{
  "best_metric": 0.018313532695174217,
  "best_model_checkpoint": "runs/deepseek_lora_20240423-133229/checkpoint-2000",
  "epoch": 1.5665387326701654,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 1.5461480617523193,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 1.5239,
      "step": 10
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.797250747680664,
      "learning_rate": 8.000000000000001e-07,
      "loss": 1.4966,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.274460315704346,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.4788,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9981650114059448,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.4001,
      "step": 40
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2322547435760498,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.4683,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.250878930091858,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.1961,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.917982578277588,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 1.4261,
      "step": 70
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.885340929031372,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.2742,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.170296311378479,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 1.253,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.992724895477295,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.2521,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9540568590164185,
      "learning_rate": 4.4e-06,
      "loss": 1.0828,
      "step": 110
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9779148697853088,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.1255,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2271130084991455,
      "learning_rate": 5.2e-06,
      "loss": 0.9496,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.1939122676849365,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.953,
      "step": 140
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.5918376445770264,
      "learning_rate": 6e-06,
      "loss": 0.9348,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.772602379322052,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.7762,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8226258158683777,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.6798,
      "step": 170
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0945135354995728,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.5936,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6972842812538147,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.4161,
      "step": 190
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.9560859203338623,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.3959,
      "step": 200
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4297197759151459,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.3166,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.4494841992855072,
      "learning_rate": 8.8e-06,
      "loss": 0.2056,
      "step": 220
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6981032490730286,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.3024,
      "step": 230
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5783653259277344,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.4392,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.33407163619995117,
      "learning_rate": 1e-05,
      "loss": 0.3708,
      "step": 250
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.3096323311328888,
      "learning_rate": 1.04e-05,
      "loss": 0.1637,
      "step": 260
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.6187375783920288,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.1078,
      "step": 270
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.27954748272895813,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.2668,
      "step": 280
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4082590937614441,
      "learning_rate": 1.16e-05,
      "loss": 0.1843,
      "step": 290
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.2993428707122803,
      "learning_rate": 1.2e-05,
      "loss": 0.1969,
      "step": 300
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.0335004329681396,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.122,
      "step": 310
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.0972365066409111,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.1354,
      "step": 320
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.26203760504722595,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.17,
      "step": 330
    },
    {
      "epoch": 0.11,
      "grad_norm": 5.179258346557617,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.1409,
      "step": 340
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.032190825790166855,
      "learning_rate": 1.4e-05,
      "loss": 0.076,
      "step": 350
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.02556140348315239,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.0468,
      "step": 360
    },
    {
      "epoch": 0.12,
      "grad_norm": 5.667189121246338,
      "learning_rate": 1.48e-05,
      "loss": 0.1345,
      "step": 370
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.07244217395782471,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0411,
      "step": 380
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.11255475878715515,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.1907,
      "step": 390
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.9602845907211304,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0717,
      "step": 400
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.013551376760005951,
      "learning_rate": 1.64e-05,
      "loss": 0.0808,
      "step": 410
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.14421305060386658,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0604,
      "step": 420
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.01705547235906124,
      "learning_rate": 1.72e-05,
      "loss": 0.1742,
      "step": 430
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.021360402926802635,
      "learning_rate": 1.76e-05,
      "loss": 0.0886,
      "step": 440
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.04369840770959854,
      "learning_rate": 1.8e-05,
      "loss": 0.1063,
      "step": 450
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.2816083431243896,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.1061,
      "step": 460
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.00989863183349371,
      "learning_rate": 1.88e-05,
      "loss": 0.2317,
      "step": 470
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.014644264243543148,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.1054,
      "step": 480
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.13385985791683197,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.1196,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.099637985229492,
      "learning_rate": 2e-05,
      "loss": 0.1584,
      "step": 500
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.05546905845403671,
      "eval_runtime": 62.1757,
      "eval_samples_per_second": 16.083,
      "eval_steps_per_second": 16.083,
      "step": 500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.013259253464639187,
      "learning_rate": 1.9955555555555557e-05,
      "loss": 0.0373,
      "step": 510
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.1753215193748474,
      "learning_rate": 1.9911111111111112e-05,
      "loss": 0.0986,
      "step": 520
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.006998735014349222,
      "learning_rate": 1.9866666666666667e-05,
      "loss": 0.1372,
      "step": 530
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3990759253501892,
      "learning_rate": 1.9822222222222226e-05,
      "loss": 0.127,
      "step": 540
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8248735666275024,
      "learning_rate": 1.977777777777778e-05,
      "loss": 0.1239,
      "step": 550
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.005832708440721035,
      "learning_rate": 1.9733333333333336e-05,
      "loss": 0.1181,
      "step": 560
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.10705375671386719,
      "learning_rate": 1.968888888888889e-05,
      "loss": 0.0972,
      "step": 570
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.7429046630859375,
      "learning_rate": 1.9644444444444447e-05,
      "loss": 0.1543,
      "step": 580
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.010058793239295483,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.1146,
      "step": 590
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.012124909088015556,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 0.1582,
      "step": 600
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.012700803577899933,
      "learning_rate": 1.9511111111111113e-05,
      "loss": 0.0711,
      "step": 610
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2153177261352539,
      "learning_rate": 1.9466666666666668e-05,
      "loss": 0.0452,
      "step": 620
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.01289213914424181,
      "learning_rate": 1.9422222222222223e-05,
      "loss": 0.067,
      "step": 630
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.007021366152912378,
      "learning_rate": 1.9377777777777778e-05,
      "loss": 0.0776,
      "step": 640
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.005274974275380373,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 0.2159,
      "step": 650
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6235843896865845,
      "learning_rate": 1.928888888888889e-05,
      "loss": 0.0664,
      "step": 660
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.009710369631648064,
      "learning_rate": 1.9244444444444444e-05,
