{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 12.5,
      "learning_rate": 1e-05,
      "loss": 4.8178,
      "step": 10
    },
    {
      "epoch": 0.002,
      "grad_norm": 15.0,
      "learning_rate": 1e-05,
      "loss": 3.5825,
      "step": 20
    },
    {
      "epoch": 0.003,
      "grad_norm": 12.0625,
      "learning_rate": 1e-05,
      "loss": 3.7442,
      "step": 30
    },
    {
      "epoch": 0.004,
      "grad_norm": 14.125,
      "learning_rate": 1e-05,
      "loss": 3.384,
      "step": 40
    },
    {
      "epoch": 0.005,
      "grad_norm": 13.6875,
      "learning_rate": 1e-05,
      "loss": 3.131,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 2.5038,
      "step": 60
    },
    {
      "epoch": 0.007,
      "grad_norm": 11.75,
      "learning_rate": 1e-05,
      "loss": 3.1993,
      "step": 70
    },
    {
      "epoch": 0.008,
      "grad_norm": 8.5625,
      "learning_rate": 1e-05,
      "loss": 2.7646,
      "step": 80
    },
    {
      "epoch": 0.009,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 2.9096,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 2.318,
      "step": 100
    },
    {
      "epoch": 0.011,
      "grad_norm": 13.625,
      "learning_rate": 1e-05,
      "loss": 2.4156,
      "step": 110
    },
    {
      "epoch": 0.012,
      "grad_norm": 13.1875,
      "learning_rate": 1e-05,
      "loss": 2.5023,
      "step": 120
    },
    {
      "epoch": 0.013,
      "grad_norm": 10.75,
      "learning_rate": 1e-05,
      "loss": 2.526,
      "step": 130
    },
    {
      "epoch": 0.014,
      "grad_norm": 10.875,
      "learning_rate": 1e-05,
      "loss": 2.5059,
      "step": 140
    },
    {
      "epoch": 0.015,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 2.5216,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 10.4375,
      "learning_rate": 1e-05,
      "loss": 2.0107,
      "step": 160
    },
    {
      "epoch": 0.017,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 2.6329,
      "step": 170
    },
    {
      "epoch": 0.018,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 2.5261,
      "step": 180
    },
    {
      "epoch": 0.019,
      "grad_norm": 10.6875,
      "learning_rate": 1e-05,
      "loss": 2.3679,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.5,
      "learning_rate": 1e-05,
      "loss": 2.5779,
      "step": 200
    },
    {
      "epoch": 0.021,
      "grad_norm": 13.9375,
      "learning_rate": 1e-05,
      "loss": 2.3589,
      "step": 210
    },
    {
      "epoch": 0.022,
      "grad_norm": 10.125,
      "learning_rate": 1e-05,
      "loss": 2.2825,
      "step": 220
    },
    {
      "epoch": 0.023,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 2.5285,
      "step": 230
    },
    {
      "epoch": 0.024,
      "grad_norm": 13.8125,
      "learning_rate": 1e-05,
      "loss": 2.1764,
      "step": 240
    },
    {
      "epoch": 0.025,
      "grad_norm": 10.8125,
      "learning_rate": 1e-05,
      "loss": 2.0058,
      "step": 250
    },
    {
      "epoch": 0.026,
      "grad_norm": 23.5,
      "learning_rate": 1e-05,
      "loss": 2.177,
      "step": 260
    },
    {
      "epoch": 0.027,
      "grad_norm": 15.0,
      "learning_rate": 1e-05,
      "loss": 2.5206,
      "step": 270
    },
    {
      "epoch": 0.028,
      "grad_norm": 11.5625,
      "learning_rate": 1e-05,
      "loss": 2.2534,
      "step": 280
    },
    {
      "epoch": 0.029,
      "grad_norm": 13.125,
      "learning_rate": 1e-05,
      "loss": 2.0127,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.75,
      "learning_rate": 1e-05,
      "loss": 2.1765,
      "step": 300
    },
    {
      "epoch": 0.031,
      "grad_norm": 8.75,
      "learning_rate": 1e-05,
      "loss": 1.7118,
      "step": 310
    },
    {
      "epoch": 0.032,
      "grad_norm": 12.75,
      "learning_rate": 1e-05,
      "loss": 1.9264,
      "step": 320
    },
    {
      "epoch": 0.033,
      "grad_norm": 10.875,
      "learning_rate": 1e-05,
      "loss": 2.1289,
      "step": 330
    },
    {
      "epoch": 0.034,
      "grad_norm": 9.875,
      "learning_rate": 1e-05,
      "loss": 2.1149,
      "step": 340
    },
    {
      "epoch": 0.035,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 2.1333,
      "step": 350
    },
    {
      "epoch": 0.036,
      "grad_norm": 20.125,
      "learning_rate": 1e-05,
      "loss": 2.3275,
      "step": 360
    },
    {
      "epoch": 0.037,
      "grad_norm": 17.0,
      "learning_rate": 1e-05,
      "loss": 2.6152,
      "step": 370
    },
    {
      "epoch": 0.038,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 2.0291,
      "step": 380
    },
    {
      "epoch": 0.039,
      "grad_norm": 21.0,
      "learning_rate": 1e-05,
      "loss": 2.1323,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.5,
      "learning_rate": 1e-05,
      "loss": 1.9322,
      "step": 400
    },
    {
      "epoch": 0.041,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 4.5483,
      "step": 410
    },
    {
      "epoch": 0.042,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 2.315,
      "step": 420
    },
    {
      "epoch": 0.043,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 1.9925,
      "step": 430
    },
    {
      "epoch": 0.044,
      "grad_norm": 15.8125,
      "learning_rate": 1e-05,
      "loss": 2.2832,
      "step": 440
    },
    {
      "epoch": 0.045,
      "grad_norm": 9.5625,
      "learning_rate": 1e-05,
      "loss": 1.8034,
      "step": 450
    },
    {
      "epoch": 0.046,
      "grad_norm": 16.125,
      "learning_rate": 1e-05,
      "loss": 2.2797,
      "step": 460
    },
    {
      "epoch": 0.047,
      "grad_norm": 16.5,
      "learning_rate": 1e-05,
      "loss": 2.1466,
      "step": 470
    },
    {
      "epoch": 0.048,
      "grad_norm": 11.5625,
      "learning_rate": 1e-05,
      "loss": 2.2831,
      "step": 480
    },
    {
      "epoch": 0.049,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 1.9358,
      "step": 490
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 1.8661,
      "step": 500
    },
    {
      "epoch": 0.051,
      "grad_norm": 26.0,
      "learning_rate": 1e-05,
      "loss": 2.1335,
      "step": 510
    },
    {
      "epoch": 0.052,
      "grad_norm": 10.6875,
      "learning_rate": 1e-05,
      "loss": 1.9704,
      "step": 520
    },
    {
      "epoch": 0.053,
      "grad_norm": 17.25,
      "learning_rate": 1e-05,
      "loss": 2.4475,
      "step": 530
    },
    {
      "epoch": 0.054,
      "grad_norm": 12.5,
      "learning_rate": 1e-05,
      "loss": 2.0418,
      "step": 540
    },
    {
      "epoch": 0.055,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 2.3471,
      "step": 550
    },
    {
      "epoch": 0.056,
      "grad_norm": 10.5625,
      "learning_rate": 1e-05,
      "loss": 2.1779,
      "step": 560
    },
    {
      "epoch": 0.057,
      "grad_norm": 13.625,
      "learning_rate": 1e-05,
      "loss": 1.9877,
      "step": 570
    },
    {
      "epoch": 0.058,
      "grad_norm": 12.375,
      "learning_rate": 1e-05,
      "loss": 1.9859,
      "step": 580
    },
    {
      "epoch": 0.059,
      "grad_norm": 13.5625,
      "learning_rate": 1e-05,
      "loss": 2.2656,
      "step": 590
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.375,
      "learning_rate": 1e-05,
      "loss": 2.1405,
      "step": 600
    },
    {
      "epoch": 0.061,
      "grad_norm": 11.125,
      "learning_rate": 1e-05,
      "loss": 1.8149,
      "step": 610
    },
    {
      "epoch": 0.062,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 2.0527,
      "step": 620
    },
    {
      "epoch": 0.063,
      "grad_norm": 9.8125,
      "learning_rate": 1e-05,
      "loss": 1.9026,
      "step": 630
    },
    {
      "epoch": 0.064,
      "grad_norm": 10.4375,
      "learning_rate": 1e-05,
      "loss": 2.1107,
      "step": 640
    },
    {
      "epoch": 0.065,
      "grad_norm": 13.0625,
      "learning_rate": 1e-05,
      "loss": 1.7943,
      "step": 650
