 0.8091,
      "step": 2170
    },
    {
      "epoch": 21.37,
      "learning_rate": 0.00031288,
      "loss": 0.7911,
      "step": 2180
    },
    {
      "epoch": 21.47,
      "learning_rate": 0.00031248,
      "loss": 0.8679,
      "step": 2190
    },
    {
      "epoch": 21.57,
      "learning_rate": 0.00031208,
      "loss": 0.893,
      "step": 2200
    },
    {
      "epoch": 21.67,
      "learning_rate": 0.00031168,
      "loss": 0.9048,
      "step": 2210
    },
    {
      "epoch": 21.76,
      "learning_rate": 0.00031128000000000004,
      "loss": 0.9283,
      "step": 2220
    },
    {
      "epoch": 21.86,
      "learning_rate": 0.00031088000000000003,
      "loss": 0.9238,
      "step": 2230
    },
    {
      "epoch": 21.96,
      "learning_rate": 0.00031048,
      "loss": 0.8698,
      "step": 2240
    },
    {
      "epoch": 22.06,
      "learning_rate": 0.00031008,
      "loss": 0.794,
      "step": 2250
    },
    {
      "epoch": 22.16,
      "learning_rate": 0.00030968,
      "loss": 0.8646,
      "step": 2260
    },
    {
      "epoch": 22.25,
      "learning_rate": 0.00030928,
      "loss": 0.7995,
      "step": 2270
    },
    {
      "epoch": 22.35,
      "learning_rate": 0.00030888000000000004,
      "loss": 0.8045,
      "step": 2280
    },
    {
      "epoch": 22.45,
      "learning_rate": 0.00030848000000000003,
      "loss": 0.8186,
      "step": 2290
    },
    {
      "epoch": 22.55,
      "learning_rate": 0.00030808,
      "loss": 0.8085,
      "step": 2300
    },
    {
      "epoch": 22.65,
      "learning_rate": 0.00030768,
      "loss": 0.8098,
      "step": 2310
    },
    {
      "epoch": 22.75,
      "learning_rate": 0.00030728,
      "loss": 0.8741,
      "step": 2320
    },
    {
      "epoch": 22.84,
      "learning_rate": 0.00030688,
      "loss": 0.8602,
      "step": 2330
    },
    {
      "epoch": 22.94,
      "learning_rate": 0.00030648000000000003,
      "loss": 0.7975,
      "step": 2340
    },
    {
      "epoch": 23.04,
      "learning_rate": 0.00030608,
      "loss": 0.6979,
      "step": 2350
    },
    {
      "epoch": 23.14,
      "learning_rate": 0.00030568,
      "loss": 0.8089,
      "step": 2360
    },
    {
      "epoch": 23.24,
      "learning_rate": 0.00030528,
      "loss": 0.7418,
      "step": 2370
    },
    {
      "epoch": 23.33,
      "learning_rate": 0.00030488,
      "loss": 0.752,
      "step": 2380
    },
    {
      "epoch": 23.43,
      "learning_rate": 0.00030448,
      "loss": 0.7302,
      "step": 2390
    },
    {
      "epoch": 23.53,
      "learning_rate": 0.00030408000000000003,
      "loss": 0.7431,
      "step": 2400
    },
    {
      "epoch": 23.63,
      "learning_rate": 0.00030368,
      "loss": 0.7443,
      "step": 2410
    },
    {
      "epoch": 23.73,
      "learning_rate": 0.00030328,
      "loss": 0.7954,
      "step": 2420
    },
    {
      "epoch": 23.82,
      "learning_rate": 0.00030288,
      "loss": 0.8135,
      "step": 2430
    },
    {
      "epoch": 23.92,
      "learning_rate": 0.00030248,
      "loss": 0.7517,
      "step": 2440
    },
    {
      "epoch": 24.02,
      "learning_rate": 0.00030208,
      "loss": 0.6313,
      "step": 2450
    },
    {
      "epoch": 24.12,
      "learning_rate": 0.00030168,
      "loss": 0.7276,
      "step": 2460
    },
    {
      "epoch": 24.22,
      "learning_rate": 0.00030128,
      "loss": 0.6859,
      "step": 2470
    },
    {
      "epoch": 24.31,
      "learning_rate": 0.00030088,
      "loss": 0.6701,
      "step": 2480
    },
    {
      "epoch": 24.41,
      "learning_rate": 0.00030048,
      "loss": 0.6492,
      "step": 2490
    },
    {
      "epoch": 24.51,
      "learning_rate": 0.00030008,
      "loss": 0.7006,
      "step": 2500
    },
    {
      "epoch": 24.61,
      "learning_rate": 0.00029968000000000003,
      "loss": 0.7105,
      "step": 2510
    },
    {
      "epoch": 24.71,
      "learning_rate": 0.00029928,
      "loss": 0.7175,
      "step": 2520
    },
    {
      "epoch": 24.8,
      "learning_rate": 0.00029888,
      "loss": 0.7297,
      "step": 2530