      "loss": 0.1125,
      "step": 670
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.0918051227927208,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0755,
      "step": 680
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.012988925911486149,
      "learning_rate": 1.9155555555555558e-05,
      "loss": 0.095,
      "step": 690
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.09063850343227386,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 0.0419,
      "step": 700
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.1540229320526123,
      "learning_rate": 1.9066666666666668e-05,
      "loss": 0.0338,
      "step": 710
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.009780442342162132,
      "learning_rate": 1.9022222222222223e-05,
      "loss": 0.0935,
      "step": 720
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.011248274706304073,
      "learning_rate": 1.897777777777778e-05,
      "loss": 0.017,
      "step": 730
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.03275947645306587,
      "learning_rate": 1.8933333333333334e-05,
      "loss": 0.1128,
      "step": 740
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.224815607070923,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.1213,
      "step": 750
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.9777376651763916,
      "learning_rate": 1.8844444444444444e-05,
      "loss": 0.1128,
      "step": 760
    },
    {
      "epoch": 0.24,
      "grad_norm": 7.6119561195373535,
      "learning_rate": 1.88e-05,
      "loss": 0.0557,
      "step": 770
    },
    {
      "epoch": 0.24,
      "grad_norm": 3.7092978954315186,
      "learning_rate": 1.8755555555555558e-05,
      "loss": 0.1196,
      "step": 780
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.08959495276212692,
      "learning_rate": 1.8711111111111113e-05,
      "loss": 0.0231,
      "step": 790
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.716322183609009,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.1253,
      "step": 800
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.006216592621058226,
      "learning_rate": 1.8622222222222224e-05,
      "loss": 0.0377,
      "step": 810
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.005503402091562748,
      "learning_rate": 1.857777777777778e-05,
      "loss": 0.1072,
      "step": 820
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4769572913646698,
      "learning_rate": 1.8533333333333334e-05,
      "loss": 0.0888,
      "step": 830
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.034854818135499954,
      "learning_rate": 1.848888888888889e-05,
      "loss": 0.0424,
      "step": 840
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.00536635797470808,
      "learning_rate": 1.8444444444444448e-05,
      "loss": 0.0161,
      "step": 850
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.002790963975712657,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.025,
      "step": 860
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.003568426240235567,
      "learning_rate": 1.835555555555556e-05,
      "loss": 0.0597,
      "step": 870
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.005008562933653593,
      "learning_rate": 1.8311111111111114e-05,
      "loss": 0.1817,
      "step": 880
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.006756258197128773,
      "learning_rate": 1.826666666666667e-05,
      "loss": 0.0084,
      "step": 890
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.3732385635375977,
      "learning_rate": 1.8222222222222224e-05,
      "loss": 0.072,
      "step": 900
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.003554440103471279,
      "learning_rate": 1.817777777777778e-05,
      "loss": 0.0507,
      "step": 910
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.06785377860069275,
      "learning_rate": 1.8133333333333335e-05,
      "loss": 0.0428,
      "step": 920
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.009546133689582348,
      "learning_rate": 1.808888888888889e-05,
      "loss": 0.0179,
      "step": 930
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.595461368560791,
      "learning_rate": 1.8044444444444445e-05,
      "loss": 0.0506,
      "step": 940
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.310696601867676,
      "learning_rate": 1.8e-05,
      "loss": 0.0719,
      "step": 950
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.8953649997711182,
      "learning_rate": 1.7955555555555556e-05,
      "loss": 0.1718,
      "step": 960
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.706213116645813,
      "learning_rate": 1.791111111111111e-05,
      "loss": 0.0807,
      "step": 970
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.06022366136312485,
      "learning_rate": 1.7866666666666666e-05,
      "loss": 0.0792,
      "step": 980
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.033781811594963074,
      "learning_rate": 1.782222222222222e-05,
      "loss": 0.0864,
      "step": 990
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.014879009686410427,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 0.0527,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.03033074364066124,
      "eval_runtime": 62.1716,
      "eval_samples_per_second": 16.085,
      "eval_steps_per_second": 16.085,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.003431650809943676,
      "learning_rate": 1.7733333333333335e-05,
      "loss": 0.0425,
      "step": 1010
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.005604119040071964,
      "learning_rate": 1.768888888888889e-05,
      "loss": 0.0727,
      "step": 1020
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.058546032756567,
      "learning_rate": 1.7644444444444446e-05,
      "loss": 0.1304,
      "step": 1030
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0267212390899658,
      "learning_rate": 1.76e-05,
      "loss": 0.0496,
      "step": 1040
    },
    {
      "epoch": 0.33,
      "grad_norm": 3.5381667613983154,
      "learning_rate": 1.7555555555555556e-05,
      "loss": 0.0935,
      "step": 1050
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.006851164624094963,
      "learning_rate": 1.751111111111111e-05,
      "loss": 0.0344,
      "step": 1060
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.005180824548006058,
      "learning_rate": 1.7466666666666667e-05,
      "loss": 0.0195,
      "step": 1070
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.586403489112854,
      "learning_rate": 1.7422222222222222e-05,
      "loss": 0.0868,
      "step": 1080
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.14490236341953278,
      "learning_rate": 1.737777777777778e-05,
      "loss": 0.0489,
      "step": 1090
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.005752285942435265,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.0625,
      "step": 1100
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.8000261187553406,
      "learning_rate": 1.728888888888889e-05,
      "loss": 0.0923,
      "step": 1110
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.024827931076288223,
      "learning_rate": 1.7244444444444446e-05,
      "loss": 0.0355,
      "step": 1120
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.424243927001953,
      "learning_rate": 1.72e-05,
      "loss": 0.0659,
      "step": 1130
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.4055392742156982,
      "learning_rate": 1.7155555555555557e-05,
      "loss": 0.0681,
      "step": 1140
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.0077874124981462955,
      "learning_rate": 1.7111111111111112e-05,
      "loss": 0.106,
      "step": 1150
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.006299824919551611,
      "learning_rate": 1.706666666666667e-05,
      "loss": 0.0438,
      "step": 1160
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.004886701237410307,
      "learning_rate": 1.7022222222222226e-05,
      "loss": 0.0403,
      "step": 1170
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.0028305843006819487,
      "learning_rate": 1.697777777777778e-05,
      "loss": 0.0299,
      "step": 1180
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.36411717534065247,
      "learning_rate": 1.6933333333333336e-05,
      "loss": 0.0639,
      "step": 1190
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.004984024912118912,
      "learning_rate": 1.688888888888889e-05,
      "loss": 0.1363,
      "step": 1200
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.4117214679718018,
      "learning_rate": 1.6844444444444447e-05,
      "loss": 0.018,
      "step": 1210
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2183905839920044,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0062,
      "step": 1220
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.05338621512055397,
      "learning_rate": 1.6755555555555557e-05,
      "loss": 0.0343,
      "step": 1230
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.009272832423448563,
      "learning_rate": 1.6711111111111112e-05,
      "loss": 0.073,
      "step": 1240
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.0033628540113568306,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0272,
      "step": 1250
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.055562809109687805,
      "learning_rate": 1.6622222222222223e-05,
      "loss": 0.0821,
      "step": 1260
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6196753978729248,
      "learning_rate": 1.6577777777777778e-05,
      "loss": 0.0153,
      "step": 1270
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.004278264008462429,
      "learning_rate": 1.6533333333333333e-05,
      "loss": 0.022,
      "step": 1280
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.044023897498846054,
      "learning_rate": 1.648888888888889e-05,
      "loss": 0.0652,
      "step": 1290
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.08157920837402344,