    },
    {
      "epoch": 0.066,
      "grad_norm": 12.25,
      "learning_rate": 1e-05,
      "loss": 1.8904,
      "step": 660
    },
    {
      "epoch": 0.067,
      "grad_norm": 19.25,
      "learning_rate": 1e-05,
      "loss": 2.0774,
      "step": 670
    },
    {
      "epoch": 0.068,
      "grad_norm": 9.1875,
      "learning_rate": 1e-05,
      "loss": 2.0132,
      "step": 680
    },
    {
      "epoch": 0.069,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 3.4386,
      "step": 690
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.3125,
      "learning_rate": 1e-05,
      "loss": 1.7073,
      "step": 700
    },
    {
      "epoch": 0.071,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.9587,
      "step": 710
    },
    {
      "epoch": 0.072,
      "grad_norm": 10.5625,
      "learning_rate": 1e-05,
      "loss": 1.706,
      "step": 720
    },
    {
      "epoch": 0.073,
      "grad_norm": 13.9375,
      "learning_rate": 1e-05,
      "loss": 1.9715,
      "step": 730
    },
    {
      "epoch": 0.074,
      "grad_norm": 24.125,
      "learning_rate": 1e-05,
      "loss": 2.1728,
      "step": 740
    },
    {
      "epoch": 0.075,
      "grad_norm": 15.9375,
      "learning_rate": 1e-05,
      "loss": 1.8296,
      "step": 750
    },
    {
      "epoch": 0.076,
      "grad_norm": 19.0,
      "learning_rate": 1e-05,
      "loss": 2.0436,
      "step": 760
    },
    {
      "epoch": 0.077,
      "grad_norm": 11.8125,
      "learning_rate": 1e-05,
      "loss": 1.8851,
      "step": 770
    },
    {
      "epoch": 0.078,
      "grad_norm": 12.6875,
      "learning_rate": 1e-05,
      "loss": 1.8506,
      "step": 780
    },
    {
      "epoch": 0.079,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 2.0522,
      "step": 790
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.5,
      "learning_rate": 1e-05,
      "loss": 2.1956,
      "step": 800
    },
    {
      "epoch": 0.081,
      "grad_norm": 9.25,
      "learning_rate": 1e-05,
      "loss": 2.5578,
      "step": 810
    },
    {
      "epoch": 0.082,
      "grad_norm": 12.25,
      "learning_rate": 1e-05,
      "loss": 1.9882,
      "step": 820
    },
    {
      "epoch": 0.083,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 2.0607,
      "step": 830
    },
    {
      "epoch": 0.084,
      "grad_norm": 8.5625,
      "learning_rate": 1e-05,
      "loss": 1.7497,
      "step": 840
    },
    {
      "epoch": 0.085,
      "grad_norm": 15.625,
      "learning_rate": 1e-05,
      "loss": 2.1156,
      "step": 850
    },
    {
      "epoch": 0.086,
      "grad_norm": 24.5,
      "learning_rate": 1e-05,
      "loss": 2.3917,
      "step": 860
    },
    {
      "epoch": 0.087,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 1.8928,
      "step": 870
    },
    {
      "epoch": 0.088,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.8971,
      "step": 880
    },
    {
      "epoch": 0.089,
      "grad_norm": 12.5,
      "learning_rate": 1e-05,
      "loss": 1.964,
      "step": 890
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 1.7416,
      "step": 900
    },
    {
      "epoch": 0.091,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.9847,
      "step": 910
    },
    {
      "epoch": 0.092,
      "grad_norm": 15.0,
      "learning_rate": 1e-05,
      "loss": 2.0122,
      "step": 920
    },
    {
      "epoch": 0.093,
      "grad_norm": 15.4375,
      "learning_rate": 1e-05,
      "loss": 2.0215,
      "step": 930
    },
    {
      "epoch": 0.094,
      "grad_norm": 17.0,
      "learning_rate": 1e-05,
      "loss": 1.8966,
      "step": 940
    },
    {
      "epoch": 0.095,
      "grad_norm": 10.875,
      "learning_rate": 1e-05,
      "loss": 2.1329,
      "step": 950
    },
    {
      "epoch": 0.096,
      "grad_norm": 15.3125,
      "learning_rate": 1e-05,
      "loss": 2.0491,
      "step": 960
    },
    {
      "epoch": 0.097,
      "grad_norm": 15.875,
      "learning_rate": 1e-05,
      "loss": 2.0018,
      "step": 970
    },
    {
      "epoch": 0.098,
      "grad_norm": 14.25,
      "learning_rate": 1e-05,
      "loss": 2.0352,
      "step": 980
    },
    {
      "epoch": 0.099,
      "grad_norm": 8.75,
      "learning_rate": 1e-05,
      "loss": 1.6287,
      "step": 990
    },
    {
      "epoch": 0.1,
      "grad_norm": 14.8125,
      "learning_rate": 1e-05,
      "loss": 2.1023,
      "step": 1000
    },
    {
      "epoch": 0.101,
      "grad_norm": 13.75,
      "learning_rate": 1e-05,
      "loss": 2.2457,
      "step": 1010
    },
    {
      "epoch": 0.102,
      "grad_norm": 14.0,
      "learning_rate": 1e-05,
      "loss": 2.0374,
      "step": 1020
    },
    {
      "epoch": 0.103,
      "grad_norm": 14.125,
      "learning_rate": 1e-05,
      "loss": 2.1338,
      "step": 1030
    },
    {
      "epoch": 0.104,
      "grad_norm": 16.75,
      "learning_rate": 1e-05,
      "loss": 1.8552,
      "step": 1040
    },
    {
      "epoch": 0.105,
      "grad_norm": 8.8125,
      "learning_rate": 1e-05,
      "loss": 1.7178,
      "step": 1050
    },
    {
      "epoch": 0.106,
      "grad_norm": 14.3125,
      "learning_rate": 1e-05,
      "loss": 1.9404,
      "step": 1060
    },
    {
      "epoch": 0.107,
      "grad_norm": 17.625,
      "learning_rate": 1e-05,
      "loss": 1.9794,
      "step": 1070
    },
    {
      "epoch": 0.108,
      "grad_norm": 13.4375,
      "learning_rate": 1e-05,
      "loss": 1.7349,
      "step": 1080
    },
    {
      "epoch": 0.109,
      "grad_norm": 13.8125,
      "learning_rate": 1e-05,
      "loss": 1.9692,
      "step": 1090
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.25,
      "learning_rate": 1e-05,
      "loss": 1.9093,
      "step": 1100
    },
    {
      "epoch": 0.111,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.729,
      "step": 1110
    },
    {
      "epoch": 0.112,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 1.8809,
      "step": 1120
    },
    {
      "epoch": 0.113,
      "grad_norm": 13.75,
      "learning_rate": 1e-05,
      "loss": 2.2044,
      "step": 1130
    },
    {
      "epoch": 0.114,
      "grad_norm": 10.8125,
      "learning_rate": 1e-05,
      "loss": 1.5764,
      "step": 1140
    },
    {
      "epoch": 0.115,
      "grad_norm": 12.9375,
      "learning_rate": 1e-05,
      "loss": 2.0671,
      "step": 1150
    },
    {
      "epoch": 0.116,
      "grad_norm": 12.0625,
      "learning_rate": 1e-05,
      "loss": 1.6256,
      "step": 1160
    },
    {
      "epoch": 0.117,
      "grad_norm": 35.75,
      "learning_rate": 1e-05,
      "loss": 2.0677,
      "step": 1170
    },
    {
      "epoch": 0.118,
      "grad_norm": 20.0,
      "learning_rate": 1e-05,
      "loss": 2.332,
      "step": 1180
    },
    {
      "epoch": 0.119,
      "grad_norm": 11.25,
      "learning_rate": 1e-05,
      "loss": 2.0185,
      "step": 1190
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.8125,
      "learning_rate": 1e-05,
      "loss": 1.7844,
      "step": 1200
    },
    {
      "epoch": 0.121,
      "grad_norm": 16.75,
      "learning_rate": 1e-05,
      "loss": 1.9551,
      "step": 1210
    },
    {
      "epoch": 0.122,
      "grad_norm": 9.5625,
      "learning_rate": 1e-05,
      "loss": 1.6201,
      "step": 1220
    },
    {
      "epoch": 0.123,
      "grad_norm": 18.875,
      "learning_rate": 1e-05,
      "loss": 2.091,
      "step": 1230
    },
    {
      "epoch": 0.124,
      "grad_norm": 16.375,
      "learning_rate": 1e-05,
      "loss": 2.3187,
      "step": 1240
    },
    {
      "epoch": 0.125,
      "grad_norm": 9.5625,
      "learning_rate": 1e-05,
      "loss": 1.9329,
      "step": 1250
    },
    {
      "epoch": 0.126,
      "grad_norm": 13.0625,
      "learning_rate": 1e-05,
      "loss": 2.8992,
      "step": 1260
    },
    {
      "epoch": 0.127,
      "grad_norm": 10.625,
      "learning_rate": 1e-05,
      "loss": 1.9234,
      "step": 1270
    },
    {
      "epoch": 0.128,
      "grad_norm": 15.0625,
      "learning_rate": 1e-05,
