{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0002,
      "grad_norm": 8.95457649230957,
      "learning_rate": 0.0,
      "loss": 4.5624,
      "step": 1
    },
    {
      "epoch": 0.0004,
      "grad_norm": 8.470853805541992,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.5771,
      "step": 2
    },
    {
      "epoch": 0.0006,
      "grad_norm": 7.9080610275268555,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.5221,
      "step": 3
    },
    {
      "epoch": 0.0008,
      "grad_norm": 7.897581100463867,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 4.5801,
      "step": 4
    },
    {
      "epoch": 0.001,
      "grad_norm": 7.688938617706299,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 4.3139,
      "step": 5
    },
    {
      "epoch": 0.0012,
      "grad_norm": 6.975260257720947,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 4.3402,
      "step": 6
    },
    {
      "epoch": 0.0014,
      "grad_norm": 7.0634765625,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 4.1657,
      "step": 7
    },
    {
      "epoch": 0.0016,
      "grad_norm": 7.310269832611084,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 4.3021,
      "step": 8
    },
    {
      "epoch": 0.0018,
      "grad_norm": 8.602141380310059,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 4.4401,
      "step": 9
    },
    {
      "epoch": 0.002,
      "grad_norm": 8.004490852355957,
      "learning_rate": 3.6e-06,
      "loss": 4.559,
      "step": 10
    },
    {
      "epoch": 0.0022,
      "grad_norm": 8.732573509216309,
      "learning_rate": 4.000000000000001e-06,
      "loss": 4.6551,
      "step": 11
    },
    {
      "epoch": 0.0024,
      "grad_norm": 8.629362106323242,
      "learning_rate": 4.4e-06,
      "loss": 4.4654,
      "step": 12
    },
    {
      "epoch": 0.0026,
      "grad_norm": 6.896476745605469,
      "learning_rate": 4.800000000000001e-06,
      "loss": 4.1281,
      "step": 13
    },
    {
      "epoch": 0.0028,
      "grad_norm": 6.789773464202881,
      "learning_rate": 5.2e-06,
      "loss": 4.2926,
      "step": 14
    },
    {
      "epoch": 0.003,
      "grad_norm": 6.726924896240234,
      "learning_rate": 5.600000000000001e-06,
      "loss": 4.1422,
      "step": 15
    },
    {
      "epoch": 0.0032,
      "grad_norm": 7.723042964935303,
      "learning_rate": 6e-06,
      "loss": 4.2752,
      "step": 16
    },
    {
      "epoch": 0.0034,
      "grad_norm": 7.3740315437316895,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 4.0774,
      "step": 17
    },
    {
      "epoch": 0.0036,
      "grad_norm": 6.809215545654297,
      "learning_rate": 6.800000000000001e-06,
      "loss": 4.0093,
      "step": 18
    },
    {
      "epoch": 0.0038,
      "grad_norm": 6.224776268005371,
      "learning_rate": 7.2e-06,
      "loss": 3.8365,
      "step": 19
    },
    {
      "epoch": 0.004,
      "grad_norm": 6.402893543243408,
      "learning_rate": 7.6e-06,
      "loss": 3.9231,
      "step": 20
    },
    {
      "epoch": 0.0042,
      "grad_norm": 5.9624786376953125,
      "learning_rate": 8.000000000000001e-06,
      "loss": 3.7886,
      "step": 21
    },
    {
      "epoch": 0.0044,
      "grad_norm": 5.4187164306640625,
      "learning_rate": 8.400000000000001e-06,
      "loss": 3.6696,
      "step": 22
    },
    {
      "epoch": 0.0046,
      "grad_norm": 4.8634185791015625,
      "learning_rate": 8.8e-06,
      "loss": 3.5485,
      "step": 23
    },
    {
      "epoch": 0.0048,
      "grad_norm": 4.493101596832275,
      "learning_rate": 9.2e-06,
      "loss": 3.4394,
      "step": 24
    },
    {
      "epoch": 0.005,
      "grad_norm": 3.4735097885131836,
      "learning_rate": 9.600000000000001e-06,
      "loss": 3.3979,
      "step": 25
    },
    {
      "epoch": 0.0052,
      "grad_norm": 4.174388408660889,
      "learning_rate": 1e-05,
      "loss": 3.4608,
      "step": 26
    },
    {
      "epoch": 0.0054,
      "grad_norm": 3.293051242828369,
      "learning_rate": 1.04e-05,
      "loss": 3.34,
      "step": 27
    },
    {
      "epoch": 0.0056,
      "grad_norm": 2.3477447032928467,
      "learning_rate": 1.08e-05,
      "loss": 3.2005,
      "step": 28
    },
    {
      "epoch": 0.0058,
      "grad_norm": 2.0154621601104736,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 3.2013,
      "step": 29
    },
    {
      "epoch": 0.006,
      "grad_norm": 1.6802643537521362,
      "learning_rate": 1.16e-05,
      "loss": 3.172,
      "step": 30
    },
    {
      "epoch": 0.0062,
      "grad_norm": 1.0652613639831543,
      "learning_rate": 1.2e-05,
      "loss": 3.1434,
      "step": 31
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.5900894403457642,
      "learning_rate": 1.24e-05,
      "loss": 3.0914,
      "step": 32
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.6736801266670227,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 3.0713,
      "step": 33
    },
    {
      "epoch": 0.0068,
      "grad_norm": 1.0225924253463745,
      "learning_rate": 1.32e-05,
      "loss": 3.0961,
      "step": 34
    },
    {
      "epoch": 0.007,
      "grad_norm": 1.0093497037887573,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 3.1308,
      "step": 35
    },
    {
      "epoch": 0.0072,
      "grad_norm": 1.1132439374923706,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 3.0991,
      "step": 36
    },
    {
      "epoch": 0.0074,
      "grad_norm": 1.2905092239379883,
      "learning_rate": 1.44e-05,
      "loss": 3.1089,
      "step": 37
    },
    {
      "epoch": 0.0076,
      "grad_norm": 1.4385223388671875,
      "learning_rate": 1.48e-05,
      "loss": 3.121,
      "step": 38
    },
    {
      "epoch": 0.0078,
      "grad_norm": 1.3655472993850708,
      "learning_rate": 1.52e-05,
      "loss": 3.0636,
      "step": 39
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.511385202407837,
      "learning_rate": 1.56e-05,
      "loss": 3.0514,
      "step": 40
    },
    {
      "epoch": 0.0082,
      "grad_norm": 1.1128480434417725,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.0254,
      "step": 41
    },
    {
      "epoch": 0.0084,
      "grad_norm": 1.201245665550232,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 3.0313,
      "step": 42
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.8725724816322327,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 3.0252,
      "step": 43
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.9620676040649414,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 2.9875,
      "step": 44
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.6491302847862244,
      "learning_rate": 1.76e-05,
      "loss": 2.9711,
      "step": 45
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.7817885875701904,
      "learning_rate": 1.8e-05,
      "loss": 2.987,
      "step": 46
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.6245918273925781,
      "learning_rate": 1.84e-05,
      "loss": 2.9419,
      "step": 47
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.6139571666717529,
      "learning_rate": 1.88e-05,
      "loss": 2.9391,
      "step": 48
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.7280861139297485,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 2.9108,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7247921228408813,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 2.9133,
      "step": 50
    },
    {
      "epoch": 0.0102,
      "grad_norm": 0.6968711018562317,
      "learning_rate": 2e-05,
      "loss": 2.8768,
      "step": 51
    },
    {
      "epoch": 0.0104,
      "grad_norm": 1.0155657529830933,
      "learning_rate": 2.04e-05,
      "loss": 2.8616,
      "step": 52
    },
    {
      "epoch": 0.0106,
      "grad_norm": 0.8728637099266052,
      "learning_rate": 2.08e-05,
      "loss": 2.8471,
      "step": 53
    },
    {
      "epoch": 0.0108,
      "grad_norm": 0.5890777707099915,
      "learning_rate": 2.12e-05,
      "loss": 2.82,
      "step": 54
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.6640511155128479,
      "learning_rate": 2.16e-05,
      "loss": 2.8101,
      "step": 55
    },
    {
      "epoch": 0.0112,
      "grad_norm": 0.7456818222999573,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 2.8045,
      "step": 56
    },
    {
      "epoch": 0.0114,
      "grad_norm": 0.6434550881385803,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 2.7796,
      "step": 57
    },
    {
      "epoch": 0.0116,
      "grad_norm": 0.4944389760494232,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 2.7658,
      "step": 58
    },
    {
      "epoch": 0.0118,
      "grad_norm": 0.5783756971359253,
      "learning_rate": 2.32e-05,
      "loss": 2.7495,
      "step": 59
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.7428807020187378,
      "learning_rate": 2.36e-05,
      "loss": 2.7502,
      "step": 60
    },
    {
      "epoch": 0.0122,
      "grad_norm": 0.7576146721839905,
      "learning_rate": 2.4e-05,
      "loss": 2.72,
      "step": 61
    },
    {
      "epoch": 0.0124,
      "grad_norm": 0.6714867353439331,
      "learning_rate": 2.44e-05,
      "loss": 2.7029,
      "step": 62
    },
    {
      "epoch": 0.0126,
      "grad_norm": 0.504632294178009,
      "learning_rate": 2.48e-05,
      "loss": 2.6855,
      "step": 63
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.40775713324546814,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 2.6721,
      "step": 64
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.4155912399291992,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 2.6453,
      "step": 65
    },
    {
      "epoch": 0.0132,
      "grad_norm": 0.44520696997642517,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 2.6547,
      "step": 66
    },
    {
      "epoch": 0.0134,
      "grad_norm": 1.3109104633331299,
      "learning_rate": 2.64e-05,
      "loss": 2.6535,
      "step": 67
    },
    {
      "epoch": 0.0136,
      "grad_norm": 1.090923547744751,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 2.6385,
      "step": 68
    },
    {
      "epoch": 0.0138,
      "grad_norm": 0.7457869648933411,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 2.6265,
      "step": 69
    },
    {
      "epoch": 0.014,
      "grad_norm": 0.4242352843284607,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 2.604,
      "step": 70
    },
    {
      "epoch": 0.0142,
      "grad_norm": 0.6218993067741394,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 2.5996,
      "step": 71
    },
    {
      "epoch": 0.0144,
      "grad_norm": 0.8016015291213989,
      "learning_rate": 2.84e-05,
      "loss": 2.5974,
      "step": 72
    },
    {
      "epoch": 0.0146,
      "grad_norm": 0.5211424231529236,
      "learning_rate": 2.88e-05,
      "loss": 2.6243,
      "step": 73
    },
    {
      "epoch": 0.0148,
      "grad_norm": 0.5598169565200806,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 2.5764,
      "step": 74
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.22318528592586517,
      "learning_rate": 2.96e-05,
      "loss": 2.5743,
      "step": 75
    },
    {
      "epoch": 0.0152,
      "grad_norm": 0.43333470821380615,
      "learning_rate": 3e-05,
      "loss": 2.5738,
      "step": 76
    },
    {
      "epoch": 0.0154,
      "grad_norm": 0.7180959582328796,
      "learning_rate": 3.04e-05,
      "loss": 2.5691,
      "step": 77
    },
    {
      "epoch": 0.0156,
      "grad_norm": 0.8009986281394958,
      "learning_rate": 3.08e-05,
      "loss": 2.5667,
      "step": 78
    },
    {
      "epoch": 0.0158,
      "grad_norm": 0.2783987820148468,
      "learning_rate": 3.12e-05,
      "loss": 2.5543,
      "step": 79
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.24904108047485352,
      "learning_rate": 3.16e-05,
      "loss": 2.5633,
      "step": 80
    },
    {
      "epoch": 0.0162,
      "grad_norm": 0.2789011597633362,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 2.5521,
      "step": 81
    },
    {
      "epoch": 0.0164,
      "grad_norm": 0.5622010231018066,
      "learning_rate": 3.24e-05,
      "loss": 2.5483,
      "step": 82
    },
    {
      "epoch": 0.0166,
      "grad_norm": 0.2821291983127594,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 2.545,
      "step": 83
    },
    {
      "epoch": 0.0168,
      "grad_norm": 0.6145974397659302,
      "learning_rate": 3.32e-05,
      "loss": 2.5327,
      "step": 84
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.49089381098747253,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 2.5241,
      "step": 85
    },
    {
      "epoch": 0.0172,
      "grad_norm": 0.18112079799175262,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 2.5361,
      "step": 86
    },
    {
      "epoch": 0.0174,
      "grad_norm": 0.2559221684932709,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 2.5259,
      "step": 87
    },
    {
      "epoch": 0.0176,
      "grad_norm": 0.17089028656482697,
      "learning_rate": 3.48e-05,
      "loss": 2.5221,
      "step": 88
    },
    {
      "epoch": 0.0178,
      "grad_norm": 0.2147020697593689,
      "learning_rate": 3.52e-05,
      "loss": 2.5151,
      "step": 89
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.43801289796829224,
      "learning_rate": 3.56e-05,
      "loss": 2.5209,
      "step": 90
    },
    {
      "epoch": 0.0182,
      "grad_norm": 0.1812380701303482,
      "learning_rate": 3.6e-05,
      "loss": 2.5099,
      "step": 91
    },
    {
      "epoch": 0.0184,
      "grad_norm": 0.4475123882293701,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 2.5167,
      "step": 92
    },
    {
      "epoch": 0.0186,
      "grad_norm": 0.46092352271080017,
      "learning_rate": 3.68e-05,
      "loss": 2.515,
      "step": 93
    },
    {
      "epoch": 0.0188,
      "grad_norm": 0.2914877235889435,
      "learning_rate": 3.72e-05,
      "loss": 2.5033,
      "step": 94
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.17350858449935913,
      "learning_rate": 3.76e-05,
      "loss": 2.4994,
      "step": 95
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.17111025750637054,
      "learning_rate": 3.8e-05,
      "loss": 2.5026,
      "step": 96
    },
    {
      "epoch": 0.0194,
      "grad_norm": 0.33729806542396545,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 2.5015,
      "step": 97
    },
    {
      "epoch": 0.0196,
      "grad_norm": 0.6496284008026123,
      "learning_rate": 3.88e-05,
      "loss": 2.4982,
      "step": 98
    },
    {
      "epoch": 0.0198,
      "grad_norm": 0.1803646683692932,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 2.4913,
      "step": 99
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.24662846326828003,
      "learning_rate": 3.960000000000001e-05,
      "loss": 2.4879,
      "step": 100
    },
    {
      "epoch": 0.0202,
      "grad_norm": 0.8761454224586487,
      "learning_rate": 4e-05,
      "loss": 2.4845,
      "step": 101
    },
    {
      "epoch": 0.0204,
      "grad_norm": 0.31369855999946594,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 2.4817,
      "step": 102
    },
    {
      "epoch": 0.0206,
      "grad_norm": 0.14703480899333954,
      "learning_rate": 4.08e-05,
      "loss": 2.4787,
      "step": 103
    },
    {
      "epoch": 0.0208,
      "grad_norm": 0.4149743914604187,
      "learning_rate": 4.12e-05,
      "loss": 2.4749,
      "step": 104
    },
    {
      "epoch": 0.021,
      "grad_norm": 0.2869059443473816,
      "learning_rate": 4.16e-05,
      "loss": 2.4741,
      "step": 105
    },
    {
      "epoch": 0.0212,
      "grad_norm": 0.2535330355167389,
      "learning_rate": 4.2e-05,
      "loss": 2.4695,
      "step": 106
    },
    {
      "epoch": 0.0214,
      "grad_norm": 0.24170562624931335,
      "learning_rate": 4.24e-05,
      "loss": 2.4688,
      "step": 107
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.12546133995056152,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 2.4654,
      "step": 108
    },
    {
      "epoch": 0.0218,
      "grad_norm": 0.19572150707244873,
      "learning_rate": 4.32e-05,
      "loss": 2.4634,
      "step": 109
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.17278055846691132,
      "learning_rate": 4.36e-05,
      "loss": 2.4581,
      "step": 110
    },
    {
      "epoch": 0.0222,
      "grad_norm": 0.12620271742343903,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 2.4551,
      "step": 111
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.42709383368492126,
      "learning_rate": 4.44e-05,
      "loss": 2.4486,
      "step": 112
    },
    {
      "epoch": 0.0226,
      "grad_norm": 0.1260664016008377,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 2.4448,
      "step": 113
    },
    {
      "epoch": 0.0228,
      "grad_norm": 0.14292410016059875,
      "learning_rate": 4.52e-05,
      "loss": 2.446,
      "step": 114
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.2574887275695801,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 2.4399,
      "step": 115
    },
    {
      "epoch": 0.0232,
      "grad_norm": 0.13797590136528015,
      "learning_rate": 4.600000000000001e-05,
      "loss": 2.4418,
      "step": 116
    },
    {
      "epoch": 0.0234,
      "grad_norm": 0.3942067325115204,
      "learning_rate": 4.64e-05,
      "loss": 2.4355,
      "step": 117
    },
    {
      "epoch": 0.0236,
      "grad_norm": 0.3321908116340637,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 2.4268,
      "step": 118
    },
    {
      "epoch": 0.0238,
      "grad_norm": 0.1167830303311348,
      "learning_rate": 4.72e-05,
      "loss": 2.4313,
      "step": 119
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.47900086641311646,
      "learning_rate": 4.76e-05,
      "loss": 2.4323,
      "step": 120
    },
    {
      "epoch": 0.0242,
      "grad_norm": 0.47437116503715515,
      "learning_rate": 4.8e-05,
      "loss": 2.4243,
      "step": 121
    },
    {
      "epoch": 0.0244,
      "grad_norm": 0.13677360117435455,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 2.4219,
      "step": 122
    },
    {
      "epoch": 0.0246,
      "grad_norm": 0.7528557181358337,
      "learning_rate": 4.88e-05,
      "loss": 2.4235,
      "step": 123
    },
    {
      "epoch": 0.0248,
      "grad_norm": 0.5480535626411438,
      "learning_rate": 4.92e-05,
      "loss": 2.4207,
      "step": 124
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.16372530162334442,
      "learning_rate": 4.96e-05,
      "loss": 2.4149,
      "step": 125
    },
    {
      "epoch": 0.0252,
      "grad_norm": 0.6893525123596191,
      "learning_rate": 5e-05,
      "loss": 2.4142,
      "step": 126
    },
    {
      "epoch": 0.0254,
      "grad_norm": 0.4446101784706116,
      "learning_rate": 5.0400000000000005e-05,
      "loss": 2.4096,
      "step": 127
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.11107442528009415,
      "learning_rate": 5.08e-05,
      "loss": 2.4052,
      "step": 128
    },
    {
      "epoch": 0.0258,
      "grad_norm": 0.585650622844696,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 2.4035,
      "step": 129
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.545045018196106,
      "learning_rate": 5.16e-05,
      "loss": 2.4051,
      "step": 130
    },
    {
      "epoch": 0.0262,
      "grad_norm": 0.21546784043312073,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 2.398,
      "step": 131
    },
    {
      "epoch": 0.0264,
      "grad_norm": 0.37488865852355957,
      "learning_rate": 5.2400000000000007e-05,
      "loss": 2.3987,
      "step": 132
    },
    {
      "epoch": 0.0266,
      "grad_norm": 0.33907943964004517,
      "learning_rate": 5.28e-05,
      "loss": 2.3988,
      "step": 133
    },
    {
      "epoch": 0.0268,
      "grad_norm": 0.35165393352508545,
      "learning_rate": 5.3200000000000006e-05,
      "loss": 2.39,
      "step": 134
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.14002704620361328,
      "learning_rate": 5.360000000000001e-05,
      "loss": 2.3882,
      "step": 135
    },
    {
      "epoch": 0.0272,
      "grad_norm": 0.49009817838668823,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 2.3875,
      "step": 136
    },
    {
      "epoch": 0.0274,
      "grad_norm": 0.3474384844303131,
      "learning_rate": 5.440000000000001e-05,
      "loss": 2.3871,
      "step": 137
    },
    {
      "epoch": 0.0276,
      "grad_norm": 0.12790752947330475,
      "learning_rate": 5.4800000000000004e-05,
      "loss": 2.3836,
      "step": 138
    },
    {
      "epoch": 0.0278,
      "grad_norm": 0.32718437910079956,
      "learning_rate": 5.520000000000001e-05,
      "loss": 2.384,
      "step": 139
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.3266403377056122,
      "learning_rate": 5.560000000000001e-05,
      "loss": 2.3823,
      "step": 140
    },
    {
      "epoch": 0.0282,
      "grad_norm": 0.16459183394908905,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 2.3778,
      "step": 141
    },
    {
      "epoch": 0.0284,
      "grad_norm": 0.3552015423774719,
      "learning_rate": 5.6399999999999995e-05,
      "loss": 2.3779,
      "step": 142
    },
    {
      "epoch": 0.0286,
      "grad_norm": 0.4122103452682495,
      "learning_rate": 5.68e-05,
      "loss": 2.3742,
      "step": 143
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.10815047472715378,
      "learning_rate": 5.72e-05,
      "loss": 2.3718,
      "step": 144
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.5638622641563416,
      "learning_rate": 5.76e-05,
      "loss": 2.3725,
      "step": 145
    },
    {
      "epoch": 0.0292,
      "grad_norm": 0.4082678556442261,
      "learning_rate": 5.8e-05,
      "loss": 2.3682,
      "step": 146
    },
    {
      "epoch": 0.0294,
      "grad_norm": 0.3196413218975067,
      "learning_rate": 5.8399999999999997e-05,
      "loss": 2.3654,
      "step": 147
    },
    {
      "epoch": 0.0296,
      "grad_norm": 0.42633146047592163,
      "learning_rate": 5.88e-05,
      "loss": 2.3643,
      "step": 148
    },
    {
      "epoch": 0.0298,
      "grad_norm": 0.26536187529563904,
      "learning_rate": 5.92e-05,
      "loss": 2.3629,
      "step": 149
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.0865107923746109,
      "learning_rate": 5.96e-05,
      "loss": 2.3587,
      "step": 150
    },
    {
      "epoch": 0.0302,
      "grad_norm": 0.5003491044044495,
      "learning_rate": 6e-05,
      "loss": 2.3626,
      "step": 151
    },
    {
      "epoch": 0.0304,
      "grad_norm": 0.3530963361263275,
      "learning_rate": 6.04e-05,
      "loss": 2.3605,
      "step": 152
    },
    {
      "epoch": 0.0306,
      "grad_norm": 0.2627134621143341,
      "learning_rate": 6.08e-05,
      "loss": 2.3562,
      "step": 153
    },
    {
      "epoch": 0.0308,
      "grad_norm": 0.40086090564727783,
      "learning_rate": 6.12e-05,
      "loss": 2.3546,
      "step": 154
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.4638344943523407,
      "learning_rate": 6.16e-05,
      "loss": 2.3566,
      "step": 155
    },
    {
      "epoch": 0.0312,
      "grad_norm": 0.3598960340023041,
      "learning_rate": 6.2e-05,
      "loss": 2.3541,
      "step": 156
    },
    {
      "epoch": 0.0314,
      "grad_norm": 0.5982425808906555,
      "learning_rate": 6.24e-05,
      "loss": 2.3547,
      "step": 157
    },
    {
      "epoch": 0.0316,
      "grad_norm": 0.09935256093740463,
      "learning_rate": 6.280000000000001e-05,
      "loss": 2.3486,
      "step": 158
    },
    {
      "epoch": 0.0318,
      "grad_norm": 0.4716227352619171,
      "learning_rate": 6.32e-05,
      "loss": 2.3486,
      "step": 159
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.37303298711776733,
      "learning_rate": 6.36e-05,
      "loss": 2.3466,
      "step": 160
    },
    {
      "epoch": 0.0322,
      "grad_norm": 0.24282042682170868,
      "learning_rate": 6.400000000000001e-05,
      "loss": 2.3455,
      "step": 161
    },
    {
      "epoch": 0.0324,
      "grad_norm": 0.46144938468933105,
      "learning_rate": 6.440000000000001e-05,
      "loss": 2.3467,
      "step": 162
    },
    {
      "epoch": 0.0326,
      "grad_norm": 0.12032832205295563,
      "learning_rate": 6.48e-05,
      "loss": 2.3432,
      "step": 163
    },
    {
      "epoch": 0.0328,
      "grad_norm": 0.47953519225120544,
      "learning_rate": 6.52e-05,
      "loss": 2.3422,
      "step": 164
    },
    {
      "epoch": 0.033,
      "grad_norm": 0.270146906375885,
      "learning_rate": 6.560000000000001e-05,
      "loss": 2.3407,
      "step": 165
    },
    {
      "epoch": 0.0332,
      "grad_norm": 0.10707657039165497,
      "learning_rate": 6.6e-05,
      "loss": 2.3422,
      "step": 166
    },
    {
      "epoch": 0.0334,
      "grad_norm": 0.412306547164917,
      "learning_rate": 6.64e-05,
      "loss": 2.3425,
      "step": 167
    },
    {
      "epoch": 0.0336,
      "grad_norm": 0.09122971445322037,
      "learning_rate": 6.680000000000001e-05,
      "loss": 2.3409,
      "step": 168
    },
    {
      "epoch": 0.0338,
      "grad_norm": 0.2821985185146332,
      "learning_rate": 6.720000000000001e-05,
      "loss": 2.3414,
      "step": 169
    },
    {
      "epoch": 0.034,
      "grad_norm": 0.28761163353919983,
      "learning_rate": 6.76e-05,
      "loss": 2.3394,
      "step": 170
    },
    {
      "epoch": 0.0342,
      "grad_norm": 0.08848586678504944,
      "learning_rate": 6.800000000000001e-05,
      "loss": 2.3382,
      "step": 171
    },
    {
      "epoch": 0.0344,
      "grad_norm": 0.2998575270175934,
      "learning_rate": 6.840000000000001e-05,
      "loss": 2.3375,
      "step": 172
    },
    {
      "epoch": 0.0346,
      "grad_norm": 0.1200309544801712,
      "learning_rate": 6.879999999999999e-05,
      "loss": 2.337,
      "step": 173
    },
    {
      "epoch": 0.0348,
      "grad_norm": 0.2690351605415344,
      "learning_rate": 6.92e-05,
      "loss": 2.3349,
      "step": 174
    },
    {
      "epoch": 0.035,
      "grad_norm": 0.14040672779083252,
      "learning_rate": 6.96e-05,
      "loss": 2.3319,
      "step": 175
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.07546723634004593,
      "learning_rate": 7e-05,
      "loss": 2.3316,
      "step": 176
    },
    {
      "epoch": 0.0354,
      "grad_norm": 0.11018352955579758,
      "learning_rate": 7.04e-05,
      "loss": 2.3318,
      "step": 177
    },
    {
      "epoch": 0.0356,
      "grad_norm": 0.1324177235364914,
      "learning_rate": 7.08e-05,
      "loss": 2.3307,
      "step": 178
    },
    {
      "epoch": 0.0358,
      "grad_norm": 0.08276952058076859,
      "learning_rate": 7.12e-05,
      "loss": 2.3297,
      "step": 179
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.0959027037024498,
      "learning_rate": 7.16e-05,
      "loss": 2.3283,
      "step": 180
    },
    {
      "epoch": 0.0362,
      "grad_norm": 0.18042200803756714,
      "learning_rate": 7.2e-05,
      "loss": 2.3289,
      "step": 181
    },
    {
      "epoch": 0.0364,
      "grad_norm": 0.06525571644306183,
      "learning_rate": 7.24e-05,
      "loss": 2.3264,
      "step": 182
    },
    {
      "epoch": 0.0366,
      "grad_norm": 0.14458802342414856,
      "learning_rate": 7.280000000000001e-05,
      "loss": 2.3263,
      "step": 183
    },
    {
      "epoch": 0.0368,
      "grad_norm": 0.0720377042889595,
      "learning_rate": 7.32e-05,
      "loss": 2.3253,
      "step": 184
    },
    {
      "epoch": 0.037,
      "grad_norm": 0.1172044649720192,
      "learning_rate": 7.36e-05,
      "loss": 2.3244,
      "step": 185
    },
    {
      "epoch": 0.0372,
      "grad_norm": 0.09052009880542755,
      "learning_rate": 7.4e-05,
      "loss": 2.3232,
      "step": 186
    },
    {
      "epoch": 0.0374,
      "grad_norm": 0.06917941570281982,
      "learning_rate": 7.44e-05,
      "loss": 2.3217,
      "step": 187
    },
    {
      "epoch": 0.0376,
      "grad_norm": 0.06567921489477158,
      "learning_rate": 7.48e-05,
      "loss": 2.3226,
      "step": 188
    },
    {
      "epoch": 0.0378,
      "grad_norm": 0.0569104366004467,
      "learning_rate": 7.52e-05,
      "loss": 2.3218,
      "step": 189
    },
    {
      "epoch": 0.038,
      "grad_norm": 0.10261756181716919,
      "learning_rate": 7.560000000000001e-05,
      "loss": 2.3218,
      "step": 190
    },
    {
      "epoch": 0.0382,
      "grad_norm": 0.08057704567909241,
      "learning_rate": 7.6e-05,
      "loss": 2.3205,
      "step": 191
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.12401796132326126,
      "learning_rate": 7.64e-05,
      "loss": 2.3201,
      "step": 192
    },
    {
      "epoch": 0.0386,
      "grad_norm": 0.061899054795503616,
      "learning_rate": 7.680000000000001e-05,
      "loss": 2.3193,
      "step": 193
    },
    {
      "epoch": 0.0388,
      "grad_norm": 0.09013772010803223,
      "learning_rate": 7.72e-05,
      "loss": 2.3189,
      "step": 194
    },
    {
      "epoch": 0.039,
      "grad_norm": 0.07368515431880951,
      "learning_rate": 7.76e-05,
      "loss": 2.3171,
      "step": 195
    },
    {
      "epoch": 0.0392,
      "grad_norm": 0.10526822507381439,
      "learning_rate": 7.800000000000001e-05,
      "loss": 2.3162,
      "step": 196
    },
    {
      "epoch": 0.0394,
      "grad_norm": 0.122613325715065,
      "learning_rate": 7.840000000000001e-05,
      "loss": 2.3161,
      "step": 197
    },
    {
      "epoch": 0.0396,
      "grad_norm": 0.07493767887353897,
      "learning_rate": 7.88e-05,
      "loss": 2.3163,
      "step": 198
    },
    {
      "epoch": 0.0398,
      "grad_norm": 0.08528421074151993,
      "learning_rate": 7.920000000000001e-05,
      "loss": 2.3169,
      "step": 199
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.09068898111581802,
      "learning_rate": 7.960000000000001e-05,
      "loss": 2.3155,
      "step": 200
    },
    {
      "epoch": 0.0402,
      "grad_norm": 0.08893673866987228,
      "learning_rate": 8e-05,
      "loss": 2.3147,
      "step": 201
    },
    {
      "epoch": 0.0404,
      "grad_norm": 0.1164683923125267,
      "learning_rate": 8.04e-05,
      "loss": 2.315,
      "step": 202
    },
    {
      "epoch": 0.0406,
      "grad_norm": 0.0811949223279953,
      "learning_rate": 8.080000000000001e-05,
      "loss": 2.3136,
      "step": 203
    },
    {
      "epoch": 0.0408,
      "grad_norm": 0.18147842586040497,
      "learning_rate": 8.120000000000001e-05,
      "loss": 2.3126,
      "step": 204
    },
    {
      "epoch": 0.041,
      "grad_norm": 0.07677672058343887,
      "learning_rate": 8.16e-05,
      "loss": 2.3122,
      "step": 205
    },
    {
      "epoch": 0.0412,
      "grad_norm": 0.2573532462120056,
      "learning_rate": 8.2e-05,
      "loss": 2.3121,
      "step": 206
    },
    {
      "epoch": 0.0414,
      "grad_norm": 0.09082259237766266,
      "learning_rate": 8.24e-05,
      "loss": 2.3112,
      "step": 207
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.2877911627292633,
      "learning_rate": 8.28e-05,
      "loss": 2.311,
      "step": 208
    },
    {
      "epoch": 0.0418,
      "grad_norm": 0.09990523755550385,
      "learning_rate": 8.32e-05,
      "loss": 2.31,
      "step": 209
    },
    {
      "epoch": 0.042,
      "grad_norm": 0.2876688838005066,
      "learning_rate": 8.36e-05,
      "loss": 2.3108,
      "step": 210
    },
    {
      "epoch": 0.0422,
      "grad_norm": 0.06088617816567421,
      "learning_rate": 8.4e-05,
      "loss": 2.3093,
      "step": 211
    },
    {
      "epoch": 0.0424,
      "grad_norm": 0.30590614676475525,
      "learning_rate": 8.44e-05,
      "loss": 2.3104,
      "step": 212
    },
    {
      "epoch": 0.0426,
      "grad_norm": 0.08738838136196136,
      "learning_rate": 8.48e-05,
      "loss": 2.3083,
      "step": 213
    },
    {
      "epoch": 0.0428,
      "grad_norm": 0.2884235680103302,
      "learning_rate": 8.52e-05,
      "loss": 2.3085,
      "step": 214
    },
    {
      "epoch": 0.043,
      "grad_norm": 0.1016777753829956,
      "learning_rate": 8.560000000000001e-05,
      "loss": 2.3075,
      "step": 215
    },
    {
      "epoch": 0.0432,
      "grad_norm": 0.4281441569328308,
      "learning_rate": 8.6e-05,
      "loss": 2.3091,
      "step": 216
    },
    {
      "epoch": 0.0434,
      "grad_norm": 0.22431500256061554,
      "learning_rate": 8.64e-05,
      "loss": 2.306,
      "step": 217
    },
    {
      "epoch": 0.0436,
      "grad_norm": 0.35231178998947144,
      "learning_rate": 8.680000000000001e-05,
      "loss": 2.3065,
      "step": 218
    },
    {
      "epoch": 0.0438,
      "grad_norm": 0.19906333088874817,
      "learning_rate": 8.72e-05,
      "loss": 2.3063,
      "step": 219
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.4406142234802246,
      "learning_rate": 8.76e-05,
      "loss": 2.3075,
      "step": 220
    },
    {
      "epoch": 0.0442,
      "grad_norm": 0.2965104579925537,
      "learning_rate": 8.800000000000001e-05,
      "loss": 2.3055,
      "step": 221
    },
    {
      "epoch": 0.0444,
      "grad_norm": 0.2796325385570526,
      "learning_rate": 8.840000000000001e-05,
      "loss": 2.3046,
      "step": 222
    },
    {
      "epoch": 0.0446,
      "grad_norm": 0.3280910551548004,
      "learning_rate": 8.88e-05,
      "loss": 2.3042,
      "step": 223
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.22922825813293457,
      "learning_rate": 8.92e-05,
      "loss": 2.3035,
      "step": 224
    },
    {
      "epoch": 0.045,
      "grad_norm": 0.36704570055007935,
      "learning_rate": 8.960000000000001e-05,
      "loss": 2.3038,
      "step": 225
    },
    {
      "epoch": 0.0452,
      "grad_norm": 0.26988229155540466,
      "learning_rate": 9e-05,
      "loss": 2.3029,
      "step": 226
    },
    {
      "epoch": 0.0454,
      "grad_norm": 0.7730883359909058,
      "learning_rate": 9.04e-05,
      "loss": 2.3048,
      "step": 227
    },
    {
      "epoch": 0.0456,
      "grad_norm": 0.19631028175354004,
      "learning_rate": 9.080000000000001e-05,
      "loss": 2.3014,
      "step": 228
    },
    {
      "epoch": 0.0458,
      "grad_norm": 0.46636050939559937,
      "learning_rate": 9.120000000000001e-05,
      "loss": 2.3029,
      "step": 229
    },
    {
      "epoch": 0.046,
      "grad_norm": 0.2230181246995926,
      "learning_rate": 9.16e-05,
      "loss": 2.3014,
      "step": 230
    },
    {
      "epoch": 0.0462,
      "grad_norm": 0.3283796012401581,
      "learning_rate": 9.200000000000001e-05,
      "loss": 2.3008,
      "step": 231
    },
    {
      "epoch": 0.0464,
      "grad_norm": 0.26169705390930176,
      "learning_rate": 9.240000000000001e-05,
      "loss": 2.3005,
      "step": 232
    },
    {
      "epoch": 0.0466,
      "grad_norm": 0.06816647946834564,
      "learning_rate": 9.28e-05,
      "loss": 2.2999,
      "step": 233
    },
    {
      "epoch": 0.0468,
      "grad_norm": 0.060220878571271896,