      "learning_rate": 1.6444444444444444e-05,
      "loss": 0.0745,
      "step": 1300
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.1145012378692627,
      "learning_rate": 1.64e-05,
      "loss": 0.0497,
      "step": 1310
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.00495466822758317,
      "learning_rate": 1.6355555555555557e-05,
      "loss": 0.0688,
      "step": 1320
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.055875778198242,
      "learning_rate": 1.6311111111111113e-05,
      "loss": 0.0456,
      "step": 1330
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.260204315185547,
      "learning_rate": 1.6266666666666668e-05,
      "loss": 0.0968,
      "step": 1340
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.00300275138579309,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 0.0031,
      "step": 1350
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.6148861050605774,
      "learning_rate": 1.617777777777778e-05,
      "loss": 0.1008,
      "step": 1360
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.005163553170859814,
      "learning_rate": 1.6133333333333334e-05,
      "loss": 0.0454,
      "step": 1370
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.313544988632202,
      "learning_rate": 1.608888888888889e-05,
      "loss": 0.0843,
      "step": 1380
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.061028964817523956,
      "learning_rate": 1.6044444444444444e-05,
      "loss": 0.0467,
      "step": 1390
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.011183706112205982,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0539,
      "step": 1400
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.00391357159242034,
      "learning_rate": 1.5955555555555558e-05,
      "loss": 0.007,
      "step": 1410
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.21193568408489227,
      "learning_rate": 1.5911111111111113e-05,
      "loss": 0.0723,
      "step": 1420
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.344232201576233,
      "learning_rate": 1.586666666666667e-05,
      "loss": 0.0093,
      "step": 1430
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.125556707382202,
      "learning_rate": 1.5822222222222224e-05,
      "loss": 0.1089,
      "step": 1440
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.9358251094818115,
      "learning_rate": 1.577777777777778e-05,
      "loss": 0.129,
      "step": 1450
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.25835612416267395,
      "learning_rate": 1.5733333333333334e-05,
      "loss": 0.0225,
      "step": 1460
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.004413430579006672,
      "learning_rate": 1.5688888888888893e-05,
      "loss": 0.0184,
      "step": 1470
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.6624996662139893,
      "learning_rate": 1.5644444444444448e-05,
      "loss": 0.0924,
      "step": 1480
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.023669052869081497,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.0468,
      "step": 1490
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9958231449127197,
      "learning_rate": 1.555555555555556e-05,
      "loss": 0.0602,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.023086324334144592,
      "eval_runtime": 62.044,
      "eval_samples_per_second": 16.118,
      "eval_steps_per_second": 16.118,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.0457289032638073,
      "learning_rate": 1.5511111111111114e-05,
      "loss": 0.0725,
      "step": 1510
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.0417847633361816,
      "learning_rate": 1.546666666666667e-05,
      "loss": 0.0446,
      "step": 1520
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.007491718512028456,
      "learning_rate": 1.5422222222222224e-05,
      "loss": 0.0693,
      "step": 1530
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.0023406276013702154,
      "learning_rate": 1.537777777777778e-05,
      "loss": 0.0398,
      "step": 1540
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.03808426856994629,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 0.0827,
      "step": 1550
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.04878142476081848,
      "learning_rate": 1.528888888888889e-05,
      "loss": 0.0094,
      "step": 1560
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.002342456951737404,
      "learning_rate": 1.5244444444444447e-05,
      "loss": 0.0308,
      "step": 1570
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.002332750940695405,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0703,
      "step": 1580
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.027806026861071587,
      "learning_rate": 1.5155555555555557e-05,
      "loss": 0.0577,
      "step": 1590
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.853770136833191,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 0.0872,
      "step": 1600
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.02809343859553337,
      "learning_rate": 1.5066666666666668e-05,
      "loss": 0.0651,
      "step": 1610
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.29635584354400635,
      "learning_rate": 1.5022222222222223e-05,
      "loss": 0.0784,
      "step": 1620
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.006036743987351656,
      "learning_rate": 1.497777777777778e-05,
      "loss": 0.0702,
      "step": 1630
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.0020562768913805485,
      "learning_rate": 1.4933333333333335e-05,
      "loss": 0.1133,
      "step": 1640
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.0853707790374756,
      "learning_rate": 1.488888888888889e-05,
      "loss": 0.0717,
      "step": 1650
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.024653434753418,
      "learning_rate": 1.4844444444444445e-05,
      "loss": 0.0978,
      "step": 1660
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.0029632914811372757,
      "learning_rate": 1.48e-05,
      "loss": 0.0269,
      "step": 1670
    },
    {
      "epoch": 0.53,
      "grad_norm": 3.2619192600250244,
      "learning_rate": 1.4755555555555556e-05,
      "loss": 0.088,
      "step": 1680
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9590596556663513,
      "learning_rate": 1.4711111111111111e-05,
      "loss": 0.0348,
      "step": 1690
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.01058279536664486,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 0.0473,
      "step": 1700
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.378236293792725,
      "learning_rate": 1.4622222222222225e-05,
      "loss": 0.0839,
      "step": 1710
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.04728490114212036,
      "learning_rate": 1.457777777777778e-05,
      "loss": 0.1131,
      "step": 1720
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.03082725591957569,
      "learning_rate": 1.4533333333333335e-05,
      "loss": 0.058,
      "step": 1730
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4159008264541626,
      "learning_rate": 1.448888888888889e-05,
      "loss": 0.027,
      "step": 1740
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.09299231320619583,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 0.0152,
      "step": 1750
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.0018398365937173367,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 0.0005,
      "step": 1760
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.0032970942556858063,
      "learning_rate": 1.4355555555555556e-05,
      "loss": 0.068,
      "step": 1770
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.9869989156723022,
      "learning_rate": 1.4311111111111111e-05,
      "loss": 0.0802,
      "step": 1780
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.4349244236946106,
      "learning_rate": 1.4266666666666668e-05,
      "loss": 0.0022,
      "step": 1790
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.195014476776123,
      "learning_rate": 1.4222222222222224e-05,
      "loss": 0.075,
      "step": 1800
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.658276557922363,
      "learning_rate": 1.4177777777777779e-05,
      "loss": 0.0341,
      "step": 1810
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.3556618690490723,
      "learning_rate": 1.4133333333333334e-05,
      "loss": 0.0735,
      "step": 1820
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.00235918490216136,
      "learning_rate": 1.408888888888889e-05,
      "loss": 0.0133,
      "step": 1830
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.0019710322376340628,
      "learning_rate": 1.4044444444444445e-05,
      "loss": 0.0714,
      "step": 1840
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.0038756057620048523,
      "learning_rate": 1.4e-05,
      "loss": 0.058,
      "step": 1850
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.13870251178741455,
      "learning_rate": 1.3955555555555558e-05,
      "loss": 0.0598,
      "step": 1860
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.0035449981223791838,
      "learning_rate": 1.3911111111111114e-05,
      "loss": 0.0825,
      "step": 1870
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.005458540748804808,
      "learning_rate": 1.3866666666666669e-05,
      "loss": 0.0485,
      "step": 1880
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.0037788543850183487,
      "learning_rate": 1.3822222222222224e-05,
      "loss": 0.0279,
      "step": 1890
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.4127650260925293,
      "learning_rate": 1.377777777777778e-05,
      "loss": 0.0414,
      "step": 1900
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.9315361976623535,
      "learning_rate": 1.3733333333333335e-05,
      "loss": 0.0541,
      "step": 1910
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.001827953034080565,
      "learning_rate": 1.368888888888889e-05,
      "loss": 0.0607,