    },
    {
      "epoch": 24.9,
      "learning_rate": 0.00029848,
      "loss": 0.7295,
      "step": 2540
    },
    {
      "epoch": 25.0,
      "learning_rate": 0.00029808,
      "loss": 0.6337,
      "step": 2550
    },
    {
      "epoch": 25.1,
      "learning_rate": 0.00029768,
      "loss": 0.6452,
      "step": 2560
    },
    {
      "epoch": 25.2,
      "learning_rate": 0.00029728,
      "loss": 0.6282,
      "step": 2570
    },
    {
      "epoch": 25.29,
      "learning_rate": 0.00029688,
      "loss": 0.6218,
      "step": 2580
    },
    {
      "epoch": 25.39,
      "learning_rate": 0.00029648,
      "loss": 0.5982,
      "step": 2590
    },
    {
      "epoch": 25.49,
      "learning_rate": 0.00029608,
      "loss": 0.6601,
      "step": 2600
    },
    {
      "epoch": 25.59,
      "learning_rate": 0.00029568,
      "loss": 0.6761,
      "step": 2610
    },
    {
      "epoch": 25.69,
      "learning_rate": 0.00029528,
      "loss": 0.6685,
      "step": 2620
    },
    {
      "epoch": 25.78,
      "learning_rate": 0.00029488,
      "loss": 0.6506,
      "step": 2630
    },
    {
      "epoch": 25.88,
      "learning_rate": 0.00029448,
      "loss": 0.6804,
      "step": 2640
    },
    {
      "epoch": 25.98,
      "learning_rate": 0.00029408,
      "loss": 0.6387,
      "step": 2650
    },
    {
      "epoch": 26.08,
      "learning_rate": 0.00029368,
      "loss": 0.5516,
      "step": 2660
    },
    {
      "epoch": 26.18,
      "learning_rate": 0.00029328,
      "loss": 0.5891,
      "step": 2670
    },
    {
      "epoch": 26.27,
      "learning_rate": 0.00029288,
      "loss": 0.5907,
      "step": 2680
    },
    {
      "epoch": 26.37,
      "learning_rate": 0.00029248,
      "loss": 0.5559,
      "step": 2690
    },
    {
      "epoch": 26.47,
      "learning_rate": 0.00029208,
      "loss": 0.5942,
      "step": 2700
    },
    {
      "epoch": 26.57,
      "learning_rate": 0.00029168,
      "loss": 0.625,
      "step": 2710
    },
    {
      "epoch": 26.67,
      "learning_rate": 0.00029128,
      "loss": 0.621,
      "step": 2720
    },
    {
      "epoch": 26.76,
      "learning_rate": 0.00029088,
      "loss": 0.5987,
      "step": 2730
    },
    {
      "epoch": 26.86,
      "learning_rate": 0.00029047999999999997,
      "loss": 0.6137,
      "step": 2740
    },
    {
      "epoch": 26.96,
      "learning_rate": 0.00029008,
      "loss": 0.5928,
      "step": 2750
    },
    {
      "epoch": 27.06,
      "learning_rate": 0.00028968,
      "loss": 0.5423,
      "step": 2760
    },
    {
      "epoch": 27.16,
      "learning_rate": 0.00028928,
      "loss": 0.5484,
      "step": 2770
    },
    {
      "epoch": 27.25,
      "learning_rate": 0.00028888,
      "loss": 0.563,
      "step": 2780
    },
    {
      "epoch": 27.35,
      "learning_rate": 0.00028848,
      "loss": 0.5161,
      "step": 2790
    },
    {
      "epoch": 27.45,
      "learning_rate": 0.00028808,
      "loss": 0.5432,
      "step": 2800
    },
    {
      "epoch": 27.55,
      "learning_rate": 0.00028768,
      "loss": 0.5607,
      "step": 2810
    },
    {
      "epoch": 27.65,
      "learning_rate": 0.00028728,
      "loss": 0.557,
      "step": 2820
    },
    {
      "epoch": 27.75,
      "learning_rate": 0.00028688,
      "loss": 0.5698,
      "step": 2830
    },
    {
      "epoch": 27.84,
      "learning_rate": 0.00028648,
      "loss": 0.5571,
      "step": 2840
    },
    {
      "epoch": 27.94,
      "learning_rate": 0.00028607999999999997,
      "loss": 0.5525,
      "step": 2850
    },
    {
      "epoch": 28.04,
      "learning_rate": 0.00028568,
      "loss": 0.5017,
      "step": 2860
    },
    {
      "epoch": 28.14,
      "learning_rate": 0.00028528,
      "loss": 0.5348,
      "step": 2870
    },
    {
      "epoch": 28.24,
      "learning_rate": 0.00028488000000000005,
      "loss": 0.4948,
      "step": 2880
    },
    {
      "epoch": 28.33,
      "learning_rate": 0.00028448000000000004,
      "loss": 0.4791,
      "step": 2890
    },
    {
      "epoch": 28.43,