      "learning_rate": 9.320000000000002e-05,
      "loss": 2.3003,
      "step": 234
    },
    {
      "epoch": 0.047,
      "grad_norm": 0.18859657645225525,
      "learning_rate": 9.360000000000001e-05,
      "loss": 2.2994,
      "step": 235
    },
    {
      "epoch": 0.0472,
      "grad_norm": 0.09130974858999252,
      "learning_rate": 9.4e-05,
      "loss": 2.2991,
      "step": 236
    },
    {
      "epoch": 0.0474,
      "grad_norm": 0.20326317846775055,
      "learning_rate": 9.44e-05,
      "loss": 2.2985,
      "step": 237
    },
    {
      "epoch": 0.0476,
      "grad_norm": 0.2140786349773407,
      "learning_rate": 9.48e-05,
      "loss": 2.2986,
      "step": 238
    },
    {
      "epoch": 0.0478,
      "grad_norm": 0.143418088555336,
      "learning_rate": 9.52e-05,
      "loss": 2.2983,
      "step": 239
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.2983095347881317,
      "learning_rate": 9.56e-05,
      "loss": 2.2984,
      "step": 240
    },
    {
      "epoch": 0.0482,
      "grad_norm": 0.08695440739393234,
      "learning_rate": 9.6e-05,
      "loss": 2.2978,
      "step": 241
    },
    {
      "epoch": 0.0484,
      "grad_norm": 0.25095242261886597,
      "learning_rate": 9.64e-05,
      "loss": 2.2964,
      "step": 242
    },
    {
      "epoch": 0.0486,
      "grad_norm": 0.18982312083244324,
      "learning_rate": 9.680000000000001e-05,
      "loss": 2.2966,
      "step": 243
    },
    {
      "epoch": 0.0488,
      "grad_norm": 0.12339793145656586,
      "learning_rate": 9.72e-05,
      "loss": 2.2962,
      "step": 244
    },
    {
      "epoch": 0.049,
      "grad_norm": 0.23478183150291443,
      "learning_rate": 9.76e-05,
      "loss": 2.2958,
      "step": 245
    },
    {
      "epoch": 0.0492,
      "grad_norm": 0.07955659925937653,
      "learning_rate": 9.8e-05,
      "loss": 2.2968,
      "step": 246
    },
    {
      "epoch": 0.0494,
      "grad_norm": 0.1298326700925827,
      "learning_rate": 9.84e-05,
      "loss": 2.2954,
      "step": 247
    },
    {
      "epoch": 0.0496,
      "grad_norm": 0.220513716340065,
      "learning_rate": 9.88e-05,
      "loss": 2.2962,
      "step": 248
    },
    {
      "epoch": 0.0498,
      "grad_norm": 0.048263467848300934,
      "learning_rate": 9.92e-05,
      "loss": 2.2943,
      "step": 249
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.197544664144516,
      "learning_rate": 9.960000000000001e-05,
      "loss": 2.2952,
      "step": 250
    },
    {
      "epoch": 0.0502,
      "grad_norm": 0.21965883672237396,
      "learning_rate": 0.0001,
      "loss": 2.295,
      "step": 251
    },
    {
      "epoch": 0.0504,
      "grad_norm": 0.03777375817298889,
      "learning_rate": 0.0001004,
      "loss": 2.2941,
      "step": 252
    },
    {
      "epoch": 0.0506,
      "grad_norm": 0.2192188948392868,
      "learning_rate": 0.00010080000000000001,
      "loss": 2.2935,
      "step": 253
    },
    {
      "epoch": 0.0508,
      "grad_norm": 0.22267603874206543,
      "learning_rate": 0.00010120000000000001,
      "loss": 2.2937,
      "step": 254
    },
    {
      "epoch": 0.051,
      "grad_norm": 0.038216084241867065,
      "learning_rate": 0.0001016,
      "loss": 2.293,
      "step": 255
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.09798715263605118,
      "learning_rate": 0.00010200000000000001,
      "loss": 2.2928,
      "step": 256
    },
    {
      "epoch": 0.0514,
      "grad_norm": 0.06784392893314362,
      "learning_rate": 0.00010240000000000001,
      "loss": 2.2923,
      "step": 257
    },
    {
      "epoch": 0.0516,
      "grad_norm": 0.03425344452261925,
      "learning_rate": 0.0001028,
      "loss": 2.292,
      "step": 258
    },
    {
      "epoch": 0.0518,
      "grad_norm": 0.0423613004386425,
      "learning_rate": 0.0001032,
      "loss": 2.2922,
      "step": 259
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.03666224703192711,
      "learning_rate": 0.00010360000000000001,
      "loss": 2.2912,
      "step": 260
    },
    {
      "epoch": 0.0522,
      "grad_norm": 0.039242714643478394,
      "learning_rate": 0.00010400000000000001,
      "loss": 2.2915,
      "step": 261
    },
    {
      "epoch": 0.0524,
      "grad_norm": 0.16176769137382507,
      "learning_rate": 0.0001044,
      "loss": 2.2912,
      "step": 262
    },
    {
      "epoch": 0.0526,
      "grad_norm": 0.2723488211631775,
      "learning_rate": 0.00010480000000000001,
      "loss": 2.2912,
      "step": 263
    },
    {
      "epoch": 0.0528,
      "grad_norm": 0.24308478832244873,
      "learning_rate": 0.00010520000000000001,
      "loss": 2.2911,
      "step": 264
    },
    {
      "epoch": 0.053,
      "grad_norm": 0.05739400163292885,
      "learning_rate": 0.0001056,
      "loss": 2.2903,
      "step": 265
    },
    {
      "epoch": 0.0532,
      "grad_norm": 0.16673007607460022,
      "learning_rate": 0.00010600000000000002,
      "loss": 2.2905,
      "step": 266
    },
    {
      "epoch": 0.0534,
      "grad_norm": 0.1820259988307953,
      "learning_rate": 0.00010640000000000001,
      "loss": 2.2911,
      "step": 267
    },
    {
      "epoch": 0.0536,
      "grad_norm": 0.04189649969339371,
      "learning_rate": 0.00010680000000000001,
      "loss": 2.2903,
      "step": 268
    },
    {
      "epoch": 0.0538,
      "grad_norm": 0.1938054859638214,
      "learning_rate": 0.00010720000000000002,
      "loss": 2.2898,
      "step": 269
    },
    {
      "epoch": 0.054,
      "grad_norm": 0.08549925684928894,
      "learning_rate": 0.00010760000000000001,
      "loss": 2.2899,
      "step": 270
    },
    {
      "epoch": 0.0542,
      "grad_norm": 0.326080322265625,
      "learning_rate": 0.00010800000000000001,
      "loss": 2.2898,
      "step": 271
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.511880099773407,
      "learning_rate": 0.00010840000000000002,
      "loss": 2.2909,
      "step": 272
    },
    {
      "epoch": 0.0546,
      "grad_norm": 0.2264321744441986,
      "learning_rate": 0.00010880000000000002,
      "loss": 2.2897,
      "step": 273
    },
    {
      "epoch": 0.0548,
      "grad_norm": 0.18147310614585876,
      "learning_rate": 0.00010920000000000001,
      "loss": 2.2898,
      "step": 274
    },
    {
      "epoch": 0.055,
      "grad_norm": 0.24681095778942108,
      "learning_rate": 0.00010960000000000001,
      "loss": 2.2891,
      "step": 275
    },
    {
      "epoch": 0.0552,
      "grad_norm": 0.03224201500415802,
      "learning_rate": 0.00011000000000000002,
      "loss": 2.288,
      "step": 276
    },
    {
      "epoch": 0.0554,
      "grad_norm": 0.3088394105434418,
      "learning_rate": 0.00011040000000000001,
      "loss": 2.2885,
      "step": 277
    },
    {
      "epoch": 0.0556,
      "grad_norm": 0.31912025809288025,
      "learning_rate": 0.00011080000000000001,
      "loss": 2.2886,
      "step": 278
    },
    {
      "epoch": 0.0558,
      "grad_norm": 0.0444965735077858,
      "learning_rate": 0.00011120000000000002,
      "loss": 2.2885,
      "step": 279
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.43468108773231506,
      "learning_rate": 0.00011160000000000002,
      "loss": 2.2888,
      "step": 280
    },
    {
      "epoch": 0.0562,
      "grad_norm": 0.46987664699554443,
      "learning_rate": 0.00011200000000000001,
      "loss": 2.2891,
      "step": 281
    },
    {
      "epoch": 0.0564,
      "grad_norm": 0.06332080811262131,
      "learning_rate": 0.00011240000000000002,
      "loss": 2.2875,
      "step": 282
    },
    {
      "epoch": 0.0566,
      "grad_norm": 0.3977407217025757,
      "learning_rate": 0.00011279999999999999,
      "loss": 2.2885,
      "step": 283
    },
    {
      "epoch": 0.0568,
      "grad_norm": 0.36916759610176086,
      "learning_rate": 0.0001132,
      "loss": 2.2874,
      "step": 284
    },
    {
      "epoch": 0.057,
      "grad_norm": 0.06653298437595367,
      "learning_rate": 0.0001136,
      "loss": 2.2869,
      "step": 285
    },
    {
      "epoch": 0.0572,
      "grad_norm": 0.3812803626060486,
      "learning_rate": 0.00011399999999999999,
      "loss": 2.2879,
      "step": 286
    },
    {
      "epoch": 0.0574,
      "grad_norm": 0.2889416217803955,
      "learning_rate": 0.0001144,
      "loss": 2.2872,
      "step": 287
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.06423268467187881,
      "learning_rate": 0.0001148,
      "loss": 2.2864,
      "step": 288
    },
    {
      "epoch": 0.0578,
      "grad_norm": 0.3131747841835022,
      "learning_rate": 0.0001152,
      "loss": 2.2873,
      "step": 289
    },
    {
      "epoch": 0.058,
      "grad_norm": 0.2854478657245636,
      "learning_rate": 0.00011559999999999999,
      "loss": 2.2871,
      "step": 290
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.060094915330410004,
      "learning_rate": 0.000116,
      "loss": 2.2864,
      "step": 291
    },
    {
      "epoch": 0.0584,
      "grad_norm": 0.15963460505008698,
      "learning_rate": 0.0001164,
      "loss": 2.2869,
      "step": 292
    },
    {
      "epoch": 0.0586,
      "grad_norm": 0.18786492943763733,
      "learning_rate": 0.00011679999999999999,
      "loss": 2.2868,
      "step": 293
    },
    {
      "epoch": 0.0588,
      "grad_norm": 0.1825259029865265,
      "learning_rate": 0.0001172,
      "loss": 2.2862,
      "step": 294
    },
    {
      "epoch": 0.059,
      "grad_norm": 0.14359527826309204,
      "learning_rate": 0.0001176,
      "loss": 2.2864,
      "step": 295
    },
    {
      "epoch": 0.0592,
      "grad_norm": 0.046451471745967865,
      "learning_rate": 0.000118,
      "loss": 2.2859,
      "step": 296
    },
    {
      "epoch": 0.0594,
      "grad_norm": 0.20154978334903717,
      "learning_rate": 0.0001184,
      "loss": 2.2865,
      "step": 297
    },
    {
      "epoch": 0.0596,
      "grad_norm": 0.16975022852420807,
      "learning_rate": 0.0001188,
      "loss": 2.286,
      "step": 298
    },
    {
      "epoch": 0.0598,
      "grad_norm": 0.032744865864515305,
      "learning_rate": 0.0001192,
      "loss": 2.286,
      "step": 299
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.08945978432893753,
      "learning_rate": 0.00011960000000000001,
      "loss": 2.2858,
      "step": 300
    },
    {
      "epoch": 0.0602,
      "grad_norm": 0.058683257550001144,
      "learning_rate": 0.00012,
      "loss": 2.2857,
      "step": 301
    },
    {
      "epoch": 0.0604,
      "grad_norm": 0.08702412992715836,
      "learning_rate": 0.0001204,
      "loss": 2.2856,
      "step": 302
    },
    {
      "epoch": 0.0606,
      "grad_norm": 0.23545832931995392,
      "learning_rate": 0.0001208,
      "loss": 2.2855,
      "step": 303
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.2892269194126129,
      "learning_rate": 0.0001212,
      "loss": 2.2857,
      "step": 304
    },
    {
      "epoch": 0.061,
      "grad_norm": 0.1562483012676239,
      "learning_rate": 0.0001216,
      "loss": 2.2857,
      "step": 305
    },
    {
      "epoch": 0.0612,
      "grad_norm": 0.16105563938617706,
      "learning_rate": 0.000122,
      "loss": 2.2852,
      "step": 306
    },
    {
      "epoch": 0.0614,
      "grad_norm": 0.3446987569332123,
      "learning_rate": 0.0001224,
      "loss": 2.2858,
      "step": 307
    },
    {
      "epoch": 0.0616,
      "grad_norm": 0.16802255809307098,
      "learning_rate": 0.0001228,
      "loss": 2.285,
      "step": 308
    },
    {
      "epoch": 0.0618,
      "grad_norm": 0.173993781208992,
      "learning_rate": 0.0001232,
      "loss": 2.2853,
      "step": 309
    },
    {
      "epoch": 0.062,
      "grad_norm": 0.47523239254951477,
      "learning_rate": 0.0001236,
      "loss": 2.2865,
      "step": 310
    },
    {
      "epoch": 0.0622,
      "grad_norm": 0.609849214553833,
      "learning_rate": 0.000124,
      "loss": 2.2859,
      "step": 311
    },
    {
      "epoch": 0.0624,
      "grad_norm": 0.3749130666255951,
      "learning_rate": 0.00012440000000000002,
      "loss": 2.2853,
      "step": 312
    },
    {
      "epoch": 0.0626,
      "grad_norm": 0.13642680644989014,
      "learning_rate": 0.0001248,
      "loss": 2.2848,
      "step": 313
    },
    {
      "epoch": 0.0628,
      "grad_norm": 0.511991024017334,
      "learning_rate": 0.0001252,
      "loss": 2.286,
      "step": 314
    },
    {
      "epoch": 0.063,
      "grad_norm": 0.40094050765037537,
      "learning_rate": 0.00012560000000000002,
      "loss": 2.2854,
      "step": 315
    },
    {
      "epoch": 0.0632,
      "grad_norm": 0.06968576461076736,
      "learning_rate": 0.000126,
      "loss": 2.2842,
      "step": 316
    },
    {
      "epoch": 0.0634,
      "grad_norm": 0.388194739818573,
      "learning_rate": 0.0001264,
      "loss": 2.2853,
      "step": 317
    },
    {
      "epoch": 0.0636,
      "grad_norm": 0.33618077635765076,
      "learning_rate": 0.00012680000000000002,
      "loss": 2.2849,
      "step": 318
    },
    {
      "epoch": 0.0638,
      "grad_norm": 0.043540507555007935,
      "learning_rate": 0.0001272,
      "loss": 2.2842,
      "step": 319
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.2992262840270996,
      "learning_rate": 0.0001276,
      "loss": 2.2833,
      "step": 320
    },
    {
      "epoch": 0.0642,
      "grad_norm": 0.42918068170547485,
      "learning_rate": 0.00012800000000000002,
      "loss": 2.2851,
      "step": 321
    },
    {
      "epoch": 0.0644,
      "grad_norm": 0.23817582428455353,
      "learning_rate": 0.0001284,
      "loss": 2.2843,
      "step": 322
    },
    {
      "epoch": 0.0646,
      "grad_norm": 0.17172348499298096,
      "learning_rate": 0.00012880000000000001,
      "loss": 2.2841,
      "step": 323
    },
    {
      "epoch": 0.0648,
      "grad_norm": 0.48059627413749695,
      "learning_rate": 0.00012920000000000002,
      "loss": 2.2855,
      "step": 324
    },
    {
      "epoch": 0.065,
      "grad_norm": 0.3884505331516266,
      "learning_rate": 0.0001296,
      "loss": 2.2846,
      "step": 325
    },
    {
      "epoch": 0.0652,
      "grad_norm": 0.07214824110269547,
      "learning_rate": 0.00013000000000000002,
      "loss": 2.2837,
      "step": 326
    },
    {
      "epoch": 0.0654,
      "grad_norm": 0.4373411536216736,
      "learning_rate": 0.0001304,
      "loss": 2.285,
      "step": 327
    },
    {
      "epoch": 0.0656,
      "grad_norm": 0.4188523590564728,
      "learning_rate": 0.0001308,
      "loss": 2.2846,
      "step": 328
    },
    {
      "epoch": 0.0658,
      "grad_norm": 0.1420585811138153,
      "learning_rate": 0.00013120000000000002,
      "loss": 2.2838,
      "step": 329
    },
    {
      "epoch": 0.066,
      "grad_norm": 0.15714919567108154,
      "learning_rate": 0.0001316,
      "loss": 2.2835,
      "step": 330
    },
    {
      "epoch": 0.0662,
      "grad_norm": 0.23266355693340302,
      "learning_rate": 0.000132,
      "loss": 2.2843,
      "step": 331
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.035295356065034866,
      "learning_rate": 0.00013240000000000002,
      "loss": 2.2838,
      "step": 332
    },
    {
      "epoch": 0.0666,
      "grad_norm": 0.2461249679327011,
      "learning_rate": 0.0001328,
      "loss": 2.2837,
      "step": 333
    },
    {
      "epoch": 0.0668,
      "grad_norm": 0.37101948261260986,
      "learning_rate": 0.0001332,
      "loss": 2.2837,
      "step": 334
    },
    {
      "epoch": 0.067,
      "grad_norm": 0.26202964782714844,
      "learning_rate": 0.00013360000000000002,
      "loss": 2.2834,
      "step": 335
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.04416802152991295,
      "learning_rate": 0.000134,
      "loss": 2.2835,
      "step": 336
    },
    {
      "epoch": 0.0674,
      "grad_norm": 0.282391220331192,
      "learning_rate": 0.00013440000000000001,
      "loss": 2.2838,
      "step": 337
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.19795246422290802,
      "learning_rate": 0.00013480000000000002,
      "loss": 2.2834,
      "step": 338
    },
    {
      "epoch": 0.0678,
      "grad_norm": 0.0777612179517746,
      "learning_rate": 0.0001352,
      "loss": 2.2831,
      "step": 339
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.21641649305820465,
      "learning_rate": 0.00013560000000000002,
      "loss": 2.2829,
      "step": 340
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.14512544870376587,
      "learning_rate": 0.00013600000000000003,
      "loss": 2.283,
      "step": 341
    },
    {
      "epoch": 0.0684,
      "grad_norm": 0.028863878920674324,
      "learning_rate": 0.0001364,
      "loss": 2.283,
      "step": 342
    },
    {
      "epoch": 0.0686,
      "grad_norm": 0.12887701392173767,
      "learning_rate": 0.00013680000000000002,
      "loss": 2.2834,
      "step": 343
    },
    {
      "epoch": 0.0688,
      "grad_norm": 0.12595301866531372,
      "learning_rate": 0.00013720000000000003,
      "loss": 2.2833,
      "step": 344
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.060561392456293106,
      "learning_rate": 0.00013759999999999998,
      "loss": 2.2831,
      "step": 345
    },
    {
      "epoch": 0.0692,
      "grad_norm": 0.01911124587059021,
      "learning_rate": 0.000138,
      "loss": 2.283,
      "step": 346
    },
    {
      "epoch": 0.0694,
      "grad_norm": 0.021568462252616882,
      "learning_rate": 0.0001384,
      "loss": 2.2826,
      "step": 347
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.019971834495663643,
      "learning_rate": 0.00013879999999999999,
      "loss": 2.2824,
      "step": 348
    },
    {
      "epoch": 0.0698,
      "grad_norm": 0.02116389386355877,
      "learning_rate": 0.0001392,
      "loss": 2.2825,
      "step": 349
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.031318120658397675,
      "learning_rate": 0.0001396,
      "loss": 2.2837,
      "step": 350
    },
    {
      "epoch": 0.0702,
      "grad_norm": 0.08550713956356049,
      "learning_rate": 0.00014,
      "loss": 2.2825,
      "step": 351
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.09932991117238998,
      "learning_rate": 0.0001404,
      "loss": 2.2827,
      "step": 352
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.06826907396316528,
      "learning_rate": 0.0001408,
      "loss": 2.2829,
      "step": 353
    },
    {
      "epoch": 0.0708,
      "grad_norm": 0.04434693604707718,
      "learning_rate": 0.0001412,
      "loss": 2.283,
      "step": 354
    },
    {
      "epoch": 0.071,
      "grad_norm": 0.06225087493658066,
      "learning_rate": 0.0001416,
      "loss": 2.2829,
      "step": 355
    },
    {
      "epoch": 0.0712,
      "grad_norm": 0.10398422181606293,
      "learning_rate": 0.000142,
      "loss": 2.2828,
      "step": 356
    },
    {
      "epoch": 0.0714,
      "grad_norm": 0.06721615046262741,
      "learning_rate": 0.0001424,
      "loss": 2.2827,
      "step": 357
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.022309327498078346,
      "learning_rate": 0.0001428,
      "loss": 2.2826,
      "step": 358
    },
    {
      "epoch": 0.0718,
      "grad_norm": 0.031445086002349854,
      "learning_rate": 0.0001432,
      "loss": 2.2829,
      "step": 359
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.024583250284194946,
      "learning_rate": 0.0001436,
      "loss": 2.2829,
      "step": 360
    },
    {
      "epoch": 0.0722,
      "grad_norm": 0.03366629034280777,
      "learning_rate": 0.000144,
      "loss": 2.2828,
      "step": 361
    },
    {
      "epoch": 0.0724,
      "grad_norm": 0.0188742708414793,
      "learning_rate": 0.0001444,
      "loss": 2.2824,
      "step": 362
    },
    {
      "epoch": 0.0726,
      "grad_norm": 0.04724787920713425,
      "learning_rate": 0.0001448,
      "loss": 2.2829,
      "step": 363
    },
    {
      "epoch": 0.0728,
      "grad_norm": 0.017444439232349396,
      "learning_rate": 0.0001452,
      "loss": 2.2821,
      "step": 364
    },
    {
      "epoch": 0.073,
      "grad_norm": 0.08171804249286652,
      "learning_rate": 0.00014560000000000002,
      "loss": 2.2826,
      "step": 365
    },
    {
      "epoch": 0.0732,
      "grad_norm": 0.1425766795873642,
      "learning_rate": 0.000146,
      "loss": 2.2828,
      "step": 366
    },
    {
      "epoch": 0.0734,
      "grad_norm": 0.1702090948820114,
      "learning_rate": 0.0001464,
      "loss": 2.2821,
      "step": 367
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.12447052448987961,
      "learning_rate": 0.00014680000000000002,
      "loss": 2.2824,
      "step": 368
    },
    {
      "epoch": 0.0738,
      "grad_norm": 0.06968312710523605,
      "learning_rate": 0.0001472,
      "loss": 2.2845,
      "step": 369
    },
    {
      "epoch": 0.074,
      "grad_norm": 0.016947397962212563,
      "learning_rate": 0.0001476,
      "loss": 2.2826,
      "step": 370
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.01601753756403923,
      "learning_rate": 0.000148,
      "loss": 2.2828,
      "step": 371
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.032339781522750854,
      "learning_rate": 0.0001484,
      "loss": 2.2818,
      "step": 372
    },
    {
      "epoch": 0.0746,
      "grad_norm": 0.04929586127400398,
      "learning_rate": 0.0001488,
      "loss": 2.2817,
      "step": 373
    },
    {
      "epoch": 0.0748,
      "grad_norm": 0.03784392774105072,
      "learning_rate": 0.0001492,
      "loss": 2.2815,
      "step": 374
    },
    {
      "epoch": 0.075,
      "grad_norm": 0.019078148528933525,
      "learning_rate": 0.0001496,
      "loss": 2.2822,
      "step": 375
    },
    {
      "epoch": 0.0752,
      "grad_norm": 0.023124821484088898,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.2817,
      "step": 376
    },
    {
      "epoch": 0.0754,
      "grad_norm": 0.07792982459068298,
      "learning_rate": 0.0001504,
      "loss": 2.282,
      "step": 377
    },
    {
      "epoch": 0.0756,
      "grad_norm": 0.17439039051532745,
      "learning_rate": 0.0001508,
      "loss": 2.2817,
      "step": 378
    },
    {
      "epoch": 0.0758,
      "grad_norm": 0.24744313955307007,
      "learning_rate": 0.00015120000000000002,
      "loss": 2.2816,
      "step": 379
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.2442767322063446,
      "learning_rate": 0.0001516,
      "loss": 2.2824,
      "step": 380
    },
    {
      "epoch": 0.0762,
      "grad_norm": 0.25204893946647644,
      "learning_rate": 0.000152,
      "loss": 2.2827,
      "step": 381
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.2619691789150238,
      "learning_rate": 0.00015240000000000002,
      "loss": 2.2825,
      "step": 382
    },
    {
      "epoch": 0.0766,
      "grad_norm": 0.2018255591392517,
      "learning_rate": 0.0001528,
      "loss": 2.2823,
      "step": 383
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.09385447204113007,
      "learning_rate": 0.0001532,
      "loss": 2.2818,
      "step": 384
    },
    {
      "epoch": 0.077,
      "grad_norm": 0.031563010066747665,
      "learning_rate": 0.00015360000000000002,
      "loss": 2.2822,
      "step": 385
    },
    {
      "epoch": 0.0772,
      "grad_norm": 0.14690834283828735,
      "learning_rate": 0.000154,
      "loss": 2.2818,
      "step": 386
    },
    {
      "epoch": 0.0774,
      "grad_norm": 0.29819247126579285,
      "learning_rate": 0.0001544,
      "loss": 2.2823,
      "step": 387
    },
    {
      "epoch": 0.0776,
      "grad_norm": 0.39595529437065125,
      "learning_rate": 0.00015480000000000002,
      "loss": 2.283,
      "step": 388
    },
    {
      "epoch": 0.0778,
      "grad_norm": 0.3982302248477936,
      "learning_rate": 0.0001552,
      "loss": 2.2826,
      "step": 389
    },
    {
      "epoch": 0.078,
      "grad_norm": 0.32614031434059143,
      "learning_rate": 0.00015560000000000001,
      "loss": 2.2816,
      "step": 390
    },
    {
      "epoch": 0.0782,
      "grad_norm": 0.1265558898448944,
      "learning_rate": 0.00015600000000000002,
      "loss": 2.2814,
      "step": 391
    },
    {
      "epoch": 0.0784,
      "grad_norm": 0.1388477236032486,
      "learning_rate": 0.0001564,
      "loss": 2.2819,
      "step": 392
    },
    {
      "epoch": 0.0786,
      "grad_norm": 0.3296907842159271,
      "learning_rate": 0.00015680000000000002,
      "loss": 2.2819,
      "step": 393
    },
    {
      "epoch": 0.0788,
      "grad_norm": 0.407669335603714,
      "learning_rate": 0.00015720000000000003,
      "loss": 2.2827,
      "step": 394
    },
    {
      "epoch": 0.079,
      "grad_norm": 0.35353323817253113,
      "learning_rate": 0.0001576,
      "loss": 2.282,
      "step": 395
    },
    {
      "epoch": 0.0792,
      "grad_norm": 0.21347156167030334,
      "learning_rate": 0.00015800000000000002,
      "loss": 2.2822,
      "step": 396
    },
    {
      "epoch": 0.0794,
      "grad_norm": 0.055684641003608704,
      "learning_rate": 0.00015840000000000003,
      "loss": 2.2818,
      "step": 397
    },
    {
      "epoch": 0.0796,
      "grad_norm": 0.0994354709982872,
      "learning_rate": 0.0001588,
      "loss": 2.2816,
      "step": 398
    },
    {
      "epoch": 0.0798,
      "grad_norm": 0.16275428235530853,
      "learning_rate": 0.00015920000000000002,
      "loss": 2.2822,
      "step": 399
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.10144132375717163,
      "learning_rate": 0.0001596,
      "loss": 2.2823,
      "step": 400
    },
    {
      "epoch": 0.0802,
      "grad_norm": 0.02480926737189293,
      "learning_rate": 0.00016,
      "loss": 2.2819,
      "step": 401
    },
    {
      "epoch": 0.0804,
      "grad_norm": 0.09799647331237793,
      "learning_rate": 0.00016040000000000002,
      "loss": 2.2819,
      "step": 402
    },
    {
      "epoch": 0.0806,
      "grad_norm": 0.11750014871358871,
      "learning_rate": 0.0001608,
      "loss": 2.2816,
      "step": 403
    },
    {
      "epoch": 0.0808,
      "grad_norm": 0.08890483528375626,
      "learning_rate": 0.00016120000000000002,
      "loss": 2.2822,
      "step": 404
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.047295525670051575,
      "learning_rate": 0.00016160000000000002,
      "loss": 2.2818,
      "step": 405
    },
    {
      "epoch": 0.0812,
      "grad_norm": 0.02024911716580391,
      "learning_rate": 0.000162,
      "loss": 2.2816,
      "step": 406
    },
    {
      "epoch": 0.0814,
      "grad_norm": 0.07361439615488052,
      "learning_rate": 0.00016240000000000002,
      "loss": 2.2818,
      "step": 407
    },
    {
      "epoch": 0.0816,
      "grad_norm": 0.1828252524137497,
      "learning_rate": 0.0001628,
      "loss": 2.2821,
      "step": 408
    },
    {
      "epoch": 0.0818,
      "grad_norm": 0.23529990017414093,
      "learning_rate": 0.0001632,
      "loss": 2.2823,
      "step": 409
    },
    {
      "epoch": 0.082,
      "grad_norm": 0.21864911913871765,
      "learning_rate": 0.0001636,
      "loss": 2.2823,
      "step": 410
    },
    {
      "epoch": 0.0822,
      "grad_norm": 0.1620609611272812,
      "learning_rate": 0.000164,
      "loss": 2.2816,
      "step": 411
    },
    {
      "epoch": 0.0824,
      "grad_norm": 0.05750500410795212,
      "learning_rate": 0.0001644,
      "loss": 2.2818,
      "step": 412
    },
    {
      "epoch": 0.0826,
      "grad_norm": 0.10753228515386581,
      "learning_rate": 0.0001648,
      "loss": 2.2824,
      "step": 413
    },
    {
      "epoch": 0.0828,
      "grad_norm": 0.30785852670669556,
      "learning_rate": 0.0001652,
      "loss": 2.2821,
      "step": 414
    },
    {
      "epoch": 0.083,
      "grad_norm": 0.49560871720314026,
      "learning_rate": 0.0001656,
      "loss": 2.2831,
      "step": 415
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.5682402849197388,
      "learning_rate": 0.000166,
      "loss": 2.2832,
      "step": 416
    },
    {
      "epoch": 0.0834,
      "grad_norm": 0.4686771631240845,
      "learning_rate": 0.0001664,
      "loss": 2.2829,
      "step": 417
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.2733679413795471,
      "learning_rate": 0.0001668,
      "loss": 2.2818,
      "step": 418
    },
    {
      "epoch": 0.0838,
      "grad_norm": 0.05353335291147232,
      "learning_rate": 0.0001672,
      "loss": 2.2814,
      "step": 419
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.17410078644752502,
      "learning_rate": 0.0001676,
      "loss": 2.2816,
      "step": 420
    },
    {
      "epoch": 0.0842,
      "grad_norm": 0.3676564395427704,
      "learning_rate": 0.000168,
      "loss": 2.2818,
      "step": 421
    },
    {
      "epoch": 0.0844,
      "grad_norm": 0.46087363362312317,
      "learning_rate": 0.0001684,
      "loss": 2.2828,
      "step": 422
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.40455594658851624,
      "learning_rate": 0.0001688,
      "loss": 2.2825,
      "step": 423
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.1831531524658203,
      "learning_rate": 0.0001692,
      "loss": 2.2815,
      "step": 424
    },
    {
      "epoch": 0.085,
      "grad_norm": 0.10726162046194077,
      "learning_rate": 0.0001696,
      "loss": 2.2813,
      "step": 425
    },
    {
      "epoch": 0.0852,
      "grad_norm": 0.30440735816955566,
      "learning_rate": 0.00017,
      "loss": 2.2821,
      "step": 426
    },
    {
      "epoch": 0.0854,
      "grad_norm": 0.39361372590065,
      "learning_rate": 0.0001704,
      "loss": 2.2822,
      "step": 427
    },
    {
      "epoch": 0.0856,
      "grad_norm": 0.3583301901817322,
      "learning_rate": 0.0001708,
      "loss": 2.2819,
      "step": 428
    },
    {
      "epoch": 0.0858,
      "grad_norm": 0.16054250299930573,
      "learning_rate": 0.00017120000000000001,
      "loss": 2.2814,
      "step": 429
    },
    {
      "epoch": 0.086,
      "grad_norm": 0.11523816734552383,
      "learning_rate": 0.0001716,
      "loss": 2.2811,
      "step": 430
    },
    {
      "epoch": 0.0862,
      "grad_norm": 0.3343912363052368,
      "learning_rate": 0.000172,
      "loss": 2.2813,
      "step": 431
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.4087778329849243,
      "learning_rate": 0.00017240000000000002,
      "loss": 2.2822,
      "step": 432
    },
    {
      "epoch": 0.0866,
      "grad_norm": 0.3660861849784851,
      "learning_rate": 0.0001728,
      "loss": 2.2822,
      "step": 433
    },
    {
      "epoch": 0.0868,
      "grad_norm": 0.24853339791297913,
      "learning_rate": 0.0001732,
      "loss": 2.2816,
      "step": 434
    },
    {
      "epoch": 0.087,
      "grad_norm": 0.07815829664468765,
      "learning_rate": 0.00017360000000000002,
      "loss": 2.2809,
      "step": 435
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.11082848906517029,
      "learning_rate": 0.000174,
      "loss": 2.2818,
      "step": 436
    },
    {
      "epoch": 0.0874,
      "grad_norm": 0.25432634353637695,
      "learning_rate": 0.0001744,
      "loss": 2.2815,
      "step": 437
    },
    {
      "epoch": 0.0876,
      "grad_norm": 0.3096747100353241,
      "learning_rate": 0.00017480000000000002,
      "loss": 2.2816,
      "step": 438
    },
    {
      "epoch": 0.0878,
      "grad_norm": 0.253441721200943,
      "learning_rate": 0.0001752,
      "loss": 2.2815,
      "step": 439
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.10816055536270142,
      "learning_rate": 0.0001756,
      "loss": 2.2809,
      "step": 440
    },
    {
      "epoch": 0.0882,
      "grad_norm": 0.059746723622083664,
      "learning_rate": 0.00017600000000000002,
      "loss": 2.2817,
      "step": 441
    },
    {
      "epoch": 0.0884,
      "grad_norm": 0.14908798038959503,
      "learning_rate": 0.0001764,
      "loss": 2.2814,
      "step": 442
    },
    {
      "epoch": 0.0886,
      "grad_norm": 0.15861541032791138,
      "learning_rate": 0.00017680000000000001,
      "loss": 2.2817,
      "step": 443
    },
    {
      "epoch": 0.0888,
      "grad_norm": 0.11781571805477142,
      "learning_rate": 0.0001772,
      "loss": 2.2811,
      "step": 444
    },
    {
      "epoch": 0.089,
      "grad_norm": 0.04747286066412926,
      "learning_rate": 0.0001776,
      "loss": 2.2814,
      "step": 445
    },
    {
      "epoch": 0.0892,
      "grad_norm": 0.017331674695014954,
      "learning_rate": 0.00017800000000000002,
      "loss": 2.2809,
      "step": 446
    },
    {
      "epoch": 0.0894,
      "grad_norm": 0.04285869002342224,
      "learning_rate": 0.0001784,
      "loss": 2.281,
      "step": 447
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.11782748997211456,
      "learning_rate": 0.0001788,
      "loss": 2.2806,
      "step": 448
    },
    {
      "epoch": 0.0898,
      "grad_norm": 0.23084361851215363,
      "learning_rate": 0.00017920000000000002,
      "loss": 2.2813,
      "step": 449
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.31252241134643555,
      "learning_rate": 0.0001796,
      "loss": 2.2818,
      "step": 450
    },
    {
      "epoch": 0.0902,
      "grad_norm": 0.3025452494621277,
      "learning_rate": 0.00018,
      "loss": 2.2813,
      "step": 451
    },
    {
      "epoch": 0.0904,
      "grad_norm": 0.21126776933670044,
      "learning_rate": 0.00018040000000000002,
      "loss": 2.2812,
      "step": 452
    },
    {
      "epoch": 0.0906,
      "grad_norm": 0.07749689370393753,
      "learning_rate": 0.0001808,
      "loss": 2.2812,
      "step": 453
    },
    {
      "epoch": 0.0908,
      "grad_norm": 0.04666844755411148,
      "learning_rate": 0.0001812,
      "loss": 2.2807,
      "step": 454
    },
    {
      "epoch": 0.091,
      "grad_norm": 0.10374990105628967,
      "learning_rate": 0.00018160000000000002,
      "loss": 2.281,
      "step": 455
    },
    {
      "epoch": 0.0912,
      "grad_norm": 0.11828792095184326,
      "learning_rate": 0.000182,
      "loss": 2.2813,
      "step": 456
    },
    {
      "epoch": 0.0914,
      "grad_norm": 0.11796707659959793,
      "learning_rate": 0.00018240000000000002,
      "loss": 2.2812,
      "step": 457
    },
    {
      "epoch": 0.0916,
      "grad_norm": 0.08526508510112762,
      "learning_rate": 0.00018280000000000003,
      "loss": 2.2809,
      "step": 458
    },
    {
      "epoch": 0.0918,
      "grad_norm": 0.0656171441078186,
      "learning_rate": 0.0001832,
      "loss": 2.2808,
      "step": 459
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.09492330253124237,
      "learning_rate": 0.00018360000000000002,
      "loss": 2.281,
      "step": 460
    },
    {
      "epoch": 0.0922,
      "grad_norm": 0.15699227154254913,
      "learning_rate": 0.00018400000000000003,
      "loss": 2.281,
      "step": 461
    },
    {
      "epoch": 0.0924,
      "grad_norm": 0.2185496687889099,
      "learning_rate": 0.0001844,
      "loss": 2.281,
      "step": 462
    },
    {
      "epoch": 0.0926,
      "grad_norm": 0.24260897934436798,
      "learning_rate": 0.00018480000000000002,