      "loss": 1.8989,
      "step": 1280
    },
    {
      "epoch": 0.129,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 1.9102,
      "step": 1290
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.6875,
      "learning_rate": 1e-05,
      "loss": 1.9919,
      "step": 1300
    },
    {
      "epoch": 0.131,
      "grad_norm": 16.75,
      "learning_rate": 1e-05,
      "loss": 1.8764,
      "step": 1310
    },
    {
      "epoch": 0.132,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 1.88,
      "step": 1320
    },
    {
      "epoch": 0.133,
      "grad_norm": 15.875,
      "learning_rate": 1e-05,
      "loss": 2.1248,
      "step": 1330
    },
    {
      "epoch": 0.134,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.8875,
      "step": 1340
    },
    {
      "epoch": 0.135,
      "grad_norm": 10.75,
      "learning_rate": 1e-05,
      "loss": 1.7,
      "step": 1350
    },
    {
      "epoch": 0.136,
      "grad_norm": 22.0,
      "learning_rate": 1e-05,
      "loss": 1.8537,
      "step": 1360
    },
    {
      "epoch": 0.137,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 2.0803,
      "step": 1370
    },
    {
      "epoch": 0.138,
      "grad_norm": 16.25,
      "learning_rate": 1e-05,
      "loss": 1.7608,
      "step": 1380
    },
    {
      "epoch": 0.139,
      "grad_norm": 12.75,
      "learning_rate": 1e-05,
      "loss": 2.0586,
      "step": 1390
    },
    {
      "epoch": 0.14,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 1.9522,
      "step": 1400
    },
    {
      "epoch": 0.141,
      "grad_norm": 13.8125,
      "learning_rate": 1e-05,
      "loss": 1.8574,
      "step": 1410
    },
    {
      "epoch": 0.142,
      "grad_norm": 12.1875,
      "learning_rate": 1e-05,
      "loss": 1.9955,
      "step": 1420
    },
    {
      "epoch": 0.143,
      "grad_norm": 13.25,
      "learning_rate": 1e-05,
      "loss": 1.7986,
      "step": 1430
    },
    {
      "epoch": 0.144,
      "grad_norm": 13.8125,
      "learning_rate": 1e-05,
      "loss": 1.808,
      "step": 1440
    },
    {
      "epoch": 0.145,
      "grad_norm": 15.125,
      "learning_rate": 1e-05,
      "loss": 1.8473,
      "step": 1450
    },
    {
      "epoch": 0.146,
      "grad_norm": 18.5,
      "learning_rate": 1e-05,
      "loss": 1.859,
      "step": 1460
    },
    {
      "epoch": 0.147,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 1.9265,
      "step": 1470
    },
    {
      "epoch": 0.148,
      "grad_norm": 26.75,
      "learning_rate": 1e-05,
      "loss": 2.0496,
      "step": 1480
    },
    {
      "epoch": 0.149,
      "grad_norm": 14.8125,
      "learning_rate": 1e-05,
      "loss": 1.6431,
      "step": 1490
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.875,
      "learning_rate": 1e-05,
      "loss": 1.8198,
      "step": 1500
    },
    {
      "epoch": 0.151,
      "grad_norm": 15.625,
      "learning_rate": 1e-05,
      "loss": 1.8431,
      "step": 1510
    },
    {
      "epoch": 0.152,
      "grad_norm": 13.625,
      "learning_rate": 1e-05,
      "loss": 2.1109,
      "step": 1520
    },
    {
      "epoch": 0.153,
      "grad_norm": 17.75,
      "learning_rate": 1e-05,
      "loss": 1.8236,
      "step": 1530
    },
    {
      "epoch": 0.154,
      "grad_norm": 19.375,
      "learning_rate": 1e-05,
      "loss": 1.898,
      "step": 1540
    },
    {
      "epoch": 0.155,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.6926,
      "step": 1550
    },
    {
      "epoch": 0.156,
      "grad_norm": 21.625,
      "learning_rate": 1e-05,
      "loss": 2.2279,
      "step": 1560
    },
    {
      "epoch": 0.157,
      "grad_norm": 8.625,
      "learning_rate": 1e-05,
      "loss": 1.6541,
      "step": 1570
    },
    {
      "epoch": 0.158,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 2.0563,
      "step": 1580
    },
    {
      "epoch": 0.159,
      "grad_norm": 11.25,
      "learning_rate": 1e-05,
      "loss": 1.63,
      "step": 1590
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.0,
      "learning_rate": 1e-05,
      "loss": 1.9665,
      "step": 1600
    },
    {
      "epoch": 0.161,
      "grad_norm": 13.1875,
      "learning_rate": 1e-05,
      "loss": 1.7257,
      "step": 1610
    },
    {
      "epoch": 0.162,
      "grad_norm": 8.8125,
      "learning_rate": 1e-05,
      "loss": 1.6863,
      "step": 1620
    },
    {
      "epoch": 0.163,
      "grad_norm": 17.625,
      "learning_rate": 1e-05,
      "loss": 1.8712,
      "step": 1630
    },
    {
      "epoch": 0.164,
      "grad_norm": 12.375,
      "learning_rate": 1e-05,
      "loss": 1.8015,
      "step": 1640
    },
    {
      "epoch": 0.165,
      "grad_norm": 13.6875,
      "learning_rate": 1e-05,
      "loss": 1.8289,
      "step": 1650
    },
    {
      "epoch": 0.166,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.8839,
      "step": 1660
    },
    {
      "epoch": 0.167,
      "grad_norm": 12.375,
      "learning_rate": 1e-05,
      "loss": 1.7506,
      "step": 1670
    },
    {
      "epoch": 0.168,
      "grad_norm": 22.0,
      "learning_rate": 1e-05,
      "loss": 1.9137,
      "step": 1680
    },
    {
      "epoch": 0.169,
      "grad_norm": 12.125,
      "learning_rate": 1e-05,
      "loss": 1.7776,
      "step": 1690
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.5625,
      "learning_rate": 1e-05,
      "loss": 1.8953,
      "step": 1700
    },
    {
      "epoch": 0.171,
      "grad_norm": 15.0,
      "learning_rate": 1e-05,
      "loss": 2.0353,
      "step": 1710
    },
    {
      "epoch": 0.172,
      "grad_norm": 8.75,
      "learning_rate": 1e-05,
      "loss": 1.666,
      "step": 1720
    },
    {
      "epoch": 0.173,
      "grad_norm": 10.625,
      "learning_rate": 1e-05,
      "loss": 1.5912,
      "step": 1730
    },
    {
      "epoch": 0.174,
      "grad_norm": 13.875,
      "learning_rate": 1e-05,
      "loss": 1.9534,
      "step": 1740
    },
    {
      "epoch": 0.175,
      "grad_norm": 17.625,
      "learning_rate": 1e-05,
      "loss": 1.76,
      "step": 1750
    },
    {
      "epoch": 0.176,
      "grad_norm": 12.625,
      "learning_rate": 1e-05,
      "loss": 2.2381,
      "step": 1760
    },
    {
      "epoch": 0.177,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.9322,
      "step": 1770
    },
    {
      "epoch": 0.178,
      "grad_norm": 13.5,
      "learning_rate": 1e-05,
      "loss": 1.8795,
      "step": 1780
    },
    {
      "epoch": 0.179,
      "grad_norm": 14.375,
      "learning_rate": 1e-05,
      "loss": 1.8898,
      "step": 1790
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 1.5216,
      "step": 1800
    },
    {
      "epoch": 0.181,
      "grad_norm": 13.375,
      "learning_rate": 1e-05,
      "loss": 1.7606,
      "step": 1810
    },
    {
      "epoch": 0.182,
      "grad_norm": 15.3125,
      "learning_rate": 1e-05,
      "loss": 2.3597,
      "step": 1820
    },
    {
      "epoch": 0.183,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 1.7536,
      "step": 1830
    },
    {
      "epoch": 0.184,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.9664,
      "step": 1840
    },
    {
      "epoch": 0.185,
      "grad_norm": 11.0625,
      "learning_rate": 1e-05,
      "loss": 1.9236,
      "step": 1850
    },
    {
      "epoch": 0.186,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 1.9873,
      "step": 1860
    },
    {
      "epoch": 0.187,
      "grad_norm": 9.3125,
      "learning_rate": 1e-05,
      "loss": 1.5427,
      "step": 1870
    },
    {
      "epoch": 0.188,
      "grad_norm": 14.625,
      "learning_rate": 1e-05,
      "loss": 1.9574,
      "step": 1880
    },
    {
      "epoch": 0.189,
      "grad_norm": 15.8125,
      "learning_rate": 1e-05,
      "loss": 1.8403,
      "step": 1890
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.5,
      "learning_rate": 1e-05,
      "loss": 1.817,
      "step": 1900
    },
    {
      "epoch": 0.191,
      "grad_norm": 12.0625,