      "learning_rate": 0.00028408000000000003,
      "loss": 0.4908,
      "step": 2900
    },
    {
      "epoch": 28.53,
      "learning_rate": 0.00028368,
      "loss": 0.4943,
      "step": 2910
    },
    {
      "epoch": 28.63,
      "learning_rate": 0.00028328,
      "loss": 0.4995,
      "step": 2920
    },
    {
      "epoch": 28.73,
      "learning_rate": 0.00028288000000000006,
      "loss": 0.5252,
      "step": 2930
    },
    {
      "epoch": 28.82,
      "learning_rate": 0.00028248000000000005,
      "loss": 0.5126,
      "step": 2940
    },
    {
      "epoch": 28.92,
      "learning_rate": 0.00028208000000000004,
      "loss": 0.5077,
      "step": 2950
    },
    {
      "epoch": 29.02,
      "learning_rate": 0.00028168000000000003,
      "loss": 0.4462,
      "step": 2960
    },
    {
      "epoch": 29.12,
      "learning_rate": 0.00028128,
      "loss": 0.4924,
      "step": 2970
    },
    {
      "epoch": 29.22,
      "learning_rate": 0.00028088,
      "loss": 0.4502,
      "step": 2980
    },
    {
      "epoch": 29.31,
      "learning_rate": 0.00028048000000000005,
      "loss": 0.4494,
      "step": 2990
    },
    {
      "epoch": 29.41,
      "learning_rate": 0.00028008000000000004,
      "loss": 0.4389,
      "step": 3000
    },
    {
      "epoch": 29.51,
      "learning_rate": 0.00027968000000000003,
      "loss": 0.4506,
      "step": 3010
    },
    {
      "epoch": 29.61,
      "learning_rate": 0.00027928,
      "loss": 0.4621,
      "step": 3020
    },
    {
      "epoch": 29.71,
      "learning_rate": 0.00027888,
      "loss": 0.4678,
      "step": 3030
    },
    {
      "epoch": 29.8,
      "learning_rate": 0.00027848,
      "loss": 0.4587,
      "step": 3040
    },
    {
      "epoch": 29.9,
      "learning_rate": 0.00027808000000000005,
      "loss": 0.4719,
      "step": 3050
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.00027768000000000004,
      "loss": 0.4271,
      "step": 3060
    },
    {
      "epoch": 30.1,
      "learning_rate": 0.00027728000000000003,
      "loss": 0.4325,
      "step": 3070
    },
    {
      "epoch": 30.2,
      "learning_rate": 0.00027688,
      "loss": 0.4136,
      "step": 3080
    },
    {
      "epoch": 30.29,
      "learning_rate": 0.00027648,
      "loss": 0.4362,
      "step": 3090
    },
    {
      "epoch": 30.39,
      "learning_rate": 0.00027608000000000005,
      "loss": 0.4099,
      "step": 3100
    },
    {
      "epoch": 30.49,
      "learning_rate": 0.00027568000000000004,
      "loss": 0.4238,
      "step": 3110
    },
    {
      "epoch": 30.59,
      "learning_rate": 0.00027528000000000003,
      "loss": 0.4295,
      "step": 3120
    },
    {
      "epoch": 30.69,
      "learning_rate": 0.00027488,
      "loss": 0.4264,
      "step": 3130
    },
    {
      "epoch": 30.78,
      "learning_rate": 0.00027448,
      "loss": 0.417,
      "step": 3140
    },
    {
      "epoch": 30.88,
      "learning_rate": 0.00027408,
      "loss": 0.4304,
      "step": 3150
    },
    {
      "epoch": 30.98,
      "learning_rate": 0.00027368000000000005,
      "loss": 0.4272,
      "step": 3160
    },
    {
      "epoch": 31.08,
      "learning_rate": 0.00027328000000000004,
      "loss": 0.3643,
      "step": 3170
    },
    {
      "epoch": 31.18,
      "learning_rate": 0.00027288000000000003,
      "loss": 0.3797,
      "step": 3180
    },
    {
      "epoch": 31.27,
      "learning_rate": 0.00027248,
      "loss": 0.4051,
      "step": 3190
    },
    {
      "epoch": 31.37,
      "learning_rate": 0.00027208,
      "loss": 0.3869,
      "step": 3200
    },
    {
      "epoch": 31.47,
      "learning_rate": 0.00027168,
      "loss": 0.3997,
      "step": 3210
    },
    {
      "epoch": 31.57,
      "learning_rate": 0.00027128000000000005,
      "loss": 0.4037,
      "step": 3220
    },
    {
      "epoch": 31.67,
      "learning_rate": 0.00027088000000000004,
      "loss": 0.385,
      "step": 3230
    },
    {
      "epoch": 31.76,
      "learning_rate": 0.00027048,
      "loss": 0.3919,
      "step": 3240
    },
    {
      "epoch": 31.86,