      "loss": 2.2813,
      "step": 463
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.23366333544254303,
      "learning_rate": 0.00018520000000000003,
      "loss": 2.2812,
      "step": 464
    },
    {
      "epoch": 0.093,
      "grad_norm": 0.22165921330451965,
      "learning_rate": 0.0001856,
      "loss": 2.2811,
      "step": 465
    },
    {
      "epoch": 0.0932,
      "grad_norm": 0.21015428006649017,
      "learning_rate": 0.00018600000000000002,
      "loss": 2.2808,
      "step": 466
    },
    {
      "epoch": 0.0934,
      "grad_norm": 0.18176808953285217,
      "learning_rate": 0.00018640000000000003,
      "loss": 2.281,
      "step": 467
    },
    {
      "epoch": 0.0936,
      "grad_norm": 0.16521356999874115,
      "learning_rate": 0.00018680000000000001,
      "loss": 2.281,
      "step": 468
    },
    {
      "epoch": 0.0938,
      "grad_norm": 0.1595943123102188,
      "learning_rate": 0.00018720000000000002,
      "loss": 2.2812,
      "step": 469
    },
    {
      "epoch": 0.094,
      "grad_norm": 0.14017139375209808,
      "learning_rate": 0.0001876,
      "loss": 2.2814,
      "step": 470
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.09117370843887329,
      "learning_rate": 0.000188,
      "loss": 2.2809,
      "step": 471
    },
    {
      "epoch": 0.0944,
      "grad_norm": 0.057835906744003296,
      "learning_rate": 0.0001884,
      "loss": 2.281,
      "step": 472
    },
    {
      "epoch": 0.0946,
      "grad_norm": 0.26034852862358093,
      "learning_rate": 0.0001888,
      "loss": 2.2815,
      "step": 473
    },
    {
      "epoch": 0.0948,
      "grad_norm": 0.4506315290927887,
      "learning_rate": 0.0001892,
      "loss": 2.282,
      "step": 474
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.6257368922233582,
      "learning_rate": 0.0001896,
      "loss": 2.2833,
      "step": 475
    },
    {
      "epoch": 0.0952,
      "grad_norm": 0.7502185702323914,
      "learning_rate": 0.00019,
      "loss": 2.2834,
      "step": 476
    },
    {
      "epoch": 0.0954,
      "grad_norm": 0.7108635306358337,
      "learning_rate": 0.0001904,
      "loss": 2.2834,
      "step": 477
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.4756382405757904,
      "learning_rate": 0.0001908,
      "loss": 2.2821,
      "step": 478
    },
    {
      "epoch": 0.0958,
      "grad_norm": 0.1173788383603096,
      "learning_rate": 0.0001912,
      "loss": 2.2811,
      "step": 479
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.20908167958259583,
      "learning_rate": 0.0001916,
      "loss": 2.2812,
      "step": 480
    },
    {
      "epoch": 0.0962,
      "grad_norm": 0.4263923466205597,
      "learning_rate": 0.000192,
      "loss": 2.2819,
      "step": 481
    },
    {
      "epoch": 0.0964,
      "grad_norm": 0.5138103365898132,
      "learning_rate": 0.00019240000000000001,
      "loss": 2.2824,
      "step": 482
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.4378727376461029,
      "learning_rate": 0.0001928,
      "loss": 2.2817,
      "step": 483
    },
    {
      "epoch": 0.0968,
      "grad_norm": 0.20226015150547028,
      "learning_rate": 0.0001932,
      "loss": 2.2809,
      "step": 484
    },
    {
      "epoch": 0.097,
      "grad_norm": 0.04938016086816788,
      "learning_rate": 0.00019360000000000002,
      "loss": 2.2809,
      "step": 485
    },
    {
      "epoch": 0.0972,
      "grad_norm": 0.19651667773723602,
      "learning_rate": 0.000194,
      "loss": 2.2801,
      "step": 486
    },
    {
      "epoch": 0.0974,
      "grad_norm": 0.24592383205890656,
      "learning_rate": 0.0001944,
      "loss": 2.2813,
      "step": 487
    },
    {
      "epoch": 0.0976,
      "grad_norm": 0.2074068784713745,
      "learning_rate": 0.0001948,
      "loss": 2.2808,
      "step": 488
    },
    {
      "epoch": 0.0978,
      "grad_norm": 0.10713998973369598,
      "learning_rate": 0.0001952,
      "loss": 2.2799,
      "step": 489
    },
    {
      "epoch": 0.098,
      "grad_norm": 0.027950560674071312,
      "learning_rate": 0.0001956,
      "loss": 2.2805,
      "step": 490
    },
    {
      "epoch": 0.0982,
      "grad_norm": 0.1463172733783722,
      "learning_rate": 0.000196,
      "loss": 2.2809,
      "step": 491
    },
    {
      "epoch": 0.0984,
      "grad_norm": 0.216153085231781,
      "learning_rate": 0.0001964,
      "loss": 2.2811,
      "step": 492
    },
    {
      "epoch": 0.0986,
      "grad_norm": 0.23782171308994293,
      "learning_rate": 0.0001968,
      "loss": 2.281,
      "step": 493
    },
    {
      "epoch": 0.0988,
      "grad_norm": 0.1842999905347824,
      "learning_rate": 0.0001972,
      "loss": 2.2806,
      "step": 494
    },
    {
      "epoch": 0.099,
      "grad_norm": 0.06852799654006958,
      "learning_rate": 0.0001976,
      "loss": 2.2806,
      "step": 495
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.025989985093474388,
      "learning_rate": 0.00019800000000000002,
      "loss": 2.2803,
      "step": 496
    },
    {
      "epoch": 0.0994,
      "grad_norm": 0.053139783442020416,
      "learning_rate": 0.0001984,
      "loss": 2.2804,
      "step": 497
    },
    {
      "epoch": 0.0996,
      "grad_norm": 0.0570230595767498,
      "learning_rate": 0.0001988,
      "loss": 2.2807,
      "step": 498
    },
    {
      "epoch": 0.0998,
      "grad_norm": 0.053285494446754456,
      "learning_rate": 0.00019920000000000002,
      "loss": 2.2804,
      "step": 499
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.036276284605264664,
      "learning_rate": 0.0001996,
      "loss": 2.2803,
      "step": 500
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.014786849729716778,
      "learning_rate": 0.0002,
      "loss": 2.2803,
      "step": 501
    },
    {
      "epoch": 0.1004,
      "grad_norm": 0.021120330318808556,
      "learning_rate": 0.0001999999756306074,
      "loss": 2.2807,
      "step": 502
    },
    {
      "epoch": 0.1006,
      "grad_norm": 0.04206474497914314,
      "learning_rate": 0.00019999990252244151,
      "loss": 2.2802,
      "step": 503
    },
    {
      "epoch": 0.1008,
      "grad_norm": 0.08569806069135666,
      "learning_rate": 0.00019999978067553796,
      "loss": 2.2808,
      "step": 504
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.13088926672935486,
      "learning_rate": 0.00019999961008995606,
      "loss": 2.2805,
      "step": 505
    },
    {
      "epoch": 0.1012,
      "grad_norm": 0.14949537813663483,
      "learning_rate": 0.00019999939076577905,
      "loss": 2.28,
      "step": 506
    },
    {
      "epoch": 0.1014,
      "grad_norm": 0.14203059673309326,
      "learning_rate": 0.00019999912270311375,
      "loss": 2.2809,
      "step": 507
    },
    {
      "epoch": 0.1016,
      "grad_norm": 0.08080507069826126,
      "learning_rate": 0.0001999988059020909,
      "loss": 2.2807,
      "step": 508
    },
    {
      "epoch": 0.1018,
      "grad_norm": 0.031237328425049782,
      "learning_rate": 0.00019999844036286483,
      "loss": 2.28,
      "step": 509
    },
    {
      "epoch": 0.102,
      "grad_norm": 0.1011975109577179,
      "learning_rate": 0.0001999980260856137,
      "loss": 2.2808,
      "step": 510
    },
    {
      "epoch": 0.1022,
      "grad_norm": 0.10643411427736282,
      "learning_rate": 0.00019999756307053948,
      "loss": 2.2796,
      "step": 511
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.10320140421390533,
      "learning_rate": 0.0001999970513178678,
      "loss": 2.2803,
      "step": 512
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.16870427131652832,
      "learning_rate": 0.0001999964908278481,
      "loss": 2.2806,
      "step": 513
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.21562083065509796,
      "learning_rate": 0.00019999588160075348,
      "loss": 2.2809,
      "step": 514
    },
    {
      "epoch": 0.103,
      "grad_norm": 0.09428993612527847,
      "learning_rate": 0.000199995223636881,
      "loss": 2.2803,
      "step": 515
    },
    {
      "epoch": 0.1032,
      "grad_norm": 0.11836577206850052,
      "learning_rate": 0.00019999451693655123,
      "loss": 2.2805,
      "step": 516
    },
    {
      "epoch": 0.1034,
      "grad_norm": 0.26793423295021057,
      "learning_rate": 0.00019999376150010866,
      "loss": 2.2811,
      "step": 517
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.31984058022499084,
      "learning_rate": 0.00019999295732792146,
      "loss": 2.2808,
      "step": 518
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.29297932982444763,
      "learning_rate": 0.00019999210442038162,
      "loss": 2.2806,
      "step": 519
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.20576144754886627,
      "learning_rate": 0.00019999120277790477,
      "loss": 2.2802,
      "step": 520
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.0873214527964592,
      "learning_rate": 0.00019999025240093044,
      "loss": 2.28,
      "step": 521
    },
    {
      "epoch": 0.1044,
      "grad_norm": 0.01677442155778408,
      "learning_rate": 0.00019998925328992175,
      "loss": 2.2802,
      "step": 522
    },
    {
      "epoch": 0.1046,
      "grad_norm": 0.028931522741913795,
      "learning_rate": 0.0001999882054453657,
      "loss": 2.2802,
      "step": 523
    },
    {
      "epoch": 0.1048,
      "grad_norm": 0.02204977534711361,
      "learning_rate": 0.000199987108867773,
      "loss": 2.2797,
      "step": 524
    },
    {
      "epoch": 0.105,
      "grad_norm": 0.014201335608959198,
      "learning_rate": 0.00019998596355767805,
      "loss": 2.2795,
      "step": 525
    },
    {
      "epoch": 0.1052,
      "grad_norm": 0.05695422366261482,
      "learning_rate": 0.00019998476951563915,
      "loss": 2.2806,
      "step": 526
    },
    {
      "epoch": 0.1054,
      "grad_norm": 0.15905390679836273,
      "learning_rate": 0.00019998352674223816,
      "loss": 2.2807,
      "step": 527
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.25310465693473816,
      "learning_rate": 0.0001999822352380809,
      "loss": 2.2809,
      "step": 528
    },
    {
      "epoch": 0.1058,
      "grad_norm": 0.3224073648452759,
      "learning_rate": 0.00019998089500379676,
      "loss": 2.281,
      "step": 529
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.33755892515182495,
      "learning_rate": 0.000199979506040039,
      "loss": 2.2806,
      "step": 530
    },
    {
      "epoch": 0.1062,
      "grad_norm": 0.2882723808288574,
      "learning_rate": 0.00019997806834748456,
      "loss": 2.281,
      "step": 531
    },
    {
      "epoch": 0.1064,
      "grad_norm": 0.15627703070640564,
      "learning_rate": 0.00019997658192683412,
      "loss": 2.28,
      "step": 532
    },
    {
      "epoch": 0.1066,
      "grad_norm": 0.06614156067371368,
      "learning_rate": 0.00019997504677881224,
      "loss": 2.2804,
      "step": 533
    },
    {
      "epoch": 0.1068,
      "grad_norm": 0.2877042889595032,
      "learning_rate": 0.000199973462904167,
      "loss": 2.2809,
      "step": 534
    },
    {
      "epoch": 0.107,
      "grad_norm": 0.3881016969680786,
      "learning_rate": 0.00019997183030367048,
      "loss": 2.281,
      "step": 535
    },
    {
      "epoch": 0.1072,
      "grad_norm": 0.3455062806606293,
      "learning_rate": 0.00019997014897811833,
      "loss": 2.2809,
      "step": 536
    },
    {
      "epoch": 0.1074,
      "grad_norm": 0.2451910674571991,
      "learning_rate": 0.00019996841892833,
      "loss": 2.2805,
      "step": 537
    },
    {
      "epoch": 0.1076,
      "grad_norm": 0.12973299622535706,
      "learning_rate": 0.00019996664015514871,
      "loss": 2.2804,
      "step": 538
    },
    {
      "epoch": 0.1078,
      "grad_norm": 0.01612965762615204,
      "learning_rate": 0.00019996481265944146,
      "loss": 2.2802,
      "step": 539
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.10599194467067719,
      "learning_rate": 0.00019996293644209887,
      "loss": 2.2802,
      "step": 540
    },
    {
      "epoch": 0.1082,
      "grad_norm": 0.17717047035694122,
      "learning_rate": 0.00019996101150403543,
      "loss": 2.2804,
      "step": 541
    },
    {
      "epoch": 0.1084,
      "grad_norm": 0.19096246361732483,
      "learning_rate": 0.00019995903784618936,
      "loss": 2.2806,
      "step": 542
    },
    {
      "epoch": 0.1086,
      "grad_norm": 0.17166586220264435,
      "learning_rate": 0.0001999570154695225,
      "loss": 2.2802,
      "step": 543
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.10972718149423599,
      "learning_rate": 0.00019995494437502064,
      "loss": 2.2801,
      "step": 544
    },
    {
      "epoch": 0.109,
      "grad_norm": 0.02123638242483139,
      "learning_rate": 0.0001999528245636931,
      "loss": 2.2799,
      "step": 545
    },
    {
      "epoch": 0.1092,
      "grad_norm": 0.0867324024438858,
      "learning_rate": 0.00019995065603657316,
      "loss": 2.2801,
      "step": 546
    },
    {
      "epoch": 0.1094,
      "grad_norm": 0.17642538249492645,
      "learning_rate": 0.0001999484387947177,
      "loss": 2.28,
      "step": 547
    },
    {
      "epoch": 0.1096,
      "grad_norm": 0.20906378328800201,
      "learning_rate": 0.0001999461728392073,
      "loss": 2.2802,
      "step": 548
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.19347107410430908,
      "learning_rate": 0.00019994385817114646,
      "loss": 2.2804,
      "step": 549
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.10389434546232224,
      "learning_rate": 0.00019994149479166324,
      "loss": 2.2798,
      "step": 550
    },
    {
      "epoch": 0.1102,
      "grad_norm": 0.04363143444061279,
      "learning_rate": 0.0001999390827019096,
      "loss": 2.2805,
      "step": 551
    },
    {
      "epoch": 0.1104,
      "grad_norm": 0.1606559157371521,
      "learning_rate": 0.0001999366219030611,
      "loss": 2.2801,
      "step": 552
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.2399071753025055,
      "learning_rate": 0.0001999341123963171,
      "loss": 2.2802,
      "step": 553
    },
    {
      "epoch": 0.1108,
      "grad_norm": 0.2415113002061844,
      "learning_rate": 0.0001999315541829008,
      "loss": 2.2806,
      "step": 554
    },
    {
      "epoch": 0.111,
      "grad_norm": 0.2092633843421936,
      "learning_rate": 0.00019992894726405893,
      "loss": 2.2794,
      "step": 555
    },
    {
      "epoch": 0.1112,
      "grad_norm": 0.13680899143218994,
      "learning_rate": 0.0001999262916410621,
      "loss": 2.2798,
      "step": 556
    },
    {
      "epoch": 0.1114,
      "grad_norm": 0.03662730008363724,
      "learning_rate": 0.00019992358731520468,
      "loss": 2.2803,
      "step": 557
    },
    {
      "epoch": 0.1116,
      "grad_norm": 0.21957595646381378,
      "learning_rate": 0.0001999208342878047,
      "loss": 2.2803,
      "step": 558
    },
    {
      "epoch": 0.1118,
      "grad_norm": 0.3251246511936188,
      "learning_rate": 0.00019991803256020393,
      "loss": 2.2807,
      "step": 559
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.3059234023094177,
      "learning_rate": 0.00019991518213376787,
      "loss": 2.2809,
      "step": 560
    },
    {
      "epoch": 0.1122,
      "grad_norm": 0.11267164349555969,
      "learning_rate": 0.00019991228300988585,
      "loss": 2.2797,
      "step": 561
    },
    {
      "epoch": 0.1124,
      "grad_norm": 0.22132998704910278,
      "learning_rate": 0.00019990933518997084,
      "loss": 2.2806,
      "step": 562
    },
    {
      "epoch": 0.1126,
      "grad_norm": 0.5312188863754272,
      "learning_rate": 0.00019990633867545955,
      "loss": 2.2815,
      "step": 563
    },
    {
      "epoch": 0.1128,
      "grad_norm": 0.6166166663169861,
      "learning_rate": 0.0001999032934678125,
      "loss": 2.2819,
      "step": 564
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.38915199041366577,
      "learning_rate": 0.00019990019956851382,
      "loss": 2.2807,
      "step": 565
    },
    {
      "epoch": 0.1132,
      "grad_norm": 0.06809045374393463,
      "learning_rate": 0.00019989705697907149,
      "loss": 2.2791,
      "step": 566
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.48107489943504333,
      "learning_rate": 0.00019989386570101714,
      "loss": 2.2807,
      "step": 567
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.6274041533470154,
      "learning_rate": 0.00019989062573590616,
      "loss": 2.2819,
      "step": 568
    },
    {
      "epoch": 0.1138,
      "grad_norm": 0.44906163215637207,
      "learning_rate": 0.0001998873370853177,
      "loss": 2.2813,
      "step": 569
    },
    {
      "epoch": 0.114,
      "grad_norm": 0.023275399580597878,
      "learning_rate": 0.0001998839997508546,
      "loss": 2.2792,
      "step": 570
    },
    {
      "epoch": 0.1142,
      "grad_norm": 0.47907164692878723,
      "learning_rate": 0.0001998806137341434,
      "loss": 2.2807,
      "step": 571
    },
    {
      "epoch": 0.1144,
      "grad_norm": 0.749935507774353,
      "learning_rate": 0.00019987717903683448,
      "loss": 2.2832,
      "step": 572
    },
    {
      "epoch": 0.1146,
      "grad_norm": 0.6068918704986572,
      "learning_rate": 0.00019987369566060176,
      "loss": 2.282,
      "step": 573
    },
    {
      "epoch": 0.1148,
      "grad_norm": 0.1824079304933548,
      "learning_rate": 0.00019987016360714307,
      "loss": 2.2803,
      "step": 574
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.20601218938827515,
      "learning_rate": 0.00019986658287817987,
      "loss": 2.2799,
      "step": 575
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.35462623834609985,
      "learning_rate": 0.0001998629534754574,
      "loss": 2.2805,
      "step": 576
    },
    {
      "epoch": 0.1154,
      "grad_norm": 0.2860219478607178,
      "learning_rate": 0.00019985927540074454,
      "loss": 2.2802,
      "step": 577
    },
    {
      "epoch": 0.1156,
      "grad_norm": 0.10487033426761627,
      "learning_rate": 0.00019985554865583393,
      "loss": 2.2802,
      "step": 578
    },
    {
      "epoch": 0.1158,
      "grad_norm": 0.10130726546049118,
      "learning_rate": 0.000199851773242542,
      "loss": 2.2794,
      "step": 579
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.23929226398468018,
      "learning_rate": 0.00019984794916270876,
      "loss": 2.2803,
      "step": 580
    },
    {
      "epoch": 0.1162,
      "grad_norm": 0.34940221905708313,
      "learning_rate": 0.00019984407641819812,
      "loss": 2.2802,
      "step": 581
    },
    {
      "epoch": 0.1164,
      "grad_norm": 0.34953010082244873,
      "learning_rate": 0.00019984015501089752,
      "loss": 2.2805,
      "step": 582
    },
    {
      "epoch": 0.1166,
      "grad_norm": 0.08668052405118942,
      "learning_rate": 0.00019983618494271826,
      "loss": 2.2802,
      "step": 583
    },
    {
      "epoch": 0.1168,
      "grad_norm": 0.2837238311767578,
      "learning_rate": 0.00019983216621559524,
      "loss": 2.2802,
      "step": 584
    },
    {
      "epoch": 0.117,
      "grad_norm": 0.44326290488243103,
      "learning_rate": 0.00019982809883148722,
      "loss": 2.2807,
      "step": 585
    },
    {
      "epoch": 0.1172,
      "grad_norm": 0.24276161193847656,
      "learning_rate": 0.00019982398279237655,
      "loss": 2.2803,
      "step": 586
    },
    {
      "epoch": 0.1174,
      "grad_norm": 0.15362048149108887,
      "learning_rate": 0.00019981981810026934,
      "loss": 2.2793,
      "step": 587
    },
    {
      "epoch": 0.1176,
      "grad_norm": 0.4177349805831909,
      "learning_rate": 0.00019981560475719538,
      "loss": 2.28,
      "step": 588
    },
    {
      "epoch": 0.1178,
      "grad_norm": 0.3652116358280182,
      "learning_rate": 0.00019981134276520827,
      "loss": 2.2803,
      "step": 589
    },
    {
      "epoch": 0.118,
      "grad_norm": 0.034599728882312775,
      "learning_rate": 0.00019980703212638522,
      "loss": 2.2788,
      "step": 590
    },
    {
      "epoch": 0.1182,
      "grad_norm": 0.2750101685523987,
      "learning_rate": 0.00019980267284282717,
      "loss": 2.2797,
      "step": 591
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.2902490794658661,
      "learning_rate": 0.00019979826491665881,
      "loss": 2.2796,
      "step": 592
    },
    {
      "epoch": 0.1186,
      "grad_norm": 0.060035012662410736,
      "learning_rate": 0.00019979380835002846,
      "loss": 2.2792,
      "step": 593
    },
    {
      "epoch": 0.1188,
      "grad_norm": 0.20929494500160217,
      "learning_rate": 0.00019978930314510824,
      "loss": 2.2797,
      "step": 594
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.28507983684539795,
      "learning_rate": 0.00019978474930409395,
      "loss": 2.2799,
      "step": 595
    },
    {
      "epoch": 0.1192,
      "grad_norm": 0.1167522519826889,
      "learning_rate": 0.000199780146829205,
      "loss": 2.2795,
      "step": 596
    },
    {
      "epoch": 0.1194,
      "grad_norm": 0.13687732815742493,
      "learning_rate": 0.00019977549572268468,
      "loss": 2.2795,
      "step": 597
    },
    {
      "epoch": 0.1196,
      "grad_norm": 0.3001672327518463,
      "learning_rate": 0.00019977079598679977,
      "loss": 2.2805,
      "step": 598
    },
    {
      "epoch": 0.1198,
      "grad_norm": 0.2636941075325012,
      "learning_rate": 0.000199766047623841,
      "loss": 2.2799,
      "step": 599
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.04676492512226105,
      "learning_rate": 0.00019976125063612252,
      "loss": 2.2795,
      "step": 600
    },
    {
      "epoch": 0.1202,
      "grad_norm": 0.20666885375976562,
      "learning_rate": 0.00019975640502598244,
      "loss": 2.279,
      "step": 601
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.31128838658332825,
      "learning_rate": 0.00019975151079578237,
      "loss": 2.2803,
      "step": 602
    },
    {
      "epoch": 0.1206,
      "grad_norm": 0.2112381011247635,
      "learning_rate": 0.00019974656794790775,
      "loss": 2.2802,
      "step": 603
    },
    {
      "epoch": 0.1208,
      "grad_norm": 0.03735370188951492,
      "learning_rate": 0.00019974157648476766,
      "loss": 2.2796,
      "step": 604
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.2768845856189728,
      "learning_rate": 0.00019973653640879485,
      "loss": 2.2798,
      "step": 605
    },
    {
      "epoch": 0.1212,
      "grad_norm": 0.315682053565979,
      "learning_rate": 0.00019973144772244582,
      "loss": 2.2796,
      "step": 606
    },
    {
      "epoch": 0.1214,
      "grad_norm": 0.16628865897655487,
      "learning_rate": 0.0001997263104282007,
      "loss": 2.2798,
      "step": 607
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.02315085008740425,
      "learning_rate": 0.00019972112452856339,
      "loss": 2.2794,
      "step": 608
    },
    {
      "epoch": 0.1218,
      "grad_norm": 0.15766003727912903,
      "learning_rate": 0.0001997158900260614,
      "loss": 2.2795,
      "step": 609
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.2139901965856552,
      "learning_rate": 0.00019971060692324598,
      "loss": 2.2796,
      "step": 610
    },
    {
      "epoch": 0.1222,
      "grad_norm": 0.15274262428283691,
      "learning_rate": 0.00019970527522269205,
      "loss": 2.2794,
      "step": 611
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.057756345719099045,
      "learning_rate": 0.0001996998949269982,
      "loss": 2.2798,
      "step": 612
    },
    {
      "epoch": 0.1226,
      "grad_norm": 0.2794957458972931,
      "learning_rate": 0.00019969446603878673,
      "loss": 2.2797,
      "step": 613
    },
    {
      "epoch": 0.1228,
      "grad_norm": 0.33388832211494446,
      "learning_rate": 0.0001996889885607036,
      "loss": 2.2804,
      "step": 614
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.14143064618110657,
      "learning_rate": 0.00019968346249541846,
      "loss": 2.2795,
      "step": 615
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.19406208395957947,
      "learning_rate": 0.00019967788784562473,
      "loss": 2.2793,
      "step": 616
    },
    {
      "epoch": 0.1234,
      "grad_norm": 0.40118327736854553,
      "learning_rate": 0.00019967226461403933,
      "loss": 2.2802,
      "step": 617
    },
    {
      "epoch": 0.1236,
      "grad_norm": 0.30361640453338623,
      "learning_rate": 0.00019966659280340297,
      "loss": 2.2798,
      "step": 618
    },
    {
      "epoch": 0.1238,
      "grad_norm": 0.016054242849349976,
      "learning_rate": 0.0001996608724164801,
      "loss": 2.2789,
      "step": 619
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.18487444519996643,
      "learning_rate": 0.00019965510345605866,
      "loss": 2.2798,
      "step": 620
    },
    {
      "epoch": 0.1242,
      "grad_norm": 0.1438528299331665,
      "learning_rate": 0.00019964928592495045,
      "loss": 2.279,
      "step": 621
    },
    {
      "epoch": 0.1244,
      "grad_norm": 0.0232748631387949,
      "learning_rate": 0.00019964341982599078,
      "loss": 2.2796,
      "step": 622
    },
    {
      "epoch": 0.1246,
      "grad_norm": 0.07437591254711151,
      "learning_rate": 0.00019963750516203884,
      "loss": 2.2793,
      "step": 623
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.1418762505054474,
      "learning_rate": 0.00019963154193597727,
      "loss": 2.2793,
      "step": 624
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.15046685934066772,
      "learning_rate": 0.0001996255301507125,
      "loss": 2.2797,
      "step": 625
    },
    {
      "epoch": 0.1252,
      "grad_norm": 0.04566844180226326,
      "learning_rate": 0.00019961946980917456,
      "loss": 2.2792,
      "step": 626
    },
    {
      "epoch": 0.1254,
      "grad_norm": 0.14824572205543518,
      "learning_rate": 0.00019961336091431727,
      "loss": 2.2788,
      "step": 627
    },
    {
      "epoch": 0.1256,
      "grad_norm": 0.2914900779724121,
      "learning_rate": 0.00019960720346911797,
      "loss": 2.2796,
      "step": 628
    },
    {
      "epoch": 0.1258,
      "grad_norm": 0.2843387722969055,
      "learning_rate": 0.00019960099747657774,
      "loss": 2.2794,
      "step": 629
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.10247065871953964,
      "learning_rate": 0.00019959474293972129,
      "loss": 2.279,
      "step": 630
    },
    {
      "epoch": 0.1262,
      "grad_norm": 0.18605265021324158,
      "learning_rate": 0.00019958843986159704,
      "loss": 2.2792,
      "step": 631
    },
    {
      "epoch": 0.1264,
      "grad_norm": 0.35206976532936096,
      "learning_rate": 0.000199582088245277,
      "loss": 2.2789,
      "step": 632
    },
    {
      "epoch": 0.1266,
      "grad_norm": 0.2526860237121582,
      "learning_rate": 0.00019957568809385694,
      "loss": 2.2791,
      "step": 633
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.017734436318278313,
      "learning_rate": 0.0001995692394104561,
      "loss": 2.2784,
      "step": 634
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.27641531825065613,
      "learning_rate": 0.00019956274219821757,
      "loss": 2.2791,
      "step": 635
    },
    {
      "epoch": 0.1272,
      "grad_norm": 0.42820021510124207,
      "learning_rate": 0.00019955619646030802,
      "loss": 2.2805,
      "step": 636
    },
    {
      "epoch": 0.1274,
      "grad_norm": 0.3758969008922577,
      "learning_rate": 0.0001995496021999177,
      "loss": 2.2802,
      "step": 637
    },
    {
      "epoch": 0.1276,
      "grad_norm": 0.10973133146762848,
      "learning_rate": 0.00019954295942026064,
      "loss": 2.2786,
      "step": 638
    },
    {
      "epoch": 0.1278,
      "grad_norm": 0.15050821006298065,
      "learning_rate": 0.0001995362681245744,
      "loss": 2.2797,
      "step": 639
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.19409900903701782,
      "learning_rate": 0.00019952952831612026,
      "loss": 2.2792,
      "step": 640
    },
    {
      "epoch": 0.1282,
      "grad_norm": 0.0778803750872612,
      "learning_rate": 0.0001995227399981831,
      "loss": 2.2795,
      "step": 641
    },
    {
      "epoch": 0.1284,
      "grad_norm": 0.08566217869520187,
      "learning_rate": 0.00019951590317407152,
      "loss": 2.2794,
      "step": 642
    },
    {
      "epoch": 0.1286,
      "grad_norm": 0.14525559544563293,
      "learning_rate": 0.00019950901784711764,
      "loss": 2.2786,
      "step": 643
    },
    {
      "epoch": 0.1288,
      "grad_norm": 0.046612586826086044,
      "learning_rate": 0.00019950208402067733,
      "loss": 2.2794,
      "step": 644
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.0732223391532898,
      "learning_rate": 0.00019949510169813003,
      "loss": 2.2794,
      "step": 645
    },
    {
      "epoch": 0.1292,
      "grad_norm": 0.12084516137838364,
      "learning_rate": 0.00019948807088287883,
      "loss": 2.2788,
      "step": 646
    },
    {
      "epoch": 0.1294,
      "grad_norm": 0.09829748421907425,
      "learning_rate": 0.00019948099157835047,
      "loss": 2.2795,
      "step": 647
    },
    {
      "epoch": 0.1296,
      "grad_norm": 0.023321496322751045,
      "learning_rate": 0.00019947386378799532,
      "loss": 2.2782,
      "step": 648
    },
    {
      "epoch": 0.1298,
      "grad_norm": 0.18650691211223602,
      "learning_rate": 0.00019946668751528744,
      "loss": 2.2795,
      "step": 649
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2712702751159668,
      "learning_rate": 0.00019945946276372434,
      "loss": 2.2782,
      "step": 650
    },
    {
      "epoch": 0.1302,
      "grad_norm": 0.15516462922096252,
      "learning_rate": 0.00019945218953682734,
      "loss": 2.2789,
      "step": 651
    },
    {
      "epoch": 0.1304,
      "grad_norm": 0.13319337368011475,
      "learning_rate": 0.00019944486783814134,
      "loss": 2.2795,
      "step": 652
    },
    {
      "epoch": 0.1306,
      "grad_norm": 0.37324148416519165,
      "learning_rate": 0.0001994374976712348,
      "loss": 2.2799,
      "step": 653
    },
    {
      "epoch": 0.1308,
      "grad_norm": 0.3489644527435303,
      "learning_rate": 0.0001994300790396999,
      "loss": 2.2799,
      "step": 654
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.06834349036216736,
      "learning_rate": 0.00019942261194715236,
      "loss": 2.277,
      "step": 655
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.2406308501958847,
      "learning_rate": 0.00019941509639723155,
      "loss": 2.2797,
      "step": 656
    },
    {
      "epoch": 0.1314,
      "grad_norm": 0.33240246772766113,
      "learning_rate": 0.00019940753239360047,
      "loss": 2.2797,
      "step": 657
    },
    {
      "epoch": 0.1316,
      "grad_norm": 0.18410098552703857,
      "learning_rate": 0.0001993999199399457,
      "loss": 2.2769,
      "step": 658
    },
    {
      "epoch": 0.1318,
      "grad_norm": 0.06506457179784775,
      "learning_rate": 0.0001993922590399775,
      "loss": 2.2777,
      "step": 659
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.22411693632602692,
      "learning_rate": 0.00019938454969742968,
      "loss": 2.2784,
      "step": 660
    },
    {
      "epoch": 0.1322,
      "grad_norm": 0.19858789443969727,
      "learning_rate": 0.00019937679191605963,
      "loss": 2.2787,
      "step": 661
    },
    {
      "epoch": 0.1324,
      "grad_norm": 0.03587792068719864,
      "learning_rate": 0.00019936898569964848,
      "loss": 2.279,
      "step": 662
    },
    {
      "epoch": 0.1326,
      "grad_norm": 0.1602770835161209,
      "learning_rate": 0.00019936113105200085,
      "loss": 2.2777,
      "step": 663
    },
    {
      "epoch": 0.1328,
      "grad_norm": 0.2118166983127594,
      "learning_rate": 0.000199353227976945,
      "loss": 2.279,
      "step": 664
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.05836513265967369,
      "learning_rate": 0.00019934527647833276,
      "loss": 2.2792,
      "step": 665
    },
    {
      "epoch": 0.1332,
      "grad_norm": 0.13915060460567474,
      "learning_rate": 0.00019933727656003963,
      "loss": 2.2788,
      "step": 666
    },
    {
      "epoch": 0.1334,
      "grad_norm": 0.2071356177330017,
      "learning_rate": 0.00019932922822596473,
      "loss": 2.2794,
      "step": 667
    },
    {
      "epoch": 0.1336,
      "grad_norm": 0.1331482082605362,
      "learning_rate": 0.00019932113148003058,
      "loss": 2.2788,
      "step": 668
    },
    {
      "epoch": 0.1338,
      "grad_norm": 0.035213541239500046,
      "learning_rate": 0.00019931298632618356,
      "loss": 2.2768,
      "step": 669
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.15252339839935303,
      "learning_rate": 0.00019930479276839344,
      "loss": 2.2782,
      "step": 670
    },
    {
      "epoch": 0.1342,
      "grad_norm": 0.11144735664129257,
      "learning_rate": 0.0001992965508106537,
      "loss": 2.2793,
      "step": 671
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.042043495923280716,
      "learning_rate": 0.00019928826045698136,
      "loss": 2.2787,
      "step": 672
    },
    {
      "epoch": 0.1346,
      "grad_norm": 0.1480739414691925,
      "learning_rate": 0.00019927992171141708,
      "loss": 2.278,
      "step": 673
    },
    {
      "epoch": 0.1348,
      "grad_norm": 0.14725221693515778,
      "learning_rate": 0.000199271534578025,
      "loss": 2.2777,
      "step": 674
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.05970553681254387,
      "learning_rate": 0.0001992630990608929,
      "loss": 2.2776,
      "step": 675
    },
    {
      "epoch": 0.1352,
      "grad_norm": 0.034894682466983795,
      "learning_rate": 0.00019925461516413223,
      "loss": 2.2791,
      "step": 676
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.04656535014510155,
      "learning_rate": 0.00019924608289187786,
      "loss": 2.2786,
      "step": 677
    },
    {
      "epoch": 0.1356,
      "grad_norm": 0.01604841649532318,
      "learning_rate": 0.00019923750224828832,
      "loss": 2.2788,
      "step": 678
    },
    {
      "epoch": 0.1358,
      "grad_norm": 0.03293028473854065,
      "learning_rate": 0.00019922887323754577,
      "loss": 2.2779,
      "step": 679
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.04404636472463608,
      "learning_rate": 0.00019922019586385585,
      "loss": 2.2787,
      "step": 680
    },
    {
      "epoch": 0.1362,
      "grad_norm": 0.025399748235940933,
      "learning_rate": 0.0001992114701314478,
      "loss": 2.2786,
      "step": 681
    },
    {
      "epoch": 0.1364,
      "grad_norm": 0.04569943621754646,
      "learning_rate": 0.00019920269604457446,
      "loss": 2.2786,
      "step": 682
    },
    {
      "epoch": 0.1366,
      "grad_norm": 0.11472634226083755,
      "learning_rate": 0.00019919387360751217,
      "loss": 2.2784,