      "learning_rate": 1e-05,
      "loss": 1.7169,
      "step": 1910
    },
    {
      "epoch": 0.192,
      "grad_norm": 7.15625,
      "learning_rate": 1e-05,
      "loss": 1.7619,
      "step": 1920
    },
    {
      "epoch": 0.193,
      "grad_norm": 16.25,
      "learning_rate": 1e-05,
      "loss": 1.8253,
      "step": 1930
    },
    {
      "epoch": 0.194,
      "grad_norm": 11.1875,
      "learning_rate": 1e-05,
      "loss": 1.6775,
      "step": 1940
    },
    {
      "epoch": 0.195,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.7772,
      "step": 1950
    },
    {
      "epoch": 0.196,
      "grad_norm": 8.625,
      "learning_rate": 1e-05,
      "loss": 1.5344,
      "step": 1960
    },
    {
      "epoch": 0.197,
      "grad_norm": 19.5,
      "learning_rate": 1e-05,
      "loss": 1.7716,
      "step": 1970
    },
    {
      "epoch": 0.198,
      "grad_norm": 8.25,
      "learning_rate": 1e-05,
      "loss": 1.6532,
      "step": 1980
    },
    {
      "epoch": 0.199,
      "grad_norm": 12.4375,
      "learning_rate": 1e-05,
      "loss": 1.729,
      "step": 1990
    },
    {
      "epoch": 0.2,
      "grad_norm": 16.75,
      "learning_rate": 1e-05,
      "loss": 2.298,
      "step": 2000
    },
    {
      "epoch": 0.201,
      "grad_norm": 14.3125,
      "learning_rate": 1e-05,
      "loss": 2.0102,
      "step": 2010
    },
    {
      "epoch": 0.202,
      "grad_norm": 14.6875,
      "learning_rate": 1e-05,
      "loss": 1.7934,
      "step": 2020
    },
    {
      "epoch": 0.203,
      "grad_norm": 14.0625,
      "learning_rate": 1e-05,
      "loss": 1.7483,
      "step": 2030
    },
    {
      "epoch": 0.204,
      "grad_norm": 14.4375,
      "learning_rate": 1e-05,
      "loss": 1.6826,
      "step": 2040
    },
    {
      "epoch": 0.205,
      "grad_norm": 13.875,
      "learning_rate": 1e-05,
      "loss": 1.882,
      "step": 2050
    },
    {
      "epoch": 0.206,
      "grad_norm": 8.1875,
      "learning_rate": 1e-05,
      "loss": 1.5249,
      "step": 2060
    },
    {
      "epoch": 0.207,
      "grad_norm": 13.3125,
      "learning_rate": 1e-05,
      "loss": 1.5253,
      "step": 2070
    },
    {
      "epoch": 0.208,
      "grad_norm": 9.6875,
      "learning_rate": 1e-05,
      "loss": 1.6956,
      "step": 2080
    },
    {
      "epoch": 0.209,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.7893,
      "step": 2090
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.875,
      "learning_rate": 1e-05,
      "loss": 1.669,
      "step": 2100
    },
    {
      "epoch": 0.211,
      "grad_norm": 11.8125,
      "learning_rate": 1e-05,
      "loss": 1.6619,
      "step": 2110
    },
    {
      "epoch": 0.212,
      "grad_norm": 13.0,
      "learning_rate": 1e-05,
      "loss": 1.6914,
      "step": 2120
    },
    {
      "epoch": 0.213,
      "grad_norm": 13.1875,
      "learning_rate": 1e-05,
      "loss": 1.7975,
      "step": 2130
    },
    {
      "epoch": 0.214,
      "grad_norm": 11.375,
      "learning_rate": 1e-05,
      "loss": 1.7764,
      "step": 2140
    },
    {
      "epoch": 0.215,
      "grad_norm": 14.3125,
      "learning_rate": 1e-05,
      "loss": 1.6417,
      "step": 2150
    },
    {
      "epoch": 0.216,
      "grad_norm": 11.8125,
      "learning_rate": 1e-05,
      "loss": 1.9006,
      "step": 2160
    },
    {
      "epoch": 0.217,
      "grad_norm": 23.5,
      "learning_rate": 1e-05,
      "loss": 1.9608,
      "step": 2170
    },
    {
      "epoch": 0.218,
      "grad_norm": 13.375,
      "learning_rate": 1e-05,
      "loss": 1.6029,
      "step": 2180
    },
    {
      "epoch": 0.219,
      "grad_norm": 14.8125,
      "learning_rate": 1e-05,
      "loss": 1.6106,
      "step": 2190
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.6875,
      "learning_rate": 1e-05,
      "loss": 1.6695,
      "step": 2200
    },
    {
      "epoch": 0.221,
      "grad_norm": 10.25,
      "learning_rate": 1e-05,
      "loss": 1.9153,
      "step": 2210
    },
    {
      "epoch": 0.222,
      "grad_norm": 9.5625,
      "learning_rate": 1e-05,
      "loss": 1.9047,
      "step": 2220
    },
    {
      "epoch": 0.223,
      "grad_norm": 17.875,
      "learning_rate": 1e-05,
      "loss": 1.7705,
      "step": 2230
    },
    {
      "epoch": 0.224,
      "grad_norm": 14.125,
      "learning_rate": 1e-05,
      "loss": 1.7962,
      "step": 2240
    },
    {
      "epoch": 0.225,
      "grad_norm": 17.625,
      "learning_rate": 1e-05,
      "loss": 2.0007,
      "step": 2250
    },
    {
      "epoch": 0.226,
      "grad_norm": 16.375,
      "learning_rate": 1e-05,
      "loss": 1.8476,
      "step": 2260
    },
    {
      "epoch": 0.227,
      "grad_norm": 13.875,
      "learning_rate": 1e-05,
      "loss": 1.7328,
      "step": 2270
    },
    {
      "epoch": 0.228,
      "grad_norm": 10.3125,
      "learning_rate": 1e-05,
      "loss": 1.7836,
      "step": 2280
    },
    {
      "epoch": 0.229,
      "grad_norm": 10.875,
      "learning_rate": 1e-05,
      "loss": 1.6794,
      "step": 2290
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.6353,
      "step": 2300
    },
    {
      "epoch": 0.231,
      "grad_norm": 17.125,
      "learning_rate": 1e-05,
      "loss": 2.062,
      "step": 2310
    },
    {
      "epoch": 0.232,
      "grad_norm": 14.875,
      "learning_rate": 1e-05,
      "loss": 1.6459,
      "step": 2320
    },
    {
      "epoch": 0.233,
      "grad_norm": 12.125,
      "learning_rate": 1e-05,
      "loss": 2.0505,
      "step": 2330
    },
    {
      "epoch": 0.234,
      "grad_norm": 16.875,
      "learning_rate": 1e-05,
      "loss": 1.7258,
      "step": 2340
    },
    {
      "epoch": 0.235,
      "grad_norm": 20.875,
      "learning_rate": 1e-05,
      "loss": 2.3883,
      "step": 2350
    },
    {
      "epoch": 0.236,
      "grad_norm": 16.625,
      "learning_rate": 1e-05,
      "loss": 2.0195,
      "step": 2360
    },
    {
      "epoch": 0.237,
      "grad_norm": 13.6875,
      "learning_rate": 1e-05,
      "loss": 1.8855,
      "step": 2370
    },
    {
      "epoch": 0.238,
      "grad_norm": 12.0625,
      "learning_rate": 1e-05,
      "loss": 1.811,
      "step": 2380
    },
    {
      "epoch": 0.239,
      "grad_norm": 15.5625,
      "learning_rate": 1e-05,
      "loss": 1.4883,
      "step": 2390
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.3125,
      "learning_rate": 1e-05,
      "loss": 1.8279,
      "step": 2400
    },
    {
      "epoch": 0.241,
      "grad_norm": 16.25,
      "learning_rate": 1e-05,
      "loss": 1.6529,
      "step": 2410
    },
    {
      "epoch": 0.242,
      "grad_norm": 15.25,
      "learning_rate": 1e-05,
      "loss": 1.6827,
      "step": 2420
    },
    {
      "epoch": 0.243,
      "grad_norm": 9.875,
      "learning_rate": 1e-05,
      "loss": 1.9552,
      "step": 2430
    },
    {
      "epoch": 0.244,
      "grad_norm": 12.4375,
      "learning_rate": 1e-05,
      "loss": 1.6529,
      "step": 2440
    },
    {
      "epoch": 0.245,
      "grad_norm": 12.25,
      "learning_rate": 1e-05,
      "loss": 2.1241,
      "step": 2450
    },
    {
      "epoch": 0.246,
      "grad_norm": 13.3125,
      "learning_rate": 1e-05,
      "loss": 1.9516,
      "step": 2460
    },
    {
      "epoch": 0.247,
      "grad_norm": 17.0,
      "learning_rate": 1e-05,
      "loss": 1.9497,
      "step": 2470
    },
    {
      "epoch": 0.248,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.9704,
      "step": 2480
    },
    {
      "epoch": 0.249,
      "grad_norm": 18.25,
      "learning_rate": 1e-05,
      "loss": 1.5859,
      "step": 2490
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.0,
      "learning_rate": 1e-05,
      "loss": 2.1096,
      "step": 2500
    },
    {
      "epoch": 0.251,
      "grad_norm": 9.375,
      "learning_rate": 1e-05,
      "loss": 1.9469,
      "step": 2510
    },
    {
      "epoch": 0.252,
      "grad_norm": 16.75,
      "learning_rate": 1e-05,
      "loss": 1.8282,
      "step": 2520