      "step": 1920
    },
    {
      "epoch": 0.6,
      "grad_norm": 3.0561037063598633,
      "learning_rate": 1.3644444444444445e-05,
      "loss": 0.0489,
      "step": 1930
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7366918325424194,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.0107,
      "step": 1940
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.0025562909431755543,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 0.0305,
      "step": 1950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.006456757429987192,
      "learning_rate": 1.3511111111111112e-05,
      "loss": 0.0741,
      "step": 1960
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.003276415169239044,
      "learning_rate": 1.3466666666666668e-05,
      "loss": 0.0276,
      "step": 1970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.0036139176227152348,
      "learning_rate": 1.3422222222222223e-05,
      "loss": 0.0886,
      "step": 1980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.006649693939834833,
      "learning_rate": 1.3377777777777778e-05,
      "loss": 0.0179,
      "step": 1990
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.002991742454469204,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0372,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.018313532695174217,
      "eval_runtime": 62.0835,
      "eval_samples_per_second": 16.107,
      "eval_steps_per_second": 16.107,
      "step": 2000
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.0025979657657444477,
      "learning_rate": 1.3288888888888889e-05,
      "loss": 0.0485,
      "step": 2010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.00765944505110383,
      "learning_rate": 1.3244444444444447e-05,
      "loss": 0.0292,
      "step": 2020
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9524846076965332,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.0545,
      "step": 2030
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.0015762520488351583,
      "learning_rate": 1.3155555555555558e-05,
      "loss": 0.0003,
      "step": 2040
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.12439166754484177,
      "learning_rate": 1.3111111111111113e-05,
      "loss": 0.0346,
      "step": 2050
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.002841875422745943,
      "learning_rate": 1.3066666666666668e-05,
      "loss": 0.0785,
      "step": 2060
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.3388820886611938,
      "learning_rate": 1.3022222222222223e-05,
      "loss": 0.0748,
      "step": 2070
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.003183074528351426,
      "learning_rate": 1.2977777777777779e-05,
      "loss": 0.0653,
      "step": 2080
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0637716054916382,
      "learning_rate": 1.2933333333333334e-05,
      "loss": 0.0522,
      "step": 2090
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.1531606912612915,
      "learning_rate": 1.288888888888889e-05,
      "loss": 0.0968,
      "step": 2100
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.0031077053863555193,
      "learning_rate": 1.2844444444444446e-05,
      "loss": 0.0419,
      "step": 2110
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.011700381524860859,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0525,
      "step": 2120
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.0015547301154583693,
      "learning_rate": 1.2755555555555556e-05,
      "loss": 0.0473,
      "step": 2130
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.0020142821595072746,
      "learning_rate": 1.2711111111111112e-05,
      "loss": 0.044,
      "step": 2140
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.005582904908806086,
      "learning_rate": 1.2666666666666667e-05,
      "loss": 0.0789,
      "step": 2150
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0064437431283295155,
      "learning_rate": 1.2622222222222222e-05,
      "loss": 0.099,
      "step": 2160
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.0012363777495920658,
      "learning_rate": 1.257777777777778e-05,
      "loss": 0.0411,
      "step": 2170
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14603936672210693,
      "learning_rate": 1.2533333333333336e-05,
      "loss": 0.006,
      "step": 2180
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.001298425835557282,
      "learning_rate": 1.2488888888888891e-05,
      "loss": 0.1263,
      "step": 2190
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.001547910855151713,
      "learning_rate": 1.2444444444444446e-05,
      "loss": 0.0342,
      "step": 2200
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.0013805734924972057,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.0824,
      "step": 2210
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.123624563217163,
      "learning_rate": 1.2355555555555557e-05,
      "loss": 0.0202,
      "step": 2220
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.0014894693158566952,
      "learning_rate": 1.2311111111111112e-05,
      "loss": 0.0313,
      "step": 2230
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.194322109222412,
      "learning_rate": 1.2266666666666667e-05,
      "loss": 0.123,
      "step": 2240
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.4329869747161865,
      "learning_rate": 1.2222222222222224e-05,
      "loss": 0.0619,
      "step": 2250
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.0048148720525205135,
      "learning_rate": 1.217777777777778e-05,
      "loss": 0.0105,
      "step": 2260
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8393607139587402,
      "learning_rate": 1.2133333333333335e-05,
      "loss": 0.0763,
      "step": 2270
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.0013370034284889698,
      "learning_rate": 1.208888888888889e-05,
      "loss": 0.0263,
      "step": 2280
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0011019165394827724,
      "learning_rate": 1.2044444444444445e-05,
      "loss": 0.0451,
      "step": 2290
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0017341958591714501,
      "learning_rate": 1.2e-05,
      "loss": 0.0612,
      "step": 2300
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.0012070784578099847,
      "learning_rate": 1.1955555555555556e-05,
      "loss": 0.0383,
      "step": 2310
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.004143261816352606,
      "learning_rate": 1.191111111111111e-05,
      "loss": 0.0523,
      "step": 2320
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.0022921899799257517,
      "learning_rate": 1.186666666666667e-05,
      "loss": 0.054,
      "step": 2330
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.052677869796753,
      "learning_rate": 1.1822222222222225e-05,
      "loss": 0.0329,
      "step": 2340
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5296663641929626,
      "learning_rate": 1.177777777777778e-05,
      "loss": 0.075,
      "step": 2350
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.001526731881313026,
      "learning_rate": 1.1733333333333335e-05,
      "loss": 0.0386,
      "step": 2360
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.293755531311035,
      "learning_rate": 1.168888888888889e-05,
      "loss": 0.0576,
      "step": 2370
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.002120732795447111,
      "learning_rate": 1.1644444444444446e-05,
      "loss": 0.0398,
      "step": 2380
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.001393395708873868,
      "learning_rate": 1.16e-05,
      "loss": 0.0335,
      "step": 2390
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.001357159111648798,
      "learning_rate": 1.1555555555555556e-05,
      "loss": 0.0186,
      "step": 2400
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.845559597015381,
      "learning_rate": 1.1511111111111113e-05,
      "loss": 0.0231,
      "step": 2410
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.001933318912051618,
      "learning_rate": 1.1466666666666668e-05,
      "loss": 0.0572,
      "step": 2420
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.08605944365262985,
      "learning_rate": 1.1422222222222223e-05,
      "loss": 0.0361,
      "step": 2430
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.6697850227355957,
      "learning_rate": 1.1377777777777779e-05,
      "loss": 0.1098,
      "step": 2440
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6600360870361328,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 0.0599,
      "step": 2450
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.8962271213531494,
      "learning_rate": 1.1288888888888889e-05,
      "loss": 0.0333,
      "step": 2460
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3499470055103302,
      "learning_rate": 1.1244444444444444e-05,
      "loss": 0.0276,
      "step": 2470
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0011357500916346908,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0662,
      "step": 2480
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9571266174316406,
      "learning_rate": 1.1155555555555556e-05,
      "loss": 0.0403,
      "step": 2490
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0018001727294176817,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.0549,
      "step": 2500
    },
    {
      "epoch": 0.78,
      "eval_loss": 0.019961679354310036,
      "eval_runtime": 62.0505,
      "eval_samples_per_second": 16.116,
      "eval_steps_per_second": 16.116,
      "step": 2500
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.002932826289907098,
      "learning_rate": 1.1066666666666669e-05,
      "loss": 0.0785,
      "step": 2510
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.0028327910695225,
      "learning_rate": 1.1022222222222224e-05,
      "loss": 0.0283,
      "step": 2520
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.9044318199157715,
      "learning_rate": 1.0977777777777779e-05,