      "step": 683
    },
    {
      "epoch": 0.1368,
      "grad_norm": 0.08987665176391602,
      "learning_rate": 0.0001991850028245609,
      "loss": 2.2787,
      "step": 684
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.028263477608561516,
      "learning_rate": 0.00019917608370004417,
      "loss": 2.2783,
      "step": 685
    },
    {
      "epoch": 0.1372,
      "grad_norm": 0.13331487774848938,
      "learning_rate": 0.00019916711623830903,
      "loss": 2.2792,
      "step": 686
    },
    {
      "epoch": 0.1374,
      "grad_norm": 0.22757762670516968,
      "learning_rate": 0.00019915810044372618,
      "loss": 2.2789,
      "step": 687
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.23724430799484253,
      "learning_rate": 0.00019914903632068973,
      "loss": 2.2794,
      "step": 688
    },
    {
      "epoch": 0.1378,
      "grad_norm": 0.12243066728115082,
      "learning_rate": 0.00019913992387361745,
      "loss": 2.2792,
      "step": 689
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.031778763979673386,
      "learning_rate": 0.00019913076310695068,
      "loss": 2.2773,
      "step": 690
    },
    {
      "epoch": 0.1382,
      "grad_norm": 0.14335907995700836,
      "learning_rate": 0.00019912155402515417,
      "loss": 2.2779,
      "step": 691
    },
    {
      "epoch": 0.1384,
      "grad_norm": 0.16564233601093292,
      "learning_rate": 0.0001991122966327164,
      "loss": 2.2789,
      "step": 692
    },
    {
      "epoch": 0.1386,
      "grad_norm": 0.07848090678453445,
      "learning_rate": 0.0001991029909341493,
      "loss": 2.2792,
      "step": 693
    },
    {
      "epoch": 0.1388,
      "grad_norm": 0.04365299269556999,
      "learning_rate": 0.00019909363693398828,
      "loss": 2.2784,
      "step": 694
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.07511486113071442,
      "learning_rate": 0.00019908423463679248,
      "loss": 2.2783,
      "step": 695
    },
    {
      "epoch": 0.1392,
      "grad_norm": 0.015732429921627045,
      "learning_rate": 0.00019907478404714436,
      "loss": 2.2787,
      "step": 696
    },
    {
      "epoch": 0.1394,
      "grad_norm": 0.07734601944684982,
      "learning_rate": 0.00019906528516965008,
      "loss": 2.2789,
      "step": 697
    },
    {
      "epoch": 0.1396,
      "grad_norm": 0.0667557492852211,
      "learning_rate": 0.0001990557380089393,
      "loss": 2.2782,
      "step": 698
    },
    {
      "epoch": 0.1398,
      "grad_norm": 0.020859936252236366,
      "learning_rate": 0.00019904614256966512,
      "loss": 2.2785,
      "step": 699
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.11919737607240677,
      "learning_rate": 0.0001990364988565043,
      "loss": 2.2787,
      "step": 700
    },
    {
      "epoch": 0.1402,
      "grad_norm": 0.1453191637992859,
      "learning_rate": 0.00019902680687415705,
      "loss": 2.279,
      "step": 701
    },
    {
      "epoch": 0.1404,
      "grad_norm": 0.03443368151783943,
      "learning_rate": 0.00019901706662734712,
      "loss": 2.2785,
      "step": 702
    },
    {
      "epoch": 0.1406,
      "grad_norm": 0.1568976193666458,
      "learning_rate": 0.00019900727812082177,
      "loss": 2.2785,
      "step": 703
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.22339050471782684,
      "learning_rate": 0.0001989974413593518,
      "loss": 2.2783,
      "step": 704
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.06081041693687439,
      "learning_rate": 0.00019898755634773158,
      "loss": 2.2783,
      "step": 705
    },
    {
      "epoch": 0.1412,
      "grad_norm": 0.20864734053611755,
      "learning_rate": 0.0001989776230907789,
      "loss": 2.2779,
      "step": 706
    },
    {
      "epoch": 0.1414,
      "grad_norm": 0.3483375608921051,
      "learning_rate": 0.0001989676415933351,
      "loss": 2.2774,
      "step": 707
    },
    {
      "epoch": 0.1416,
      "grad_norm": 0.2625489830970764,
      "learning_rate": 0.0001989576118602651,
      "loss": 2.2792,
      "step": 708
    },
    {
      "epoch": 0.1418,
      "grad_norm": 0.015311597846448421,
      "learning_rate": 0.00019894753389645723,
      "loss": 2.278,
      "step": 709
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.27625328302383423,
      "learning_rate": 0.00019893740770682335,
      "loss": 2.2787,
      "step": 710
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.3835970163345337,
      "learning_rate": 0.00019892723329629887,
      "loss": 2.2785,
      "step": 711
    },
    {
      "epoch": 0.1424,
      "grad_norm": 0.20415307581424713,
      "learning_rate": 0.00019891701066984262,
      "loss": 2.279,
      "step": 712
    },
    {
      "epoch": 0.1426,
      "grad_norm": 0.12386977672576904,
      "learning_rate": 0.00019890673983243706,
      "loss": 2.2786,
      "step": 713
    },
    {
      "epoch": 0.1428,
      "grad_norm": 0.30907508730888367,
      "learning_rate": 0.00019889642078908804,
      "loss": 2.279,
      "step": 714
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.215351402759552,
      "learning_rate": 0.0001988860535448249,
      "loss": 2.2785,
      "step": 715
    },
    {
      "epoch": 0.1432,
      "grad_norm": 0.03182166442275047,
      "learning_rate": 0.0001988756381047006,
      "loss": 2.2788,
      "step": 716
    },
    {
      "epoch": 0.1434,
      "grad_norm": 0.18799608945846558,
      "learning_rate": 0.0001988651744737914,
      "loss": 2.2785,
      "step": 717
    },
    {
      "epoch": 0.1436,
      "grad_norm": 0.1575077623128891,
      "learning_rate": 0.0001988546626571972,
      "loss": 2.2789,
      "step": 718
    },
    {
      "epoch": 0.1438,
      "grad_norm": 0.03640049323439598,
      "learning_rate": 0.00019884410266004135,
      "loss": 2.2781,
      "step": 719
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.08687899261713028,
      "learning_rate": 0.00019883349448747062,
      "loss": 2.2791,
      "step": 720
    },
    {
      "epoch": 0.1442,
      "grad_norm": 0.17488446831703186,
      "learning_rate": 0.0001988228381446553,
      "loss": 2.2781,
      "step": 721
    },
    {
      "epoch": 0.1444,
      "grad_norm": 0.16622193157672882,
      "learning_rate": 0.0001988121336367892,
      "loss": 2.2785,
      "step": 722
    },
    {
      "epoch": 0.1446,
      "grad_norm": 0.02572149597108364,
      "learning_rate": 0.00019880138096908952,
      "loss": 2.2779,
      "step": 723
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.1588442623615265,
      "learning_rate": 0.00019879058014679704,
      "loss": 2.2782,
      "step": 724
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.20796070992946625,
      "learning_rate": 0.0001987797311751759,
      "loss": 2.2784,
      "step": 725
    },
    {
      "epoch": 0.1452,
      "grad_norm": 0.07014245539903641,
      "learning_rate": 0.00019876883405951377,
      "loss": 2.2785,
      "step": 726
    },
    {
      "epoch": 0.1454,
      "grad_norm": 0.1403411626815796,
      "learning_rate": 0.0001987578888051218,
      "loss": 2.2788,
      "step": 727
    },
    {
      "epoch": 0.1456,
      "grad_norm": 0.2576597034931183,
      "learning_rate": 0.00019874689541733457,
      "loss": 2.2774,
      "step": 728
    },
    {
      "epoch": 0.1458,
      "grad_norm": 0.19393837451934814,
      "learning_rate": 0.00019873585390151003,
      "loss": 2.2776,
      "step": 729
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.015364548191428185,
      "learning_rate": 0.00019872476426302982,
      "loss": 2.2787,
      "step": 730
    },
    {
      "epoch": 0.1462,
      "grad_norm": 0.14269497990608215,
      "learning_rate": 0.0001987136265072988,
      "loss": 2.2778,
      "step": 731
    },
    {
      "epoch": 0.1464,
      "grad_norm": 0.1443844586610794,
      "learning_rate": 0.0001987024406397454,
      "loss": 2.2786,
      "step": 732
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.04474532604217529,
      "learning_rate": 0.00019869120666582153,
      "loss": 2.2782,
      "step": 733
    },
    {
      "epoch": 0.1468,
      "grad_norm": 0.055939119309186935,
      "learning_rate": 0.0001986799245910024,
      "loss": 2.2782,
      "step": 734
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.054671481251716614,
      "learning_rate": 0.0001986685944207868,
      "loss": 2.2782,
      "step": 735
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.032003253698349,
      "learning_rate": 0.00019865721616069696,
      "loss": 2.2786,
      "step": 736
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.10768909752368927,
      "learning_rate": 0.00019864578981627844,
      "loss": 2.2779,
      "step": 737
    },
    {
      "epoch": 0.1476,
      "grad_norm": 0.11895795911550522,
      "learning_rate": 0.0001986343153931003,
      "loss": 2.2778,
      "step": 738
    },
    {
      "epoch": 0.1478,
      "grad_norm": 0.058790113776922226,
      "learning_rate": 0.00019862279289675509,
      "loss": 2.2774,
      "step": 739
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.035196930170059204,
      "learning_rate": 0.0001986112223328587,
      "loss": 2.2774,
      "step": 740
    },
    {
      "epoch": 0.1482,
      "grad_norm": 0.0628749206662178,
      "learning_rate": 0.0001985996037070505,
      "loss": 2.2786,
      "step": 741
    },
    {
      "epoch": 0.1484,
      "grad_norm": 0.016495222225785255,
      "learning_rate": 0.00019858793702499323,
      "loss": 2.2781,
      "step": 742
    },
    {
      "epoch": 0.1486,
      "grad_norm": 0.1120016798377037,
      "learning_rate": 0.00019857622229237313,
      "loss": 2.2787,
      "step": 743
    },
    {
      "epoch": 0.1488,
      "grad_norm": 0.16403645277023315,
      "learning_rate": 0.00019856445951489982,
      "loss": 2.2782,
      "step": 744
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.11986328661441803,
      "learning_rate": 0.00019855264869830629,
      "loss": 2.2786,
      "step": 745
    },
    {
      "epoch": 0.1492,
      "grad_norm": 0.019817916676402092,
      "learning_rate": 0.00019854078984834903,
      "loss": 2.2784,
      "step": 746
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.14692340791225433,
      "learning_rate": 0.00019852888297080786,
      "loss": 2.2777,
      "step": 747
    },
    {
      "epoch": 0.1496,
      "grad_norm": 0.2007637321949005,
      "learning_rate": 0.0001985169280714861,
      "loss": 2.2778,
      "step": 748
    },
    {
      "epoch": 0.1498,
      "grad_norm": 0.1401161104440689,
      "learning_rate": 0.00019850492515621038,
      "loss": 2.2778,
      "step": 749
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.034117113798856735,
      "learning_rate": 0.00019849287423083078,
      "loss": 2.2786,
      "step": 750
    },
    {
      "epoch": 0.1502,
      "grad_norm": 0.2212217152118683,
      "learning_rate": 0.00019848077530122083,
      "loss": 2.2777,
      "step": 751
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.26467886567115784,
      "learning_rate": 0.0001984686283732773,
      "loss": 2.279,
      "step": 752
    },
    {
      "epoch": 0.1506,
      "grad_norm": 0.12230066955089569,
      "learning_rate": 0.00019845643345292054,
      "loss": 2.2786,
      "step": 753
    },
    {
      "epoch": 0.1508,
      "grad_norm": 0.06449053436517715,
      "learning_rate": 0.0001984441905460942,
      "loss": 2.2774,
      "step": 754
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.17709341645240784,
      "learning_rate": 0.00019843189965876526,
      "loss": 2.2788,
      "step": 755
    },
    {
      "epoch": 0.1512,
      "grad_norm": 0.2080894112586975,
      "learning_rate": 0.0001984195607969242,
      "loss": 2.2778,
      "step": 756
    },
    {
      "epoch": 0.1514,
      "grad_norm": 0.14670182764530182,
      "learning_rate": 0.00019840717396658484,
      "loss": 2.2777,
      "step": 757
    },
    {
      "epoch": 0.1516,
      "grad_norm": 0.02199646830558777,
      "learning_rate": 0.00019839473917378434,
      "loss": 2.278,
      "step": 758
    },
    {
      "epoch": 0.1518,
      "grad_norm": 0.09907554090023041,
      "learning_rate": 0.00019838225642458327,
      "loss": 2.2777,
      "step": 759
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.1592327356338501,
      "learning_rate": 0.00019836972572506557,
      "loss": 2.2781,
      "step": 760
    },
    {
      "epoch": 0.1522,
      "grad_norm": 0.14565294981002808,
      "learning_rate": 0.00019835714708133862,
      "loss": 2.278,
      "step": 761
    },
    {
      "epoch": 0.1524,
      "grad_norm": 0.02589743584394455,
      "learning_rate": 0.00019834452049953297,
      "loss": 2.2783,
      "step": 762
    },
    {
      "epoch": 0.1526,
      "grad_norm": 0.13925807178020477,
      "learning_rate": 0.00019833184598580276,
      "loss": 2.2781,
      "step": 763
    },
    {
      "epoch": 0.1528,
      "grad_norm": 0.20186848938465118,
      "learning_rate": 0.00019831912354632535,
      "loss": 2.2782,
      "step": 764
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.12890882790088654,
      "learning_rate": 0.00019830635318730154,
      "loss": 2.2777,
      "step": 765
    },
    {
      "epoch": 0.1532,
      "grad_norm": 0.028270110487937927,
      "learning_rate": 0.00019829353491495545,
      "loss": 2.2783,
      "step": 766
    },
    {
      "epoch": 0.1534,
      "grad_norm": 0.1533058136701584,
      "learning_rate": 0.00019828066873553448,
      "loss": 2.2785,
      "step": 767
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.17202724516391754,
      "learning_rate": 0.0001982677546553095,
      "loss": 2.2785,
      "step": 768
    },
    {
      "epoch": 0.1538,
      "grad_norm": 0.0957600325345993,
      "learning_rate": 0.00019825479268057467,
      "loss": 2.2784,
      "step": 769
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.018903283402323723,
      "learning_rate": 0.00019824178281764753,
      "loss": 2.278,
      "step": 770
    },
    {
      "epoch": 0.1542,
      "grad_norm": 0.08487782627344131,
      "learning_rate": 0.0001982287250728689,
      "loss": 2.2776,
      "step": 771
    },
    {
      "epoch": 0.1544,
      "grad_norm": 0.09570179134607315,
      "learning_rate": 0.0001982156194526029,
      "loss": 2.2779,
      "step": 772
    },
    {
      "epoch": 0.1546,
      "grad_norm": 0.049390994012355804,
      "learning_rate": 0.0001982024659632372,
      "loss": 2.2782,
      "step": 773
    },
    {
      "epoch": 0.1548,
      "grad_norm": 0.022077850997447968,
      "learning_rate": 0.00019818926461118253,
      "loss": 2.2773,
      "step": 774
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.06410589069128036,
      "learning_rate": 0.00019817601540287306,
      "loss": 2.277,
      "step": 775
    },
    {
      "epoch": 0.1552,
      "grad_norm": 0.07010999321937561,
      "learning_rate": 0.00019816271834476642,
      "loss": 2.2778,
      "step": 776
    },
    {
      "epoch": 0.1554,
      "grad_norm": 0.03410787507891655,
      "learning_rate": 0.0001981493734433433,
      "loss": 2.2771,
      "step": 777
    },
    {
      "epoch": 0.1556,
      "grad_norm": 0.0164511539041996,
      "learning_rate": 0.00019813598070510792,
      "loss": 2.2778,
      "step": 778
    },
    {
      "epoch": 0.1558,
      "grad_norm": 0.01690712384879589,
      "learning_rate": 0.00019812254013658768,
      "loss": 2.2775,
      "step": 779
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.025936225429177284,
      "learning_rate": 0.0001981090517443334,
      "loss": 2.2783,
      "step": 780
    },
    {
      "epoch": 0.1562,
      "grad_norm": 0.03492847457528114,
      "learning_rate": 0.00019809551553491916,
      "loss": 2.2768,
      "step": 781
    },
    {
      "epoch": 0.1564,
      "grad_norm": 0.014003599062561989,
      "learning_rate": 0.00019808193151494232,
      "loss": 2.2757,
      "step": 782
    },
    {
      "epoch": 0.1566,
      "grad_norm": 0.027292242273688316,
      "learning_rate": 0.00019806829969102357,
      "loss": 2.2776,
      "step": 783
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.022686228156089783,
      "learning_rate": 0.00019805462006980689,
      "loss": 2.278,
      "step": 784
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.03363360837101936,
      "learning_rate": 0.0001980408926579596,
      "loss": 2.2778,
      "step": 785
    },
    {
      "epoch": 0.1572,
      "grad_norm": 0.07329356670379639,
      "learning_rate": 0.00019802711746217218,
      "loss": 2.2778,
      "step": 786
    },
    {
      "epoch": 0.1574,
      "grad_norm": 0.06073550879955292,
      "learning_rate": 0.00019801329448915862,
      "loss": 2.2778,
      "step": 787
    },
    {
      "epoch": 0.1576,
      "grad_norm": 0.02863689884543419,
      "learning_rate": 0.00019799942374565597,
      "loss": 2.2773,
      "step": 788
    },
    {
      "epoch": 0.1578,
      "grad_norm": 0.07947874814271927,
      "learning_rate": 0.0001979855052384247,
      "loss": 2.2782,
      "step": 789
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.07919802516698837,
      "learning_rate": 0.00019797153897424852,
      "loss": 2.2777,
      "step": 790
    },
    {
      "epoch": 0.1582,
      "grad_norm": 0.05466971918940544,
      "learning_rate": 0.0001979575249599344,
      "loss": 2.2777,
      "step": 791
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.019060084596276283,
      "learning_rate": 0.00019794346320231265,
      "loss": 2.2778,
      "step": 792
    },
    {
      "epoch": 0.1586,
      "grad_norm": 0.043811772018671036,
      "learning_rate": 0.00019792935370823675,
      "loss": 2.2769,
      "step": 793
    },
    {
      "epoch": 0.1588,
      "grad_norm": 0.09096509218215942,
      "learning_rate": 0.00019791519648458352,
      "loss": 2.2774,
      "step": 794
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.08400837332010269,
      "learning_rate": 0.000197900991538253,
      "loss": 2.2773,
      "step": 795
    },
    {
      "epoch": 0.1592,
      "grad_norm": 0.03962600603699684,
      "learning_rate": 0.0001978867388761685,
      "loss": 2.2777,
      "step": 796
    },
    {
      "epoch": 0.1594,
      "grad_norm": 0.025669757276773453,
      "learning_rate": 0.00019787243850527664,
      "loss": 2.2773,
      "step": 797
    },
    {
      "epoch": 0.1596,
      "grad_norm": 0.08377869427204132,
      "learning_rate": 0.00019785809043254722,
      "loss": 2.2769,
      "step": 798
    },
    {
      "epoch": 0.1598,
      "grad_norm": 0.09462933242321014,
      "learning_rate": 0.0001978436946649733,
      "loss": 2.2779,
      "step": 799
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.04146955534815788,
      "learning_rate": 0.00019782925120957124,
      "loss": 2.2776,
      "step": 800
    },
    {
      "epoch": 0.1602,
      "grad_norm": 0.020734094083309174,
      "learning_rate": 0.00019781476007338058,
      "loss": 2.2771,
      "step": 801
    },
    {
      "epoch": 0.1604,
      "grad_norm": 0.022373395040631294,
      "learning_rate": 0.0001978002212634641,
      "loss": 2.2772,
      "step": 802
    },
    {
      "epoch": 0.1606,
      "grad_norm": 0.027219176292419434,
      "learning_rate": 0.0001977856347869079,
      "loss": 2.277,
      "step": 803
    },
    {
      "epoch": 0.1608,
      "grad_norm": 0.05763978138566017,
      "learning_rate": 0.00019777100065082118,
      "loss": 2.2768,
      "step": 804
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.04742516949772835,
      "learning_rate": 0.00019775631886233654,
      "loss": 2.2771,
      "step": 805
    },
    {
      "epoch": 0.1612,
      "grad_norm": 0.02655097283422947,
      "learning_rate": 0.0001977415894286096,
      "loss": 2.2772,
      "step": 806
    },
    {
      "epoch": 0.1614,
      "grad_norm": 0.1149844229221344,
      "learning_rate": 0.00019772681235681936,
      "loss": 2.2773,
      "step": 807
    },
    {
      "epoch": 0.1616,
      "grad_norm": 0.18978655338287354,
      "learning_rate": 0.000197711987654168,
      "loss": 2.2775,
      "step": 808
    },
    {
      "epoch": 0.1618,
      "grad_norm": 0.21619023382663727,
      "learning_rate": 0.00019769711532788083,
      "loss": 2.2777,
      "step": 809
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.14237523078918457,
      "learning_rate": 0.0001976821953852065,
      "loss": 2.2778,
      "step": 810
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.026844708248972893,
      "learning_rate": 0.0001976672278334168,
      "loss": 2.2775,
      "step": 811
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.1642199605703354,
      "learning_rate": 0.00019765221267980675,
      "loss": 2.2781,
      "step": 812
    },
    {
      "epoch": 0.1626,
      "grad_norm": 0.17302243411540985,
      "learning_rate": 0.00019763714993169452,
      "loss": 2.2776,
      "step": 813
    },
    {
      "epoch": 0.1628,
      "grad_norm": 0.043105266988277435,
      "learning_rate": 0.0001976220395964215,
      "loss": 2.2774,
      "step": 814
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.122604601085186,
      "learning_rate": 0.00019760688168135232,
      "loss": 2.2774,
      "step": 815
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.22143201529979706,
      "learning_rate": 0.00019759167619387476,
      "loss": 2.2784,
      "step": 816
    },
    {
      "epoch": 0.1634,
      "grad_norm": 0.2061559110879898,
      "learning_rate": 0.00019757642314139977,
      "loss": 2.2776,
      "step": 817
    },
    {
      "epoch": 0.1636,
      "grad_norm": 0.07330457121133804,
      "learning_rate": 0.0001975611225313615,
      "loss": 2.2777,
      "step": 818
    },
    {
      "epoch": 0.1638,
      "grad_norm": 0.09442374110221863,
      "learning_rate": 0.00019754577437121733,
      "loss": 2.2771,
      "step": 819
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.16002941131591797,
      "learning_rate": 0.00019753037866844771,
      "loss": 2.2775,
      "step": 820
    },
    {
      "epoch": 0.1642,
      "grad_norm": 0.07086198031902313,
      "learning_rate": 0.00019751493543055632,
      "loss": 2.2776,
      "step": 821
    },
    {
      "epoch": 0.1644,
      "grad_norm": 0.0960012897849083,
      "learning_rate": 0.00019749944466507008,
      "loss": 2.2777,
      "step": 822
    },
    {
      "epoch": 0.1646,
      "grad_norm": 0.2062522917985916,
      "learning_rate": 0.0001974839063795389,
      "loss": 2.2777,
      "step": 823
    },
    {
      "epoch": 0.1648,
      "grad_norm": 0.22492897510528564,
      "learning_rate": 0.00019746832058153602,
      "loss": 2.2781,
      "step": 824
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.13917921483516693,
      "learning_rate": 0.00019745268727865774,
      "loss": 2.2779,
      "step": 825
    },
    {
      "epoch": 0.1652,
      "grad_norm": 0.0389823243021965,
      "learning_rate": 0.00019743700647852354,
      "loss": 2.2766,
      "step": 826
    },
    {
      "epoch": 0.1654,
      "grad_norm": 0.20714466273784637,
      "learning_rate": 0.00019742127818877606,
      "loss": 2.278,
      "step": 827
    },
    {
      "epoch": 0.1656,
      "grad_norm": 0.267648845911026,
      "learning_rate": 0.00019740550241708108,
      "loss": 2.277,
      "step": 828
    },
    {
      "epoch": 0.1658,
      "grad_norm": 0.20970462262630463,
      "learning_rate": 0.0001973896791711275,
      "loss": 2.2778,
      "step": 829
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.07883492857217789,
      "learning_rate": 0.00019737380845862745,
      "loss": 2.2774,
      "step": 830
    },
    {
      "epoch": 0.1662,
      "grad_norm": 0.057048212736845016,
      "learning_rate": 0.00019735789028731604,
      "loss": 2.278,
      "step": 831
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.11497144401073456,
      "learning_rate": 0.00019734192466495162,
      "loss": 2.2774,
      "step": 832
    },
    {
      "epoch": 0.1666,
      "grad_norm": 0.08913909643888474,
      "learning_rate": 0.0001973259115993156,
      "loss": 2.2777,
      "step": 833
    },
    {
      "epoch": 0.1668,
      "grad_norm": 0.015228581614792347,
      "learning_rate": 0.00019730985109821266,
      "loss": 2.2774,
      "step": 834
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.07681162655353546,
      "learning_rate": 0.0001972937431694704,
      "loss": 2.2775,
      "step": 835
    },
    {
      "epoch": 0.1672,
      "grad_norm": 0.10126935690641403,
      "learning_rate": 0.00019727758782093967,
      "loss": 2.2776,
      "step": 836
    },
    {
      "epoch": 0.1674,
      "grad_norm": 0.05074957013130188,
      "learning_rate": 0.00019726138506049438,
      "loss": 2.2777,
      "step": 837
    },
    {
      "epoch": 0.1676,
      "grad_norm": 0.05678354203701019,
      "learning_rate": 0.00019724513489603155,
      "loss": 2.2774,
      "step": 838
    },
    {
      "epoch": 0.1678,
      "grad_norm": 0.1244342029094696,
      "learning_rate": 0.00019722883733547128,
      "loss": 2.2773,
      "step": 839
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.10281989723443985,
      "learning_rate": 0.00019721249238675688,
      "loss": 2.2777,
      "step": 840
    },
    {
      "epoch": 0.1682,
      "grad_norm": 0.016955170780420303,
      "learning_rate": 0.00019719610005785465,
      "loss": 2.2771,
      "step": 841
    },
    {
      "epoch": 0.1684,
      "grad_norm": 0.07747147977352142,
      "learning_rate": 0.00019717966035675397,
      "loss": 2.2774,
      "step": 842
    },
    {
      "epoch": 0.1686,
      "grad_norm": 0.1168031096458435,
      "learning_rate": 0.0001971631732914674,
      "loss": 2.2772,
      "step": 843
    },
    {
      "epoch": 0.1688,
      "grad_norm": 0.12153472751379013,
      "learning_rate": 0.00019714663887003054,
      "loss": 2.2778,
      "step": 844
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.07379572093486786,
      "learning_rate": 0.000197130057100502,
      "loss": 2.2777,
      "step": 845
    },
    {
      "epoch": 0.1692,
      "grad_norm": 0.04453488439321518,
      "learning_rate": 0.00019711342799096361,
      "loss": 2.2772,
      "step": 846
    },
    {
      "epoch": 0.1694,
      "grad_norm": 0.14964714646339417,
      "learning_rate": 0.00019709675154952017,
      "loss": 2.2765,
      "step": 847
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.15212681889533997,
      "learning_rate": 0.00019708002778429955,
      "loss": 2.2773,
      "step": 848
    },
    {
      "epoch": 0.1698,
      "grad_norm": 0.05380436033010483,
      "learning_rate": 0.00019706325670345275,
      "loss": 2.2776,
      "step": 849
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.05452640727162361,
      "learning_rate": 0.00019704643831515374,
      "loss": 2.2773,
      "step": 850
    },
    {
      "epoch": 0.1702,
      "grad_norm": 0.11623701453208923,
      "learning_rate": 0.00019702957262759965,
      "loss": 2.2778,
      "step": 851
    },
    {
      "epoch": 0.1704,
      "grad_norm": 0.11594992876052856,
      "learning_rate": 0.0001970126596490106,
      "loss": 2.2769,
      "step": 852
    },
    {
      "epoch": 0.1706,
      "grad_norm": 0.03919559344649315,
      "learning_rate": 0.00019699569938762973,
      "loss": 2.2775,
      "step": 853
    },
    {
      "epoch": 0.1708,
      "grad_norm": 0.07343373447656631,
      "learning_rate": 0.00019697869185172331,
      "loss": 2.2772,
      "step": 854
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.13522569835186005,
      "learning_rate": 0.0001969616370495806,
      "loss": 2.2774,
      "step": 855
    },
    {
      "epoch": 0.1712,
      "grad_norm": 0.1344885379076004,
      "learning_rate": 0.0001969445349895139,
      "loss": 2.2776,
      "step": 856
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.06731849908828735,
      "learning_rate": 0.00019692738567985853,
      "loss": 2.2772,
      "step": 857
    },
    {
      "epoch": 0.1716,
      "grad_norm": 0.04796075448393822,
      "learning_rate": 0.00019691018912897286,
      "loss": 2.2771,
      "step": 858
    },
    {
      "epoch": 0.1718,
      "grad_norm": 0.12386324256658554,
      "learning_rate": 0.0001968929453452383,
      "loss": 2.2763,
      "step": 859
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.11047438532114029,
      "learning_rate": 0.00019687565433705926,
      "loss": 2.2766,
      "step": 860
    },
    {
      "epoch": 0.1722,
      "grad_norm": 0.04133928567171097,
      "learning_rate": 0.0001968583161128631,
      "loss": 2.277,
      "step": 861
    },
    {
      "epoch": 0.1724,
      "grad_norm": 0.037795182317495346,
      "learning_rate": 0.00019684093068110038,
      "loss": 2.2764,
      "step": 862
    },
    {
      "epoch": 0.1726,
      "grad_norm": 0.08269113302230835,
      "learning_rate": 0.00019682349805024446,
      "loss": 2.2774,
      "step": 863
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.09728846698999405,
      "learning_rate": 0.00019680601822879182,
      "loss": 2.2775,
      "step": 864
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.08211679011583328,
      "learning_rate": 0.00019678849122526187,
      "loss": 2.2769,
      "step": 865
    },
    {
      "epoch": 0.1732,
      "grad_norm": 0.038585077971220016,
      "learning_rate": 0.00019677091704819715,
      "loss": 2.2774,
      "step": 866
    },
    {
      "epoch": 0.1734,
      "grad_norm": 0.023618334904313087,
      "learning_rate": 0.00019675329570616298,
      "loss": 2.2768,
      "step": 867
    },
    {
      "epoch": 0.1736,
      "grad_norm": 0.04759468138217926,
      "learning_rate": 0.00019673562720774792,
      "loss": 2.2769,
      "step": 868
    },
    {
      "epoch": 0.1738,
      "grad_norm": 0.05124516040086746,
      "learning_rate": 0.0001967179115615633,
      "loss": 2.2773,
      "step": 869
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.04508763551712036,
      "learning_rate": 0.00019670014877624353,
      "loss": 2.2772,
      "step": 870
    },
    {
      "epoch": 0.1742,
      "grad_norm": 0.04163754731416702,
      "learning_rate": 0.00019668233886044597,
      "loss": 2.2768,
      "step": 871
    },
    {
      "epoch": 0.1744,
      "grad_norm": 0.040363822132349014,
      "learning_rate": 0.00019666448182285094,
      "loss": 2.2772,
      "step": 872
    },
    {
      "epoch": 0.1746,
      "grad_norm": 0.031363628804683685,
      "learning_rate": 0.00019664657767216176,
      "loss": 2.2767,
      "step": 873
    },
    {
      "epoch": 0.1748,
      "grad_norm": 0.014453503303229809,
      "learning_rate": 0.0001966286264171047,
      "loss": 2.2772,
      "step": 874
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.024039695039391518,
      "learning_rate": 0.00019661062806642903,
      "loss": 2.2772,
      "step": 875
    },
    {
      "epoch": 0.1752,
      "grad_norm": 0.04461683705449104,
      "learning_rate": 0.00019659258262890683,
      "loss": 2.2771,
      "step": 876
    },
    {
      "epoch": 0.1754,
      "grad_norm": 0.053156472742557526,
      "learning_rate": 0.00019657449011333328,
      "loss": 2.2769,
      "step": 877
    },
    {
      "epoch": 0.1756,
      "grad_norm": 0.04546341672539711,
      "learning_rate": 0.00019655635052852647,
      "loss": 2.2773,
      "step": 878
    },
    {
      "epoch": 0.1758,
      "grad_norm": 0.015409605577588081,
      "learning_rate": 0.0001965381638833274,
      "loss": 2.2769,
      "step": 879
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.03428732231259346,
      "learning_rate": 0.0001965199301866,
      "loss": 2.2768,
      "step": 880
    },
    {
      "epoch": 0.1762,
      "grad_norm": 0.036051519215106964,
      "learning_rate": 0.00019650164944723115,
      "loss": 2.2767,
      "step": 881
    },
    {
      "epoch": 0.1764,
      "grad_norm": 0.014138374477624893,
      "learning_rate": 0.00019648332167413067,
      "loss": 2.2771,
      "step": 882
    },
    {
      "epoch": 0.1766,
      "grad_norm": 0.02256808616220951,
      "learning_rate": 0.00019646494687623135,
      "loss": 2.2755,
      "step": 883
    },
    {
      "epoch": 0.1768,
      "grad_norm": 0.016448775306344032,
      "learning_rate": 0.00019644652506248874,
      "loss": 2.2763,
      "step": 884
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.04505930840969086,
      "learning_rate": 0.00019642805624188147,
      "loss": 2.2761,
      "step": 885
    },
    {
      "epoch": 0.1772,
      "grad_norm": 0.05723484605550766,
      "learning_rate": 0.00019640954042341103,
      "loss": 2.2763,
      "step": 886
    },
    {
      "epoch": 0.1774,
      "grad_norm": 0.046228133141994476,
      "learning_rate": 0.00019639097761610174,
      "loss": 2.2759,
      "step": 887
    },
    {
      "epoch": 0.1776,
      "grad_norm": 0.020313933491706848,
      "learning_rate": 0.000196372367829001,
      "loss": 2.2759,
      "step": 888
    },
    {
      "epoch": 0.1778,
      "grad_norm": 0.022386573255062103,
      "learning_rate": 0.00019635371107117888,
      "loss": 2.2755,
      "step": 889
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.03975597396492958,
      "learning_rate": 0.0001963350073517285,
      "loss": 2.277,
      "step": 890
    },
    {
      "epoch": 0.1782,
      "grad_norm": 0.027703052386641502,
      "learning_rate": 0.00019631625667976583,
      "loss": 2.2769,
      "step": 891
    },
    {
      "epoch": 0.1784,
      "grad_norm": 0.015061361715197563,
      "learning_rate": 0.0001962974590644297,
      "loss": 2.2758,
      "step": 892
    },
    {
      "epoch": 0.1786,
      "grad_norm": 0.023242395371198654,
      "learning_rate": 0.00019627861451488189,
      "loss": 2.2767,
      "step": 893
    },
    {
      "epoch": 0.1788,
      "grad_norm": 0.01741502620279789,
      "learning_rate": 0.00019625972304030697,
      "loss": 2.2755,
      "step": 894
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.01690218411386013,
      "learning_rate": 0.0001962407846499124,
      "loss": 2.2762,
      "step": 895
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.030568817630410194,
      "learning_rate": 0.00019622179935292855,
      "loss": 2.2765,
      "step": 896
    },
    {
      "epoch": 0.1794,
      "grad_norm": 0.022824786603450775,
      "learning_rate": 0.0001962027671586086,
      "loss": 2.277,
      "step": 897
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.013926653191447258,
      "learning_rate": 0.00019618368807622862,
      "loss": 2.276,
      "step": 898
    },
    {
      "epoch": 0.1798,
      "grad_norm": 0.015729809179902077,
      "learning_rate": 0.00019616456211508752,
      "loss": 2.2767,
      "step": 899
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.034621451050043106,
      "learning_rate": 0.0001961453892845071,
      "loss": 2.2758,
      "step": 900
    },
    {
      "epoch": 0.1802,