    },
    {
      "epoch": 0.253,
      "grad_norm": 12.125,
      "learning_rate": 1e-05,
      "loss": 1.6671,
      "step": 2530
    },
    {
      "epoch": 0.254,
      "grad_norm": 12.75,
      "learning_rate": 1e-05,
      "loss": 1.5986,
      "step": 2540
    },
    {
      "epoch": 0.255,
      "grad_norm": 15.1875,
      "learning_rate": 1e-05,
      "loss": 1.7509,
      "step": 2550
    },
    {
      "epoch": 0.256,
      "grad_norm": 9.875,
      "learning_rate": 1e-05,
      "loss": 1.4303,
      "step": 2560
    },
    {
      "epoch": 0.257,
      "grad_norm": 10.8125,
      "learning_rate": 1e-05,
      "loss": 1.7776,
      "step": 2570
    },
    {
      "epoch": 0.258,
      "grad_norm": 15.3125,
      "learning_rate": 1e-05,
      "loss": 1.6254,
      "step": 2580
    },
    {
      "epoch": 0.259,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 2.0549,
      "step": 2590
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.8125,
      "learning_rate": 1e-05,
      "loss": 1.76,
      "step": 2600
    },
    {
      "epoch": 0.261,
      "grad_norm": 21.125,
      "learning_rate": 1e-05,
      "loss": 1.674,
      "step": 2610
    },
    {
      "epoch": 0.262,
      "grad_norm": 29.75,
      "learning_rate": 1e-05,
      "loss": 2.1121,
      "step": 2620
    },
    {
      "epoch": 0.263,
      "grad_norm": 12.125,
      "learning_rate": 1e-05,
      "loss": 2.0997,
      "step": 2630
    },
    {
      "epoch": 0.264,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.9954,
      "step": 2640
    },
    {
      "epoch": 0.265,
      "grad_norm": 13.5,
      "learning_rate": 1e-05,
      "loss": 2.0843,
      "step": 2650
    },
    {
      "epoch": 0.266,
      "grad_norm": 14.4375,
      "learning_rate": 1e-05,
      "loss": 1.6546,
      "step": 2660
    },
    {
      "epoch": 0.267,
      "grad_norm": 12.9375,
      "learning_rate": 1e-05,
      "loss": 1.6982,
      "step": 2670
    },
    {
      "epoch": 0.268,
      "grad_norm": 18.25,
      "learning_rate": 1e-05,
      "loss": 1.662,
      "step": 2680
    },
    {
      "epoch": 0.269,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 1.7339,
      "step": 2690
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.125,
      "learning_rate": 1e-05,
      "loss": 1.8546,
      "step": 2700
    },
    {
      "epoch": 0.271,
      "grad_norm": 14.375,
      "learning_rate": 1e-05,
      "loss": 1.5251,
      "step": 2710
    },
    {
      "epoch": 0.272,
      "grad_norm": 16.625,
      "learning_rate": 1e-05,
      "loss": 2.0505,
      "step": 2720
    },
    {
      "epoch": 0.273,
      "grad_norm": 14.0,
      "learning_rate": 1e-05,
      "loss": 1.7012,
      "step": 2730
    },
    {
      "epoch": 0.274,
      "grad_norm": 15.8125,
      "learning_rate": 1e-05,
      "loss": 2.0051,
      "step": 2740
    },
    {
      "epoch": 0.275,
      "grad_norm": 18.875,
      "learning_rate": 1e-05,
      "loss": 2.079,
      "step": 2750
    },
    {
      "epoch": 0.276,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 1.8696,
      "step": 2760
    },
    {
      "epoch": 0.277,
      "grad_norm": 25.375,
      "learning_rate": 1e-05,
      "loss": 1.892,
      "step": 2770
    },
    {
      "epoch": 0.278,
      "grad_norm": 11.3125,
      "learning_rate": 1e-05,
      "loss": 1.6591,
      "step": 2780
    },
    {
      "epoch": 0.279,
      "grad_norm": 10.25,
      "learning_rate": 1e-05,
      "loss": 1.6393,
      "step": 2790
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.9375,
      "learning_rate": 1e-05,
      "loss": 1.7708,
      "step": 2800
    },
    {
      "epoch": 0.281,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.7073,
      "step": 2810
    },
    {
      "epoch": 0.282,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.6873,
      "step": 2820
    },
    {
      "epoch": 0.283,
      "grad_norm": 12.9375,
      "learning_rate": 1e-05,
      "loss": 1.6827,
      "step": 2830
    },
    {
      "epoch": 0.284,
      "grad_norm": 11.25,
      "learning_rate": 1e-05,
      "loss": 1.6091,
      "step": 2840
    },
    {
      "epoch": 0.285,
      "grad_norm": 15.0625,
      "learning_rate": 1e-05,
      "loss": 1.4785,
      "step": 2850
    },
    {
      "epoch": 0.286,
      "grad_norm": 20.0,
      "learning_rate": 1e-05,
      "loss": 1.8004,
      "step": 2860
    },
    {
      "epoch": 0.287,
      "grad_norm": 17.25,
      "learning_rate": 1e-05,
      "loss": 2.0436,
      "step": 2870
    },
    {
      "epoch": 0.288,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.8407,
      "step": 2880
    },
    {
      "epoch": 0.289,
      "grad_norm": 26.125,
      "learning_rate": 1e-05,
      "loss": 1.8658,
      "step": 2890
    },
    {
      "epoch": 0.29,
      "grad_norm": 21.0,
      "learning_rate": 1e-05,
      "loss": 1.5994,
      "step": 2900
    },
    {
      "epoch": 0.291,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.8467,
      "step": 2910
    },
    {
      "epoch": 0.292,
      "grad_norm": 7.5625,
      "learning_rate": 1e-05,
      "loss": 1.7801,
      "step": 2920
    },
    {
      "epoch": 0.293,
      "grad_norm": 14.375,
      "learning_rate": 1e-05,
      "loss": 1.759,
      "step": 2930
    },
    {
      "epoch": 0.294,
      "grad_norm": 10.9375,
      "learning_rate": 1e-05,
      "loss": 1.5657,
      "step": 2940
    },
    {
      "epoch": 0.295,
      "grad_norm": 12.75,
      "learning_rate": 1e-05,
      "loss": 1.8102,
      "step": 2950
    },
    {
      "epoch": 0.296,
      "grad_norm": 10.5625,
      "learning_rate": 1e-05,
      "loss": 1.7314,
      "step": 2960
    },
    {
      "epoch": 0.297,
      "grad_norm": 13.6875,
      "learning_rate": 1e-05,
      "loss": 1.7728,
      "step": 2970
    },
    {
      "epoch": 0.298,
      "grad_norm": 11.125,
      "learning_rate": 1e-05,
      "loss": 1.8561,
      "step": 2980
    },
    {
      "epoch": 0.299,
      "grad_norm": 15.75,
      "learning_rate": 1e-05,
      "loss": 1.9538,
      "step": 2990
    },
    {
      "epoch": 0.3,
      "grad_norm": 16.0,
      "learning_rate": 1e-05,
      "loss": 1.7902,
      "step": 3000
    },
    {
      "epoch": 0.301,
      "grad_norm": 16.875,
      "learning_rate": 1e-05,
      "loss": 1.8427,
      "step": 3010
    },
    {
      "epoch": 0.302,
      "grad_norm": 9.75,
      "learning_rate": 1e-05,
      "loss": 1.6548,
      "step": 3020
    },
    {
      "epoch": 0.303,
      "grad_norm": 12.4375,
      "learning_rate": 1e-05,
      "loss": 1.5911,
      "step": 3030
    },
    {
      "epoch": 0.304,
      "grad_norm": 16.5,
      "learning_rate": 1e-05,
      "loss": 2.1339,
      "step": 3040
    },
    {
      "epoch": 0.305,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.5936,
      "step": 3050
    },
    {
      "epoch": 0.306,
      "grad_norm": 11.8125,
      "learning_rate": 1e-05,
      "loss": 1.8171,
      "step": 3060
    },
    {
      "epoch": 0.307,
      "grad_norm": 17.75,
      "learning_rate": 1e-05,
      "loss": 1.9329,
      "step": 3070
    },
    {
      "epoch": 0.308,
      "grad_norm": 11.625,
      "learning_rate": 1e-05,
      "loss": 2.3599,
      "step": 3080
    },
    {
      "epoch": 0.309,
      "grad_norm": 14.6875,
      "learning_rate": 1e-05,
      "loss": 1.939,
      "step": 3090
    },
    {
      "epoch": 0.31,
      "grad_norm": 18.5,
      "learning_rate": 1e-05,
      "loss": 1.6498,
      "step": 3100
    },
    {
      "epoch": 0.311,
      "grad_norm": 14.25,
      "learning_rate": 1e-05,
      "loss": 2.231,
      "step": 3110
    },
    {
      "epoch": 0.312,
      "grad_norm": 10.0625,
      "learning_rate": 1e-05,
      "loss": 1.6276,
      "step": 3120
    },
    {
      "epoch": 0.313,
      "grad_norm": 20.25,
      "learning_rate": 1e-05,
      "loss": 1.6312,
      "step": 3130
    },
    {
      "epoch": 0.314,
      "grad_norm": 22.25,
      "learning_rate": 1e-05,
      "loss": 1.9693,
      "step": 3140