      "loss": 0.1637,
      "step": 2530
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0019232028862461448,
      "learning_rate": 1.0933333333333334e-05,
      "loss": 0.004,
      "step": 2540
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.002454034984111786,
      "learning_rate": 1.088888888888889e-05,
      "loss": 0.0091,
      "step": 2550
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.0015948887448757887,
      "learning_rate": 1.0844444444444446e-05,
      "loss": 0.0276,
      "step": 2560
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.5870649814605713,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.0262,
      "step": 2570
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.2778539657592773,
      "learning_rate": 1.0755555555555557e-05,
      "loss": 0.0424,
      "step": 2580
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.6451878547668457,
      "learning_rate": 1.0711111111111112e-05,
      "loss": 0.0679,
      "step": 2590
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.001484246808104217,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.0373,
      "step": 2600
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.001308325445279479,
      "learning_rate": 1.0622222222222223e-05,
      "loss": 0.0191,
      "step": 2610
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.0027939951978623867,
      "learning_rate": 1.0577777777777778e-05,
      "loss": 0.0459,
      "step": 2620
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.009125540032982826,
      "learning_rate": 1.0533333333333333e-05,
      "loss": 0.0704,
      "step": 2630
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.0012461950536817312,
      "learning_rate": 1.048888888888889e-05,
      "loss": 0.0463,
      "step": 2640
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.906261920928955,
      "learning_rate": 1.0444444444444445e-05,
      "loss": 0.0923,
      "step": 2650
    },
    {
      "epoch": 0.83,
      "grad_norm": 3.149310827255249,
      "learning_rate": 1.04e-05,
      "loss": 0.0558,
      "step": 2660
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.07791195064783096,
      "learning_rate": 1.0355555555555557e-05,
      "loss": 0.0148,
      "step": 2670
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.0020532067865133286,
      "learning_rate": 1.0311111111111113e-05,
      "loss": 0.062,
      "step": 2680
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.0010498109040781856,
      "learning_rate": 1.0266666666666668e-05,
      "loss": 0.0577,
      "step": 2690
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.1568715125322342,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 0.0025,
      "step": 2700
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.4488503932952881,
      "learning_rate": 1.0177777777777778e-05,
      "loss": 0.0395,
      "step": 2710
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.2777860164642334,
      "learning_rate": 1.0133333333333335e-05,
      "loss": 0.0293,
      "step": 2720
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.08910083770751953,
      "learning_rate": 1.008888888888889e-05,
      "loss": 0.0319,
      "step": 2730
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.0010414342395961285,
      "learning_rate": 1.0044444444444446e-05,
      "loss": 0.0082,
      "step": 2740
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.0031285579316318035,
      "learning_rate": 1e-05,
      "loss": 0.0415,
      "step": 2750
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.005843820981681347,
      "learning_rate": 9.955555555555556e-06,
      "loss": 0.0557,
      "step": 2760
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0012017801636829972,
      "learning_rate": 9.911111111111113e-06,
      "loss": 0.052,
      "step": 2770
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.001439902582205832,
      "learning_rate": 9.866666666666668e-06,
      "loss": 0.0853,
      "step": 2780
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0035338387824594975,
      "learning_rate": 9.822222222222223e-06,
      "loss": 0.0227,
      "step": 2790
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0015976278809830546,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.0647,
      "step": 2800
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0012226419057697058,
      "learning_rate": 9.733333333333334e-06,
      "loss": 0.1104,
      "step": 2810
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5588458776473999,
      "learning_rate": 9.688888888888889e-06,
      "loss": 0.0732,
      "step": 2820
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.45813611149787903,
      "learning_rate": 9.644444444444444e-06,
      "loss": 0.0565,
      "step": 2830
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5053278207778931,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0676,
      "step": 2840
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.002007542410865426,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.0256,
      "step": 2850
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.0025233286432921886,
      "learning_rate": 9.511111111111112e-06,
      "loss": 0.057,
      "step": 2860
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.0035814358852803707,
      "learning_rate": 9.466666666666667e-06,
      "loss": 0.1282,
      "step": 2870
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.0031847648788243532,
      "learning_rate": 9.422222222222222e-06,
      "loss": 0.0126,
      "step": 2880
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.0034481678158044815,
      "learning_rate": 9.377777777777779e-06,
      "loss": 0.0627,
      "step": 2890
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.18580594658851624,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.035,
      "step": 2900
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.0014439361402764916,
      "learning_rate": 9.28888888888889e-06,
      "loss": 0.0766,
      "step": 2910
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.001558059360831976,
      "learning_rate": 9.244444444444445e-06,
      "loss": 0.023,
      "step": 2920
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.0029459816869348288,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.041,
      "step": 2930
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.0017032440518960357,
      "learning_rate": 9.155555555555557e-06,
      "loss": 0.0191,
      "step": 2940
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.002062348648905754,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.0263,
      "step": 2950
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.001492173527367413,
      "learning_rate": 9.066666666666667e-06,
      "loss": 0.0419,
      "step": 2960
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.903357982635498,
      "learning_rate": 9.022222222222223e-06,
      "loss": 0.0557,
      "step": 2970
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0011335255112498999,
      "learning_rate": 8.977777777777778e-06,
      "loss": 0.0065,
      "step": 2980
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0009899291908368468,
      "learning_rate": 8.933333333333333e-06,
      "loss": 0.0339,
      "step": 2990
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0012059457367286086,
      "learning_rate": 8.888888888888888e-06,
      "loss": 0.0374,
      "step": 3000
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.022419294342398643,
      "eval_runtime": 62.1056,
      "eval_samples_per_second": 16.102,
      "eval_steps_per_second": 16.102,
      "step": 3000
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.89267098903656,
      "learning_rate": 8.844444444444445e-06,
      "loss": 0.1307,
      "step": 3010
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.0008318617474287748,
      "learning_rate": 8.8e-06,
      "loss": 0.0331,
      "step": 3020
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.00299050472676754,
      "learning_rate": 8.755555555555556e-06,
      "loss": 0.0166,
      "step": 3030
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.3594794273376465,
      "learning_rate": 8.711111111111111e-06,
      "loss": 0.1204,
      "step": 3040
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.039698030799627304,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.0397,
      "step": 3050
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0027413545176386833,
      "learning_rate": 8.622222222222223e-06,
      "loss": 0.0555,
      "step": 3060
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.7610267400741577,
      "learning_rate": 8.577777777777778e-06,
      "loss": 0.0595,
      "step": 3070
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.001599436509422958,
      "learning_rate": 8.533333333333335e-06,
      "loss": 0.0525,
      "step": 3080
    },
    {
      "epoch": 0.97,
      "grad_norm": 3.3269574642181396,
      "learning_rate": 8.48888888888889e-06,
      "loss": 0.0149,
      "step": 3090
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.0009359756950289011,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.0772,
      "step": 3100
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.002731655491515994,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.1128,
      "step": 3110
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.01761116273701191,
      "learning_rate": 8.355555555555556e-06,
      "loss": 0.0228,
      "step": 3120
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.0019919571932405233,
      "learning_rate": 8.311111111111111e-06,
      "loss": 0.0389,
      "step": 3130
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.0010057106846943498,
      "learning_rate": 8.266666666666667e-06,
      "loss": 0.1025,
      "step": 3140
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.3677994012832642,
      "learning_rate": 8.222222222222222e-06,
      "loss": 0.0159,
      "step": 3150