      "learning_rate": 0.00027008,
      "loss": 0.3823,
      "step": 3250
    },
    {
      "epoch": 31.96,
      "learning_rate": 0.00026968,
      "loss": 0.382,
      "step": 3260
    },
    {
      "epoch": 32.06,
      "learning_rate": 0.00026928,
      "loss": 0.349,
      "step": 3270
    },
    {
      "epoch": 32.16,
      "learning_rate": 0.00026888000000000004,
      "loss": 0.3416,
      "step": 3280
    },
    {
      "epoch": 32.25,
      "learning_rate": 0.00026848000000000003,
      "loss": 0.3733,
      "step": 3290
    },
    {
      "epoch": 32.35,
      "learning_rate": 0.00026808,
      "loss": 0.3589,
      "step": 3300
    },
    {
      "epoch": 32.45,
      "learning_rate": 0.00026768,
      "loss": 0.3697,
      "step": 3310
    },
    {
      "epoch": 32.55,
      "learning_rate": 0.00026728,
      "loss": 0.3787,
      "step": 3320
    },
    {
      "epoch": 32.65,
      "learning_rate": 0.00026688,
      "loss": 0.3564,
      "step": 3330
    },
    {
      "epoch": 32.75,
      "learning_rate": 0.00026648000000000004,
      "loss": 0.3636,
      "step": 3340
    },
    {
      "epoch": 32.84,
      "learning_rate": 0.00026608000000000003,
      "loss": 0.3519,
      "step": 3350
    },
    {
      "epoch": 32.94,
      "learning_rate": 0.00026568,
      "loss": 0.3526,
      "step": 3360
    },
    {
      "epoch": 33.04,
      "learning_rate": 0.00026528,
      "loss": 0.3287,
      "step": 3370
    },
    {
      "epoch": 33.14,
      "learning_rate": 0.00026488,
      "loss": 0.3138,
      "step": 3380
    },
    {
      "epoch": 33.24,
      "learning_rate": 0.00026448000000000004,
      "loss": 0.3274,
      "step": 3390
    },
    {
      "epoch": 33.33,
      "learning_rate": 0.00026408000000000003,
      "loss": 0.3302,
      "step": 3400
    },
    {
      "epoch": 33.43,
      "learning_rate": 0.00026368,
      "loss": 0.3327,
      "step": 3410
    },
    {
      "epoch": 33.53,
      "learning_rate": 0.00026328,
      "loss": 0.3383,
      "step": 3420
    },
    {
      "epoch": 33.63,
      "learning_rate": 0.00026288,
      "loss": 0.3343,
      "step": 3430
    },
    {
      "epoch": 33.73,
      "learning_rate": 0.00026248,
      "loss": 0.3375,
      "step": 3440
    },
    {
      "epoch": 33.82,
      "learning_rate": 0.00026208000000000004,
      "loss": 0.3378,
      "step": 3450
    },
    {
      "epoch": 33.92,
      "learning_rate": 0.00026168000000000003,
      "loss": 0.3298,
      "step": 3460
    },
    {
      "epoch": 34.02,
      "learning_rate": 0.00026128,
      "loss": 0.2956,
      "step": 3470
    },
    {
      "epoch": 34.12,
      "learning_rate": 0.00026088,
      "loss": 0.295,
      "step": 3480
    },
    {
      "epoch": 34.22,
      "learning_rate": 0.00026048,
      "loss": 0.2847,
      "step": 3490
    },
    {
      "epoch": 34.31,
      "learning_rate": 0.00026008,
      "loss": 0.2993,
      "step": 3500
    },
    {
      "epoch": 34.41,
      "learning_rate": 0.00025968000000000003,
      "loss": 0.2974,
      "step": 3510
    },
    {
      "epoch": 34.51,
      "learning_rate": 0.00025928,
      "loss": 0.3076,
      "step": 3520
    },
    {
      "epoch": 34.61,
      "learning_rate": 0.00025888,
      "loss": 0.3146,
      "step": 3530
    },
    {
      "epoch": 34.71,
      "learning_rate": 0.00025848,
      "loss": 0.3124,
      "step": 3540
    },
    {
      "epoch": 34.8,
      "learning_rate": 0.00025808,
      "loss": 0.3072,
      "step": 3550
    },
    {
      "epoch": 34.9,
      "learning_rate": 0.00025768,
      "loss": 0.3078,
      "step": 3560
    },
    {
      "epoch": 35.0,
      "learning_rate": 0.00025728000000000003,
      "loss": 0.2738,
      "step": 3570
    },
    {
      "epoch": 35.1,
      "learning_rate": 0.00025688,
      "loss": 0.2736,
      "step": 3580
    },
    {
      "epoch": 35.2,
      "learning_rate": 0.00025648,
      "loss": 0.257,
      "step": 3590
    },
    {
      "epoch": 35.29,
      "learning_rate": 0.00025608,
      "loss": 0.2671,
      "step": 3600
    },
    {
      "epoch": 35.39,