    },
    {
      "epoch": 0.315,
      "grad_norm": 8.75,
      "learning_rate": 1e-05,
      "loss": 1.7655,
      "step": 3150
    },
    {
      "epoch": 0.316,
      "grad_norm": 10.6875,
      "learning_rate": 1e-05,
      "loss": 1.7087,
      "step": 3160
    },
    {
      "epoch": 0.317,
      "grad_norm": 11.0,
      "learning_rate": 1e-05,
      "loss": 1.8512,
      "step": 3170
    },
    {
      "epoch": 0.318,
      "grad_norm": 15.0625,
      "learning_rate": 1e-05,
      "loss": 1.5766,
      "step": 3180
    },
    {
      "epoch": 0.319,
      "grad_norm": 10.3125,
      "learning_rate": 1e-05,
      "loss": 1.654,
      "step": 3190
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 1.8305,
      "step": 3200
    },
    {
      "epoch": 0.321,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.3816,
      "step": 3210
    },
    {
      "epoch": 0.322,
      "grad_norm": 11.125,
      "learning_rate": 1e-05,
      "loss": 1.6938,
      "step": 3220
    },
    {
      "epoch": 0.323,
      "grad_norm": 12.1875,
      "learning_rate": 1e-05,
      "loss": 1.7627,
      "step": 3230
    },
    {
      "epoch": 0.324,
      "grad_norm": 10.625,
      "learning_rate": 1e-05,
      "loss": 1.7272,
      "step": 3240
    },
    {
      "epoch": 0.325,
      "grad_norm": 9.3125,
      "learning_rate": 1e-05,
      "loss": 1.7999,
      "step": 3250
    },
    {
      "epoch": 0.326,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 1.6183,
      "step": 3260
    },
    {
      "epoch": 0.327,
      "grad_norm": 15.6875,
      "learning_rate": 1e-05,
      "loss": 1.5071,
      "step": 3270
    },
    {
      "epoch": 0.328,
      "grad_norm": 16.5,
      "learning_rate": 1e-05,
      "loss": 1.7405,
      "step": 3280
    },
    {
      "epoch": 0.329,
      "grad_norm": 14.4375,
      "learning_rate": 1e-05,
      "loss": 1.7118,
      "step": 3290
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.875,
      "learning_rate": 1e-05,
      "loss": 1.6516,
      "step": 3300
    },
    {
      "epoch": 0.331,
      "grad_norm": 9.75,
      "learning_rate": 1e-05,
      "loss": 1.5256,
      "step": 3310
    },
    {
      "epoch": 0.332,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 1.7742,
      "step": 3320
    },
    {
      "epoch": 0.333,
      "grad_norm": 19.75,
      "learning_rate": 1e-05,
      "loss": 1.4777,
      "step": 3330
    },
    {
      "epoch": 0.334,
      "grad_norm": 15.9375,
      "learning_rate": 1e-05,
      "loss": 1.6461,
      "step": 3340
    },
    {
      "epoch": 0.335,
      "grad_norm": 13.9375,
      "learning_rate": 1e-05,
      "loss": 1.9839,
      "step": 3350
    },
    {
      "epoch": 0.336,
      "grad_norm": 13.0625,
      "learning_rate": 1e-05,
      "loss": 1.9362,
      "step": 3360
    },
    {
      "epoch": 0.337,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.8488,
      "step": 3370
    },
    {
      "epoch": 0.338,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 1.6072,
      "step": 3380
    },
    {
      "epoch": 0.339,
      "grad_norm": 14.625,
      "learning_rate": 1e-05,
      "loss": 1.9736,
      "step": 3390
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.375,
      "learning_rate": 1e-05,
      "loss": 1.7284,
      "step": 3400
    },
    {
      "epoch": 0.341,
      "grad_norm": 18.625,
      "learning_rate": 1e-05,
      "loss": 1.7378,
      "step": 3410
    },
    {
      "epoch": 0.342,
      "grad_norm": 17.5,
      "learning_rate": 1e-05,
      "loss": 1.6532,
      "step": 3420
    },
    {
      "epoch": 0.343,
      "grad_norm": 11.5625,
      "learning_rate": 1e-05,
      "loss": 1.5446,
      "step": 3430
    },
    {
      "epoch": 0.344,
      "grad_norm": 12.875,
      "learning_rate": 1e-05,
      "loss": 1.7806,
      "step": 3440
    },
    {
      "epoch": 0.345,
      "grad_norm": 14.0625,
      "learning_rate": 1e-05,
      "loss": 1.8175,
      "step": 3450
    },
    {
      "epoch": 0.346,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.5735,
      "step": 3460
    },
    {
      "epoch": 0.347,
      "grad_norm": 14.625,
      "learning_rate": 1e-05,
      "loss": 1.5978,
      "step": 3470
    },
    {
      "epoch": 0.348,
      "grad_norm": 13.125,
      "learning_rate": 1e-05,
      "loss": 1.8075,
      "step": 3480
    },
    {
      "epoch": 0.349,
      "grad_norm": 15.0625,
      "learning_rate": 1e-05,
      "loss": 1.849,
      "step": 3490
    },
    {
      "epoch": 0.35,
      "grad_norm": 15.875,
      "learning_rate": 1e-05,
      "loss": 2.3398,
      "step": 3500
    },
    {
      "epoch": 0.351,
      "grad_norm": 15.3125,
      "learning_rate": 1e-05,
      "loss": 1.5815,
      "step": 3510
    },
    {
      "epoch": 0.352,
      "grad_norm": 15.5625,
      "learning_rate": 1e-05,
      "loss": 1.7524,
      "step": 3520
    },
    {
      "epoch": 0.353,
      "grad_norm": 19.75,
      "learning_rate": 1e-05,
      "loss": 1.9377,
      "step": 3530
    },
    {
      "epoch": 0.354,
      "grad_norm": 13.25,
      "learning_rate": 1e-05,
      "loss": 1.6627,
      "step": 3540
    },
    {
      "epoch": 0.355,
      "grad_norm": 25.5,
      "learning_rate": 1e-05,
      "loss": 2.068,
      "step": 3550
    },
    {
      "epoch": 0.356,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.8114,
      "step": 3560
    },
    {
      "epoch": 0.357,
      "grad_norm": 11.0625,
      "learning_rate": 1e-05,
      "loss": 1.9164,
      "step": 3570
    },
    {
      "epoch": 0.358,
      "grad_norm": 14.5,
      "learning_rate": 1e-05,
      "loss": 1.8936,
      "step": 3580
    },
    {
      "epoch": 0.359,
      "grad_norm": 15.1875,
      "learning_rate": 1e-05,
      "loss": 1.6262,
      "step": 3590
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 1.5741,
      "step": 3600
    },
    {
      "epoch": 0.361,
      "grad_norm": 15.9375,
      "learning_rate": 1e-05,
      "loss": 1.8585,
      "step": 3610
    },
    {
      "epoch": 0.362,
      "grad_norm": 14.0,
      "learning_rate": 1e-05,
      "loss": 1.8171,
      "step": 3620
    },
    {
      "epoch": 0.363,
      "grad_norm": 15.6875,
      "learning_rate": 1e-05,
      "loss": 1.7571,
      "step": 3630
    },
    {
      "epoch": 0.364,
      "grad_norm": 13.5,
      "learning_rate": 1e-05,
      "loss": 1.6693,
      "step": 3640
    },
    {
      "epoch": 0.365,
      "grad_norm": 16.125,
      "learning_rate": 1e-05,
      "loss": 1.9003,
      "step": 3650
    },
    {
      "epoch": 0.366,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 1.4876,
      "step": 3660
    },
    {
      "epoch": 0.367,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 1.9428,
      "step": 3670
    },
    {
      "epoch": 0.368,
      "grad_norm": 17.5,
      "learning_rate": 1e-05,
      "loss": 1.9743,
      "step": 3680
    },
    {
      "epoch": 0.369,
      "grad_norm": 10.625,
      "learning_rate": 1e-05,
      "loss": 1.9147,
      "step": 3690
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.125,
      "learning_rate": 1e-05,
      "loss": 1.7948,
      "step": 3700
    },
    {
      "epoch": 0.371,
      "grad_norm": 14.0,
      "learning_rate": 1e-05,
      "loss": 1.7385,
      "step": 3710
    },
    {
      "epoch": 0.372,
      "grad_norm": 14.25,
      "learning_rate": 1e-05,
      "loss": 1.9731,
      "step": 3720
    },
    {
      "epoch": 0.373,
      "grad_norm": 12.1875,
      "learning_rate": 1e-05,
      "loss": 1.5938,
      "step": 3730
    },
    {
      "epoch": 0.374,
      "grad_norm": 11.6875,
      "learning_rate": 1e-05,
      "loss": 1.7843,
      "step": 3740
    },
    {
      "epoch": 0.375,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 1.7282,
      "step": 3750
    },
    {
      "epoch": 0.376,
      "grad_norm": 25.0,
      "learning_rate": 1e-05,
      "loss": 1.5946,
      "step": 3760
    },
    {
      "epoch": 0.377,
      "grad_norm": 10.8125,
      "learning_rate": 1e-05,
      "loss": 1.6076,