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.0024271418806165457,
      "learning_rate": 8.177777777777779e-06,
      "loss": 0.031,
      "step": 3160
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.0010294296080246568,
      "learning_rate": 8.133333333333334e-06,
      "loss": 0.0168,
      "step": 3170
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0011234580306336284,
      "learning_rate": 8.08888888888889e-06,
      "loss": 0.058,
      "step": 3180
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0011441799579188228,
      "learning_rate": 8.044444444444444e-06,
      "loss": 0.0243,
      "step": 3190
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.860286295413971,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0128,
      "step": 3200
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.7325726747512817,
      "learning_rate": 7.955555555555557e-06,
      "loss": 0.0156,
      "step": 3210
    },
    {
      "epoch": 1.01,
      "grad_norm": 2.548293352127075,
      "learning_rate": 7.911111111111112e-06,
      "loss": 0.0724,
      "step": 3220
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.0010266635799780488,
      "learning_rate": 7.866666666666667e-06,
      "loss": 0.0194,
      "step": 3230
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.34962719678878784,
      "learning_rate": 7.822222222222224e-06,
      "loss": 0.0086,
      "step": 3240
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.0028716479428112507,
      "learning_rate": 7.77777777777778e-06,
      "loss": 0.0143,
      "step": 3250
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.1319417804479599,
      "learning_rate": 7.733333333333334e-06,
      "loss": 0.0231,
      "step": 3260
    },
    {
      "epoch": 1.02,
      "grad_norm": 2.1257731914520264,
      "learning_rate": 7.68888888888889e-06,
      "loss": 0.0767,
      "step": 3270
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.001123911701142788,
      "learning_rate": 7.644444444444445e-06,
      "loss": 0.0082,
      "step": 3280
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.0013617794029414654,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.0003,
      "step": 3290
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.0023631304502487183,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.0154,
      "step": 3300
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.145058512687683,
      "learning_rate": 7.511111111111111e-06,
      "loss": 0.0369,
      "step": 3310
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.0006993517745286226,
      "learning_rate": 7.4666666666666675e-06,
      "loss": 0.0195,
      "step": 3320
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.5191490650177002,
      "learning_rate": 7.422222222222223e-06,
      "loss": 0.1223,
      "step": 3330
    },
    {
      "epoch": 1.05,
      "grad_norm": 4.084077835083008,
      "learning_rate": 7.377777777777778e-06,
      "loss": 0.0498,
      "step": 3340
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.01493588462471962,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.0161,
      "step": 3350
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.0015668733976781368,
      "learning_rate": 7.28888888888889e-06,
      "loss": 0.0276,
      "step": 3360
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.0019036026205867529,
      "learning_rate": 7.244444444444445e-06,
      "loss": 0.0464,
      "step": 3370
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.003628369653597474,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.0636,
      "step": 3380
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.771296501159668,
      "learning_rate": 7.155555555555556e-06,
      "loss": 0.0138,
      "step": 3390
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.001274996087886393,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.0074,
      "step": 3400
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.4918831586837769,
      "learning_rate": 7.066666666666667e-06,
      "loss": 0.0227,
      "step": 3410
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.0008646890637464821,
      "learning_rate": 7.022222222222222e-06,
      "loss": 0.0054,
      "step": 3420
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.0009093726403079927,
      "learning_rate": 6.977777777777779e-06,
      "loss": 0.0434,
      "step": 3430
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0018424297450110316,
      "learning_rate": 6.9333333333333344e-06,
      "loss": 0.0601,
      "step": 3440
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0018895555986091495,
      "learning_rate": 6.88888888888889e-06,
      "loss": 0.0056,
      "step": 3450
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.0008156482363119721,
      "learning_rate": 6.844444444444445e-06,
      "loss": 0.0229,
      "step": 3460
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.46498626470565796,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0192,
      "step": 3470
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.8943893909454346,
      "learning_rate": 6.755555555555556e-06,
      "loss": 0.0321,
      "step": 3480
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.5523977279663086,
      "learning_rate": 6.711111111111111e-06,
      "loss": 0.0447,
      "step": 3490
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.000732199230697006,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.032,
      "step": 3500
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.02055978961288929,
      "eval_runtime": 62.0715,
      "eval_samples_per_second": 16.11,
      "eval_steps_per_second": 16.11,
      "step": 3500
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.002656770870089531,
      "learning_rate": 6.6222222222222236e-06,
      "loss": 0.0632,
      "step": 3510
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.0010300944559276104,
      "learning_rate": 6.577777777777779e-06,
      "loss": 0.0144,
      "step": 3520
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.3620741367340088,
      "learning_rate": 6.533333333333334e-06,
      "loss": 0.0497,
      "step": 3530
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.04097941517829895,
      "learning_rate": 6.488888888888889e-06,
      "loss": 0.025,
      "step": 3540
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.0011906519066542387,
      "learning_rate": 6.444444444444445e-06,
      "loss": 0.0059,
      "step": 3550
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0008852950995787978,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0354,
      "step": 3560
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0006289037992246449,
      "learning_rate": 6.355555555555556e-06,
      "loss": 0.0315,
      "step": 3570
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.001330237602815032,
      "learning_rate": 6.311111111111111e-06,
      "loss": 0.0161,
      "step": 3580
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.0021519146393984556,
      "learning_rate": 6.266666666666668e-06,
      "loss": 0.0105,
      "step": 3590
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5696661472320557,
      "learning_rate": 6.222222222222223e-06,
      "loss": 0.0244,
      "step": 3600
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.001931514940224588,
      "learning_rate": 6.177777777777778e-06,
      "loss": 0.0001,
      "step": 3610
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.0009438694687560201,
      "learning_rate": 6.133333333333334e-06,
      "loss": 0.0366,
      "step": 3620
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.6622722148895264,
      "learning_rate": 6.08888888888889e-06,
      "loss": 0.0113,
      "step": 3630
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.6738483309745789,
      "learning_rate": 6.044444444444445e-06,
      "loss": 0.0314,
      "step": 3640
    },
    {
      "epoch": 1.14,
      "grad_norm": 2.3022451400756836,
      "learning_rate": 6e-06,
      "loss": 0.0252,
      "step": 3650
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.823756694793701,
      "learning_rate": 5.955555555555555e-06,
      "loss": 0.0468,
      "step": 3660
    },
    {
      "epoch": 1.15,
      "grad_norm": 1.8479762077331543,
      "learning_rate": 5.911111111111112e-06,
      "loss": 0.0275,
      "step": 3670
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.0007701431168243289,
      "learning_rate": 5.8666666666666675e-06,
      "loss": 0.032,
      "step": 3680
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.1393382549285889,
      "learning_rate": 5.822222222222223e-06,
      "loss": 0.0791,
      "step": 3690
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.9080690145492554,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.0259,
      "step": 3700
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.001581453951075673,
      "learning_rate": 5.733333333333334e-06,
      "loss": 0.0409,
      "step": 3710
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.03290742263197899,
      "learning_rate": 5.688888888888889e-06,
      "loss": 0.0929,
      "step": 3720
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.0012888371711596847,
      "learning_rate": 5.6444444444444445e-06,
      "loss": 0.0168,
      "step": 3730
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.0007963385432958603,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0525,
      "step": 3740
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.015161274001002312,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.0375,
      "step": 3750
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.0009869037894532084,
      "learning_rate": 5.511111111111112e-06,
      "loss": 0.0246,
      "step": 3760
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.002037523780018091,
      "learning_rate": 5.466666666666667e-06,
      "loss": 0.0934,
      "step": 3770
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.002192340325564146,