      "grad_norm": 0.04504714533686638,
      "learning_rate": 0.0001961261695938319,
      "loss": 2.2754,
      "step": 901
    },
    {
      "epoch": 0.1804,
      "grad_norm": 0.023847922682762146,
      "learning_rate": 0.0001961069030524294,
      "loss": 2.2766,
      "step": 902
    },
    {
      "epoch": 0.1806,
      "grad_norm": 0.02119390107691288,
      "learning_rate": 0.00019608758966968988,
      "loss": 2.2768,
      "step": 903
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.03997572511434555,
      "learning_rate": 0.0001960682294550264,
      "loss": 2.2768,
      "step": 904
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.06063878536224365,
      "learning_rate": 0.00019604882241787498,
      "loss": 2.2761,
      "step": 905
    },
    {
      "epoch": 0.1812,
      "grad_norm": 0.05068110302090645,
      "learning_rate": 0.0001960293685676943,
      "loss": 2.2771,
      "step": 906
    },
    {
      "epoch": 0.1814,
      "grad_norm": 0.015309859998524189,
      "learning_rate": 0.000196009867913966,
      "loss": 2.2771,
      "step": 907
    },
    {
      "epoch": 0.1816,
      "grad_norm": 0.061038605868816376,
      "learning_rate": 0.00019599032046619438,
      "loss": 2.277,
      "step": 908
    },
    {
      "epoch": 0.1818,
      "grad_norm": 0.11889402568340302,
      "learning_rate": 0.00019597072623390668,
      "loss": 2.277,
      "step": 909
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.13131162524223328,
      "learning_rate": 0.0001959510852266529,
      "loss": 2.2767,
      "step": 910
    },
    {
      "epoch": 0.1822,
      "grad_norm": 0.06728943437337875,
      "learning_rate": 0.00019593139745400576,
      "loss": 2.275,
      "step": 911
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.0315927192568779,
      "learning_rate": 0.0001959116629255609,
      "loss": 2.2766,
      "step": 912
    },
    {
      "epoch": 0.1826,
      "grad_norm": 0.09226155281066895,
      "learning_rate": 0.0001958918816509367,
      "loss": 2.2766,
      "step": 913
    },
    {
      "epoch": 0.1828,
      "grad_norm": 0.1228950247168541,
      "learning_rate": 0.00019587205363977427,
      "loss": 2.2766,
      "step": 914
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.11624473333358765,
      "learning_rate": 0.0001958521789017376,
      "loss": 2.2769,
      "step": 915
    },
    {
      "epoch": 0.1832,
      "grad_norm": 0.043423689901828766,
      "learning_rate": 0.00019583225744651333,
      "loss": 2.2773,
      "step": 916
    },
    {
      "epoch": 0.1834,
      "grad_norm": 0.037720419466495514,
      "learning_rate": 0.00019581228928381098,
      "loss": 2.2773,
      "step": 917
    },
    {
      "epoch": 0.1836,
      "grad_norm": 0.05741436406970024,
      "learning_rate": 0.00019579227442336278,
      "loss": 2.2762,
      "step": 918
    },
    {
      "epoch": 0.1838,
      "grad_norm": 0.039177343249320984,
      "learning_rate": 0.00019577221287492367,
      "loss": 2.2765,
      "step": 919
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.01452620793133974,
      "learning_rate": 0.00019575210464827149,
      "loss": 2.2769,
      "step": 920
    },
    {
      "epoch": 0.1842,
      "grad_norm": 0.033221155405044556,
      "learning_rate": 0.00019573194975320673,
      "loss": 2.2769,
      "step": 921
    },
    {
      "epoch": 0.1844,
      "grad_norm": 0.04238063469529152,
      "learning_rate": 0.00019571174819955263,
      "loss": 2.2767,
      "step": 922
    },
    {
      "epoch": 0.1846,
      "grad_norm": 0.048513952642679214,
      "learning_rate": 0.00019569149999715515,
      "loss": 2.2771,
      "step": 923
    },
    {
      "epoch": 0.1848,
      "grad_norm": 0.05425655469298363,
      "learning_rate": 0.00019567120515588308,
      "loss": 2.2773,
      "step": 924
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.05590939149260521,
      "learning_rate": 0.0001956508636856278,
      "loss": 2.2771,
      "step": 925
    },
    {
      "epoch": 0.1852,
      "grad_norm": 0.0721336305141449,
      "learning_rate": 0.00019563047559630357,
      "loss": 2.2769,
      "step": 926
    },
    {
      "epoch": 0.1854,
      "grad_norm": 0.0844496414065361,
      "learning_rate": 0.00019561004089784723,
      "loss": 2.2763,
      "step": 927
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.04649016261100769,
      "learning_rate": 0.00019558955960021849,
      "loss": 2.276,
      "step": 928
    },
    {
      "epoch": 0.1858,
      "grad_norm": 0.028873570263385773,
      "learning_rate": 0.00019556903171339963,
      "loss": 2.2761,
      "step": 929
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.09972953051328659,
      "learning_rate": 0.00019554845724739566,
      "loss": 2.2764,
      "step": 930
    },
    {
      "epoch": 0.1862,
      "grad_norm": 0.14999909698963165,
      "learning_rate": 0.00019552783621223436,
      "loss": 2.2766,
      "step": 931
    },
    {
      "epoch": 0.1864,
      "grad_norm": 0.14394615590572357,
      "learning_rate": 0.00019550716861796623,
      "loss": 2.2751,
      "step": 932
    },
    {
      "epoch": 0.1866,
      "grad_norm": 0.07258053869009018,
      "learning_rate": 0.00019548645447466431,
      "loss": 2.2763,
      "step": 933
    },
    {
      "epoch": 0.1868,
      "grad_norm": 0.03492654487490654,
      "learning_rate": 0.00019546569379242444,
      "loss": 2.2763,
      "step": 934
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.11813312768936157,
      "learning_rate": 0.00019544488658136523,
      "loss": 2.2766,
      "step": 935
    },
    {
      "epoch": 0.1872,
      "grad_norm": 0.14365655183792114,
      "learning_rate": 0.0001954240328516277,
      "loss": 2.2751,
      "step": 936
    },
    {
      "epoch": 0.1874,
      "grad_norm": 0.10949747264385223,
      "learning_rate": 0.0001954031326133758,
      "loss": 2.2762,
      "step": 937
    },
    {
      "epoch": 0.1876,
      "grad_norm": 0.04050913825631142,
      "learning_rate": 0.00019538218587679605,
      "loss": 2.2754,
      "step": 938
    },
    {
      "epoch": 0.1878,
      "grad_norm": 0.044760603457689285,
      "learning_rate": 0.0001953611926520976,
      "loss": 2.2757,
      "step": 939
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.09899906069040298,
      "learning_rate": 0.00019534015294951233,
      "loss": 2.2762,
      "step": 940
    },
    {
      "epoch": 0.1882,
      "grad_norm": 0.09773119539022446,
      "learning_rate": 0.0001953190667792947,
      "loss": 2.2763,
      "step": 941
    },
    {
      "epoch": 0.1884,
      "grad_norm": 0.05877260863780975,
      "learning_rate": 0.00019529793415172192,
      "loss": 2.2769,
      "step": 942
    },
    {
      "epoch": 0.1886,
      "grad_norm": 0.017347227782011032,
      "learning_rate": 0.00019527675507709366,
      "loss": 2.2767,
      "step": 943
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.030595362186431885,
      "learning_rate": 0.00019525552956573244,
      "loss": 2.2767,
      "step": 944
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.04094453901052475,
      "learning_rate": 0.00019523425762798329,
      "loss": 2.2765,
      "step": 945
    },
    {
      "epoch": 0.1892,
      "grad_norm": 0.030689526349306107,
      "learning_rate": 0.00019521293927421388,
      "loss": 2.2758,
      "step": 946
    },
    {
      "epoch": 0.1894,
      "grad_norm": 0.01976362243294716,
      "learning_rate": 0.00019519157451481454,
      "loss": 2.2765,
      "step": 947
    },
    {
      "epoch": 0.1896,
      "grad_norm": 0.016150450333952904,
      "learning_rate": 0.0001951701633601982,
      "loss": 2.2764,
      "step": 948
    },
    {
      "epoch": 0.1898,
      "grad_norm": 0.040354691445827484,
      "learning_rate": 0.00019514870582080032,
      "loss": 2.2758,
      "step": 949
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.07688114047050476,
      "learning_rate": 0.00019512720190707913,
      "loss": 2.2766,
      "step": 950
    },
    {
      "epoch": 0.1902,
      "grad_norm": 0.09586871415376663,
      "learning_rate": 0.00019510565162951537,
      "loss": 2.2762,
      "step": 951
    },
    {
      "epoch": 0.1904,
      "grad_norm": 0.07364441454410553,
      "learning_rate": 0.00019508405499861232,
      "loss": 2.2761,
      "step": 952
    },
    {
      "epoch": 0.1906,
      "grad_norm": 0.02756970003247261,
      "learning_rate": 0.00019506241202489602,
      "loss": 2.2765,
      "step": 953
    },
    {
      "epoch": 0.1908,
      "grad_norm": 0.027276957407593727,
      "learning_rate": 0.00019504072271891488,
      "loss": 2.2761,
      "step": 954
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.05717033892869949,
      "learning_rate": 0.00019501898709124008,
      "loss": 2.2765,
      "step": 955
    },
    {
      "epoch": 0.1912,
      "grad_norm": 0.06790348887443542,
      "learning_rate": 0.00019499720515246525,
      "loss": 2.2764,
      "step": 956
    },
    {
      "epoch": 0.1914,
      "grad_norm": 0.05845238268375397,
      "learning_rate": 0.00019497537691320668,
      "loss": 2.2766,
      "step": 957
    },
    {
      "epoch": 0.1916,
      "grad_norm": 0.03277203068137169,
      "learning_rate": 0.0001949535023841032,
      "loss": 2.2765,
      "step": 958
    },
    {
      "epoch": 0.1918,
      "grad_norm": 0.016472307965159416,
      "learning_rate": 0.00019493158157581615,
      "loss": 2.2753,
      "step": 959
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.026944037526845932,
      "learning_rate": 0.00019490961449902946,
      "loss": 2.276,
      "step": 960
    },
    {
      "epoch": 0.1922,
      "grad_norm": 0.05147011578083038,
      "learning_rate": 0.00019488760116444966,
      "loss": 2.2752,
      "step": 961
    },
    {
      "epoch": 0.1924,
      "grad_norm": 0.045549385249614716,
      "learning_rate": 0.00019486554158280574,
      "loss": 2.2761,
      "step": 962
    },
    {
      "epoch": 0.1926,
      "grad_norm": 0.021716147661209106,
      "learning_rate": 0.00019484343576484933,
      "loss": 2.2759,
      "step": 963
    },
    {
      "epoch": 0.1928,
      "grad_norm": 0.019031288102269173,
      "learning_rate": 0.00019482128372135446,
      "loss": 2.2753,
      "step": 964
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.02211000584065914,
      "learning_rate": 0.00019479908546311781,
      "loss": 2.2756,
      "step": 965
    },
    {
      "epoch": 0.1932,
      "grad_norm": 0.016877086833119392,
      "learning_rate": 0.0001947768410009586,
      "loss": 2.2756,
      "step": 966
    },
    {
      "epoch": 0.1934,
      "grad_norm": 0.026217222213745117,
      "learning_rate": 0.00019475455034571838,
      "loss": 2.2759,
      "step": 967
    },
    {
      "epoch": 0.1936,
      "grad_norm": 0.06791318207979202,
      "learning_rate": 0.00019473221350826142,
      "loss": 2.2758,
      "step": 968
    },
    {
      "epoch": 0.1938,
      "grad_norm": 0.10495511442422867,
      "learning_rate": 0.00019470983049947444,
      "loss": 2.2764,
      "step": 969
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.10060416907072067,
      "learning_rate": 0.0001946874013302666,
      "loss": 2.2759,
      "step": 970
    },
    {
      "epoch": 0.1942,
      "grad_norm": 0.06912078708410263,
      "learning_rate": 0.00019466492601156966,
      "loss": 2.2751,
      "step": 971
    },
    {
      "epoch": 0.1944,
      "grad_norm": 0.04309812933206558,
      "learning_rate": 0.00019464240455433775,
      "loss": 2.2763,
      "step": 972
    },
    {
      "epoch": 0.1946,
      "grad_norm": 0.02277280017733574,
      "learning_rate": 0.00019461983696954758,
      "loss": 2.2768,
      "step": 973
    },
    {
      "epoch": 0.1948,
      "grad_norm": 0.025801392272114754,
      "learning_rate": 0.00019459722326819838,
      "loss": 2.2753,
      "step": 974
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.03906439617276192,
      "learning_rate": 0.0001945745634613117,
      "loss": 2.2762,
      "step": 975
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.03675737977027893,
      "learning_rate": 0.0001945518575599317,
      "loss": 2.2763,
      "step": 976
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.051766298711299896,
      "learning_rate": 0.00019452910557512496,
      "loss": 2.2757,
      "step": 977
    },
    {
      "epoch": 0.1956,
      "grad_norm": 0.0709015429019928,
      "learning_rate": 0.00019450630751798048,
      "loss": 2.2762,
      "step": 978
    },
    {
      "epoch": 0.1958,
      "grad_norm": 0.07053941488265991,
      "learning_rate": 0.00019448346339960982,
      "loss": 2.2761,
      "step": 979
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.048651594668626785,
      "learning_rate": 0.0001944605732311469,
      "loss": 2.2763,
      "step": 980
    },
    {
      "epoch": 0.1962,
      "grad_norm": 0.014794922433793545,
      "learning_rate": 0.00019443763702374812,
      "loss": 2.2753,
      "step": 981
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.05528155341744423,
      "learning_rate": 0.00019441465478859228,
      "loss": 2.2761,
      "step": 982
    },
    {
      "epoch": 0.1966,
      "grad_norm": 0.08325508236885071,
      "learning_rate": 0.00019439162653688065,
      "loss": 2.2757,
      "step": 983
    },
    {
      "epoch": 0.1968,
      "grad_norm": 0.07241332530975342,
      "learning_rate": 0.00019436855227983695,
      "loss": 2.2762,
      "step": 984
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.05063369870185852,
      "learning_rate": 0.00019434543202870725,
      "loss": 2.2765,
      "step": 985
    },
    {
      "epoch": 0.1972,
      "grad_norm": 0.01754833571612835,
      "learning_rate": 0.0001943222657947601,
      "loss": 2.2763,
      "step": 986
    },
    {
      "epoch": 0.1974,
      "grad_norm": 0.08200036734342575,
      "learning_rate": 0.00019429905358928646,
      "loss": 2.2753,
      "step": 987
    },
    {
      "epoch": 0.1976,
      "grad_norm": 0.17413122951984406,
      "learning_rate": 0.00019427579542359965,
      "loss": 2.2757,
      "step": 988
    },
    {
      "epoch": 0.1978,
      "grad_norm": 0.23842550814151764,
      "learning_rate": 0.00019425249130903543,
      "loss": 2.2761,
      "step": 989
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.24179133772850037,
      "learning_rate": 0.0001942291412569519,
      "loss": 2.2762,
      "step": 990
    },
    {
      "epoch": 0.1982,
      "grad_norm": 0.1548772156238556,
      "learning_rate": 0.00019420574527872968,
      "loss": 2.2759,
      "step": 991
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.02866436541080475,
      "learning_rate": 0.0001941823033857716,
      "loss": 2.2758,
      "step": 992
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.09405525028705597,
      "learning_rate": 0.00019415881558950302,
      "loss": 2.2762,
      "step": 993
    },
    {
      "epoch": 0.1988,
      "grad_norm": 0.17289823293685913,
      "learning_rate": 0.0001941352819013716,
      "loss": 2.2755,
      "step": 994
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.21196381747722626,
      "learning_rate": 0.00019411170233284727,
      "loss": 2.2766,
      "step": 995
    },
    {
      "epoch": 0.1992,
      "grad_norm": 0.20563484728336334,
      "learning_rate": 0.00019408807689542257,
      "loss": 2.276,
      "step": 996
    },
    {
      "epoch": 0.1994,
      "grad_norm": 0.14772312343120575,
      "learning_rate": 0.00019406440560061216,
      "loss": 2.2762,
      "step": 997
    },
    {
      "epoch": 0.1996,
      "grad_norm": 0.04696567356586456,
      "learning_rate": 0.00019404068845995317,
      "loss": 2.2754,
      "step": 998
    },
    {
      "epoch": 0.1998,
      "grad_norm": 0.060809534043073654,
      "learning_rate": 0.00019401692548500502,
      "loss": 2.2759,
      "step": 999
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.13851693272590637,
      "learning_rate": 0.00019399311668734956,
      "loss": 2.2757,
      "step": 1000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 0.17955000698566437,
      "learning_rate": 0.00019396926207859084,
      "loss": 2.2758,
      "step": 1001
    },
    {
      "epoch": 0.2004,
      "grad_norm": 0.17860695719718933,
      "learning_rate": 0.00019394536167035534,
      "loss": 2.2761,
      "step": 1002
    },
    {
      "epoch": 0.2006,
      "grad_norm": 0.14230038225650787,
      "learning_rate": 0.00019392141547429183,
      "loss": 2.2756,
      "step": 1003
    },
    {
      "epoch": 0.2008,
      "grad_norm": 0.07779045403003693,
      "learning_rate": 0.00019389742350207141,
      "loss": 2.2756,
      "step": 1004
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.023380029946565628,
      "learning_rate": 0.00019387338576538744,
      "loss": 2.2748,
      "step": 1005
    },
    {
      "epoch": 0.2012,
      "grad_norm": 0.09567804634571075,
      "learning_rate": 0.0001938493022759556,
      "loss": 2.276,
      "step": 1006
    },
    {
      "epoch": 0.2014,
      "grad_norm": 0.12242486327886581,
      "learning_rate": 0.00019382517304551396,
      "loss": 2.2756,
      "step": 1007
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.10110235959291458,
      "learning_rate": 0.00019380099808582278,
      "loss": 2.276,
      "step": 1008
    },
    {
      "epoch": 0.2018,
      "grad_norm": 0.037292394787073135,
      "learning_rate": 0.0001937767774086646,
      "loss": 2.2754,
      "step": 1009
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.05538066104054451,
      "learning_rate": 0.0001937525110258444,
      "loss": 2.2744,
      "step": 1010
    },
    {
      "epoch": 0.2022,
      "grad_norm": 0.12209614366292953,
      "learning_rate": 0.00019372819894918915,
      "loss": 2.2756,
      "step": 1011
    },
    {
      "epoch": 0.2024,
      "grad_norm": 0.1468183398246765,
      "learning_rate": 0.0001937038411905484,
      "loss": 2.2753,
      "step": 1012
    },
    {
      "epoch": 0.2026,
      "grad_norm": 0.11472291499376297,
      "learning_rate": 0.0001936794377617938,
      "loss": 2.276,
      "step": 1013
    },
    {
      "epoch": 0.2028,
      "grad_norm": 0.054645273834466934,
      "learning_rate": 0.00019365498867481923,
      "loss": 2.2756,
      "step": 1014
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.017564281821250916,
      "learning_rate": 0.00019363049394154094,
      "loss": 2.2746,
      "step": 1015
    },
    {
      "epoch": 0.2032,
      "grad_norm": 0.06848401576280594,
      "learning_rate": 0.00019360595357389735,
      "loss": 2.275,
      "step": 1016
    },
    {
      "epoch": 0.2034,
      "grad_norm": 0.11628009378910065,
      "learning_rate": 0.00019358136758384912,
      "loss": 2.2758,
      "step": 1017
    },
    {
      "epoch": 0.2036,
      "grad_norm": 0.12202831357717514,
      "learning_rate": 0.00019355673598337914,
      "loss": 2.2756,
      "step": 1018
    },
    {
      "epoch": 0.2038,
      "grad_norm": 0.08781236410140991,
      "learning_rate": 0.00019353205878449258,
      "loss": 2.2752,
      "step": 1019
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.03883212059736252,
      "learning_rate": 0.00019350733599921683,
      "loss": 2.276,
      "step": 1020
    },
    {
      "epoch": 0.2042,
      "grad_norm": 0.018991349264979362,
      "learning_rate": 0.00019348256763960145,
      "loss": 2.2746,
      "step": 1021
    },
    {
      "epoch": 0.2044,
      "grad_norm": 0.05409194156527519,
      "learning_rate": 0.00019345775371771824,
      "loss": 2.275,
      "step": 1022
    },
    {
      "epoch": 0.2046,
      "grad_norm": 0.10024632513523102,
      "learning_rate": 0.00019343289424566122,
      "loss": 2.2754,
      "step": 1023
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.1387016773223877,
      "learning_rate": 0.00019340798923554657,
      "loss": 2.2761,
      "step": 1024
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.12565483152866364,
      "learning_rate": 0.00019338303869951269,
      "loss": 2.2756,
      "step": 1025
    },
    {
      "epoch": 0.2052,
      "grad_norm": 0.07241412997245789,
      "learning_rate": 0.00019335804264972018,
      "loss": 2.2753,
      "step": 1026
    },
    {
      "epoch": 0.2054,
      "grad_norm": 0.020123500376939774,
      "learning_rate": 0.0001933330010983518,
      "loss": 2.2753,
      "step": 1027
    },
    {
      "epoch": 0.2056,
      "grad_norm": 0.050656650215387344,
      "learning_rate": 0.00019330791405761252,
      "loss": 2.2761,
      "step": 1028
    },
    {
      "epoch": 0.2058,
      "grad_norm": 0.12171557545661926,
      "learning_rate": 0.00019328278153972947,
      "loss": 2.2751,
      "step": 1029
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.16283775866031647,
      "learning_rate": 0.00019325760355695188,
      "loss": 2.2757,
      "step": 1030
    },
    {
      "epoch": 0.2062,
      "grad_norm": 0.15986451506614685,
      "learning_rate": 0.00019323238012155123,
      "loss": 2.2755,
      "step": 1031
    },
    {
      "epoch": 0.2064,
      "grad_norm": 0.11766772717237473,
      "learning_rate": 0.0001932071112458211,
      "loss": 2.2758,
      "step": 1032
    },
    {
      "epoch": 0.2066,
      "grad_norm": 0.07162011414766312,
      "learning_rate": 0.00019318179694207725,
      "loss": 2.2758,
      "step": 1033
    },
    {
      "epoch": 0.2068,
      "grad_norm": 0.032585836946964264,
      "learning_rate": 0.00019315643722265757,
      "loss": 2.2752,
      "step": 1034
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.02295178361237049,
      "learning_rate": 0.00019313103209992204,
      "loss": 2.2748,
      "step": 1035
    },
    {
      "epoch": 0.2072,
      "grad_norm": 0.0483243390917778,
      "learning_rate": 0.00019310558158625285,
      "loss": 2.2747,
      "step": 1036
    },
    {
      "epoch": 0.2074,
      "grad_norm": 0.04004925489425659,
      "learning_rate": 0.00019308008569405422,
      "loss": 2.2755,
      "step": 1037
    },
    {
      "epoch": 0.2076,
      "grad_norm": 0.01955476403236389,
      "learning_rate": 0.00019305454443575262,
      "loss": 2.2755,
      "step": 1038
    },
    {
      "epoch": 0.2078,
      "grad_norm": 0.01889149099588394,
      "learning_rate": 0.0001930289578237965,
      "loss": 2.2748,
      "step": 1039
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.04028429463505745,
      "learning_rate": 0.0001930033258706564,
      "loss": 2.2751,
      "step": 1040
    },
    {
      "epoch": 0.2082,
      "grad_norm": 0.058958880603313446,
      "learning_rate": 0.00019297764858882514,
      "loss": 2.2759,
      "step": 1041
    },
    {
      "epoch": 0.2084,
      "grad_norm": 0.0684417188167572,
      "learning_rate": 0.00019295192599081746,
      "loss": 2.2755,
      "step": 1042
    },
    {
      "epoch": 0.2086,
      "grad_norm": 0.07066122442483902,
      "learning_rate": 0.00019292615808917026,
      "loss": 2.2745,
      "step": 1043
    },
    {
      "epoch": 0.2088,
      "grad_norm": 0.06518283486366272,
      "learning_rate": 0.00019290034489644246,
      "loss": 2.2758,
      "step": 1044
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.05169698968529701,
      "learning_rate": 0.00019287448642521513,
      "loss": 2.2757,
      "step": 1045
    },
    {
      "epoch": 0.2092,
      "grad_norm": 0.029688889160752296,
      "learning_rate": 0.00019284858268809137,
      "loss": 2.2753,
      "step": 1046
    },
    {
      "epoch": 0.2094,
      "grad_norm": 0.017427178099751472,
      "learning_rate": 0.00019282263369769633,
      "loss": 2.2753,
      "step": 1047
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.03195904567837715,
      "learning_rate": 0.00019279663946667727,
      "loss": 2.2754,
      "step": 1048
    },
    {
      "epoch": 0.2098,
      "grad_norm": 0.04563708230853081,
      "learning_rate": 0.00019277060000770342,
      "loss": 2.2747,
      "step": 1049
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.047819167375564575,
      "learning_rate": 0.00019274451533346615,
      "loss": 2.2755,
      "step": 1050
    },
    {
      "epoch": 0.2102,
      "grad_norm": 0.0444926954805851,
      "learning_rate": 0.00019271838545667876,
      "loss": 2.2753,
      "step": 1051
    },
    {
      "epoch": 0.2104,
      "grad_norm": 0.03862835094332695,
      "learning_rate": 0.00019269221039007665,
      "loss": 2.2756,
      "step": 1052
    },
    {
      "epoch": 0.2106,
      "grad_norm": 0.04083564132452011,
      "learning_rate": 0.0001926659901464172,
      "loss": 2.2753,
      "step": 1053
    },
    {
      "epoch": 0.2108,
      "grad_norm": 0.03982382267713547,
      "learning_rate": 0.00019263972473847993,
      "loss": 2.2755,
      "step": 1054
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.02759697660803795,
      "learning_rate": 0.00019261341417906621,
      "loss": 2.2752,
      "step": 1055
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.0167767945677042,
      "learning_rate": 0.0001925870584809995,
      "loss": 2.2755,
      "step": 1056
    },
    {
      "epoch": 0.2114,
      "grad_norm": 0.0164140984416008,
      "learning_rate": 0.00019256065765712522,
      "loss": 2.2756,
      "step": 1057
    },
    {
      "epoch": 0.2116,
      "grad_norm": 0.024128392338752747,
      "learning_rate": 0.00019253421172031086,
      "loss": 2.2747,
      "step": 1058
    },
    {
      "epoch": 0.2118,
      "grad_norm": 0.04931613802909851,
      "learning_rate": 0.0001925077206834458,
      "loss": 2.2752,
      "step": 1059
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.07539703696966171,
      "learning_rate": 0.0001924811845594415,
      "loss": 2.2749,
      "step": 1060
    },
    {
      "epoch": 0.2122,
      "grad_norm": 0.08213331550359726,
      "learning_rate": 0.00019245460336123134,
      "loss": 2.2751,
      "step": 1061
    },
    {
      "epoch": 0.2124,
      "grad_norm": 0.07719892263412476,
      "learning_rate": 0.0001924279771017706,
      "loss": 2.2754,
      "step": 1062
    },
    {
      "epoch": 0.2126,
      "grad_norm": 0.0431867279112339,
      "learning_rate": 0.0001924013057940367,
      "loss": 2.2749,
      "step": 1063
    },
    {
      "epoch": 0.2128,
      "grad_norm": 0.0295414999127388,
      "learning_rate": 0.00019237458945102882,
      "loss": 2.2756,
      "step": 1064
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.09315326809883118,
      "learning_rate": 0.00019234782808576824,
      "loss": 2.2751,
      "step": 1065
    },
    {
      "epoch": 0.2132,
      "grad_norm": 0.15217578411102295,
      "learning_rate": 0.00019232102171129811,
      "loss": 2.2745,
      "step": 1066
    },
    {
      "epoch": 0.2134,
      "grad_norm": 0.18975219130516052,
      "learning_rate": 0.0001922941703406835,
      "loss": 2.2755,
      "step": 1067
    },
    {
      "epoch": 0.2136,
      "grad_norm": 0.18854458630084991,
      "learning_rate": 0.0001922672739870115,
      "loss": 2.2741,
      "step": 1068
    },
    {
      "epoch": 0.2138,
      "grad_norm": 0.17744384706020355,
      "learning_rate": 0.00019224033266339102,
      "loss": 2.2754,
      "step": 1069
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.15391021966934204,
      "learning_rate": 0.00019221334638295294,
      "loss": 2.2743,
      "step": 1070
    },
    {
      "epoch": 0.2142,
      "grad_norm": 0.10269510746002197,
      "learning_rate": 0.00019218631515885006,
      "loss": 2.2752,
      "step": 1071
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.03851315751671791,
      "learning_rate": 0.00019215923900425707,
      "loss": 2.2751,
      "step": 1072
    },
    {
      "epoch": 0.2146,
      "grad_norm": 0.03018578700721264,
      "learning_rate": 0.00019213211793237057,
      "loss": 2.2749,
      "step": 1073
    },
    {
      "epoch": 0.2148,
      "grad_norm": 0.09061158448457718,
      "learning_rate": 0.00019210495195640895,
      "loss": 2.275,
      "step": 1074
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.1153443455696106,
      "learning_rate": 0.00019207774108961272,
      "loss": 2.2748,
      "step": 1075
    },
    {
      "epoch": 0.2152,
      "grad_norm": 0.08077684789896011,
      "learning_rate": 0.00019205048534524406,
      "loss": 2.2747,
      "step": 1076
    },
    {
      "epoch": 0.2154,
      "grad_norm": 0.037089720368385315,
      "learning_rate": 0.00019202318473658705,
      "loss": 2.2752,
      "step": 1077
    },
    {
      "epoch": 0.2156,
      "grad_norm": 0.01737387292087078,
      "learning_rate": 0.00019199583927694772,
      "loss": 2.2753,
      "step": 1078
    },
    {
      "epoch": 0.2158,
      "grad_norm": 0.025888439267873764,
      "learning_rate": 0.00019196844897965393,
      "loss": 2.2752,
      "step": 1079
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.0315244123339653,
      "learning_rate": 0.0001919410138580553,
      "loss": 2.2748,
      "step": 1080
    },
    {
      "epoch": 0.2162,
      "grad_norm": 0.034914616495370865,
      "learning_rate": 0.00019191353392552344,
      "loss": 2.2744,
      "step": 1081
    },
    {
      "epoch": 0.2164,
      "grad_norm": 0.038672249764204025,
      "learning_rate": 0.00019188600919545174,
      "loss": 2.2744,
      "step": 1082
    },
    {
      "epoch": 0.2166,
      "grad_norm": 0.0472988523542881,
      "learning_rate": 0.0001918584396812554,
      "loss": 2.2747,
      "step": 1083
    },
    {
      "epoch": 0.2168,
      "grad_norm": 0.05503524839878082,
      "learning_rate": 0.00019183082539637146,
      "loss": 2.2751,
      "step": 1084
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.06091729551553726,
      "learning_rate": 0.0001918031663542588,
      "loss": 2.2748,
      "step": 1085
    },
    {
      "epoch": 0.2172,
      "grad_norm": 0.06968719512224197,
      "learning_rate": 0.00019177546256839812,
      "loss": 2.275,
      "step": 1086
    },
    {
      "epoch": 0.2174,
      "grad_norm": 0.07177258282899857,
      "learning_rate": 0.00019174771405229186,
      "loss": 2.275,
      "step": 1087
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.06453506648540497,
      "learning_rate": 0.00019171992081946435,
      "loss": 2.2747,
      "step": 1088
    },
    {
      "epoch": 0.2178,
      "grad_norm": 0.05271800979971886,
      "learning_rate": 0.00019169208288346166,
      "loss": 2.2743,
      "step": 1089
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.03303663805127144,
      "learning_rate": 0.00019166420025785164,
      "loss": 2.2745,
      "step": 1090
    },
    {
      "epoch": 0.2182,
      "grad_norm": 0.022610699757933617,
      "learning_rate": 0.00019163627295622397,
      "loss": 2.274,
      "step": 1091
    },
    {
      "epoch": 0.2184,
      "grad_norm": 0.017644556239247322,
      "learning_rate": 0.00019160830099219006,
      "loss": 2.2745,
      "step": 1092
    },
    {
      "epoch": 0.2186,
      "grad_norm": 0.018855497241020203,
      "learning_rate": 0.00019158028437938317,
      "loss": 2.2751,
      "step": 1093
    },
    {
      "epoch": 0.2188,
      "grad_norm": 0.01959771290421486,
      "learning_rate": 0.00019155222313145816,
      "loss": 2.275,
      "step": 1094
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.019097471609711647,
      "learning_rate": 0.00019152411726209176,
      "loss": 2.275,
      "step": 1095
    },
    {
      "epoch": 0.2192,
      "grad_norm": 0.01670154184103012,
      "learning_rate": 0.0001914959667849825,
      "loss": 2.2748,
      "step": 1096
    },
    {
      "epoch": 0.2194,
      "grad_norm": 0.022499823942780495,
      "learning_rate": 0.0001914677717138505,
      "loss": 2.2748,
      "step": 1097
    },
    {
      "epoch": 0.2196,
      "grad_norm": 0.028221387416124344,
      "learning_rate": 0.00019143953206243776,
      "loss": 2.2748,
      "step": 1098
    },
    {
      "epoch": 0.2198,
      "grad_norm": 0.057778384536504745,
      "learning_rate": 0.0001914112478445079,
      "loss": 2.2751,
      "step": 1099
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.10118427127599716,
      "learning_rate": 0.0001913829190738463,
      "loss": 2.275,
      "step": 1100
    },
    {
      "epoch": 0.2202,
      "grad_norm": 0.1314593106508255,
      "learning_rate": 0.0001913545457642601,
      "loss": 2.2744,
      "step": 1101
    },
    {
      "epoch": 0.2204,
      "grad_norm": 0.1647198349237442,
      "learning_rate": 0.00019132612792957808,
      "loss": 2.2751,
      "step": 1102
    },
    {
      "epoch": 0.2206,
      "grad_norm": 0.19469502568244934,
      "learning_rate": 0.00019129766558365074,
      "loss": 2.2737,
      "step": 1103
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.205699160695076,
      "learning_rate": 0.0001912691587403503,
      "loss": 2.2749,
      "step": 1104
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.18581464886665344,
      "learning_rate": 0.00019124060741357063,
      "loss": 2.2749,
      "step": 1105
    },
    {
      "epoch": 0.2212,
      "grad_norm": 0.15892480313777924,
      "learning_rate": 0.0001912120116172273,
      "loss": 2.2745,
      "step": 1106
    },
    {
      "epoch": 0.2214,
      "grad_norm": 0.11141704767942429,
      "learning_rate": 0.0001911833713652576,
      "loss": 2.2746,
      "step": 1107
    },
    {
      "epoch": 0.2216,
      "grad_norm": 0.03141669183969498,
      "learning_rate": 0.00019115468667162038,
      "loss": 2.275,
      "step": 1108
    },
    {
      "epoch": 0.2218,
      "grad_norm": 0.0681605413556099,
      "learning_rate": 0.00019112595755029624,
      "loss": 2.2735,
      "step": 1109
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.1491374522447586,
      "learning_rate": 0.0001910971840152874,
      "loss": 2.2745,
      "step": 1110
    },
    {
      "epoch": 0.2222,
      "grad_norm": 0.21494059264659882,
      "learning_rate": 0.00019106836608061772,
      "loss": 2.2742,
      "step": 1111
    },
    {
      "epoch": 0.2224,
      "grad_norm": 0.24070721864700317,
      "learning_rate": 0.00019103950376033276,
      "loss": 2.2748,
      "step": 1112
    },
    {
      "epoch": 0.2226,
      "grad_norm": 0.238128662109375,
      "learning_rate": 0.00019101059706849957,
      "loss": 2.2746,
      "step": 1113
    },
    {
      "epoch": 0.2228,
      "grad_norm": 0.24070072174072266,
      "learning_rate": 0.000190981646019207,
      "loss": 2.2745,
      "step": 1114
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.24531258642673492,
      "learning_rate": 0.00019095265062656544,
      "loss": 2.275,
      "step": 1115
    },
    {
      "epoch": 0.2232,
      "grad_norm": 0.2224995344877243,
      "learning_rate": 0.00019092361090470688,
      "loss": 2.275,
      "step": 1116
    },
    {
      "epoch": 0.2234,
      "grad_norm": 0.1655479371547699,
      "learning_rate": 0.00019089452686778488,