      "learning_rate": 0.00025568,
      "loss": 0.2638,
      "step": 3610
    },
    {
      "epoch": 35.49,
      "learning_rate": 0.00025528,
      "loss": 0.2707,
      "step": 3620
    },
    {
      "epoch": 35.59,
      "learning_rate": 0.00025488,
      "loss": 0.2795,
      "step": 3630
    },
    {
      "epoch": 35.69,
      "learning_rate": 0.00025448,
      "loss": 0.2765,
      "step": 3640
    },
    {
      "epoch": 35.78,
      "learning_rate": 0.00025408,
      "loss": 0.2757,
      "step": 3650
    },
    {
      "epoch": 35.88,
      "learning_rate": 0.00025368,
      "loss": 0.2779,
      "step": 3660
    },
    {
      "epoch": 35.98,
      "learning_rate": 0.00025328,
      "loss": 0.2628,
      "step": 3670
    },
    {
      "epoch": 36.08,
      "learning_rate": 0.00025288000000000003,
      "loss": 0.2403,
      "step": 3680
    },
    {
      "epoch": 36.18,
      "learning_rate": 0.00025248,
      "loss": 0.2361,
      "step": 3690
    },
    {
      "epoch": 36.27,
      "learning_rate": 0.00025208,
      "loss": 0.2394,
      "step": 3700
    },
    {
      "epoch": 36.37,
      "learning_rate": 0.00025168,
      "loss": 0.2309,
      "step": 3710
    },
    {
      "epoch": 36.47,
      "learning_rate": 0.00025128,
      "loss": 0.233,
      "step": 3720
    },
    {
      "epoch": 36.57,
      "learning_rate": 0.00025088,
      "loss": 0.2446,
      "step": 3730
    },
    {
      "epoch": 36.67,
      "learning_rate": 0.00025048000000000003,
      "loss": 0.2385,
      "step": 3740
    },
    {
      "epoch": 36.76,
      "learning_rate": 0.00025008,
      "loss": 0.239,
      "step": 3750
    },
    {
      "epoch": 36.86,
      "learning_rate": 0.00024968,
      "loss": 0.2392,
      "step": 3760
    },
    {
      "epoch": 36.96,
      "learning_rate": 0.00024928,
      "loss": 0.2318,
      "step": 3770
    },
    {
      "epoch": 37.06,
      "learning_rate": 0.00024888,
      "loss": 0.218,
      "step": 3780
    },
    {
      "epoch": 37.16,
      "learning_rate": 0.00024848,
      "loss": 0.2059,
      "step": 3790
    },
    {
      "epoch": 37.25,
      "learning_rate": 0.00024808,
      "loss": 0.2082,
      "step": 3800
    },
    {
      "epoch": 37.35,
      "learning_rate": 0.00024768,
      "loss": 0.2058,
      "step": 3810
    },
    {
      "epoch": 37.45,
      "learning_rate": 0.00024728,
      "loss": 0.2037,
      "step": 3820
    },
    {
      "epoch": 37.55,
      "learning_rate": 0.00024688,
      "loss": 0.2083,
      "step": 3830
    },
    {
      "epoch": 37.65,
      "learning_rate": 0.00024648,
      "loss": 0.2007,
      "step": 3840
    },
    {
      "epoch": 37.75,
      "learning_rate": 0.00024608,
      "loss": 0.2012,
      "step": 3850
    },
    {
      "epoch": 37.84,
      "learning_rate": 0.00024568,
      "loss": 0.1972,
      "step": 3860
    },
    {
      "epoch": 37.94,
      "learning_rate": 0.00024528,
      "loss": 0.1996,
      "step": 3870
    },
    {
      "epoch": 38.04,
      "learning_rate": 0.00024488,
      "loss": 0.1859,
      "step": 3880
    },
    {
      "epoch": 38.14,
      "learning_rate": 0.00024448,
      "loss": 0.1801,
      "step": 3890
    },
    {
      "epoch": 38.24,
      "learning_rate": 0.00024408,
      "loss": 0.1708,
      "step": 3900
    },
    {
      "epoch": 38.33,
      "learning_rate": 0.00024368,
      "loss": 0.1709,
      "step": 3910
    },
    {
      "epoch": 38.43,
      "learning_rate": 0.00024328,
      "loss": 0.1736,
      "step": 3920
    },
    {
      "epoch": 38.53,
      "learning_rate": 0.00024288,
      "loss": 0.1726,
      "step": 3930
    },
    {
      "epoch": 38.63,
      "learning_rate": 0.00024248,
      "loss": 0.1681,
      "step": 3940
    },
    {
      "epoch": 38.73,
      "learning_rate": 0.00024207999999999999,
      "loss": 0.1674,
      "step": 3950
    },
    {
      "epoch": 38.82,
      "learning_rate": 0.00024168,
      "loss": 0.1601,
      "step": 3960
    },
    {
      "epoch": 38.92,
      "learning_rate": 0.00024128,
      "loss": 0.1552,
      "step": 3970