      "step": 3770
    },
    {
      "epoch": 0.378,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.6829,
      "step": 3780
    },
    {
      "epoch": 0.379,
      "grad_norm": 21.625,
      "learning_rate": 1e-05,
      "loss": 1.8562,
      "step": 3790
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.6875,
      "learning_rate": 1e-05,
      "loss": 1.8785,
      "step": 3800
    },
    {
      "epoch": 0.381,
      "grad_norm": 20.375,
      "learning_rate": 1e-05,
      "loss": 1.8259,
      "step": 3810
    },
    {
      "epoch": 0.382,
      "grad_norm": 12.625,
      "learning_rate": 1e-05,
      "loss": 1.5926,
      "step": 3820
    },
    {
      "epoch": 0.383,
      "grad_norm": 11.25,
      "learning_rate": 1e-05,
      "loss": 2.3374,
      "step": 3830
    },
    {
      "epoch": 0.384,
      "grad_norm": 15.25,
      "learning_rate": 1e-05,
      "loss": 1.775,
      "step": 3840
    },
    {
      "epoch": 0.385,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.4259,
      "step": 3850
    },
    {
      "epoch": 0.386,
      "grad_norm": 18.125,
      "learning_rate": 1e-05,
      "loss": 1.8814,
      "step": 3860
    },
    {
      "epoch": 0.387,
      "grad_norm": 13.1875,
      "learning_rate": 1e-05,
      "loss": 2.0477,
      "step": 3870
    },
    {
      "epoch": 0.388,
      "grad_norm": 15.375,
      "learning_rate": 1e-05,
      "loss": 2.1475,
      "step": 3880
    },
    {
      "epoch": 0.389,
      "grad_norm": 14.9375,
      "learning_rate": 1e-05,
      "loss": 1.7746,
      "step": 3890
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.75,
      "learning_rate": 1e-05,
      "loss": 2.0745,
      "step": 3900
    },
    {
      "epoch": 0.391,
      "grad_norm": 15.8125,
      "learning_rate": 1e-05,
      "loss": 1.7862,
      "step": 3910
    },
    {
      "epoch": 0.392,
      "grad_norm": 17.5,
      "learning_rate": 1e-05,
      "loss": 1.6869,
      "step": 3920
    },
    {
      "epoch": 0.393,
      "grad_norm": 16.5,
      "learning_rate": 1e-05,
      "loss": 1.8003,
      "step": 3930
    },
    {
      "epoch": 0.394,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.5467,
      "step": 3940
    },
    {
      "epoch": 0.395,
      "grad_norm": 10.75,
      "learning_rate": 1e-05,
      "loss": 1.5916,
      "step": 3950
    },
    {
      "epoch": 0.396,
      "grad_norm": 15.9375,
      "learning_rate": 1e-05,
      "loss": 1.8588,
      "step": 3960
    },
    {
      "epoch": 0.397,
      "grad_norm": 13.75,
      "learning_rate": 1e-05,
      "loss": 1.5461,
      "step": 3970
    },
    {
      "epoch": 0.398,
      "grad_norm": 9.875,
      "learning_rate": 1e-05,
      "loss": 1.4544,
      "step": 3980
    },
    {
      "epoch": 0.399,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 1.6998,
      "step": 3990
    },
    {
      "epoch": 0.4,
      "grad_norm": 34.5,
      "learning_rate": 1e-05,
      "loss": 1.7816,
      "step": 4000
    },
    {
      "epoch": 0.401,
      "grad_norm": 11.3125,
      "learning_rate": 1e-05,
      "loss": 1.7877,
      "step": 4010
    },
    {
      "epoch": 0.402,
      "grad_norm": 15.875,
      "learning_rate": 1e-05,
      "loss": 1.5258,
      "step": 4020
    },
    {
      "epoch": 0.403,
      "grad_norm": 13.0625,
      "learning_rate": 1e-05,
      "loss": 1.5588,
      "step": 4030
    },
    {
      "epoch": 0.404,
      "grad_norm": 16.875,
      "learning_rate": 1e-05,
      "loss": 1.6647,
      "step": 4040
    },
    {
      "epoch": 0.405,
      "grad_norm": 13.25,
      "learning_rate": 1e-05,
      "loss": 1.6484,
      "step": 4050
    },
    {
      "epoch": 0.406,
      "grad_norm": 14.625,
      "learning_rate": 1e-05,
      "loss": 1.7257,
      "step": 4060
    },
    {
      "epoch": 0.407,
      "grad_norm": 12.625,
      "learning_rate": 1e-05,
      "loss": 1.7299,
      "step": 4070
    },
    {
      "epoch": 0.408,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 1.731,
      "step": 4080
    },
    {
      "epoch": 0.409,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 2.0782,
      "step": 4090
    },
    {
      "epoch": 0.41,
      "grad_norm": 15.25,
      "learning_rate": 1e-05,
      "loss": 1.6928,
      "step": 4100
    },
    {
      "epoch": 0.411,
      "grad_norm": 15.5,
      "learning_rate": 1e-05,
      "loss": 1.7013,
      "step": 4110
    },
    {
      "epoch": 0.412,
      "grad_norm": 13.125,
      "learning_rate": 1e-05,
      "loss": 1.7821,
      "step": 4120
    },
    {
      "epoch": 0.413,
      "grad_norm": 12.1875,
      "learning_rate": 1e-05,
      "loss": 1.2811,
      "step": 4130
    },
    {
      "epoch": 0.414,
      "grad_norm": 16.375,
      "learning_rate": 1e-05,
      "loss": 1.534,
      "step": 4140
    },
    {
      "epoch": 0.415,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 3.9214,
      "step": 4150
    },
    {
      "epoch": 0.416,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 3.0651,
      "step": 4160
    },
    {
      "epoch": 0.417,
      "grad_norm": 9.3125,
      "learning_rate": 1e-05,
      "loss": 1.9012,
      "step": 4170
    },
    {
      "epoch": 0.418,
      "grad_norm": 8.375,
      "learning_rate": 1e-05,
      "loss": 1.3554,
      "step": 4180
    },
    {
      "epoch": 0.419,
      "grad_norm": 10.5,
      "learning_rate": 1e-05,
      "loss": 1.7087,
      "step": 4190
    },
    {
      "epoch": 0.42,
      "grad_norm": 12.9375,
      "learning_rate": 1e-05,
      "loss": 1.8635,
      "step": 4200
    },
    {
      "epoch": 0.421,
      "grad_norm": 28.5,
      "learning_rate": 1e-05,
      "loss": 2.1169,
      "step": 4210
    },
    {
      "epoch": 0.422,
      "grad_norm": 14.1875,
      "learning_rate": 1e-05,
      "loss": 1.6834,
      "step": 4220
    },
    {
      "epoch": 0.423,
      "grad_norm": 12.3125,
      "learning_rate": 1e-05,
      "loss": 1.7137,
      "step": 4230
    },
    {
      "epoch": 0.424,
      "grad_norm": 11.375,
      "learning_rate": 1e-05,
      "loss": 2.001,
      "step": 4240
    },
    {
      "epoch": 0.425,
      "grad_norm": 19.25,
      "learning_rate": 1e-05,
      "loss": 1.886,
      "step": 4250
    },
    {
      "epoch": 0.426,
      "grad_norm": 15.625,
      "learning_rate": 1e-05,
      "loss": 1.5167,
      "step": 4260
    },
    {
      "epoch": 0.427,
      "grad_norm": 14.375,
      "learning_rate": 1e-05,
      "loss": 1.8816,
      "step": 4270
    },
    {
      "epoch": 0.428,
      "grad_norm": 22.125,
      "learning_rate": 1e-05,
      "loss": 1.8211,
      "step": 4280
    },
    {
      "epoch": 0.429,
      "grad_norm": 17.75,
      "learning_rate": 1e-05,
      "loss": 1.8533,
      "step": 4290
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.8125,
      "learning_rate": 1e-05,
      "loss": 1.5333,
      "step": 4300
    },
    {
      "epoch": 0.431,
      "grad_norm": 9.75,
      "learning_rate": 1e-05,
      "loss": 1.8127,
      "step": 4310
    },
    {
      "epoch": 0.432,
      "grad_norm": 11.5625,
      "learning_rate": 1e-05,
      "loss": 1.6367,
      "step": 4320
    },
    {
      "epoch": 0.433,
      "grad_norm": 16.875,
      "learning_rate": 1e-05,
      "loss": 1.7815,
      "step": 4330
    },
    {
      "epoch": 0.434,
      "grad_norm": 11.8125,
      "learning_rate": 1e-05,
      "loss": 1.5546,
      "step": 4340
    },
    {
      "epoch": 0.435,
      "grad_norm": 11.3125,
      "learning_rate": 1e-05,
      "loss": 1.4757,
      "step": 4350
    },
    {
      "epoch": 0.436,
      "grad_norm": 14.0625,
      "learning_rate": 1e-05,
      "loss": 1.5582,
      "step": 4360
    },
    {
      "epoch": 0.437,
      "grad_norm": 11.6875,
      "learning_rate": 1e-05,
      "loss": 1.7033,
      "step": 4370
    },
    {
      "epoch": 0.438,
      "grad_norm": 13.0,
      "learning_rate": 1e-05,
      "loss": 1.6713,
      "step": 4380
    },
    {
      "epoch": 0.439,
      "grad_norm": 11.9375,