      "loss": 2.2748,
      "step": 1117
    },
    {
      "epoch": 0.2236,
      "grad_norm": 0.07127039134502411,
      "learning_rate": 0.0001908653985299747,
      "loss": 2.2745,
      "step": 1118
    },
    {
      "epoch": 0.2238,
      "grad_norm": 0.056577976793050766,
      "learning_rate": 0.00019083622590547312,
      "loss": 2.2747,
      "step": 1119
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.18091930449008942,
      "learning_rate": 0.00019080700900849851,
      "loss": 2.2743,
      "step": 1120
    },
    {
      "epoch": 0.2242,
      "grad_norm": 0.2663746476173401,
      "learning_rate": 0.00019077774785329087,
      "loss": 2.2752,
      "step": 1121
    },
    {
      "epoch": 0.2244,
      "grad_norm": 0.2962970733642578,
      "learning_rate": 0.0001907484424541117,
      "loss": 2.2747,
      "step": 1122
    },
    {
      "epoch": 0.2246,
      "grad_norm": 0.27879995107650757,
      "learning_rate": 0.00019071909282524413,
      "loss": 2.2736,
      "step": 1123
    },
    {
      "epoch": 0.2248,
      "grad_norm": 0.21643061935901642,
      "learning_rate": 0.0001906896989809927,
      "loss": 2.2748,
      "step": 1124
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.1111210286617279,
      "learning_rate": 0.00019066026093568378,
      "loss": 2.2744,
      "step": 1125
    },
    {
      "epoch": 0.2252,
      "grad_norm": 0.016307858750224113,
      "learning_rate": 0.000190630778703665,
      "loss": 2.2745,
      "step": 1126
    },
    {
      "epoch": 0.2254,
      "grad_norm": 0.09294392168521881,
      "learning_rate": 0.0001906012522993057,
      "loss": 2.2739,
      "step": 1127
    },
    {
      "epoch": 0.2256,
      "grad_norm": 0.16277514398097992,
      "learning_rate": 0.00019057168173699664,
      "loss": 2.2744,
      "step": 1128
    },
    {
      "epoch": 0.2258,
      "grad_norm": 0.1855819821357727,
      "learning_rate": 0.0001905420670311502,
      "loss": 2.2745,
      "step": 1129
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.15048329532146454,
      "learning_rate": 0.00019051240819620014,
      "loss": 2.2747,
      "step": 1130
    },
    {
      "epoch": 0.2262,
      "grad_norm": 0.0862617939710617,
      "learning_rate": 0.00019048270524660196,
      "loss": 2.2749,
      "step": 1131
    },
    {
      "epoch": 0.2264,
      "grad_norm": 0.01809128187596798,
      "learning_rate": 0.00019045295819683242,
      "loss": 2.2745,
      "step": 1132
    },
    {
      "epoch": 0.2266,
      "grad_norm": 0.05765523388981819,
      "learning_rate": 0.00019042316706138987,
      "loss": 2.274,
      "step": 1133
    },
    {
      "epoch": 0.2268,
      "grad_norm": 0.09108307212591171,
      "learning_rate": 0.00019039333185479418,
      "loss": 2.2742,
      "step": 1134
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.08757930994033813,
      "learning_rate": 0.00019036345259158667,
      "loss": 2.2743,
      "step": 1135
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.07108922302722931,
      "learning_rate": 0.0001903335292863301,
      "loss": 2.2743,
      "step": 1136
    },
    {
      "epoch": 0.2274,
      "grad_norm": 0.057079728692770004,
      "learning_rate": 0.00019030356195360874,
      "loss": 2.2739,
      "step": 1137
    },
    {
      "epoch": 0.2276,
      "grad_norm": 0.04014390707015991,
      "learning_rate": 0.0001902735506080283,
      "loss": 2.2745,
      "step": 1138
    },
    {
      "epoch": 0.2278,
      "grad_norm": 0.019262414425611496,
      "learning_rate": 0.00019024349526421594,
      "loss": 2.2744,
      "step": 1139
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.021807752549648285,
      "learning_rate": 0.00019021339593682028,
      "loss": 2.2742,
      "step": 1140
    },
    {
      "epoch": 0.2282,
      "grad_norm": 0.03829371556639671,
      "learning_rate": 0.0001901832526405114,
      "loss": 2.2743,
      "step": 1141
    },
    {
      "epoch": 0.2284,
      "grad_norm": 0.04308750107884407,
      "learning_rate": 0.0001901530653899807,
      "loss": 2.2742,
      "step": 1142
    },
    {
      "epoch": 0.2286,
      "grad_norm": 0.0361798033118248,
      "learning_rate": 0.00019012283419994115,
      "loss": 2.2743,
      "step": 1143
    },
    {
      "epoch": 0.2288,
      "grad_norm": 0.01845213770866394,
      "learning_rate": 0.000190092559085127,
      "loss": 2.2727,
      "step": 1144
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.0319996103644371,
      "learning_rate": 0.00019006224006029406,
      "loss": 2.2741,
      "step": 1145
    },
    {
      "epoch": 0.2292,
      "grad_norm": 0.0662069097161293,
      "learning_rate": 0.00019003187714021938,
      "loss": 2.2736,
      "step": 1146
    },
    {
      "epoch": 0.2294,
      "grad_norm": 0.09192172437906265,
      "learning_rate": 0.00019000147033970148,
      "loss": 2.2731,
      "step": 1147
    },
    {
      "epoch": 0.2296,
      "grad_norm": 0.10042911022901535,
      "learning_rate": 0.0001899710196735603,
      "loss": 2.274,
      "step": 1148
    },
    {
      "epoch": 0.2298,
      "grad_norm": 0.09657784551382065,
      "learning_rate": 0.0001899405251566371,
      "loss": 2.2748,
      "step": 1149
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.09045377373695374,
      "learning_rate": 0.00018990998680379456,
      "loss": 2.2746,
      "step": 1150
    },
    {
      "epoch": 0.2302,
      "grad_norm": 0.06262235343456268,
      "learning_rate": 0.0001898794046299167,
      "loss": 2.2741,
      "step": 1151
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.023098820820450783,
      "learning_rate": 0.00018984877864990888,
      "loss": 2.2745,
      "step": 1152
    },
    {
      "epoch": 0.2306,
      "grad_norm": 0.03203006088733673,
      "learning_rate": 0.00018981810887869785,
      "loss": 2.2745,
      "step": 1153
    },
    {
      "epoch": 0.2308,
      "grad_norm": 0.05601131170988083,
      "learning_rate": 0.0001897873953312317,
      "loss": 2.2731,
      "step": 1154
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.07534896582365036,
      "learning_rate": 0.00018975663802247976,
      "loss": 2.2742,
      "step": 1155
    },
    {
      "epoch": 0.2312,
      "grad_norm": 0.08611790090799332,
      "learning_rate": 0.00018972583696743285,
      "loss": 2.2743,
      "step": 1156
    },
    {
      "epoch": 0.2314,
      "grad_norm": 0.0882631242275238,
      "learning_rate": 0.000189694992181103,
      "loss": 2.2729,
      "step": 1157
    },
    {
      "epoch": 0.2316,
      "grad_norm": 0.07613847404718399,
      "learning_rate": 0.00018966410367852362,
      "loss": 2.2737,
      "step": 1158
    },
    {
      "epoch": 0.2318,
      "grad_norm": 0.053955189883708954,
      "learning_rate": 0.0001896331714747493,
      "loss": 2.274,
      "step": 1159
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.01962536759674549,
      "learning_rate": 0.0001896021955848561,
      "loss": 2.2741,
      "step": 1160
    },
    {
      "epoch": 0.2322,
      "grad_norm": 0.04247131943702698,
      "learning_rate": 0.0001895711760239413,
      "loss": 2.2739,
      "step": 1161
    },
    {
      "epoch": 0.2324,
      "grad_norm": 0.07808595895767212,
      "learning_rate": 0.0001895401128071234,
      "loss": 2.2736,
      "step": 1162
    },
    {
      "epoch": 0.2326,
      "grad_norm": 0.09102291613817215,
      "learning_rate": 0.00018950900594954227,
      "loss": 2.274,
      "step": 1163
    },
    {
      "epoch": 0.2328,
      "grad_norm": 0.08400877565145493,
      "learning_rate": 0.00018947785546635904,
      "loss": 2.2742,
      "step": 1164
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.06813614815473557,
      "learning_rate": 0.000189446661372756,
      "loss": 2.2731,
      "step": 1165
    },
    {
      "epoch": 0.2332,
      "grad_norm": 0.06691192835569382,
      "learning_rate": 0.0001894154236839368,
      "loss": 2.2733,
      "step": 1166
    },
    {
      "epoch": 0.2334,
      "grad_norm": 0.08010988682508469,
      "learning_rate": 0.0001893841424151264,
      "loss": 2.274,
      "step": 1167
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.10338600724935532,
      "learning_rate": 0.00018935281758157078,
      "loss": 2.2737,
      "step": 1168
    },
    {
      "epoch": 0.2338,
      "grad_norm": 0.11105001717805862,
      "learning_rate": 0.0001893214491985374,
      "loss": 2.2733,
      "step": 1169
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.10118139535188675,
      "learning_rate": 0.0001892900372813147,
      "loss": 2.2738,
      "step": 1170
    },
    {
      "epoch": 0.2342,
      "grad_norm": 0.0781741589307785,
      "learning_rate": 0.00018925858184521256,
      "loss": 2.2739,
      "step": 1171
    },
    {
      "epoch": 0.2344,
      "grad_norm": 0.04374557361006737,
      "learning_rate": 0.00018922708290556198,
      "loss": 2.274,
      "step": 1172
    },
    {
      "epoch": 0.2346,
      "grad_norm": 0.01621735468506813,
      "learning_rate": 0.0001891955404777151,
      "loss": 2.2736,
      "step": 1173
    },
    {
      "epoch": 0.2348,
      "grad_norm": 0.0331512987613678,
      "learning_rate": 0.00018916395457704534,
      "loss": 2.2738,
      "step": 1174
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.06558608263731003,
      "learning_rate": 0.00018913232521894732,
      "loss": 2.2744,
      "step": 1175
    },
    {
      "epoch": 0.2352,
      "grad_norm": 0.10001567006111145,
      "learning_rate": 0.0001891006524188368,
      "loss": 2.2745,
      "step": 1176
    },
    {
      "epoch": 0.2354,
      "grad_norm": 0.11139976233243942,
      "learning_rate": 0.00018906893619215066,
      "loss": 2.2739,
      "step": 1177
    },
    {
      "epoch": 0.2356,
      "grad_norm": 0.10231137275695801,
      "learning_rate": 0.00018903717655434707,
      "loss": 2.2737,
      "step": 1178
    },
    {
      "epoch": 0.2358,
      "grad_norm": 0.08980283886194229,
      "learning_rate": 0.00018900537352090524,
      "loss": 2.2734,
      "step": 1179
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.06426140666007996,
      "learning_rate": 0.00018897352710732564,
      "loss": 2.2736,
      "step": 1180
    },
    {
      "epoch": 0.2362,
      "grad_norm": 0.033428680151700974,
      "learning_rate": 0.00018894163732912977,
      "loss": 2.2741,
      "step": 1181
    },
    {
      "epoch": 0.2364,
      "grad_norm": 0.0168497022241354,
      "learning_rate": 0.00018890970420186033,
      "loss": 2.2736,
      "step": 1182
    },
    {
      "epoch": 0.2366,
      "grad_norm": 0.04611941799521446,
      "learning_rate": 0.00018887772774108116,
      "loss": 2.273,
      "step": 1183
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.08332769572734833,
      "learning_rate": 0.00018884570796237718,
      "loss": 2.2738,
      "step": 1184
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.09852485358715057,
      "learning_rate": 0.00018881364488135448,
      "loss": 2.274,
      "step": 1185
    },
    {
      "epoch": 0.2372,
      "grad_norm": 0.0877944752573967,
      "learning_rate": 0.00018878153851364013,
      "loss": 2.2732,
      "step": 1186
    },
    {
      "epoch": 0.2374,
      "grad_norm": 0.057949647307395935,
      "learning_rate": 0.00018874938887488248,
      "loss": 2.2734,
      "step": 1187
    },
    {
      "epoch": 0.2376,
      "grad_norm": 0.02228829637169838,
      "learning_rate": 0.0001887171959807508,
      "loss": 2.2736,
      "step": 1188
    },
    {
      "epoch": 0.2378,
      "grad_norm": 0.037648510187864304,
      "learning_rate": 0.0001886849598469356,
      "loss": 2.2749,
      "step": 1189
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.06360916793346405,
      "learning_rate": 0.00018865268048914828,
      "loss": 2.2738,
      "step": 1190
    },
    {
      "epoch": 0.2382,
      "grad_norm": 0.08083274960517883,
      "learning_rate": 0.00018862035792312147,
      "loss": 2.2738,
      "step": 1191
    },
    {
      "epoch": 0.2384,
      "grad_norm": 0.07407180219888687,
      "learning_rate": 0.00018858799216460881,
      "loss": 2.2736,
      "step": 1192
    },
    {
      "epoch": 0.2386,
      "grad_norm": 0.04186059907078743,
      "learning_rate": 0.00018855558322938493,
      "loss": 2.2731,
      "step": 1193
    },
    {
      "epoch": 0.2388,
      "grad_norm": 0.03256576135754585,
      "learning_rate": 0.00018852313113324552,
      "loss": 2.2737,
      "step": 1194
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.1002478376030922,
      "learning_rate": 0.00018849063589200743,
      "loss": 2.2727,
      "step": 1195
    },
    {
      "epoch": 0.2392,
      "grad_norm": 0.15222732722759247,
      "learning_rate": 0.0001884580975215084,
      "loss": 2.274,
      "step": 1196
    },
    {
      "epoch": 0.2394,
      "grad_norm": 0.17252174019813538,
      "learning_rate": 0.00018842551603760724,
      "loss": 2.2736,
      "step": 1197
    },
    {
      "epoch": 0.2396,
      "grad_norm": 0.1749628782272339,
      "learning_rate": 0.00018839289145618378,
      "loss": 2.2734,
      "step": 1198
    },
    {
      "epoch": 0.2398,
      "grad_norm": 0.1614755392074585,
      "learning_rate": 0.00018836022379313883,
      "loss": 2.2735,
      "step": 1199
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.1300250142812729,
      "learning_rate": 0.00018832751306439418,
      "loss": 2.2737,
      "step": 1200
    },
    {
      "epoch": 0.2402,
      "grad_norm": 0.09916914999485016,
      "learning_rate": 0.00018829475928589271,
      "loss": 2.273,
      "step": 1201
    },
    {
      "epoch": 0.2404,
      "grad_norm": 0.07617977261543274,
      "learning_rate": 0.00018826196247359817,
      "loss": 2.2734,
      "step": 1202
    },
    {
      "epoch": 0.2406,
      "grad_norm": 0.03953251615166664,
      "learning_rate": 0.00018822912264349534,
      "loss": 2.273,
      "step": 1203
    },
    {
      "epoch": 0.2408,
      "grad_norm": 0.03322632238268852,
      "learning_rate": 0.00018819623981158995,
      "loss": 2.2733,
      "step": 1204
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.10222387313842773,
      "learning_rate": 0.0001881633139939087,
      "loss": 2.2734,
      "step": 1205
    },
    {
      "epoch": 0.2412,
      "grad_norm": 0.1621135175228119,
      "learning_rate": 0.0001881303452064992,
      "loss": 2.2734,
      "step": 1206
    },
    {
      "epoch": 0.2414,
      "grad_norm": 0.21896812319755554,
      "learning_rate": 0.00018809733346543013,
      "loss": 2.2737,
      "step": 1207
    },
    {
      "epoch": 0.2416,
      "grad_norm": 0.25362133979797363,
      "learning_rate": 0.00018806427878679093,
      "loss": 2.2738,
      "step": 1208
    },
    {
      "epoch": 0.2418,
      "grad_norm": 0.258774071931839,
      "learning_rate": 0.00018803118118669202,
      "loss": 2.2733,
      "step": 1209
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.24900923669338226,
      "learning_rate": 0.00018799804068126485,
      "loss": 2.2727,
      "step": 1210
    },
    {
      "epoch": 0.2422,
      "grad_norm": 0.22746139764785767,
      "learning_rate": 0.00018796485728666165,
      "loss": 2.2738,
      "step": 1211
    },
    {
      "epoch": 0.2424,
      "grad_norm": 0.17471890151500702,
      "learning_rate": 0.00018793163101905563,
      "loss": 2.2733,
      "step": 1212
    },
    {
      "epoch": 0.2426,
      "grad_norm": 0.0922284871339798,
      "learning_rate": 0.00018789836189464086,
      "loss": 2.2733,
      "step": 1213
    },
    {
      "epoch": 0.2428,
      "grad_norm": 0.020702259615063667,
      "learning_rate": 0.0001878650499296323,
      "loss": 2.2733,
      "step": 1214
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.09700783342123032,
      "learning_rate": 0.00018783169514026578,
      "loss": 2.2729,
      "step": 1215
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.1591915637254715,
      "learning_rate": 0.00018779829754279805,
      "loss": 2.2724,
      "step": 1216
    },
    {
      "epoch": 0.2434,
      "grad_norm": 0.19792035222053528,
      "learning_rate": 0.00018776485715350671,
      "loss": 2.2737,
      "step": 1217
    },
    {
      "epoch": 0.2436,
      "grad_norm": 0.19761644303798676,
      "learning_rate": 0.00018773137398869015,
      "loss": 2.2726,
      "step": 1218
    },
    {
      "epoch": 0.2438,
      "grad_norm": 0.17071227729320526,
      "learning_rate": 0.0001876978480646677,
      "loss": 2.2735,
      "step": 1219
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.14282631874084473,
      "learning_rate": 0.00018766427939777945,
      "loss": 2.2733,
      "step": 1220
    },
    {
      "epoch": 0.2442,
      "grad_norm": 0.10818415880203247,
      "learning_rate": 0.00018763066800438636,
      "loss": 2.2734,
      "step": 1221
    },
    {
      "epoch": 0.2444,
      "grad_norm": 0.0702347457408905,
      "learning_rate": 0.00018759701390087027,
      "loss": 2.2736,
      "step": 1222
    },
    {
      "epoch": 0.2446,
      "grad_norm": 0.026546422392129898,
      "learning_rate": 0.00018756331710363374,
      "loss": 2.2733,
      "step": 1223
    },
    {
      "epoch": 0.2448,
      "grad_norm": 0.03539653494954109,
      "learning_rate": 0.00018752957762910018,
      "loss": 2.2729,
      "step": 1224
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.07221216708421707,
      "learning_rate": 0.0001874957954937138,
      "loss": 2.2732,
      "step": 1225
    },
    {
      "epoch": 0.2452,
      "grad_norm": 0.08833950012922287,
      "learning_rate": 0.00018746197071393958,
      "loss": 2.2736,
      "step": 1226
    },
    {
      "epoch": 0.2454,
      "grad_norm": 0.08128584921360016,
      "learning_rate": 0.00018742810330626337,
      "loss": 2.2729,
      "step": 1227
    },
    {
      "epoch": 0.2456,
      "grad_norm": 0.05104634910821915,
      "learning_rate": 0.0001873941932871917,
      "loss": 2.2723,
      "step": 1228
    },
    {
      "epoch": 0.2458,
      "grad_norm": 0.02217726595699787,
      "learning_rate": 0.00018736024067325188,
      "loss": 2.2731,
      "step": 1229
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.028875228017568588,
      "learning_rate": 0.00018732624548099204,
      "loss": 2.2733,
      "step": 1230
    },
    {
      "epoch": 0.2462,
      "grad_norm": 0.05586044862866402,
      "learning_rate": 0.00018729220772698097,
      "loss": 2.2722,
      "step": 1231
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.07743534445762634,
      "learning_rate": 0.00018725812742780834,
      "loss": 2.2727,
      "step": 1232
    },
    {
      "epoch": 0.2466,
      "grad_norm": 0.06681790947914124,
      "learning_rate": 0.0001872240046000844,
      "loss": 2.2728,
      "step": 1233
    },
    {
      "epoch": 0.2468,
      "grad_norm": 0.02199430577456951,
      "learning_rate": 0.0001871898392604402,
      "loss": 2.2727,
      "step": 1234
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.06366501748561859,
      "learning_rate": 0.00018715563142552758,
      "loss": 2.2734,
      "step": 1235
    },
    {
      "epoch": 0.2472,
      "grad_norm": 0.11104170978069305,
      "learning_rate": 0.00018712138111201895,
      "loss": 2.2732,
      "step": 1236
    },
    {
      "epoch": 0.2474,
      "grad_norm": 0.13253532350063324,
      "learning_rate": 0.00018708708833660754,
      "loss": 2.2725,
      "step": 1237
    },
    {
      "epoch": 0.2476,
      "grad_norm": 0.1241152286529541,
      "learning_rate": 0.00018705275311600722,
      "loss": 2.2723,
      "step": 1238
    },
    {
      "epoch": 0.2478,
      "grad_norm": 0.09269341826438904,
      "learning_rate": 0.0001870183754669526,
      "loss": 2.2727,
      "step": 1239
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.04431738331913948,
      "learning_rate": 0.0001869839554061988,
      "loss": 2.2723,
      "step": 1240
    },
    {
      "epoch": 0.2482,
      "grad_norm": 0.018444037064909935,
      "learning_rate": 0.0001869494929505219,
      "loss": 2.2733,
      "step": 1241
    },
    {
      "epoch": 0.2484,
      "grad_norm": 0.0628691017627716,
      "learning_rate": 0.0001869149881167184,
      "loss": 2.273,
      "step": 1242
    },
    {
      "epoch": 0.2486,
      "grad_norm": 0.11412082612514496,
      "learning_rate": 0.00018688044092160551,
      "loss": 2.2726,
      "step": 1243
    },
    {
      "epoch": 0.2488,
      "grad_norm": 0.12489474564790726,
      "learning_rate": 0.00018684585138202122,
      "loss": 2.273,
      "step": 1244
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.09026756137609482,
      "learning_rate": 0.00018681121951482393,
      "loss": 2.2724,
      "step": 1245
    },
    {
      "epoch": 0.2492,
      "grad_norm": 0.052065249532461166,
      "learning_rate": 0.00018677654533689287,
      "loss": 2.2725,
      "step": 1246
    },
    {
      "epoch": 0.2494,
      "grad_norm": 0.019952699542045593,
      "learning_rate": 0.00018674182886512774,
      "loss": 2.2726,
      "step": 1247
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.04031473398208618,
      "learning_rate": 0.000186707070116449,
      "loss": 2.2729,
      "step": 1248
    },
    {
      "epoch": 0.2498,
      "grad_norm": 0.07765551656484604,
      "learning_rate": 0.00018667226910779765,
      "loss": 2.2733,
      "step": 1249
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.09009881317615509,
      "learning_rate": 0.00018663742585613518,
      "loss": 2.2731,
      "step": 1250
    },
    {
      "epoch": 0.2502,
      "grad_norm": 0.07830610126256943,
      "learning_rate": 0.00018660254037844388,
      "loss": 2.2728,
      "step": 1251
    },
    {
      "epoch": 0.2504,
      "grad_norm": 0.05488136410713196,
      "learning_rate": 0.00018656761269172643,
      "loss": 2.2728,
      "step": 1252
    },
    {
      "epoch": 0.2506,
      "grad_norm": 0.043626558035612106,
      "learning_rate": 0.00018653264281300622,
      "loss": 2.2716,
      "step": 1253
    },
    {
      "epoch": 0.2508,
      "grad_norm": 0.038563184440135956,
      "learning_rate": 0.00018649763075932708,
      "loss": 2.2726,
      "step": 1254
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.030154559761285782,
      "learning_rate": 0.0001864625765477535,
      "loss": 2.2727,
      "step": 1255
    },
    {
      "epoch": 0.2512,
      "grad_norm": 0.019865375012159348,
      "learning_rate": 0.0001864274801953705,
      "loss": 2.2723,
      "step": 1256
    },
    {
      "epoch": 0.2514,
      "grad_norm": 0.01932448148727417,
      "learning_rate": 0.00018639234171928353,
      "loss": 2.2724,
      "step": 1257
    },
    {
      "epoch": 0.2516,
      "grad_norm": 0.02649908885359764,
      "learning_rate": 0.00018635716113661873,
      "loss": 2.273,
      "step": 1258
    },
    {
      "epoch": 0.2518,
      "grad_norm": 0.02807345986366272,
      "learning_rate": 0.0001863219384645227,
      "loss": 2.2725,
      "step": 1259
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.017531130462884903,
      "learning_rate": 0.0001862866737201625,
      "loss": 2.2719,
      "step": 1260
    },
    {
      "epoch": 0.2522,
      "grad_norm": 0.037619829177856445,
      "learning_rate": 0.00018625136692072575,
      "loss": 2.2724,
      "step": 1261
    },
    {
      "epoch": 0.2524,
      "grad_norm": 0.053380317986011505,
      "learning_rate": 0.00018621601808342056,
      "loss": 2.2728,
      "step": 1262
    },
    {
      "epoch": 0.2526,
      "grad_norm": 0.07947780191898346,
      "learning_rate": 0.0001861806272254755,
      "loss": 2.2723,
      "step": 1263
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.13442444801330566,
      "learning_rate": 0.0001861451943641397,
      "loss": 2.2728,
      "step": 1264
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.19188784062862396,
      "learning_rate": 0.00018610971951668265,
      "loss": 2.2726,
      "step": 1265
    },
    {
      "epoch": 0.2532,
      "grad_norm": 0.20482972264289856,
      "learning_rate": 0.0001860742027003944,
      "loss": 2.2734,
      "step": 1266
    },
    {
      "epoch": 0.2534,
      "grad_norm": 0.17099079489707947,
      "learning_rate": 0.00018603864393258534,
      "loss": 2.2718,
      "step": 1267
    },
    {
      "epoch": 0.2536,
      "grad_norm": 0.1212049126625061,
      "learning_rate": 0.00018600304323058647,
      "loss": 2.2727,
      "step": 1268
    },
    {
      "epoch": 0.2538,
      "grad_norm": 0.07645677775144577,
      "learning_rate": 0.0001859674006117491,
      "loss": 2.2728,
      "step": 1269
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.02665785141289234,
      "learning_rate": 0.00018593171609344503,
      "loss": 2.2715,
      "step": 1270
    },
    {
      "epoch": 0.2542,
      "grad_norm": 0.04032415524125099,
      "learning_rate": 0.00018589598969306645,
      "loss": 2.2723,
      "step": 1271
    },
    {
      "epoch": 0.2544,
      "grad_norm": 0.08855785429477692,
      "learning_rate": 0.00018586022142802597,
      "loss": 2.2718,
      "step": 1272
    },
    {
      "epoch": 0.2546,
      "grad_norm": 0.1281125545501709,
      "learning_rate": 0.0001858244113157566,
      "loss": 2.2721,
      "step": 1273
    },
    {
      "epoch": 0.2548,
      "grad_norm": 0.15618152916431427,
      "learning_rate": 0.00018578855937371173,
      "loss": 2.2716,
      "step": 1274
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.15904147922992706,
      "learning_rate": 0.00018575266561936523,
      "loss": 2.2723,
      "step": 1275
    },
    {
      "epoch": 0.2552,
      "grad_norm": 0.14155755937099457,
      "learning_rate": 0.00018571673007021123,
      "loss": 2.272,
      "step": 1276
    },
    {
      "epoch": 0.2554,
      "grad_norm": 0.10975802689790726,
      "learning_rate": 0.0001856807527437643,
      "loss": 2.2722,
      "step": 1277
    },
    {
      "epoch": 0.2556,
      "grad_norm": 0.08034799993038177,
      "learning_rate": 0.00018564473365755935,
      "loss": 2.2716,
      "step": 1278
    },
    {
      "epoch": 0.2558,
      "grad_norm": 0.06338639557361603,
      "learning_rate": 0.0001856086728291516,
      "loss": 2.2717,
      "step": 1279
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.05528232455253601,
      "learning_rate": 0.00018557257027611675,
      "loss": 2.2719,
      "step": 1280
    },
    {
      "epoch": 0.2562,
      "grad_norm": 0.050760362297296524,
      "learning_rate": 0.00018553642601605068,
      "loss": 2.272,
      "step": 1281
    },
    {
      "epoch": 0.2564,
      "grad_norm": 0.050296034663915634,
      "learning_rate": 0.00018550024006656966,
      "loss": 2.2725,
      "step": 1282
    },
    {
      "epoch": 0.2566,
      "grad_norm": 0.04248759150505066,
      "learning_rate": 0.0001854640124453103,
      "loss": 2.2716,
      "step": 1283
    },
    {
      "epoch": 0.2568,
      "grad_norm": 0.03622244670987129,
      "learning_rate": 0.0001854277431699295,
      "loss": 2.2719,
      "step": 1284
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.023918412625789642,
      "learning_rate": 0.0001853914322581045,
      "loss": 2.2725,
      "step": 1285
    },
    {
      "epoch": 0.2572,
      "grad_norm": 0.020551763474941254,
      "learning_rate": 0.00018535507972753274,
      "loss": 2.2723,
      "step": 1286
    },
    {
      "epoch": 0.2574,
      "grad_norm": 0.040450744330883026,
      "learning_rate": 0.00018531868559593204,
      "loss": 2.2721,
      "step": 1287
    },
    {
      "epoch": 0.2576,
      "grad_norm": 0.05999109894037247,
      "learning_rate": 0.00018528224988104044,
      "loss": 2.2724,
      "step": 1288
    },
    {
      "epoch": 0.2578,
      "grad_norm": 0.068489670753479,
      "learning_rate": 0.00018524577260061627,
      "loss": 2.272,
      "step": 1289
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.05962831899523735,
      "learning_rate": 0.0001852092537724381,
      "loss": 2.2721,
      "step": 1290
    },
    {
      "epoch": 0.2582,
      "grad_norm": 0.03976378217339516,
      "learning_rate": 0.00018517269341430476,
      "loss": 2.2713,
      "step": 1291
    },
    {
      "epoch": 0.2584,
      "grad_norm": 0.02035221830010414,
      "learning_rate": 0.00018513609154403534,
      "loss": 2.2719,
      "step": 1292
    },
    {
      "epoch": 0.2586,
      "grad_norm": 0.024587951600551605,
      "learning_rate": 0.00018509944817946922,
      "loss": 2.2715,
      "step": 1293
    },
    {
      "epoch": 0.2588,
      "grad_norm": 0.03585909307003021,
      "learning_rate": 0.00018506276333846579,
      "loss": 2.2714,
      "step": 1294
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.028521157801151276,
      "learning_rate": 0.00018502603703890488,
      "loss": 2.2726,
      "step": 1295
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.027302797883749008,
      "learning_rate": 0.00018498926929868642,
      "loss": 2.2716,
      "step": 1296
    },
    {
      "epoch": 0.2594,
      "grad_norm": 0.02652679570019245,
      "learning_rate": 0.00018495246013573054,
      "loss": 2.2727,
      "step": 1297
    },
    {
      "epoch": 0.2596,
      "grad_norm": 0.02862907201051712,
      "learning_rate": 0.00018491560956797765,
      "loss": 2.2718,
      "step": 1298
    },
    {
      "epoch": 0.2598,
      "grad_norm": 0.018866857513785362,
      "learning_rate": 0.0001848787176133882,
      "loss": 2.2721,
      "step": 1299
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.03688428923487663,
      "learning_rate": 0.0001848417842899429,
      "loss": 2.2716,
      "step": 1300
    },
    {
      "epoch": 0.2602,
      "grad_norm": 0.09076087921857834,
      "learning_rate": 0.0001848048096156426,
      "loss": 2.2716,
      "step": 1301
    },
    {
      "epoch": 0.2604,
      "grad_norm": 0.1420779824256897,
      "learning_rate": 0.00018476779360850832,
      "loss": 2.2718,
      "step": 1302
    },
    {
      "epoch": 0.2606,
      "grad_norm": 0.1742860972881317,
      "learning_rate": 0.0001847307362865812,
      "loss": 2.2713,
      "step": 1303
    },
    {
      "epoch": 0.2608,
      "grad_norm": 0.15622897446155548,
      "learning_rate": 0.00018469363766792255,
      "loss": 2.2718,
      "step": 1304
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.09568609297275543,
      "learning_rate": 0.0001846564977706138,
      "loss": 2.272,
      "step": 1305
    },
    {
      "epoch": 0.2612,
      "grad_norm": 0.024810127913951874,
      "learning_rate": 0.00018461931661275643,
      "loss": 2.2711,
      "step": 1306
    },
    {
      "epoch": 0.2614,
      "grad_norm": 0.059386588633060455,
      "learning_rate": 0.00018458209421247208,
      "loss": 2.2724,
      "step": 1307
    },
    {
      "epoch": 0.2616,
      "grad_norm": 0.11297091096639633,
      "learning_rate": 0.00018454483058790255,
      "loss": 2.272,
      "step": 1308
    },
    {
      "epoch": 0.2618,
      "grad_norm": 0.13969917595386505,
      "learning_rate": 0.00018450752575720967,
      "loss": 2.2712,
      "step": 1309
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.16538985073566437,
      "learning_rate": 0.00018447017973857532,
      "loss": 2.2711,
      "step": 1310
    },
    {
      "epoch": 0.2622,
      "grad_norm": 0.178890660405159,
      "learning_rate": 0.00018443279255020152,
      "loss": 2.2719,
      "step": 1311
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.1583133190870285,
      "learning_rate": 0.00018439536421031033,
      "loss": 2.2715,
      "step": 1312
    },
    {
      "epoch": 0.2626,
      "grad_norm": 0.10794474184513092,
      "learning_rate": 0.0001843578947371439,
      "loss": 2.2722,
      "step": 1313
    },
    {
      "epoch": 0.2628,
      "grad_norm": 0.058001525700092316,
      "learning_rate": 0.00018432038414896434,
      "loss": 2.2722,
      "step": 1314
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.017529336735606194,
      "learning_rate": 0.0001842828324640539,
      "loss": 2.2714,
      "step": 1315
    },
    {
      "epoch": 0.2632,
      "grad_norm": 0.06086404249072075,
      "learning_rate": 0.00018424523970071477,
      "loss": 2.272,
      "step": 1316
    },
    {
      "epoch": 0.2634,
      "grad_norm": 0.11419446021318436,
      "learning_rate": 0.00018420760587726923,
      "loss": 2.2714,
      "step": 1317
    },
    {
      "epoch": 0.2636,
      "grad_norm": 0.14242252707481384,
      "learning_rate": 0.00018416993101205958,
      "loss": 2.2724,
      "step": 1318
    },
    {
      "epoch": 0.2638,
      "grad_norm": 0.14795567095279694,
      "learning_rate": 0.00018413221512344805,
      "loss": 2.2713,
      "step": 1319
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.13342203199863434,
      "learning_rate": 0.00018409445822981693,
      "loss": 2.2715,
      "step": 1320
    },
    {
      "epoch": 0.2642,
      "grad_norm": 0.08599308878183365,
      "learning_rate": 0.00018405666034956844,
      "loss": 2.2712,
      "step": 1321
    },
    {
      "epoch": 0.2644,
      "grad_norm": 0.017697300761938095,
      "learning_rate": 0.00018401882150112484,
      "loss": 2.2712,
      "step": 1322
    },
    {
      "epoch": 0.2646,
      "grad_norm": 0.05350171774625778,
      "learning_rate": 0.0001839809417029283,
      "loss": 2.2707,
      "step": 1323
    },
    {
      "epoch": 0.2648,
      "grad_norm": 0.043255820870399475,
      "learning_rate": 0.000183943020973441,
      "loss": 2.2721,
      "step": 1324
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.024379784241318703,
      "learning_rate": 0.000183905059331145,
      "loss": 2.2747,
      "step": 1325
    },
    {
      "epoch": 0.2652,
      "grad_norm": 0.07373654097318649,
      "learning_rate": 0.00018386705679454242,
      "loss": 2.2711,
      "step": 1326
    },
    {
      "epoch": 0.2654,
      "grad_norm": 0.13008445501327515,
      "learning_rate": 0.00018382901338215516,
      "loss": 2.2707,
      "step": 1327
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.16558608412742615,
      "learning_rate": 0.00018379092911252514,
      "loss": 2.2719,
      "step": 1328
    },
    {
      "epoch": 0.2658,
      "grad_norm": 0.17904089391231537,
      "learning_rate": 0.0001837528040042142,
      "loss": 2.2722,
      "step": 1329
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.1715807020664215,
      "learning_rate": 0.000183714638075804,
      "loss": 2.2713,
      "step": 1330
    },
    {
      "epoch": 0.2662,
      "grad_norm": 0.15722256898880005,
      "learning_rate": 0.00018367643134589617,
      "loss": 2.2712,
      "step": 1331
    },
    {
      "epoch": 0.2664,
      "grad_norm": 0.13573917746543884,
      "learning_rate": 0.00018363818383311225,
      "loss": 2.2709,
      "step": 1332
    },
    {
      "epoch": 0.2666,
      "grad_norm": 0.09251439571380615,
      "learning_rate": 0.00018359989555609353,
      "loss": 2.2693,