    },
    {
      "epoch": 39.02,
      "learning_rate": 0.00024087999999999998,
      "loss": 0.1414,
      "step": 3980
    },
    {
      "epoch": 39.12,
      "learning_rate": 0.00024048,
      "loss": 0.151,
      "step": 3990
    },
    {
      "epoch": 39.22,
      "learning_rate": 0.00024008,
      "loss": 0.1418,
      "step": 4000
    },
    {
      "epoch": 39.31,
      "learning_rate": 0.00023967999999999998,
      "loss": 0.1372,
      "step": 4010
    },
    {
      "epoch": 39.41,
      "learning_rate": 0.00023928,
      "loss": 0.1334,
      "step": 4020
    },
    {
      "epoch": 39.51,
      "learning_rate": 0.00023888,
      "loss": 0.1339,
      "step": 4030
    },
    {
      "epoch": 39.61,
      "learning_rate": 0.00023847999999999998,
      "loss": 0.137,
      "step": 4040
    },
    {
      "epoch": 39.71,
      "learning_rate": 0.00023808,
      "loss": 0.1325,
      "step": 4050
    },
    {
      "epoch": 39.8,
      "learning_rate": 0.00023768,
      "loss": 0.1322,
      "step": 4060
    },
    {
      "epoch": 39.9,
      "learning_rate": 0.00023727999999999998,
      "loss": 0.1262,
      "step": 4070
    },
    {
      "epoch": 40.0,
      "learning_rate": 0.00023688,
      "loss": 0.1111,
      "step": 4080
    },
    {
      "epoch": 40.1,
      "learning_rate": 0.00023647999999999999,
      "loss": 0.1125,
      "step": 4090
    },
    {
      "epoch": 40.2,
      "learning_rate": 0.00023608,
      "loss": 0.1128,
      "step": 4100
    },
    {
      "epoch": 40.29,
      "learning_rate": 0.00023568,
      "loss": 0.1098,
      "step": 4110
    },
    {
      "epoch": 40.39,
      "learning_rate": 0.00023527999999999998,
      "loss": 0.1057,
      "step": 4120
    },
    {
      "epoch": 40.49,
      "learning_rate": 0.00023488000000000003,
      "loss": 0.1019,
      "step": 4130
    },
    {
      "epoch": 40.59,
      "learning_rate": 0.00023448000000000005,
      "loss": 0.1024,
      "step": 4140
    },
    {
      "epoch": 40.69,
      "learning_rate": 0.00023408000000000004,
      "loss": 0.1014,
      "step": 4150
    },
    {
      "epoch": 40.78,
      "learning_rate": 0.00023368000000000003,
      "loss": 0.1038,
      "step": 4160
    },
    {
      "epoch": 40.88,
      "learning_rate": 0.00023328000000000004,
      "loss": 0.096,
      "step": 4170
    },
    {
      "epoch": 40.98,
      "learning_rate": 0.00023288000000000003,
      "loss": 0.0943,
      "step": 4180
    },
    {
      "epoch": 41.08,
      "learning_rate": 0.00023248000000000002,
      "loss": 0.0821,
      "step": 4190
    },
    {
      "epoch": 41.18,
      "learning_rate": 0.00023208000000000004,
      "loss": 0.0861,
      "step": 4200
    },
    {
      "epoch": 41.27,
      "learning_rate": 0.00023168000000000003,
      "loss": 0.0854,
      "step": 4210
    },
    {
      "epoch": 41.37,
      "learning_rate": 0.00023128000000000002,
      "loss": 0.081,
      "step": 4220
    },
    {
      "epoch": 41.47,
      "learning_rate": 0.00023088000000000004,
      "loss": 0.0787,
      "step": 4230
    },
    {
      "epoch": 41.57,
      "learning_rate": 0.00023048000000000003,
      "loss": 0.0775,
      "step": 4240
    },
    {
      "epoch": 41.67,
      "learning_rate": 0.00023008000000000002,
      "loss": 0.0742,
      "step": 4250
    },
    {
      "epoch": 41.76,
      "learning_rate": 0.00022968000000000004,
      "loss": 0.0727,
      "step": 4260
    },
    {
      "epoch": 41.86,
      "learning_rate": 0.00022928000000000003,
      "loss": 0.0716,
      "step": 4270
    },
    {
      "epoch": 41.96,
      "learning_rate": 0.00022888000000000002,
      "loss": 0.0701,
      "step": 4280
    },
    {
      "epoch": 42.06,
      "learning_rate": 0.00022848000000000004,
      "loss": 0.0639,
      "step": 4290
    },
    {
      "epoch": 42.16,
      "learning_rate": 0.00022808000000000003,
      "loss": 0.0636,
      "step": 4300
    },
    {
      "epoch": 42.25,
      "learning_rate": 0.00022768000000000002,
      "loss": 0.0625,
      "step": 4310
    },
    {
      "epoch": 42.35,