      "learning_rate": 5.422222222222223e-06,
      "loss": 0.0138,
      "step": 3780
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.10088424384593964,
      "learning_rate": 5.3777777777777784e-06,
      "loss": 0.0259,
      "step": 3790
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.38258832693099976,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0266,
      "step": 3800
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.0011203467147424817,
      "learning_rate": 5.288888888888889e-06,
      "loss": 0.0069,
      "step": 3810
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.0006357720121741295,
      "learning_rate": 5.244444444444445e-06,
      "loss": 0.0272,
      "step": 3820
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.12165514379739761,
      "learning_rate": 5.2e-06,
      "loss": 0.0175,
      "step": 3830
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.0015337629010900855,
      "learning_rate": 5.155555555555556e-06,
      "loss": 0.027,
      "step": 3840
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.0012696600751951337,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.0005,
      "step": 3850
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.8595750331878662,
      "learning_rate": 5.0666666666666676e-06,
      "loss": 0.0367,
      "step": 3860
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.002280933316797018,
      "learning_rate": 5.022222222222223e-06,
      "loss": 0.0374,
      "step": 3870
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.0005754129961133003,
      "learning_rate": 4.977777777777778e-06,
      "loss": 0.0168,
      "step": 3880
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.8617763519287109,
      "learning_rate": 4.933333333333334e-06,
      "loss": 0.0155,
      "step": 3890
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.5179722905158997,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.0486,
      "step": 3900
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.0009101899340748787,
      "learning_rate": 4.8444444444444446e-06,
      "loss": 0.0027,
      "step": 3910
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.008863825350999832,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.055,
      "step": 3920
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.593704342842102,
      "learning_rate": 4.755555555555556e-06,
      "loss": 0.0128,
      "step": 3930
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.0012650558492168784,
      "learning_rate": 4.711111111111111e-06,
      "loss": 0.0,
      "step": 3940
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0025375504046678543,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0254,
      "step": 3950
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0035014720633625984,
      "learning_rate": 4.622222222222222e-06,
      "loss": 0.0254,
      "step": 3960
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.0010467343963682652,
      "learning_rate": 4.5777777777777785e-06,
      "loss": 0.0556,
      "step": 3970
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.0006358217215165496,
      "learning_rate": 4.533333333333334e-06,
      "loss": 0.0266,
      "step": 3980
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.3243739902973175,
      "learning_rate": 4.488888888888889e-06,
      "loss": 0.0079,
      "step": 3990
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.04155297204852104,
      "learning_rate": 4.444444444444444e-06,
      "loss": 0.0163,
      "step": 4000
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.02105984091758728,
      "eval_runtime": 62.0961,
      "eval_samples_per_second": 16.104,
      "eval_steps_per_second": 16.104,
      "step": 4000
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.0018766351277008653,
      "learning_rate": 4.4e-06,
      "loss": 0.0521,
      "step": 4010
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.39914241433143616,
      "learning_rate": 4.3555555555555555e-06,
      "loss": 0.0195,
      "step": 4020
    },
    {
      "epoch": 1.26,
      "grad_norm": 4.4533562660217285,
      "learning_rate": 4.3111111111111115e-06,
      "loss": 0.0588,
      "step": 4030
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.0005871405592188239,
      "learning_rate": 4.266666666666668e-06,
      "loss": 0.0162,
      "step": 4040
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.0015552936820313334,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.1069,
      "step": 4050
    },
    {
      "epoch": 1.27,
      "grad_norm": 3.2195968627929688,
      "learning_rate": 4.177777777777778e-06,
      "loss": 0.0274,
      "step": 4060
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.004314718768000603,
      "learning_rate": 4.133333333333333e-06,
      "loss": 0.0057,
      "step": 4070
    },
    {
      "epoch": 1.28,
      "grad_norm": 1.0093330144882202,
      "learning_rate": 4.088888888888889e-06,
      "loss": 0.0325,
      "step": 4080
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.0012565231882035732,
      "learning_rate": 4.044444444444445e-06,
      "loss": 0.0296,
      "step": 4090
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.032142095267772675,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0341,
      "step": 4100
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.0013831367250531912,
      "learning_rate": 3.955555555555556e-06,
      "loss": 0.0149,
      "step": 4110
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.502413511276245,
      "learning_rate": 3.911111111111112e-06,
      "loss": 0.0599,
      "step": 4120
    },
    {
      "epoch": 1.29,
      "grad_norm": 2.310913562774658,
      "learning_rate": 3.866666666666667e-06,
      "loss": 0.0421,
      "step": 4130
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.06956034153699875,
      "learning_rate": 3.8222222222222224e-06,
      "loss": 0.0155,
      "step": 4140
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.0008762804791331291,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.0559,
      "step": 4150
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.245699167251587,
      "learning_rate": 3.7333333333333337e-06,
      "loss": 0.0407,
      "step": 4160
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.3646597862243652,
      "learning_rate": 3.688888888888889e-06,
      "loss": 0.0283,
      "step": 4170
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.0010521197691559792,
      "learning_rate": 3.644444444444445e-06,
      "loss": 0.018,
      "step": 4180
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.0006315945065580308,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.006,
      "step": 4190
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.000512352678924799,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.0065,
      "step": 4200
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.3170325756073,
      "learning_rate": 3.511111111111111e-06,
      "loss": 0.017,
      "step": 4210
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.004807967692613602,
      "learning_rate": 3.4666666666666672e-06,
      "loss": 0.0202,
      "step": 4220
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0015995922731235623,
      "learning_rate": 3.4222222222222224e-06,
      "loss": 0.0436,
      "step": 4230
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0005936655215919018,
      "learning_rate": 3.377777777777778e-06,
      "loss": 0.0169,
      "step": 4240
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0005440403474494815,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0172,
      "step": 4250
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.0013694684021174908,
      "learning_rate": 3.2888888888888894e-06,
      "loss": 0.0129,
      "step": 4260
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0017236489802598953,
      "learning_rate": 3.2444444444444446e-06,
      "loss": 0.0349,
      "step": 4270
    },
    {
      "epoch": 1.34,
      "grad_norm": 2.429488182067871,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.1352,
      "step": 4280
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.0007962828385643661,
      "learning_rate": 3.1555555555555555e-06,
      "loss": 0.0051,
      "step": 4290
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.0005601115408353508,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 0.0091,
      "step": 4300
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.0022927261888980865,
      "learning_rate": 3.066666666666667e-06,
      "loss": 0.0208,
      "step": 4310
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.00264713354408741,
      "learning_rate": 3.0222222222222225e-06,
      "loss": 0.0445,
      "step": 4320
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0009310416062362492,
      "learning_rate": 2.9777777777777777e-06,
      "loss": 0.0295,
      "step": 4330
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0005691644619219005,
      "learning_rate": 2.9333333333333338e-06,
      "loss": 0.0497,
      "step": 4340
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.0012022920418530703,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.04,
      "step": 4350
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.9444104433059692,
      "learning_rate": 2.8444444444444446e-06,
      "loss": 0.063,
      "step": 4360
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.8922362327575684,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0319,
      "step": 4370
    },
    {
      "epoch": 1.37,
      "grad_norm": 1.2501660585403442,
      "learning_rate": 2.755555555555556e-06,
      "loss": 0.0716,
      "step": 4380
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.014218101277947426,
      "learning_rate": 2.7111111111111116e-06,
      "loss": 0.0219,
      "step": 4390
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.16711218655109406,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.0356,