      "step": 1333
    },
    {
      "epoch": 0.2668,
      "grad_norm": 0.02658320590853691,
      "learning_rate": 0.00018356156653350137,
      "loss": 2.2714,
      "step": 1334
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.05690321698784828,
      "learning_rate": 0.00018352319678401676,
      "loss": 2.2715,
      "step": 1335
    },
    {
      "epoch": 0.2672,
      "grad_norm": 0.12232377380132675,
      "learning_rate": 0.00018348478632634066,
      "loss": 2.2708,
      "step": 1336
    },
    {
      "epoch": 0.2674,
      "grad_norm": 0.17977692186832428,
      "learning_rate": 0.00018344633517919392,
      "loss": 2.2713,
      "step": 1337
    },
    {
      "epoch": 0.2676,
      "grad_norm": 0.22217988967895508,
      "learning_rate": 0.00018340784336131713,
      "loss": 2.2719,
      "step": 1338
    },
    {
      "epoch": 0.2678,
      "grad_norm": 0.22895099222660065,
      "learning_rate": 0.00018336931089147073,
      "loss": 2.2721,
      "step": 1339
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.19778035581111908,
      "learning_rate": 0.000183330737788435,
      "loss": 2.2717,
      "step": 1340
    },
    {
      "epoch": 0.2682,
      "grad_norm": 0.11894042044878006,
      "learning_rate": 0.00018329212407100994,
      "loss": 2.2706,
      "step": 1341
    },
    {
      "epoch": 0.2684,
      "grad_norm": 0.02068590559065342,
      "learning_rate": 0.0001832534697580155,
      "loss": 2.2708,
      "step": 1342
    },
    {
      "epoch": 0.2686,
      "grad_norm": 0.09740550816059113,
      "learning_rate": 0.00018321477486829126,
      "loss": 2.2711,
      "step": 1343
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.17174659669399261,
      "learning_rate": 0.00018317603942069664,
      "loss": 2.2711,
      "step": 1344
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.19613651931285858,
      "learning_rate": 0.00018313726343411086,
      "loss": 2.2712,
      "step": 1345
    },
    {
      "epoch": 0.2692,
      "grad_norm": 0.1781928390264511,
      "learning_rate": 0.00018309844692743283,
      "loss": 2.272,
      "step": 1346
    },
    {
      "epoch": 0.2694,
      "grad_norm": 0.13193204998970032,
      "learning_rate": 0.00018305958991958127,
      "loss": 2.2715,
      "step": 1347
    },
    {
      "epoch": 0.2696,
      "grad_norm": 0.0731089860200882,
      "learning_rate": 0.0001830206924294946,
      "loss": 2.2706,
      "step": 1348
    },
    {
      "epoch": 0.2698,
      "grad_norm": 0.031999655067920685,
      "learning_rate": 0.00018298175447613096,
      "loss": 2.2711,
      "step": 1349
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.017948748543858528,
      "learning_rate": 0.00018294277607846832,
      "loss": 2.2713,
      "step": 1350
    },
    {
      "epoch": 0.2702,
      "grad_norm": 0.03515775874257088,
      "learning_rate": 0.00018290375725550417,
      "loss": 2.2708,
      "step": 1351
    },
    {
      "epoch": 0.2704,
      "grad_norm": 0.06325741857290268,
      "learning_rate": 0.00018286469802625589,
      "loss": 2.2703,
      "step": 1352
    },
    {
      "epoch": 0.2706,
      "grad_norm": 0.09797893464565277,
      "learning_rate": 0.00018282559840976042,
      "loss": 2.2711,
      "step": 1353
    },
    {
      "epoch": 0.2708,
      "grad_norm": 0.13586437702178955,
      "learning_rate": 0.00018278645842507448,
      "loss": 2.2713,
      "step": 1354
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.16219502687454224,
      "learning_rate": 0.00018274727809127438,
      "loss": 2.2715,
      "step": 1355
    },
    {
      "epoch": 0.2712,
      "grad_norm": 0.17816513776779175,
      "learning_rate": 0.00018270805742745617,
      "loss": 2.2712,
      "step": 1356
    },
    {
      "epoch": 0.2714,
      "grad_norm": 0.15748128294944763,
      "learning_rate": 0.00018266879645273556,
      "loss": 2.2713,
      "step": 1357
    },
    {
      "epoch": 0.2716,
      "grad_norm": 0.0963616743683815,
      "learning_rate": 0.0001826294951862478,
      "loss": 2.2716,
      "step": 1358
    },
    {
      "epoch": 0.2718,
      "grad_norm": 0.028683925047516823,
      "learning_rate": 0.00018259015364714787,
      "loss": 2.2699,
      "step": 1359
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.040252722799777985,
      "learning_rate": 0.00018255077185461038,
      "loss": 2.271,
      "step": 1360
    },
    {
      "epoch": 0.2722,
      "grad_norm": 0.07407834380865097,
      "learning_rate": 0.00018251134982782952,
      "loss": 2.2711,
      "step": 1361
    },
    {
      "epoch": 0.2724,
      "grad_norm": 0.07703407108783722,
      "learning_rate": 0.0001824718875860191,
      "loss": 2.2711,
      "step": 1362
    },
    {
      "epoch": 0.2726,
      "grad_norm": 0.07326248288154602,
      "learning_rate": 0.0001824323851484126,
      "loss": 2.2713,
      "step": 1363
    },
    {
      "epoch": 0.2728,
      "grad_norm": 0.06763024628162384,
      "learning_rate": 0.00018239284253426295,
      "loss": 2.2702,
      "step": 1364
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.0630875751376152,
      "learning_rate": 0.00018235325976284275,
      "loss": 2.2707,
      "step": 1365
    },
    {
      "epoch": 0.2732,
      "grad_norm": 0.046407781541347504,
      "learning_rate": 0.0001823136368534442,
      "loss": 2.2704,
      "step": 1366
    },
    {
      "epoch": 0.2734,
      "grad_norm": 0.022101914510130882,
      "learning_rate": 0.000182273973825379,
      "loss": 2.271,
      "step": 1367
    },
    {
      "epoch": 0.2736,
      "grad_norm": 0.04648075997829437,
      "learning_rate": 0.00018223427069797844,
      "loss": 2.2706,
      "step": 1368
    },
    {
      "epoch": 0.2738,
      "grad_norm": 0.09836606681346893,
      "learning_rate": 0.0001821945274905933,
      "loss": 2.2708,
      "step": 1369
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.14717517793178558,
      "learning_rate": 0.00018215474422259402,
      "loss": 2.2702,
      "step": 1370
    },
    {
      "epoch": 0.2742,
      "grad_norm": 0.17388032376766205,
      "learning_rate": 0.00018211492091337042,
      "loss": 2.2709,
      "step": 1371
    },
    {
      "epoch": 0.2744,
      "grad_norm": 0.17207740247249603,
      "learning_rate": 0.0001820750575823319,
      "loss": 2.271,
      "step": 1372
    },
    {
      "epoch": 0.2746,
      "grad_norm": 0.15780960023403168,
      "learning_rate": 0.0001820351542489074,
      "loss": 2.2706,
      "step": 1373
    },
    {
      "epoch": 0.2748,
      "grad_norm": 0.13894428312778473,
      "learning_rate": 0.00018199521093254523,
      "loss": 2.2701,
      "step": 1374
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.09292787313461304,
      "learning_rate": 0.0001819552276527134,
      "loss": 2.2708,
      "step": 1375
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.03730470687150955,
      "learning_rate": 0.0001819152044288992,
      "loss": 2.2713,
      "step": 1376
    },
    {
      "epoch": 0.2754,
      "grad_norm": 0.030133673921227455,
      "learning_rate": 0.00018187514128060946,
      "loss": 2.2712,
      "step": 1377
    },
    {
      "epoch": 0.2756,
      "grad_norm": 0.05330245569348335,
      "learning_rate": 0.0001818350382273705,
      "loss": 2.2708,
      "step": 1378
    },
    {
      "epoch": 0.2758,
      "grad_norm": 0.05147624388337135,
      "learning_rate": 0.00018179489528872807,
      "loss": 2.2701,
      "step": 1379
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.03806859254837036,
      "learning_rate": 0.0001817547124842473,
      "loss": 2.2706,
      "step": 1380
    },
    {
      "epoch": 0.2762,
      "grad_norm": 0.04244984686374664,
      "learning_rate": 0.00018171448983351284,
      "loss": 2.271,
      "step": 1381
    },
    {
      "epoch": 0.2764,
      "grad_norm": 0.050688281655311584,
      "learning_rate": 0.00018167422735612877,
      "loss": 2.2713,
      "step": 1382
    },
    {
      "epoch": 0.2766,
      "grad_norm": 0.04466694965958595,
      "learning_rate": 0.00018163392507171842,
      "loss": 2.2708,
      "step": 1383
    },
    {
      "epoch": 0.2768,
      "grad_norm": 0.04275643080472946,
      "learning_rate": 0.00018159358299992467,
      "loss": 2.2697,
      "step": 1384
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.05379939451813698,
      "learning_rate": 0.00018155320116040982,
      "loss": 2.2707,
      "step": 1385
    },
    {
      "epoch": 0.2772,
      "grad_norm": 0.05700606107711792,
      "learning_rate": 0.00018151277957285543,
      "loss": 2.2705,
      "step": 1386
    },
    {
      "epoch": 0.2774,
      "grad_norm": 0.07316923141479492,
      "learning_rate": 0.00018147231825696252,
      "loss": 2.2705,
      "step": 1387
    },
    {
      "epoch": 0.2776,
      "grad_norm": 0.09940192103385925,
      "learning_rate": 0.0001814318172324514,
      "loss": 2.2707,
      "step": 1388
    },
    {
      "epoch": 0.2778,
      "grad_norm": 0.11867508292198181,
      "learning_rate": 0.00018139127651906184,
      "loss": 2.2707,
      "step": 1389
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.10125865042209625,
      "learning_rate": 0.0001813506961365528,
      "loss": 2.2709,
      "step": 1390
    },
    {
      "epoch": 0.2782,
      "grad_norm": 0.05360332876443863,
      "learning_rate": 0.00018131007610470276,
      "loss": 2.2699,
      "step": 1391
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.02212420664727688,
      "learning_rate": 0.0001812694164433094,
      "loss": 2.2705,
      "step": 1392
    },
    {
      "epoch": 0.2786,
      "grad_norm": 0.06939738988876343,
      "learning_rate": 0.0001812287171721897,
      "loss": 2.2702,
      "step": 1393
    },
    {
      "epoch": 0.2788,
      "grad_norm": 0.10170844942331314,
      "learning_rate": 0.0001811879783111801,
      "loss": 2.27,
      "step": 1394
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.11949747800827026,
      "learning_rate": 0.00018114719988013612,
      "loss": 2.2699,
      "step": 1395
    },
    {
      "epoch": 0.2792,
      "grad_norm": 0.14567872881889343,
      "learning_rate": 0.00018110638189893267,
      "loss": 2.2703,
      "step": 1396
    },
    {
      "epoch": 0.2794,
      "grad_norm": 0.16404619812965393,
      "learning_rate": 0.000181065524387464,
      "loss": 2.2709,
      "step": 1397
    },
    {
      "epoch": 0.2796,
      "grad_norm": 0.16571296751499176,
      "learning_rate": 0.00018102462736564355,
      "loss": 2.2706,
      "step": 1398
    },
    {
      "epoch": 0.2798,
      "grad_norm": 0.16574059426784515,
      "learning_rate": 0.00018098369085340398,
      "loss": 2.2708,
      "step": 1399
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.16112126410007477,
      "learning_rate": 0.00018094271487069735,
      "loss": 2.2701,
      "step": 1400
    },
    {
      "epoch": 0.2802,
      "grad_norm": 0.12920799851417542,
      "learning_rate": 0.00018090169943749476,
      "loss": 2.2706,
      "step": 1401
    },
    {
      "epoch": 0.2804,
      "grad_norm": 0.09794016927480698,
      "learning_rate": 0.00018086064457378665,
      "loss": 2.2703,
      "step": 1402
    },
    {
      "epoch": 0.2806,
      "grad_norm": 0.059687692672014236,
      "learning_rate": 0.00018081955029958274,
      "loss": 2.2709,
      "step": 1403
    },
    {
      "epoch": 0.2808,
      "grad_norm": 0.0197648536413908,
      "learning_rate": 0.00018077841663491175,
      "loss": 2.2697,
      "step": 1404
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.04038282111287117,
      "learning_rate": 0.00018073724359982186,
      "loss": 2.2702,
      "step": 1405
    },
    {
      "epoch": 0.2812,
      "grad_norm": 0.05086150020360947,
      "learning_rate": 0.00018069603121438022,
      "loss": 2.2697,
      "step": 1406
    },
    {
      "epoch": 0.2814,
      "grad_norm": 0.043740857392549515,
      "learning_rate": 0.00018065477949867327,
      "loss": 2.27,
      "step": 1407
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.025218386203050613,
      "learning_rate": 0.0001806134884728066,
      "loss": 2.2701,
      "step": 1408
    },
    {
      "epoch": 0.2818,
      "grad_norm": 0.026029739528894424,
      "learning_rate": 0.00018057215815690494,
      "loss": 2.2703,
      "step": 1409
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.07776573300361633,
      "learning_rate": 0.0001805307885711122,
      "loss": 2.2711,
      "step": 1410
    },
    {
      "epoch": 0.2822,
      "grad_norm": 0.13763543963432312,
      "learning_rate": 0.0001804893797355914,
      "loss": 2.2693,
      "step": 1411
    },
    {
      "epoch": 0.2824,
      "grad_norm": 0.18491984903812408,
      "learning_rate": 0.00018044793167052477,
      "loss": 2.2702,
      "step": 1412
    },
    {
      "epoch": 0.2826,
      "grad_norm": 0.20530377328395844,
      "learning_rate": 0.00018040644439611348,
      "loss": 2.2702,
      "step": 1413
    },
    {
      "epoch": 0.2828,
      "grad_norm": 0.1798400580883026,
      "learning_rate": 0.00018036491793257798,
      "loss": 2.2698,
      "step": 1414
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.09774652868509293,
      "learning_rate": 0.0001803233523001578,
      "loss": 2.2701,
      "step": 1415
    },
    {
      "epoch": 0.2832,
      "grad_norm": 0.022195078432559967,
      "learning_rate": 0.00018028174751911146,
      "loss": 2.2695,
      "step": 1416
    },
    {
      "epoch": 0.2834,
      "grad_norm": 0.06342776119709015,
      "learning_rate": 0.0001802401036097167,
      "loss": 2.2693,
      "step": 1417
    },
    {
      "epoch": 0.2836,
      "grad_norm": 0.14290398359298706,
      "learning_rate": 0.00018019842059227012,
      "loss": 2.2687,
      "step": 1418
    },
    {
      "epoch": 0.2838,
      "grad_norm": 0.20496466755867004,
      "learning_rate": 0.00018015669848708767,
      "loss": 2.2707,
      "step": 1419
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.2146347612142563,
      "learning_rate": 0.00018011493731450413,
      "loss": 2.2694,
      "step": 1420
    },
    {
      "epoch": 0.2842,
      "grad_norm": 0.18928520381450653,
      "learning_rate": 0.00018007313709487334,
      "loss": 2.2705,
      "step": 1421
    },
    {
      "epoch": 0.2844,
      "grad_norm": 0.14819419384002686,
      "learning_rate": 0.0001800312978485683,
      "loss": 2.2701,
      "step": 1422
    },
    {
      "epoch": 0.2846,
      "grad_norm": 0.12235476076602936,
      "learning_rate": 0.00017998941959598095,
      "loss": 2.2705,
      "step": 1423
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.10042469948530197,
      "learning_rate": 0.0001799475023575222,
      "loss": 2.2705,
      "step": 1424
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.05982113257050514,
      "learning_rate": 0.00017990554615362198,
      "loss": 2.2705,
      "step": 1425
    },
    {
      "epoch": 0.2852,
      "grad_norm": 0.019825218245387077,
      "learning_rate": 0.00017986355100472928,
      "loss": 2.2701,
      "step": 1426
    },
    {
      "epoch": 0.2854,
      "grad_norm": 0.0492330864071846,
      "learning_rate": 0.00017982151693131203,
      "loss": 2.2701,
      "step": 1427
    },
    {
      "epoch": 0.2856,
      "grad_norm": 0.0889987200498581,
      "learning_rate": 0.0001797794439538571,
      "loss": 2.2697,
      "step": 1428
    },
    {
      "epoch": 0.2858,
      "grad_norm": 0.09730775654315948,
      "learning_rate": 0.00017973733209287036,
      "loss": 2.2691,
      "step": 1429
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.10337181389331818,
      "learning_rate": 0.00017969518136887663,
      "loss": 2.2704,
      "step": 1430
    },
    {
      "epoch": 0.2862,
      "grad_norm": 0.10782364755868912,
      "learning_rate": 0.00017965299180241963,
      "loss": 2.2697,
      "step": 1431
    },
    {
      "epoch": 0.2864,
      "grad_norm": 0.1048455536365509,
      "learning_rate": 0.00017961076341406208,
      "loss": 2.2695,
      "step": 1432
    },
    {
      "epoch": 0.2866,
      "grad_norm": 0.09699373692274094,
      "learning_rate": 0.00017956849622438554,
      "loss": 2.2688,
      "step": 1433
    },
    {
      "epoch": 0.2868,
      "grad_norm": 0.07645592838525772,
      "learning_rate": 0.00017952619025399057,
      "loss": 2.2687,
      "step": 1434
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.040275223553180695,
      "learning_rate": 0.00017948384552349657,
      "loss": 2.2697,
      "step": 1435
    },
    {
      "epoch": 0.2872,
      "grad_norm": 0.02684357576072216,
      "learning_rate": 0.00017944146205354182,
      "loss": 2.2696,
      "step": 1436
    },
    {
      "epoch": 0.2874,
      "grad_norm": 0.10051599144935608,
      "learning_rate": 0.00017939903986478355,
      "loss": 2.269,
      "step": 1437
    },
    {
      "epoch": 0.2876,
      "grad_norm": 0.1694994568824768,
      "learning_rate": 0.0001793565789778978,
      "loss": 2.2686,
      "step": 1438
    },
    {
      "epoch": 0.2878,
      "grad_norm": 0.21824483573436737,
      "learning_rate": 0.00017931407941357947,
      "loss": 2.2695,
      "step": 1439
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.24120642244815826,
      "learning_rate": 0.00017927154119254236,
      "loss": 2.2696,
      "step": 1440
    },
    {
      "epoch": 0.2882,
      "grad_norm": 0.2295914590358734,
      "learning_rate": 0.00017922896433551907,
      "loss": 2.2693,
      "step": 1441
    },
    {
      "epoch": 0.2884,
      "grad_norm": 0.18016046285629272,
      "learning_rate": 0.00017918634886326108,
      "loss": 2.2694,
      "step": 1442
    },
    {
      "epoch": 0.2886,
      "grad_norm": 0.10218337178230286,
      "learning_rate": 0.0001791436947965386,
      "loss": 2.2686,
      "step": 1443
    },
    {
      "epoch": 0.2888,
      "grad_norm": 0.02592776156961918,
      "learning_rate": 0.0001791010021561407,
      "loss": 2.2683,
      "step": 1444
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.05823540687561035,
      "learning_rate": 0.0001790582709628753,
      "loss": 2.269,
      "step": 1445
    },
    {
      "epoch": 0.2892,
      "grad_norm": 0.11529667675495148,
      "learning_rate": 0.00017901550123756906,
      "loss": 2.2692,
      "step": 1446
    },
    {
      "epoch": 0.2894,
      "grad_norm": 0.1340433657169342,
      "learning_rate": 0.00017897269300106737,
      "loss": 2.2699,
      "step": 1447
    },
    {
      "epoch": 0.2896,
      "grad_norm": 0.11476525664329529,
      "learning_rate": 0.0001789298462742345,
      "loss": 2.2703,
      "step": 1448
    },
    {
      "epoch": 0.2898,
      "grad_norm": 0.06842014938592911,
      "learning_rate": 0.00017888696107795342,
      "loss": 2.2695,
      "step": 1449
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.020852476358413696,
      "learning_rate": 0.00017884403743312582,
      "loss": 2.2692,
      "step": 1450
    },
    {
      "epoch": 0.2902,
      "grad_norm": 0.06074566766619682,
      "learning_rate": 0.00017880107536067218,
      "loss": 2.269,
      "step": 1451
    },
    {
      "epoch": 0.2904,
      "grad_norm": 0.11151381582021713,
      "learning_rate": 0.00017875807488153175,
      "loss": 2.2692,
      "step": 1452
    },
    {
      "epoch": 0.2906,
      "grad_norm": 0.12358497828245163,
      "learning_rate": 0.00017871503601666233,
      "loss": 2.2694,
      "step": 1453
    },
    {
      "epoch": 0.2908,
      "grad_norm": 0.09900195151567459,
      "learning_rate": 0.0001786719587870406,
      "loss": 2.2684,
      "step": 1454
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.055205684155225754,
      "learning_rate": 0.00017862884321366188,
      "loss": 2.2688,
      "step": 1455
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.021744782105088234,
      "learning_rate": 0.0001785856893175402,
      "loss": 2.2686,
      "step": 1456
    },
    {
      "epoch": 0.2914,
      "grad_norm": 0.07122255861759186,
      "learning_rate": 0.00017854249711970818,
      "loss": 2.2685,
      "step": 1457
    },
    {
      "epoch": 0.2916,
      "grad_norm": 0.10865513235330582,
      "learning_rate": 0.00017849926664121726,
      "loss": 2.2687,
      "step": 1458
    },
    {
      "epoch": 0.2918,
      "grad_norm": 0.11724737286567688,
      "learning_rate": 0.00017845599790313735,
      "loss": 2.2693,
      "step": 1459
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.09726952016353607,
      "learning_rate": 0.00017841269092655715,
      "loss": 2.2691,
      "step": 1460
    },
    {
      "epoch": 0.2922,
      "grad_norm": 0.053939562290906906,
      "learning_rate": 0.000178369345732584,
      "loss": 2.2674,
      "step": 1461
    },
    {
      "epoch": 0.2924,
      "grad_norm": 0.023442771285772324,
      "learning_rate": 0.00017832596234234376,
      "loss": 2.2691,
      "step": 1462
    },
    {
      "epoch": 0.2926,
      "grad_norm": 0.08200328797101974,
      "learning_rate": 0.000178282540776981,
      "loss": 2.2688,
      "step": 1463
    },
    {
      "epoch": 0.2928,
      "grad_norm": 0.12819239497184753,
      "learning_rate": 0.0001782390810576588,
      "loss": 2.2683,
      "step": 1464
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.14113444089889526,
      "learning_rate": 0.000178195583205559,
      "loss": 2.2693,
      "step": 1465
    },
    {
      "epoch": 0.2932,
      "grad_norm": 0.11181849986314774,
      "learning_rate": 0.00017815204724188187,
      "loss": 2.2693,
      "step": 1466
    },
    {
      "epoch": 0.2934,
      "grad_norm": 0.05093465372920036,
      "learning_rate": 0.0001781084731878463,
      "loss": 2.2684,
      "step": 1467
    },
    {
      "epoch": 0.2936,
      "grad_norm": 0.04290035739541054,
      "learning_rate": 0.00017806486106468981,
      "loss": 2.2685,
      "step": 1468
    },
    {
      "epoch": 0.2938,
      "grad_norm": 0.12054619193077087,
      "learning_rate": 0.00017802121089366836,
      "loss": 2.2687,
      "step": 1469
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.17921170592308044,
      "learning_rate": 0.00017797752269605653,
      "loss": 2.269,
      "step": 1470
    },
    {
      "epoch": 0.2942,
      "grad_norm": 0.20071512460708618,
      "learning_rate": 0.00017793379649314744,
      "loss": 2.268,
      "step": 1471
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.17324388027191162,
      "learning_rate": 0.00017789003230625266,
      "loss": 2.269,
      "step": 1472
    },
    {
      "epoch": 0.2946,
      "grad_norm": 0.09632556885480881,
      "learning_rate": 0.00017784623015670238,
      "loss": 2.2691,
      "step": 1473
    },
    {
      "epoch": 0.2948,
      "grad_norm": 0.018868884071707726,
      "learning_rate": 0.00017780239006584515,
      "loss": 2.2686,
      "step": 1474
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.070726178586483,
      "learning_rate": 0.0001777585120550482,
      "loss": 2.2696,
      "step": 1475
    },
    {
      "epoch": 0.2952,
      "grad_norm": 0.12608285248279572,
      "learning_rate": 0.0001777145961456971,
      "loss": 2.2682,
      "step": 1476
    },
    {
      "epoch": 0.2954,
      "grad_norm": 0.16408991813659668,
      "learning_rate": 0.00017767064235919592,
      "loss": 2.2686,
      "step": 1477
    },
    {
      "epoch": 0.2956,
      "grad_norm": 0.15111757814884186,
      "learning_rate": 0.0001776266507169672,
      "loss": 2.2688,
      "step": 1478
    },
    {
      "epoch": 0.2958,
      "grad_norm": 0.09685707837343216,
      "learning_rate": 0.00017758262124045195,
      "loss": 2.2676,
      "step": 1479
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.022266073152422905,
      "learning_rate": 0.0001775385539511096,
      "loss": 2.268,
      "step": 1480
    },
    {
      "epoch": 0.2962,
      "grad_norm": 0.07672976702451706,
      "learning_rate": 0.00017749444887041799,
      "loss": 2.2679,
      "step": 1481
    },
    {
      "epoch": 0.2964,
      "grad_norm": 0.13811370730400085,
      "learning_rate": 0.00017745030601987337,
      "loss": 2.2689,
      "step": 1482
    },
    {
      "epoch": 0.2966,
      "grad_norm": 0.16210278868675232,
      "learning_rate": 0.00017740612542099053,
      "loss": 2.2689,
      "step": 1483
    },
    {
      "epoch": 0.2968,
      "grad_norm": 0.15763042867183685,
      "learning_rate": 0.0001773619070953025,
      "loss": 2.2685,
      "step": 1484
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.13904044032096863,
      "learning_rate": 0.00017731765106436073,
      "loss": 2.2684,
      "step": 1485
    },
    {
      "epoch": 0.2972,
      "grad_norm": 0.12343209236860275,
      "learning_rate": 0.00017727335734973512,
      "loss": 2.2681,
      "step": 1486
    },
    {
      "epoch": 0.2974,
      "grad_norm": 0.10190977901220322,
      "learning_rate": 0.00017722902597301383,
      "loss": 2.2683,
      "step": 1487
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.07033032923936844,
      "learning_rate": 0.0001771846569558035,
      "loss": 2.2694,
      "step": 1488
    },
    {
      "epoch": 0.2978,
      "grad_norm": 0.03019002638757229,
      "learning_rate": 0.00017714025031972903,
      "loss": 2.2684,
      "step": 1489
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.03277522698044777,
      "learning_rate": 0.00017709580608643363,
      "loss": 2.2676,
      "step": 1490
    },
    {
      "epoch": 0.2982,
      "grad_norm": 0.07725615054368973,
      "learning_rate": 0.00017705132427757895,
      "loss": 2.2684,
      "step": 1491
    },
    {
      "epoch": 0.2984,
      "grad_norm": 0.09689575433731079,
      "learning_rate": 0.0001770068049148448,
      "loss": 2.2682,
      "step": 1492
    },
    {
      "epoch": 0.2986,
      "grad_norm": 0.09013734012842178,
      "learning_rate": 0.00017696224801992945,
      "loss": 2.2682,
      "step": 1493
    },
    {
      "epoch": 0.2988,
      "grad_norm": 0.050008878111839294,
      "learning_rate": 0.00017691765361454938,
      "loss": 2.2679,
      "step": 1494
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.025660278275609016,
      "learning_rate": 0.00017687302172043933,
      "loss": 2.2685,
      "step": 1495
    },
    {
      "epoch": 0.2992,
      "grad_norm": 0.0732622742652893,
      "learning_rate": 0.00017682835235935236,
      "loss": 2.2674,
      "step": 1496
    },
    {
      "epoch": 0.2994,
      "grad_norm": 0.10599865764379501,
      "learning_rate": 0.00017678364555305978,
      "loss": 2.2683,
      "step": 1497
    },
    {
      "epoch": 0.2996,
      "grad_norm": 0.07630555331707001,
      "learning_rate": 0.0001767389013233511,
      "loss": 2.268,
      "step": 1498
    },
    {
      "epoch": 0.2998,
      "grad_norm": 0.036663953214883804,
      "learning_rate": 0.00017669411969203417,
      "loss": 2.2691,
      "step": 1499
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.14039383828639984,
      "learning_rate": 0.00017664930068093498,
      "loss": 2.2687,
      "step": 1500
    },
    {
      "epoch": 0.3002,
      "grad_norm": 0.20708037912845612,
      "learning_rate": 0.0001766044443118978,
      "loss": 2.2686,
      "step": 1501
    },
    {
      "epoch": 0.3004,
      "grad_norm": 0.21968142688274384,
      "learning_rate": 0.00017655955060678506,
      "loss": 2.2696,
      "step": 1502
    },
    {
      "epoch": 0.3006,
      "grad_norm": 0.19404956698417664,
      "learning_rate": 0.00017651461958747745,
      "loss": 2.2691,
      "step": 1503
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.130521759390831,
      "learning_rate": 0.0001764696512758737,
      "loss": 2.2686,
      "step": 1504
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.03800249099731445,
      "learning_rate": 0.0001764246456938909,
      "loss": 2.2684,
      "step": 1505
    },
    {
      "epoch": 0.3012,
      "grad_norm": 0.07454689592123032,
      "learning_rate": 0.00017637960286346425,
      "loss": 2.2685,
      "step": 1506
    },
    {
      "epoch": 0.3014,
      "grad_norm": 0.1437976360321045,
      "learning_rate": 0.000176334522806547,
      "loss": 2.2687,
      "step": 1507
    },
    {
      "epoch": 0.3016,
      "grad_norm": 0.18187333643436432,
      "learning_rate": 0.00017628940554511061,
      "loss": 2.2676,
      "step": 1508
    },
    {
      "epoch": 0.3018,
      "grad_norm": 0.1842782199382782,
      "learning_rate": 0.0001762442511011448,
      "loss": 2.2692,
      "step": 1509
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.14671491086483002,
      "learning_rate": 0.0001761990594966572,
      "loss": 2.2686,
      "step": 1510
    },
    {
      "epoch": 0.3022,
      "grad_norm": 0.0644034743309021,
      "learning_rate": 0.0001761538307536737,
      "loss": 2.2676,
      "step": 1511
    },
    {
      "epoch": 0.3024,
      "grad_norm": 0.05899282544851303,
      "learning_rate": 0.0001761085648942382,
      "loss": 2.2683,
      "step": 1512
    },
    {
      "epoch": 0.3026,
      "grad_norm": 0.16320548951625824,
      "learning_rate": 0.00017606326194041273,
      "loss": 2.2683,
      "step": 1513
    },
    {
      "epoch": 0.3028,
      "grad_norm": 0.20684346556663513,
      "learning_rate": 0.00017601792191427741,
      "loss": 2.2684,
      "step": 1514
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.18580101430416107,
      "learning_rate": 0.00017597254483793048,
      "loss": 2.2688,
      "step": 1515
    },
    {
      "epoch": 0.3032,
      "grad_norm": 0.11119206249713898,
      "learning_rate": 0.00017592713073348807,
      "loss": 2.2693,
      "step": 1516
    },
    {
      "epoch": 0.3034,
      "grad_norm": 0.026124753057956696,
      "learning_rate": 0.00017588167962308458,
      "loss": 2.2688,
      "step": 1517
    },
    {
      "epoch": 0.3036,
      "grad_norm": 0.06886094808578491,
      "learning_rate": 0.0001758361915288722,
      "loss": 2.2689,
      "step": 1518
    },
    {
      "epoch": 0.3038,
      "grad_norm": 0.11827419698238373,
      "learning_rate": 0.00017579066647302133,
      "loss": 2.2686,
      "step": 1519
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.12505581974983215,
      "learning_rate": 0.00017574510447772039,
      "loss": 2.2682,
      "step": 1520
    },
    {
      "epoch": 0.3042,
      "grad_norm": 0.09404702484607697,
      "learning_rate": 0.00017569950556517566,
      "loss": 2.2686,
      "step": 1521
    },
    {
      "epoch": 0.3044,
      "grad_norm": 0.04937678202986717,
      "learning_rate": 0.0001756538697576115,
      "loss": 2.2683,
      "step": 1522
    },
    {
      "epoch": 0.3046,
      "grad_norm": 0.025860706344246864,
      "learning_rate": 0.00017560819707727033,
      "loss": 2.2687,
      "step": 1523
    },
    {
      "epoch": 0.3048,
      "grad_norm": 0.08492612838745117,
      "learning_rate": 0.00017556248754641235,
      "loss": 2.2692,
      "step": 1524
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.1286468207836151,
      "learning_rate": 0.00017551674118731591,
      "loss": 2.2691,
      "step": 1525
    },
    {
      "epoch": 0.3052,
      "grad_norm": 0.1266748160123825,
      "learning_rate": 0.00017547095802227723,
      "loss": 2.2685,
      "step": 1526
    },
    {
      "epoch": 0.3054,
      "grad_norm": 0.09188909083604813,
      "learning_rate": 0.00017542513807361037,
      "loss": 2.2684,
      "step": 1527
    },
    {
      "epoch": 0.3056,
      "grad_norm": 0.028900135308504105,
      "learning_rate": 0.00017537928136364755,
      "loss": 2.2687,
      "step": 1528
    },
    {
      "epoch": 0.3058,
      "grad_norm": 0.05409521609544754,
      "learning_rate": 0.0001753333879147387,
      "loss": 2.2693,
      "step": 1529
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.08380072563886642,
      "learning_rate": 0.00017528745774925172,
      "loss": 2.2678,
      "step": 1530
    },
    {
      "epoch": 0.3062,
      "grad_norm": 0.06567096710205078,
      "learning_rate": 0.00017524149088957245,
      "loss": 2.2681,
      "step": 1531
    },
    {
      "epoch": 0.3064,
      "grad_norm": 0.02822868525981903,
      "learning_rate": 0.00017519548735810456,
      "loss": 2.2686,
      "step": 1532
    },
    {
      "epoch": 0.3066,
      "grad_norm": 0.039217423647642136,
      "learning_rate": 0.00017514944717726962,
      "loss": 2.2686,
      "step": 1533
    },
    {
      "epoch": 0.3068,
      "grad_norm": 0.07313380390405655,
      "learning_rate": 0.00017510337036950703,
      "loss": 2.2678,
      "step": 1534
    },
    {
      "epoch": 0.307,
      "grad_norm": 0.08921631425619125,
      "learning_rate": 0.00017505725695727412,
      "loss": 2.2685,
      "step": 1535
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.09417194128036499,
      "learning_rate": 0.00017501110696304596,
      "loss": 2.2681,
      "step": 1536
    },
    {
      "epoch": 0.3074,
      "grad_norm": 0.09561645239591599,
      "learning_rate": 0.00017496492040931552,
      "loss": 2.2686,
      "step": 1537
    },
    {
      "epoch": 0.3076,
      "grad_norm": 0.09420084208250046,
      "learning_rate": 0.00017491869731859353,
      "loss": 2.2675,
      "step": 1538
    },
    {
      "epoch": 0.3078,
      "grad_norm": 0.0844493880867958,
      "learning_rate": 0.0001748724377134086,
      "loss": 2.2681,
      "step": 1539
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.044156912714242935,
      "learning_rate": 0.00017482614161630714,
      "loss": 2.2685,
      "step": 1540
    },
    {
      "epoch": 0.3082,
      "grad_norm": 0.037952978163957596,
      "learning_rate": 0.0001747798090498532,
      "loss": 2.2681,
      "step": 1541
    },
    {
      "epoch": 0.3084,
      "grad_norm": 0.11068880558013916,
      "learning_rate": 0.00017473344003662877,
      "loss": 2.2686,
      "step": 1542
    },
    {
      "epoch": 0.3086,
      "grad_norm": 0.14890873432159424,
      "learning_rate": 0.00017468703459923356,
      "loss": 2.2683,
      "step": 1543
    },
    {
      "epoch": 0.3088,
      "grad_norm": 0.14448612928390503,
      "learning_rate": 0.00017464059276028497,
      "loss": 2.2687,
      "step": 1544
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.09842860698699951,
      "learning_rate": 0.00017459411454241822,
      "loss": 2.2692,
      "step": 1545
    },
    {
      "epoch": 0.3092,
      "grad_norm": 0.03161607310175896,
      "learning_rate": 0.00017454759996828623,
      "loss": 2.267,
      "step": 1546
    },
    {
      "epoch": 0.3094,
      "grad_norm": 0.04534098878502846,
      "learning_rate": 0.00017450104906055963,
      "loss": 2.2686,
      "step": 1547
    },
    {
      "epoch": 0.3096,
      "grad_norm": 0.07382204383611679,
      "learning_rate": 0.00017445446184192676,
      "loss": 2.268,
      "step": 1548
    },
    {
      "epoch": 0.3098,
      "grad_norm": 0.06652747839689255,
      "learning_rate": 0.00017440783833509366,