      "learning_rate": 0.00022728000000000003,
      "loss": 0.0607,
      "step": 4320
    },
    {
      "epoch": 42.45,
      "learning_rate": 0.00022688000000000002,
      "loss": 0.058,
      "step": 4330
    },
    {
      "epoch": 42.55,
      "learning_rate": 0.00022648000000000001,
      "loss": 0.0567,
      "step": 4340
    },
    {
      "epoch": 42.65,
      "learning_rate": 0.00022608000000000003,
      "loss": 0.0538,
      "step": 4350
    },
    {
      "epoch": 42.75,
      "learning_rate": 0.00022568000000000002,
      "loss": 0.0528,
      "step": 4360
    },
    {
      "epoch": 42.84,
      "learning_rate": 0.00022528,
      "loss": 0.0507,
      "step": 4370
    },
    {
      "epoch": 42.94,
      "learning_rate": 0.00022488000000000003,
      "loss": 0.0481,
      "step": 4380
    },
    {
      "epoch": 43.04,
      "learning_rate": 0.00022448000000000002,
      "loss": 0.0434,
      "step": 4390
    },
    {
      "epoch": 43.14,
      "learning_rate": 0.00022408000000000004,
      "loss": 0.0451,
      "step": 4400
    },
    {
      "epoch": 43.24,
      "learning_rate": 0.00022368000000000003,
      "loss": 0.0433,
      "step": 4410
    },
    {
      "epoch": 43.33,
      "learning_rate": 0.00022328000000000002,
      "loss": 0.0419,
      "step": 4420
    },
    {
      "epoch": 43.43,
      "learning_rate": 0.00022288000000000003,
      "loss": 0.0408,
      "step": 4430
    },
    {
      "epoch": 43.53,
      "learning_rate": 0.00022248000000000002,
      "loss": 0.0386,
      "step": 4440
    },
    {
      "epoch": 43.63,
      "learning_rate": 0.00022208000000000002,
      "loss": 0.038,
      "step": 4450
    },
    {
      "epoch": 43.73,
      "learning_rate": 0.00022168000000000003,
      "loss": 0.0367,
      "step": 4460
    },
    {
      "epoch": 43.82,
      "learning_rate": 0.00022128000000000002,
      "loss": 0.0362,
      "step": 4470
    },
    {
      "epoch": 43.92,
      "learning_rate": 0.00022088,
      "loss": 0.0338,
      "step": 4480
    },
    {
      "epoch": 44.02,
      "learning_rate": 0.00022048000000000003,
      "loss": 0.0303,
      "step": 4490
    },
    {
      "epoch": 44.12,
      "learning_rate": 0.00022008000000000002,
      "loss": 0.0318,
      "step": 4500
    },
    {
      "epoch": 44.22,
      "learning_rate": 0.00021968,
      "loss": 0.031,
      "step": 4510
    },
    {
      "epoch": 44.31,
      "learning_rate": 0.00021928000000000003,
      "loss": 0.0295,
      "step": 4520
    },
    {
      "epoch": 44.41,
      "learning_rate": 0.00021888000000000002,
      "loss": 0.029,
      "step": 4530
    },
    {
      "epoch": 44.51,
      "learning_rate": 0.00021848,
      "loss": 0.0275,
      "step": 4540
    },
    {
      "epoch": 44.61,
      "learning_rate": 0.00021808000000000003,
      "loss": 0.0267,
      "step": 4550
    },
    {
      "epoch": 44.71,
      "learning_rate": 0.00021768000000000002,
      "loss": 0.0269,
      "step": 4560
    },
    {
      "epoch": 44.8,
      "learning_rate": 0.00021728,
      "loss": 0.0262,
      "step": 4570
    },
    {
      "epoch": 44.9,
      "learning_rate": 0.00021688000000000002,
      "loss": 0.0259,
      "step": 4580
    },
    {
      "epoch": 45.0,
      "learning_rate": 0.00021648000000000001,
      "loss": 0.0235,
      "step": 4590
    },
    {
      "epoch": 45.1,
      "learning_rate": 0.00021608,
      "loss": 0.0225,
      "step": 4600
    },
    {
      "epoch": 45.2,
      "learning_rate": 0.00021568000000000002,
      "loss": 0.0238,
      "step": 4610
    },
    {
      "epoch": 45.29,
      "learning_rate": 0.00021528,
      "loss": 0.0233,
      "step": 4620
    },
    {
      "epoch": 45.39,
      "learning_rate": 0.00021488,
      "loss": 0.0222,
      "step": 4630
    },
    {
      "epoch": 45.49,
      "learning_rate": 0.00021448000000000002,
      "loss": 0.0218,
      "step": 4640
    },
    {
      "epoch": 45.59,
      "learning_rate": 0.00021408,
      "loss": 0.0207,
      "step": 4650
    },
    {
      "epoch": 45.69,
      "learning_rate": 0.00021368,