      "learning_rate": 1e-05,
      "loss": 1.7856,
      "step": 4390
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.4375,
      "learning_rate": 1e-05,
      "loss": 1.7393,
      "step": 4400
    },
    {
      "epoch": 0.441,
      "grad_norm": 14.25,
      "learning_rate": 1e-05,
      "loss": 1.6427,
      "step": 4410
    },
    {
      "epoch": 0.442,
      "grad_norm": 16.5,
      "learning_rate": 1e-05,
      "loss": 2.0203,
      "step": 4420
    },
    {
      "epoch": 0.443,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.8093,
      "step": 4430
    },
    {
      "epoch": 0.444,
      "grad_norm": 13.125,
      "learning_rate": 1e-05,
      "loss": 1.6276,
      "step": 4440
    },
    {
      "epoch": 0.445,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.7526,
      "step": 4450
    },
    {
      "epoch": 0.446,
      "grad_norm": 12.375,
      "learning_rate": 1e-05,
      "loss": 1.6753,
      "step": 4460
    },
    {
      "epoch": 0.447,
      "grad_norm": 25.75,
      "learning_rate": 1e-05,
      "loss": 1.7243,
      "step": 4470
    },
    {
      "epoch": 0.448,
      "grad_norm": 11.9375,
      "learning_rate": 1e-05,
      "loss": 1.6808,
      "step": 4480
    },
    {
      "epoch": 0.449,
      "grad_norm": 19.375,
      "learning_rate": 1e-05,
      "loss": 1.6972,
      "step": 4490
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.625,
      "learning_rate": 1e-05,
      "loss": 1.7271,
      "step": 4500
    },
    {
      "epoch": 0.451,
      "grad_norm": 13.0,
      "learning_rate": 1e-05,
      "loss": 1.8317,
      "step": 4510
    },
    {
      "epoch": 0.452,
      "grad_norm": 11.25,
      "learning_rate": 1e-05,
      "loss": 1.689,
      "step": 4520
    },
    {
      "epoch": 0.453,
      "grad_norm": 13.1875,
      "learning_rate": 1e-05,
      "loss": 1.7482,
      "step": 4530
    },
    {
      "epoch": 0.454,
      "grad_norm": 7.3125,
      "learning_rate": 1e-05,
      "loss": 1.8387,
      "step": 4540
    },
    {
      "epoch": 0.455,
      "grad_norm": 13.625,
      "learning_rate": 1e-05,
      "loss": 1.7703,
      "step": 4550
    },
    {
      "epoch": 0.456,
      "grad_norm": 14.4375,
      "learning_rate": 1e-05,
      "loss": 1.9912,
      "step": 4560
    },
    {
      "epoch": 0.457,
      "grad_norm": 16.125,
      "learning_rate": 1e-05,
      "loss": 1.7298,
      "step": 4570
    },
    {
      "epoch": 0.458,
      "grad_norm": 15.5,
      "learning_rate": 1e-05,
      "loss": 1.6696,
      "step": 4580
    },
    {
      "epoch": 0.459,
      "grad_norm": 10.75,
      "learning_rate": 1e-05,
      "loss": 1.5654,
      "step": 4590
    },
    {
      "epoch": 0.46,
      "grad_norm": 19.375,
      "learning_rate": 1e-05,
      "loss": 1.8688,
      "step": 4600
    },
    {
      "epoch": 0.461,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 1.8404,
      "step": 4610
    },
    {
      "epoch": 0.462,
      "grad_norm": 13.5,
      "learning_rate": 1e-05,
      "loss": 1.8966,
      "step": 4620
    },
    {
      "epoch": 0.463,
      "grad_norm": 9.4375,
      "learning_rate": 1e-05,
      "loss": 1.3358,
      "step": 4630
    },
    {
      "epoch": 0.464,
      "grad_norm": 11.125,
      "learning_rate": 1e-05,
      "loss": 1.5774,
      "step": 4640
    },
    {
      "epoch": 0.465,
      "grad_norm": 20.0,
      "learning_rate": 1e-05,
      "loss": 1.8084,
      "step": 4650
    },
    {
      "epoch": 0.466,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.5481,
      "step": 4660
    },
    {
      "epoch": 0.467,
      "grad_norm": 13.75,
      "learning_rate": 1e-05,
      "loss": 1.6949,
      "step": 4670
    },
    {
      "epoch": 0.468,
      "grad_norm": 15.125,
      "learning_rate": 1e-05,
      "loss": 1.5561,
      "step": 4680
    },
    {
      "epoch": 0.469,
      "grad_norm": 14.5625,
      "learning_rate": 1e-05,
      "loss": 2.0335,
      "step": 4690
    },
    {
      "epoch": 0.47,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 1.6344,
      "step": 4700
    },
    {
      "epoch": 0.471,
      "grad_norm": 18.625,
      "learning_rate": 1e-05,
      "loss": 1.8511,
      "step": 4710
    },
    {
      "epoch": 0.472,
      "grad_norm": 24.875,
      "learning_rate": 1e-05,
      "loss": 1.915,
      "step": 4720
    },
    {
      "epoch": 0.473,
      "grad_norm": 15.4375,
      "learning_rate": 1e-05,
      "loss": 1.8416,
      "step": 4730
    },
    {
      "epoch": 0.474,
      "grad_norm": 16.0,
      "learning_rate": 1e-05,
      "loss": 1.5345,
      "step": 4740
    },
    {
      "epoch": 0.475,
      "grad_norm": 10.875,
      "learning_rate": 1e-05,
      "loss": 1.502,
      "step": 4750
    },
    {
      "epoch": 0.476,
      "grad_norm": 9.6875,
      "learning_rate": 1e-05,
      "loss": 1.3791,
      "step": 4760
    },
    {
      "epoch": 0.477,
      "grad_norm": 10.375,
      "learning_rate": 1e-05,
      "loss": 1.9847,
      "step": 4770
    },
    {
      "epoch": 0.478,
      "grad_norm": 15.25,
      "learning_rate": 1e-05,
      "loss": 1.7426,
      "step": 4780
    },
    {
      "epoch": 0.479,
      "grad_norm": 12.0,
      "learning_rate": 1e-05,
      "loss": 1.4959,
      "step": 4790
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.375,
      "learning_rate": 1e-05,
      "loss": 1.5553,
      "step": 4800
    },
    {
      "epoch": 0.481,
      "grad_norm": 13.25,
      "learning_rate": 1e-05,
      "loss": 1.7647,
      "step": 4810
    },
    {
      "epoch": 0.482,
      "grad_norm": 12.6875,
      "learning_rate": 1e-05,
      "loss": 1.5482,
      "step": 4820
    },
    {
      "epoch": 0.483,
      "grad_norm": 19.375,
      "learning_rate": 1e-05,
      "loss": 1.6021,
      "step": 4830
    },
    {
      "epoch": 0.484,
      "grad_norm": 17.75,
      "learning_rate": 1e-05,
      "loss": 1.8294,
      "step": 4840
    },
    {
      "epoch": 0.485,
      "grad_norm": 11.5,
      "learning_rate": 1e-05,
      "loss": 1.9732,
      "step": 4850
    },
    {
      "epoch": 0.486,
      "grad_norm": 11.375,
      "learning_rate": 1e-05,
      "loss": 1.7718,
      "step": 4860
    },
    {
      "epoch": 0.487,
      "grad_norm": 11.875,
      "learning_rate": 1e-05,
      "loss": 1.7849,
      "step": 4870
    },
    {
      "epoch": 0.488,
      "grad_norm": 18.625,
      "learning_rate": 1e-05,
      "loss": 1.8882,
      "step": 4880
    },
    {
      "epoch": 0.489,
      "grad_norm": 15.625,
      "learning_rate": 1e-05,
      "loss": 2.1638,
      "step": 4890
    },
    {
      "epoch": 0.49,
      "grad_norm": 7.6875,
      "learning_rate": 1e-05,
      "loss": 1.9241,
      "step": 4900
    },
    {
      "epoch": 0.491,
      "grad_norm": 17.25,
      "learning_rate": 1e-05,
      "loss": 1.529,
      "step": 4910
    },
    {
      "epoch": 0.492,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 2.0382,
      "step": 4920
    },
    {
      "epoch": 0.493,
      "grad_norm": 14.875,
      "learning_rate": 1e-05,
      "loss": 1.7519,
      "step": 4930
    },
    {
      "epoch": 0.494,
      "grad_norm": 15.1875,
      "learning_rate": 1e-05,
      "loss": 1.6614,
      "step": 4940
    },
    {
      "epoch": 0.495,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.6375,
      "step": 4950
    },
    {
      "epoch": 0.496,
      "grad_norm": 9.6875,
      "learning_rate": 1e-05,
      "loss": 1.713,
      "step": 4960
    },
    {
      "epoch": 0.497,
      "grad_norm": 14.0,
      "learning_rate": 1e-05,
      "loss": 1.698,
      "step": 4970
    },
    {
      "epoch": 0.498,
      "grad_norm": 12.8125,
      "learning_rate": 1e-05,
      "loss": 1.6976,
      "step": 4980
    },
    {
      "epoch": 0.499,
      "grad_norm": 10.4375,
      "learning_rate": 1e-05,
      "loss": 1.7785,
      "step": 4990
    },
    {
      "epoch": 0.5,
      "grad_norm": 15.0625,
      "learning_rate": 1e-05,
      "loss": 1.5788,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}