      "step": 4400
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.0005462172557599843,
      "learning_rate": 2.6222222222222225e-06,
      "loss": 0.0503,
      "step": 4410
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.2715685367584229,
      "learning_rate": 2.577777777777778e-06,
      "loss": 0.0282,
      "step": 4420
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.0005375253385864198,
      "learning_rate": 2.5333333333333338e-06,
      "loss": 0.0346,
      "step": 4430
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.5777830481529236,
      "learning_rate": 2.488888888888889e-06,
      "loss": 0.061,
      "step": 4440
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.00046264228876680136,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.0445,
      "step": 4450
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.0011936473893001676,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0251,
      "step": 4460
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.0460318326950073,
      "learning_rate": 2.3555555555555555e-06,
      "loss": 0.0409,
      "step": 4470
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.9569495916366577,
      "learning_rate": 2.311111111111111e-06,
      "loss": 0.026,
      "step": 4480
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.012949565425515175,
      "learning_rate": 2.266666666666667e-06,
      "loss": 0.0183,
      "step": 4490
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.0013554710894823074,
      "learning_rate": 2.222222222222222e-06,
      "loss": 0.0162,
      "step": 4500
    },
    {
      "epoch": 1.41,
      "eval_loss": 0.017521824687719345,
      "eval_runtime": 62.0307,
      "eval_samples_per_second": 16.121,
      "eval_steps_per_second": 16.121,
      "step": 4500
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.001089457655325532,
      "learning_rate": 2.1777777777777777e-06,
      "loss": 0.0229,
      "step": 4510
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.5206069350242615,
      "learning_rate": 2.133333333333334e-06,
      "loss": 0.03,
      "step": 4520
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.0005083256983198225,
      "learning_rate": 2.088888888888889e-06,
      "loss": 0.0328,
      "step": 4530
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.0009927984792739153,
      "learning_rate": 2.0444444444444447e-06,
      "loss": 0.0095,
      "step": 4540
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0006405860185623169,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.009,
      "step": 4550
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0009081049938686192,
      "learning_rate": 1.955555555555556e-06,
      "loss": 0.0086,
      "step": 4560
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0004944841493852437,
      "learning_rate": 1.9111111111111112e-06,
      "loss": 0.0028,
      "step": 4570
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.0010713768424466252,
      "learning_rate": 1.8666666666666669e-06,
      "loss": 0.0038,
      "step": 4580
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0013352985261008143,
      "learning_rate": 1.8222222222222225e-06,
      "loss": 0.0517,
      "step": 4590
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0011882871622219682,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.0344,
      "step": 4600
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0005185340996831656,
      "learning_rate": 1.7333333333333336e-06,
      "loss": 0.0219,
      "step": 4610
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.0005603168392553926,
      "learning_rate": 1.688888888888889e-06,
      "loss": 0.0306,
      "step": 4620
    },
    {
      "epoch": 1.45,
      "grad_norm": 2.2889089584350586,
      "learning_rate": 1.6444444444444447e-06,
      "loss": 0.0619,
      "step": 4630
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.0005399343208409846,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0001,
      "step": 4640
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.3966524600982666,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 0.0329,
      "step": 4650
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.1912785768508911,
      "learning_rate": 1.5111111111111112e-06,
      "loss": 0.0926,
      "step": 4660
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.0009074648842215538,
      "learning_rate": 1.4666666666666669e-06,
      "loss": 0.0123,
      "step": 4670
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.0007042177603580058,
      "learning_rate": 1.4222222222222223e-06,
      "loss": 0.0029,
      "step": 4680
    },
    {
      "epoch": 1.47,
      "grad_norm": 1.4998135566711426,
      "learning_rate": 1.377777777777778e-06,
      "loss": 0.0635,
      "step": 4690
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.7605366110801697,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.0085,
      "step": 4700
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.0010083210654556751,
      "learning_rate": 1.288888888888889e-06,
      "loss": 0.0116,
      "step": 4710
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.0018564054043963552,
      "learning_rate": 1.2444444444444445e-06,
      "loss": 0.0156,
      "step": 4720
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.5900565981864929,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0436,
      "step": 4730
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.9675027132034302,
      "learning_rate": 1.1555555555555556e-06,
      "loss": 0.0535,
      "step": 4740
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.0011088999453932047,
      "learning_rate": 1.111111111111111e-06,
      "loss": 0.053,
      "step": 4750
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.0014372550649568439,
      "learning_rate": 1.066666666666667e-06,
      "loss": 0.0047,
      "step": 4760
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.9269665479660034,
      "learning_rate": 1.0222222222222223e-06,
      "loss": 0.0416,
      "step": 4770
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0020348341204226017,
      "learning_rate": 9.77777777777778e-07,
      "loss": 0.059,
      "step": 4780
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.0013614660128951073,
      "learning_rate": 9.333333333333334e-07,
      "loss": 0.0377,
      "step": 4790
    },
    {
      "epoch": 1.5,
      "grad_norm": 2.781449556350708,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.1254,
      "step": 4800
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.0010429318062961102,
      "learning_rate": 8.444444444444445e-07,
      "loss": 0.0064,
      "step": 4810
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.0012540715979412198,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.049,
      "step": 4820
    },
    {
      "epoch": 1.51,
      "grad_norm": 2.989722967147827,
      "learning_rate": 7.555555555555556e-07,
      "loss": 0.0596,
      "step": 4830
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.08106524497270584,
      "learning_rate": 7.111111111111112e-07,
      "loss": 0.0123,
      "step": 4840
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.003921021707355976,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.0313,
      "step": 4850
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.0006887316121719778,
      "learning_rate": 6.222222222222223e-07,
      "loss": 0.0162,
      "step": 4860
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.0005608502542600036,
      "learning_rate": 5.777777777777778e-07,
      "loss": 0.0579,
      "step": 4870
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.7983081936836243,
      "learning_rate": 5.333333333333335e-07,
      "loss": 0.0205,
      "step": 4880
    },
    {
      "epoch": 1.53,
      "grad_norm": 4.0297722816467285,
      "learning_rate": 4.88888888888889e-07,
      "loss": 0.0413,
      "step": 4890
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0005954677471891046,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.0443,
      "step": 4900
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0008567142067477107,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.0918,
      "step": 4910
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0158186424523592,
      "learning_rate": 3.555555555555556e-07,
      "loss": 0.0117,
      "step": 4920
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.0005192561075091362,
      "learning_rate": 3.111111111111111e-07,
      "loss": 0.0367,
      "step": 4930
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.0007436941377818584,
      "learning_rate": 2.666666666666667e-07,
      "loss": 0.0103,
      "step": 4940
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.0006840491550974548,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.025,
      "step": 4950
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.0011544113513082266,
      "learning_rate": 1.777777777777778e-07,
      "loss": 0.0408,
      "step": 4960
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.0013330102665349841,
      "learning_rate": 1.3333333333333336e-07,
      "loss": 0.0468,
      "step": 4970
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.0007117724744603038,
      "learning_rate": 8.88888888888889e-08,
      "loss": 0.0206,
      "step": 4980
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.006810193415731192,
      "learning_rate": 4.444444444444445e-08,
      "loss": 0.0489,
      "step": 4990
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.000949681329075247,
      "learning_rate": 0.0,
      "loss": 0.0389,
      "step": 5000
    },
    {
      "epoch": 1.57,
      "eval_loss": 0.02238212525844574,
      "eval_runtime": 62.0206,
      "eval_samples_per_second": 16.124,
      "eval_steps_per_second": 16.124,
      "step": 5000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "total_flos": 8.051062996992e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}