      "loss": 0.0207,
      "step": 4660
    },
    {
      "epoch": 45.78,
      "learning_rate": 0.00021328000000000002,
      "loss": 0.0214,
      "step": 4670
    },
    {
      "epoch": 45.88,
      "learning_rate": 0.00021288,
      "loss": 0.021,
      "step": 4680
    },
    {
      "epoch": 45.98,
      "learning_rate": 0.00021248000000000003,
      "loss": 0.0205,
      "step": 4690
    },
    {
      "epoch": 46.08,
      "learning_rate": 0.00021208000000000002,
      "loss": 0.0179,
      "step": 4700
    },
    {
      "epoch": 46.18,
      "learning_rate": 0.00021168,
      "loss": 0.019,
      "step": 4710
    },
    {
      "epoch": 46.27,
      "learning_rate": 0.00021128000000000002,
      "loss": 0.0192,
      "step": 4720
    },
    {
      "epoch": 46.37,
      "learning_rate": 0.00021088000000000001,
      "loss": 0.0187,
      "step": 4730
    },
    {
      "epoch": 46.47,
      "learning_rate": 0.00021048,
      "loss": 0.0183,
      "step": 4740
    },
    {
      "epoch": 46.57,
      "learning_rate": 0.00021008000000000002,
      "loss": 0.0176,
      "step": 4750
    },
    {
      "epoch": 46.67,
      "learning_rate": 0.00020968,
      "loss": 0.0169,
      "step": 4760
    },
    {
      "epoch": 46.76,
      "learning_rate": 0.00020928,
      "loss": 0.0171,
      "step": 4770
    },
    {
      "epoch": 46.86,
      "learning_rate": 0.00020888000000000002,
      "loss": 0.0172,
      "step": 4780
    },
    {
      "epoch": 46.96,
      "learning_rate": 0.00020848,
      "loss": 0.0175,
      "step": 4790
    },
    {
      "epoch": 47.06,
      "learning_rate": 0.00020808,
      "loss": 0.0153,
      "step": 4800
    },
    {
      "epoch": 47.16,
      "learning_rate": 0.00020768000000000002,
      "loss": 0.0162,
      "step": 4810
    },
    {
      "epoch": 47.25,
      "learning_rate": 0.00020728,
      "loss": 0.0161,
      "step": 4820
    },
    {
      "epoch": 47.35,
      "learning_rate": 0.00020688,
      "loss": 0.0161,
      "step": 4830
    },
    {
      "epoch": 47.45,
      "learning_rate": 0.00020648000000000002,
      "loss": 0.0157,
      "step": 4840
    },
    {
      "epoch": 47.55,
      "learning_rate": 0.00020608,
      "loss": 0.0149,
      "step": 4850
    },
    {
      "epoch": 47.65,
      "learning_rate": 0.00020568,
      "loss": 0.0143,
      "step": 4860
    },
    {
      "epoch": 47.75,
      "learning_rate": 0.00020528,
      "loss": 0.0138,
      "step": 4870
    },
    {
      "epoch": 47.84,
      "learning_rate": 0.00020488,
      "loss": 0.0136,
      "step": 4880
    },
    {
      "epoch": 47.94,
      "learning_rate": 0.00020448,
      "loss": 0.014,
      "step": 4890
    },
    {
      "epoch": 48.04,
      "learning_rate": 0.00020408,
      "loss": 0.0131,
      "step": 4900
    },
    {
      "epoch": 48.14,
      "learning_rate": 0.00020368,
      "loss": 0.0141,
      "step": 4910
    },
    {
      "epoch": 48.24,
      "learning_rate": 0.00020328,
      "loss": 0.0134,
      "step": 4920
    },
    {
      "epoch": 48.33,
      "learning_rate": 0.00020288,
      "loss": 0.013,
      "step": 4930
    },
    {
      "epoch": 48.43,
      "learning_rate": 0.00020248,
      "loss": 0.0128,
      "step": 4940
    },
    {
      "epoch": 48.53,
      "learning_rate": 0.00020208,
      "loss": 0.0127,
      "step": 4950
    },
    {
      "epoch": 48.63,
      "learning_rate": 0.00020168,
      "loss": 0.012,
      "step": 4960
    },
    {
      "epoch": 48.73,
      "learning_rate": 0.00020128,
      "loss": 0.0114,
      "step": 4970
    },
    {
      "epoch": 48.82,
      "learning_rate": 0.00020088000000000001,
      "loss": 0.0113,
      "step": 4980
    },
    {
      "epoch": 48.92,
      "learning_rate": 0.00020048,
      "loss": 0.0118,
      "step": 4990
    },
    {
      "epoch": 49.02,
      "learning_rate": 0.00020008,
      "loss": 0.011,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_train_epochs": 99,
  "save_steps": 1000,
  "total_flos": 3.3405215440896e+17,
  "trial_name": null,
  "trial_params": null
}