      "loss": 2.2675,
      "step": 1549
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.03426527976989746,
      "learning_rate": 0.0001743611785627841,
      "loss": 2.2675,
      "step": 1550
    },
    {
      "epoch": 0.3102,
      "grad_norm": 0.02080225944519043,
      "learning_rate": 0.00017431448254773944,
      "loss": 2.2675,
      "step": 1551
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.042821504175662994,
      "learning_rate": 0.00017426775031271875,
      "loss": 2.2679,
      "step": 1552
    },
    {
      "epoch": 0.3106,
      "grad_norm": 0.06657540798187256,
      "learning_rate": 0.00017422098188049883,
      "loss": 2.2675,
      "step": 1553
    },
    {
      "epoch": 0.3108,
      "grad_norm": 0.06877651065587997,
      "learning_rate": 0.00017417417727387394,
      "loss": 2.2677,
      "step": 1554
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.03293943777680397,
      "learning_rate": 0.0001741273365156561,
      "loss": 2.268,
      "step": 1555
    },
    {
      "epoch": 0.3112,
      "grad_norm": 0.041003286838531494,
      "learning_rate": 0.000174080459628675,
      "loss": 2.2683,
      "step": 1556
    },
    {
      "epoch": 0.3114,
      "grad_norm": 0.09920177608728409,
      "learning_rate": 0.00017403354663577783,
      "loss": 2.2684,
      "step": 1557
    },
    {
      "epoch": 0.3116,
      "grad_norm": 0.12667617201805115,
      "learning_rate": 0.00017398659755982936,
      "loss": 2.2676,
      "step": 1558
    },
    {
      "epoch": 0.3118,
      "grad_norm": 0.12496907263994217,
      "learning_rate": 0.00017393961242371205,
      "loss": 2.2671,
      "step": 1559
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.08745552599430084,
      "learning_rate": 0.0001738925912503259,
      "loss": 2.2668,
      "step": 1560
    },
    {
      "epoch": 0.3122,
      "grad_norm": 0.023396868258714676,
      "learning_rate": 0.00017384553406258842,
      "loss": 2.2666,
      "step": 1561
    },
    {
      "epoch": 0.3124,
      "grad_norm": 0.0780198946595192,
      "learning_rate": 0.00017379844088343468,
      "loss": 2.2672,
      "step": 1562
    },
    {
      "epoch": 0.3126,
      "grad_norm": 0.14490213990211487,
      "learning_rate": 0.0001737513117358174,
      "loss": 2.2678,
      "step": 1563
    },
    {
      "epoch": 0.3128,
      "grad_norm": 0.17309564352035522,
      "learning_rate": 0.00017370414664270674,
      "loss": 2.2685,
      "step": 1564
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.17349079251289368,
      "learning_rate": 0.00017365694562709034,
      "loss": 2.2678,
      "step": 1565
    },
    {
      "epoch": 0.3132,
      "grad_norm": 0.14511524140834808,
      "learning_rate": 0.00017360970871197346,
      "loss": 2.2674,
      "step": 1566
    },
    {
      "epoch": 0.3134,
      "grad_norm": 0.08022899925708771,
      "learning_rate": 0.00017356243592037875,
      "loss": 2.2673,
      "step": 1567
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.02902253530919552,
      "learning_rate": 0.00017351512727534644,
      "loss": 2.2688,
      "step": 1568
    },
    {
      "epoch": 0.3138,
      "grad_norm": 0.028730733320116997,
      "learning_rate": 0.00017346778279993415,
      "loss": 2.2681,
      "step": 1569
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.07114902883768082,
      "learning_rate": 0.00017342040251721702,
      "loss": 2.2669,
      "step": 1570
    },
    {
      "epoch": 0.3142,
      "grad_norm": 0.12236112356185913,
      "learning_rate": 0.00017337298645028764,
      "loss": 2.2677,
      "step": 1571
    },
    {
      "epoch": 0.3144,
      "grad_norm": 0.13281992077827454,
      "learning_rate": 0.00017332553462225602,
      "loss": 2.2675,
      "step": 1572
    },
    {
      "epoch": 0.3146,
      "grad_norm": 0.10231204330921173,
      "learning_rate": 0.00017327804705624957,
      "loss": 2.2682,
      "step": 1573
    },
    {
      "epoch": 0.3148,
      "grad_norm": 0.0645330548286438,
      "learning_rate": 0.00017323052377541317,
      "loss": 2.2687,
      "step": 1574
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.03714529052376747,
      "learning_rate": 0.0001731829648029091,
      "loss": 2.2681,
      "step": 1575
    },
    {
      "epoch": 0.3152,
      "grad_norm": 0.020050130784511566,
      "learning_rate": 0.00017313537016191706,
      "loss": 2.2674,
      "step": 1576
    },
    {
      "epoch": 0.3154,
      "grad_norm": 0.024707289412617683,
      "learning_rate": 0.00017308773987563406,
      "loss": 2.2679,
      "step": 1577
    },
    {
      "epoch": 0.3156,
      "grad_norm": 0.051821351051330566,
      "learning_rate": 0.00017304007396727448,
      "loss": 2.2675,
      "step": 1578
    },
    {
      "epoch": 0.3158,
      "grad_norm": 0.08180422335863113,
      "learning_rate": 0.00017299237246007015,
      "loss": 2.2676,
      "step": 1579
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.09667374938726425,
      "learning_rate": 0.00017294463537727024,
      "loss": 2.2671,
      "step": 1580
    },
    {
      "epoch": 0.3162,
      "grad_norm": 0.08034077286720276,
      "learning_rate": 0.00017289686274214118,
      "loss": 2.2671,
      "step": 1581
    },
    {
      "epoch": 0.3164,
      "grad_norm": 0.04502781853079796,
      "learning_rate": 0.00017284905457796675,
      "loss": 2.2677,
      "step": 1582
    },
    {
      "epoch": 0.3166,
      "grad_norm": 0.03447643294930458,
      "learning_rate": 0.00017280121090804812,
      "loss": 2.2679,
      "step": 1583
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.05283195152878761,
      "learning_rate": 0.00017275333175570368,
      "loss": 2.267,
      "step": 1584
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.0731518417596817,
      "learning_rate": 0.0001727054171442692,
      "loss": 2.2669,
      "step": 1585
    },
    {
      "epoch": 0.3172,
      "grad_norm": 0.07971451431512833,
      "learning_rate": 0.0001726574670970976,
      "loss": 2.2661,
      "step": 1586
    },
    {
      "epoch": 0.3174,
      "grad_norm": 0.05278589949011803,
      "learning_rate": 0.00017260948163755918,
      "loss": 2.266,
      "step": 1587
    },
    {
      "epoch": 0.3176,
      "grad_norm": 0.020714543759822845,
      "learning_rate": 0.00017256146078904153,
      "loss": 2.2666,
      "step": 1588
    },
    {
      "epoch": 0.3178,
      "grad_norm": 0.04756182059645653,
      "learning_rate": 0.00017251340457494934,
      "loss": 2.266,
      "step": 1589
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.07595482468605042,
      "learning_rate": 0.0001724653130187047,
      "loss": 2.2669,
      "step": 1590
    },
    {
      "epoch": 0.3182,
      "grad_norm": 0.06843958050012589,
      "learning_rate": 0.00017241718614374678,
      "loss": 2.2669,
      "step": 1591
    },
    {
      "epoch": 0.3184,
      "grad_norm": 0.03315350040793419,
      "learning_rate": 0.00017236902397353205,
      "loss": 2.2674,
      "step": 1592
    },
    {
      "epoch": 0.3186,
      "grad_norm": 0.025520555675029755,
      "learning_rate": 0.00017232082653153422,
      "loss": 2.2664,
      "step": 1593
    },
    {
      "epoch": 0.3188,
      "grad_norm": 0.04918878898024559,
      "learning_rate": 0.0001722725938412441,
      "loss": 2.2664,
      "step": 1594
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.07037054747343063,
      "learning_rate": 0.0001722243259261697,
      "loss": 2.2677,
      "step": 1595
    },
    {
      "epoch": 0.3192,
      "grad_norm": 0.097922682762146,
      "learning_rate": 0.00017217602280983623,
      "loss": 2.2676,
      "step": 1596
    },
    {
      "epoch": 0.3194,
      "grad_norm": 0.10789395123720169,
      "learning_rate": 0.00017212768451578607,
      "loss": 2.2667,
      "step": 1597
    },
    {
      "epoch": 0.3196,
      "grad_norm": 0.09213438630104065,
      "learning_rate": 0.00017207931106757868,
      "loss": 2.2668,
      "step": 1598
    },
    {
      "epoch": 0.3198,
      "grad_norm": 0.05897841602563858,
      "learning_rate": 0.0001720309024887907,
      "loss": 2.2673,
      "step": 1599
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.028660796582698822,
      "learning_rate": 0.0001719824588030159,
      "loss": 2.2676,
      "step": 1600
    },
    {
      "epoch": 0.3202,
      "grad_norm": 0.02517750859260559,
      "learning_rate": 0.0001719339800338651,
      "loss": 2.2667,
      "step": 1601
    },
    {
      "epoch": 0.3204,
      "grad_norm": 0.04347307235002518,
      "learning_rate": 0.00017188546620496635,
      "loss": 2.2668,
      "step": 1602
    },
    {
      "epoch": 0.3206,
      "grad_norm": 0.039230719208717346,
      "learning_rate": 0.00017183691733996462,
      "loss": 2.2675,
      "step": 1603
    },
    {
      "epoch": 0.3208,
      "grad_norm": 0.02984379418194294,
      "learning_rate": 0.00017178833346252206,
      "loss": 2.2672,
      "step": 1604
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.02604166604578495,
      "learning_rate": 0.00017173971459631787,
      "loss": 2.2681,
      "step": 1605
    },
    {
      "epoch": 0.3212,
      "grad_norm": 0.020855071023106575,
      "learning_rate": 0.0001716910607650483,
      "loss": 2.2673,
      "step": 1606
    },
    {
      "epoch": 0.3214,
      "grad_norm": 0.028682738542556763,
      "learning_rate": 0.0001716423719924266,
      "loss": 2.2668,
      "step": 1607
    },
    {
      "epoch": 0.3216,
      "grad_norm": 0.044038765132427216,
      "learning_rate": 0.00017159364830218312,
      "loss": 2.2671,
      "step": 1608
    },
    {
      "epoch": 0.3218,
      "grad_norm": 0.03361693024635315,
      "learning_rate": 0.00017154488971806518,
      "loss": 2.2661,
      "step": 1609
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.025394365191459656,
      "learning_rate": 0.00017149609626383717,
      "loss": 2.2669,
      "step": 1610
    },
    {
      "epoch": 0.3222,
      "grad_norm": 0.05996755138039589,
      "learning_rate": 0.00017144726796328034,
      "loss": 2.2664,
      "step": 1611
    },
    {
      "epoch": 0.3224,
      "grad_norm": 0.07253905385732651,
      "learning_rate": 0.0001713984048401931,
      "loss": 2.2661,
      "step": 1612
    },
    {
      "epoch": 0.3226,
      "grad_norm": 0.05084380879998207,
      "learning_rate": 0.00017134950691839065,
      "loss": 2.2658,
      "step": 1613
    },
    {
      "epoch": 0.3228,
      "grad_norm": 0.019744742661714554,
      "learning_rate": 0.0001713005742217053,
      "loss": 2.2641,
      "step": 1614
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.04002958536148071,
      "learning_rate": 0.00017125160677398626,
      "loss": 2.2649,
      "step": 1615
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.0547851137816906,
      "learning_rate": 0.00017120260459909967,
      "loss": 2.2668,
      "step": 1616
    },
    {
      "epoch": 0.3234,
      "grad_norm": 0.04966634139418602,
      "learning_rate": 0.00017115356772092857,
      "loss": 2.2652,
      "step": 1617
    },
    {
      "epoch": 0.3236,
      "grad_norm": 0.034773897379636765,
      "learning_rate": 0.00017110449616337289,
      "loss": 2.2661,
      "step": 1618
    },
    {
      "epoch": 0.3238,
      "grad_norm": 0.022093717008829117,
      "learning_rate": 0.00017105538995034963,
      "loss": 2.264,
      "step": 1619
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.021941518411040306,
      "learning_rate": 0.0001710062491057925,
      "loss": 2.265,
      "step": 1620
    },
    {
      "epoch": 0.3242,
      "grad_norm": 0.03455670550465584,
      "learning_rate": 0.0001709570736536521,
      "loss": 2.2631,
      "step": 1621
    },
    {
      "epoch": 0.3244,
      "grad_norm": 0.05936966463923454,
      "learning_rate": 0.000170907863617896,
      "loss": 2.2639,
      "step": 1622
    },
    {
      "epoch": 0.3246,
      "grad_norm": 0.0689873918890953,
      "learning_rate": 0.00017085861902250865,
      "loss": 2.2637,
      "step": 1623
    },
    {
      "epoch": 0.3248,
      "grad_norm": 0.05450715124607086,
      "learning_rate": 0.0001708093398914911,
      "loss": 2.2659,
      "step": 1624
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.023787522688508034,
      "learning_rate": 0.00017076002624886154,
      "loss": 2.265,
      "step": 1625
    },
    {
      "epoch": 0.3252,
      "grad_norm": 0.02959761768579483,
      "learning_rate": 0.00017071067811865476,
      "loss": 2.2628,
      "step": 1626
    },
    {
      "epoch": 0.3254,
      "grad_norm": 0.0482102707028389,
      "learning_rate": 0.0001706612955249225,
      "loss": 2.2635,
      "step": 1627
    },
    {
      "epoch": 0.3256,
      "grad_norm": 0.03318167105317116,
      "learning_rate": 0.00017061187849173317,
      "loss": 2.266,
      "step": 1628
    },
    {
      "epoch": 0.3258,
      "grad_norm": 0.02560596913099289,
      "learning_rate": 0.0001705624270431721,
      "loss": 2.2662,
      "step": 1629
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.05114148557186127,
      "learning_rate": 0.00017051294120334125,
      "loss": 2.2659,
      "step": 1630
    },
    {
      "epoch": 0.3262,
      "grad_norm": 0.061971403658390045,
      "learning_rate": 0.00017046342099635948,
      "loss": 2.2658,
      "step": 1631
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.06945698708295822,
      "learning_rate": 0.0001704138664463623,
      "loss": 2.2656,
      "step": 1632
    },
    {
      "epoch": 0.3266,
      "grad_norm": 0.0735810324549675,
      "learning_rate": 0.00017036427757750205,
      "loss": 2.2647,
      "step": 1633
    },
    {
      "epoch": 0.3268,
      "grad_norm": 0.07552054524421692,
      "learning_rate": 0.00017031465441394767,
      "loss": 2.2654,
      "step": 1634
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.0647161453962326,
      "learning_rate": 0.00017026499697988493,
      "loss": 2.2641,
      "step": 1635
    },
    {
      "epoch": 0.3272,
      "grad_norm": 0.04900818318128586,
      "learning_rate": 0.00017021530529951625,
      "loss": 2.2652,
      "step": 1636
    },
    {
      "epoch": 0.3274,
      "grad_norm": 0.03650864586234093,
      "learning_rate": 0.00017016557939706075,
      "loss": 2.2681,
      "step": 1637
    },
    {
      "epoch": 0.3276,
      "grad_norm": 0.03220511972904205,
      "learning_rate": 0.00017011581929675425,
      "loss": 2.2648,
      "step": 1638
    },
    {
      "epoch": 0.3278,
      "grad_norm": 0.02972823940217495,
      "learning_rate": 0.00017006602502284918,
      "loss": 2.265,
      "step": 1639
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.023369576781988144,
      "learning_rate": 0.00017001619659961467,
      "loss": 2.2644,
      "step": 1640
    },
    {
      "epoch": 0.3282,
      "grad_norm": 0.02194426767528057,
      "learning_rate": 0.00016996633405133655,
      "loss": 2.2648,
      "step": 1641
    },
    {
      "epoch": 0.3284,
      "grad_norm": 0.032643333077430725,
      "learning_rate": 0.00016991643740231716,
      "loss": 2.2605,
      "step": 1642
    },
    {
      "epoch": 0.3286,
      "grad_norm": 0.03854730352759361,
      "learning_rate": 0.00016986650667687552,
      "loss": 2.2664,
      "step": 1643
    },
    {
      "epoch": 0.3288,
      "grad_norm": 0.025528643280267715,
      "learning_rate": 0.00016981654189934727,
      "loss": 2.2617,
      "step": 1644
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.021104536950588226,
      "learning_rate": 0.00016976654309408464,
      "loss": 2.2613,
      "step": 1645
    },
    {
      "epoch": 0.3292,
      "grad_norm": 0.030848070979118347,
      "learning_rate": 0.00016971651028545648,
      "loss": 2.2591,
      "step": 1646
    },
    {
      "epoch": 0.3294,
      "grad_norm": 0.03728032484650612,
      "learning_rate": 0.00016966644349784808,
      "loss": 2.2643,
      "step": 1647
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.030794015154242516,
      "learning_rate": 0.00016961634275566146,
      "loss": 2.266,
      "step": 1648
    },
    {
      "epoch": 0.3298,
      "grad_norm": 0.026966778561472893,
      "learning_rate": 0.0001695662080833151,
      "loss": 2.2652,
      "step": 1649
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.0485464408993721,
      "learning_rate": 0.000169516039505244,
      "loss": 2.2661,
      "step": 1650
    },
    {
      "epoch": 0.3302,
      "grad_norm": 0.038515329360961914,
      "learning_rate": 0.00016946583704589973,
      "loss": 2.2649,
      "step": 1651
    },
    {
      "epoch": 0.3304,
      "grad_norm": 0.024815866723656654,
      "learning_rate": 0.0001694156007297504,
      "loss": 2.2656,
      "step": 1652
    },
    {
      "epoch": 0.3306,
      "grad_norm": 0.02156977728009224,
      "learning_rate": 0.0001693653305812805,
      "loss": 2.2651,
      "step": 1653
    },
    {
      "epoch": 0.3308,
      "grad_norm": 0.030948085710406303,
      "learning_rate": 0.00016931502662499118,
      "loss": 2.2649,
      "step": 1654
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.04933120682835579,
      "learning_rate": 0.0001692646888853999,
      "loss": 2.2647,
      "step": 1655
    },
    {
      "epoch": 0.3312,
      "grad_norm": 0.0698252022266388,
      "learning_rate": 0.0001692143173870407,
      "loss": 2.2651,
      "step": 1656
    },
    {
      "epoch": 0.3314,
      "grad_norm": 0.07200419902801514,
      "learning_rate": 0.000169163912154464,
      "loss": 2.2637,
      "step": 1657
    },
    {
      "epoch": 0.3316,
      "grad_norm": 0.058595020323991776,
      "learning_rate": 0.0001691134732122368,
      "loss": 2.2606,
      "step": 1658
    },
    {
      "epoch": 0.3318,
      "grad_norm": 0.03815954178571701,
      "learning_rate": 0.00016906300058494228,
      "loss": 2.2604,
      "step": 1659
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.029573658481240273,
      "learning_rate": 0.00016901249429718032,
      "loss": 2.2628,
      "step": 1660
    },
    {
      "epoch": 0.3322,
      "grad_norm": 0.03523110970854759,
      "learning_rate": 0.000168961954373567,
      "loss": 2.2653,
      "step": 1661
    },
    {
      "epoch": 0.3324,
      "grad_norm": 0.057040467858314514,
      "learning_rate": 0.00016891138083873487,
      "loss": 2.263,
      "step": 1662
    },
    {
      "epoch": 0.3326,
      "grad_norm": 0.07958883792161942,
      "learning_rate": 0.00016886077371733283,
      "loss": 2.2646,
      "step": 1663
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.08639489859342575,
      "learning_rate": 0.0001688101330340263,
      "loss": 2.2643,
      "step": 1664
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.07130776345729828,
      "learning_rate": 0.00016875945881349676,
      "loss": 2.2655,
      "step": 1665
    },
    {
      "epoch": 0.3332,
      "grad_norm": 0.05221403017640114,
      "learning_rate": 0.0001687087510804423,
      "loss": 2.2628,
      "step": 1666
    },
    {
      "epoch": 0.3334,
      "grad_norm": 0.029406961053609848,
      "learning_rate": 0.00016865800985957726,
      "loss": 2.2615,
      "step": 1667
    },
    {
      "epoch": 0.3336,
      "grad_norm": 0.040103428065776825,
      "learning_rate": 0.0001686072351756323,
      "loss": 2.2672,
      "step": 1668
    },
    {
      "epoch": 0.3338,
      "grad_norm": 0.12994462251663208,
      "learning_rate": 0.00016855642705335437,
      "loss": 2.2631,
      "step": 1669
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.25271978974342346,
      "learning_rate": 0.0001685055855175067,
      "loss": 2.2618,
      "step": 1670
    },
    {
      "epoch": 0.3342,
      "grad_norm": 0.3269751965999603,
      "learning_rate": 0.00016845471059286887,
      "loss": 2.2643,
      "step": 1671
    },
    {
      "epoch": 0.3344,
      "grad_norm": 0.30349570512771606,
      "learning_rate": 0.00016840380230423668,
      "loss": 2.265,
      "step": 1672
    },
    {
      "epoch": 0.3346,
      "grad_norm": 0.20096330344676971,
      "learning_rate": 0.00016835286067642228,
      "loss": 2.2647,
      "step": 1673
    },
    {
      "epoch": 0.3348,
      "grad_norm": 0.06626222282648087,
      "learning_rate": 0.00016830188573425389,
      "loss": 2.2646,
      "step": 1674
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.07495632022619247,
      "learning_rate": 0.0001682508775025762,
      "loss": 2.2634,
      "step": 1675
    },
    {
      "epoch": 0.3352,
      "grad_norm": 0.18910591304302216,
      "learning_rate": 0.00016819983600624986,
      "loss": 2.2639,
      "step": 1676
    },
    {
      "epoch": 0.3354,
      "grad_norm": 0.27116870880126953,
      "learning_rate": 0.000168148761270152,
      "loss": 2.2643,
      "step": 1677
    },
    {
      "epoch": 0.3356,
      "grad_norm": 0.3056391775608063,
      "learning_rate": 0.00016809765331917575,
      "loss": 2.2653,
      "step": 1678
    },
    {
      "epoch": 0.3358,
      "grad_norm": 0.27014854550361633,
      "learning_rate": 0.00016804651217823053,
      "loss": 2.2636,
      "step": 1679
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.184276282787323,
      "learning_rate": 0.00016799533787224192,
      "loss": 2.2637,
      "step": 1680
    },
    {
      "epoch": 0.3362,
      "grad_norm": 0.0719863772392273,
      "learning_rate": 0.00016794413042615168,
      "loss": 2.263,
      "step": 1681
    },
    {
      "epoch": 0.3364,
      "grad_norm": 0.05178328603506088,
      "learning_rate": 0.00016789288986491762,
      "loss": 2.2626,
      "step": 1682
    },
    {
      "epoch": 0.3366,
      "grad_norm": 0.1453728824853897,
      "learning_rate": 0.00016784161621351382,
      "loss": 2.2652,
      "step": 1683
    },
    {
      "epoch": 0.3368,
      "grad_norm": 0.18461085855960846,
      "learning_rate": 0.00016779030949693044,
      "loss": 2.2628,
      "step": 1684
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.16136996448040009,
      "learning_rate": 0.00016773896974017373,
      "loss": 2.2655,
      "step": 1685
    },
    {
      "epoch": 0.3372,
      "grad_norm": 0.08055594563484192,
      "learning_rate": 0.00016768759696826608,
      "loss": 2.2671,
      "step": 1686
    },
    {
      "epoch": 0.3374,
      "grad_norm": 0.04099695011973381,
      "learning_rate": 0.00016763619120624594,
      "loss": 2.2626,
      "step": 1687
    },
    {
      "epoch": 0.3376,
      "grad_norm": 0.1351609230041504,
      "learning_rate": 0.00016758475247916787,
      "loss": 2.2626,
      "step": 1688
    },
    {
      "epoch": 0.3378,
      "grad_norm": 0.19877922534942627,
      "learning_rate": 0.00016753328081210245,
      "loss": 2.2639,
      "step": 1689
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.20013411343097687,
      "learning_rate": 0.00016748177623013638,
      "loss": 2.2642,
      "step": 1690
    },
    {
      "epoch": 0.3382,
      "grad_norm": 0.11353179067373276,
      "learning_rate": 0.00016743023875837233,
      "loss": 2.2638,
      "step": 1691
    },
    {
      "epoch": 0.3384,
      "grad_norm": 0.03364046663045883,
      "learning_rate": 0.0001673786684219291,
      "loss": 2.2639,
      "step": 1692
    },
    {
      "epoch": 0.3386,
      "grad_norm": 0.16669687628746033,
      "learning_rate": 0.00016732706524594137,
      "loss": 2.2625,
      "step": 1693
    },
    {
      "epoch": 0.3388,
      "grad_norm": 0.23882652819156647,
      "learning_rate": 0.00016727542925555996,
      "loss": 2.2641,
      "step": 1694
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.21662646532058716,
      "learning_rate": 0.00016722376047595164,
      "loss": 2.2638,
      "step": 1695
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.10007787495851517,
      "learning_rate": 0.00016717205893229903,
      "loss": 2.2643,
      "step": 1696
    },
    {
      "epoch": 0.3394,
      "grad_norm": 0.07294706255197525,
      "learning_rate": 0.00016712032464980095,
      "loss": 2.264,
      "step": 1697
    },
    {
      "epoch": 0.3396,
      "grad_norm": 0.20780815184116364,
      "learning_rate": 0.000167068557653672,
      "loss": 2.2625,
      "step": 1698
    },
    {
      "epoch": 0.3398,
      "grad_norm": 0.24021443724632263,
      "learning_rate": 0.00016701675796914286,
      "loss": 2.2628,
      "step": 1699
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.1677653044462204,
      "learning_rate": 0.00016696492562145996,
      "loss": 2.2646,
      "step": 1700
    },
    {
      "epoch": 0.3402,
      "grad_norm": 0.02967234142124653,
      "learning_rate": 0.00016691306063588583,
      "loss": 2.2639,
      "step": 1701
    },
    {
      "epoch": 0.3404,
      "grad_norm": 0.17522132396697998,
      "learning_rate": 0.00016686116303769882,
      "loss": 2.2724,
      "step": 1702
    },
    {
      "epoch": 0.3406,
      "grad_norm": 0.32703348994255066,
      "learning_rate": 0.00016680923285219317,
      "loss": 2.264,
      "step": 1703
    },
    {
      "epoch": 0.3408,
      "grad_norm": 0.3617214262485504,
      "learning_rate": 0.00016675727010467906,
      "loss": 2.2644,
      "step": 1704
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.23754282295703888,
      "learning_rate": 0.00016670527482048246,
      "loss": 2.2632,
      "step": 1705
    },
    {
      "epoch": 0.3412,
      "grad_norm": 0.03509892150759697,
      "learning_rate": 0.00016665324702494524,
      "loss": 2.2612,
      "step": 1706
    },
    {
      "epoch": 0.3414,
      "grad_norm": 0.16950716078281403,
      "learning_rate": 0.00016660118674342517,
      "loss": 2.2634,
      "step": 1707
    },
    {
      "epoch": 0.3416,
      "grad_norm": 0.29688167572021484,
      "learning_rate": 0.00016654909400129575,
      "loss": 2.2659,
      "step": 1708
    },
    {
      "epoch": 0.3418,
      "grad_norm": 0.313616544008255,
      "learning_rate": 0.00016649696882394633,
      "loss": 2.2663,
      "step": 1709
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.17730502784252167,
      "learning_rate": 0.00016644481123678217,
      "loss": 2.2647,
      "step": 1710
    },
    {
      "epoch": 0.3422,
      "grad_norm": 0.057518161833286285,
      "learning_rate": 0.00016639262126522418,
      "loss": 2.2628,
      "step": 1711
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.24424709379673004,
      "learning_rate": 0.00016634039893470912,
      "loss": 2.2656,
      "step": 1712
    },
    {
      "epoch": 0.3426,
      "grad_norm": 0.29598942399024963,
      "learning_rate": 0.00016628814427068953,
      "loss": 2.2636,
      "step": 1713
    },
    {
      "epoch": 0.3428,
      "grad_norm": 0.17921820282936096,
      "learning_rate": 0.00016623585729863368,
      "loss": 2.2634,
      "step": 1714
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.03087906911969185,
      "learning_rate": 0.00016618353804402568,
      "loss": 2.2651,
      "step": 1715
    },
    {
      "epoch": 0.3432,
      "grad_norm": 0.1992522031068802,
      "learning_rate": 0.00016613118653236518,
      "loss": 2.2639,
      "step": 1716
    },
    {
      "epoch": 0.3434,
      "grad_norm": 0.26836884021759033,
      "learning_rate": 0.00016607880278916777,
      "loss": 2.2638,
      "step": 1717
    },
    {
      "epoch": 0.3436,
      "grad_norm": 0.2054002285003662,
      "learning_rate": 0.00016602638683996463,
      "loss": 2.2619,
      "step": 1718
    },
    {
      "epoch": 0.3438,
      "grad_norm": 0.04248137027025223,
      "learning_rate": 0.00016597393871030264,
      "loss": 2.262,
      "step": 1719
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.13817481696605682,
      "learning_rate": 0.00016592145842574433,
      "loss": 2.2611,
      "step": 1720
    },
    {
      "epoch": 0.3442,
      "grad_norm": 0.21277929842472076,
      "learning_rate": 0.00016586894601186805,
      "loss": 2.2664,
      "step": 1721
    },
    {
      "epoch": 0.3444,
      "grad_norm": 0.15461115539073944,
      "learning_rate": 0.00016581640149426768,
      "loss": 2.2641,
      "step": 1722
    },
    {
      "epoch": 0.3446,
      "grad_norm": 0.028198491781949997,
      "learning_rate": 0.00016576382489855274,
      "loss": 2.2639,
      "step": 1723
    },
    {
      "epoch": 0.3448,
      "grad_norm": 0.10733963549137115,
      "learning_rate": 0.00016571121625034847,
      "loss": 2.2623,
      "step": 1724
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.16297391057014465,
      "learning_rate": 0.00016565857557529566,
      "loss": 2.2653,
      "step": 1725
    },
    {
      "epoch": 0.3452,
      "grad_norm": 0.1324208527803421,
      "learning_rate": 0.00016560590289905073,
      "loss": 2.2648,
      "step": 1726
    },
    {
      "epoch": 0.3454,
      "grad_norm": 0.0344218946993351,
      "learning_rate": 0.00016555319824728575,
      "loss": 2.2624,
      "step": 1727
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.09017756581306458,
      "learning_rate": 0.00016550046164568827,
      "loss": 2.2639,
      "step": 1728
    },
    {
      "epoch": 0.3458,
      "grad_norm": 0.15260285139083862,
      "learning_rate": 0.00016544769311996148,
      "loss": 2.2635,
      "step": 1729
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.11849430203437805,
      "learning_rate": 0.00016539489269582416,
      "loss": 2.2648,
      "step": 1730
    },
    {
      "epoch": 0.3462,
      "grad_norm": 0.03295981511473656,
      "learning_rate": 0.00016534206039901057,
      "loss": 2.2636,
      "step": 1731
    },
    {
      "epoch": 0.3464,
      "grad_norm": 0.0597611702978611,
      "learning_rate": 0.0001652891962552705,
      "loss": 2.2631,
      "step": 1732
    },
    {
      "epoch": 0.3466,
      "grad_norm": 0.08579526096582413,
      "learning_rate": 0.00016523630029036931,
      "loss": 2.2637,
      "step": 1733
    },
    {
      "epoch": 0.3468,
      "grad_norm": 0.07938706129789352,
      "learning_rate": 0.0001651833725300879,
      "loss": 2.265,
      "step": 1734
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.053878773003816605,
      "learning_rate": 0.00016513041300022255,
      "loss": 2.2617,
      "step": 1735
    },
    {
      "epoch": 0.3472,
      "grad_norm": 0.02372097410261631,
      "learning_rate": 0.0001650774217265851,
      "loss": 2.2629,
      "step": 1736
    },
    {
      "epoch": 0.3474,
      "grad_norm": 0.0275852233171463,
      "learning_rate": 0.00016502439873500289,
      "loss": 2.2644,
      "step": 1737
    },
    {
      "epoch": 0.3476,
      "grad_norm": 0.03437388315796852,
      "learning_rate": 0.00016497134405131866,
      "loss": 2.2639,
      "step": 1738
    },
    {
      "epoch": 0.3478,
      "grad_norm": 0.03141046315431595,
      "learning_rate": 0.00016491825770139062,
      "loss": 2.263,
      "step": 1739
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.022220605984330177,
      "learning_rate": 0.00016486513971109243,
      "loss": 2.2631,
      "step": 1740
    },
    {
      "epoch": 0.3482,
      "grad_norm": 0.05327510088682175,
      "learning_rate": 0.0001648119901063131,
      "loss": 2.2637,
      "step": 1741
    },
    {
      "epoch": 0.3484,
      "grad_norm": 0.08839485049247742,
      "learning_rate": 0.00016475880891295716,
      "loss": 2.2656,
      "step": 1742
    },
    {
      "epoch": 0.3486,
      "grad_norm": 0.09865260869264603,
      "learning_rate": 0.00016470559615694446,
      "loss": 2.264,
      "step": 1743
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.08859345316886902,
      "learning_rate": 0.0001646523518642102,
      "loss": 2.2638,
      "step": 1744
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.0493168942630291,
      "learning_rate": 0.0001645990760607051,
      "loss": 2.2637,
      "step": 1745
    },
    {
      "epoch": 0.3492,
      "grad_norm": 0.023254547268152237,
      "learning_rate": 0.00016454576877239507,
      "loss": 2.2648,
      "step": 1746
    },
    {
      "epoch": 0.3494,
      "grad_norm": 0.056630924344062805,
      "learning_rate": 0.00016449243002526144,
      "loss": 2.266,
      "step": 1747
    },
    {
      "epoch": 0.3496,
      "grad_norm": 0.06723013520240784,
      "learning_rate": 0.0001644390598453009,
      "loss": 2.2645,
      "step": 1748
    },
    {
      "epoch": 0.3498,
      "grad_norm": 0.046402592211961746,
      "learning_rate": 0.0001643856582585254,
      "loss": 2.2641,
      "step": 1749
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.023152630776166916,
      "learning_rate": 0.0001643322252909622,
      "loss": 2.2645,
      "step": 1750
    },
    {
      "epoch": 0.3502,
      "grad_norm": 0.04689967259764671,
      "learning_rate": 0.00016427876096865394,
      "loss": 2.2653,
      "step": 1751
    },
    {
      "epoch": 0.3504,
      "grad_norm": 0.04938342049717903,
      "learning_rate": 0.00016422526531765846,
      "loss": 2.2648,
      "step": 1752
    },
    {
      "epoch": 0.3506,
      "grad_norm": 0.04368104413151741,
      "learning_rate": 0.00016417173836404887,
      "loss": 2.2635,
      "step": 1753
    },
    {
      "epoch": 0.3508,
      "grad_norm": 0.04534940421581268,
      "learning_rate": 0.00016411818013391355,
      "loss": 2.2658,
      "step": 1754
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.04321107268333435,
      "learning_rate": 0.00016406459065335615,
      "loss": 2.2623,
      "step": 1755
    },
    {
      "epoch": 0.3512,
      "grad_norm": 0.03263852372765541,
      "learning_rate": 0.00016401096994849557,
      "loss": 2.2642,
      "step": 1756
    },
    {
      "epoch": 0.3514,
      "grad_norm": 0.02216452546417713,
      "learning_rate": 0.0001639573180454658,
      "loss": 2.2634,
      "step": 1757
    },
    {
      "epoch": 0.3516,
      "grad_norm": 0.030919641256332397,
      "learning_rate": 0.00016390363497041622,
      "loss": 2.2639,
      "step": 1758
    },
    {
      "epoch": 0.3518,
      "grad_norm": 0.0535014383494854,
      "learning_rate": 0.00016384992074951123,
      "loss": 2.2637,
      "step": 1759
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.05596689507365227,
      "learning_rate": 0.00016379617540893056,
      "loss": 2.2635,
      "step": 1760
    },
    {
      "epoch": 0.3522,
      "grad_norm": 0.03183659166097641,
      "learning_rate": 0.000163742398974869,
      "loss": 2.2628,
      "step": 1761
    },
    {
      "epoch": 0.3524,
      "grad_norm": 0.03980323299765587,
      "learning_rate": 0.0001636885914735365,
      "loss": 2.2646,
      "step": 1762
    },
    {
      "epoch": 0.3526,
      "grad_norm": 0.08203057944774628,
      "learning_rate": 0.00016363475293115824,
      "loss": 2.263,
      "step": 1763
    },
    {
      "epoch": 0.3528,
      "grad_norm": 0.09033702313899994,
      "learning_rate": 0.00016358088337397442,
      "loss": 2.2642,
      "step": 1764
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.05512496083974838,
      "learning_rate": 0.00016352698282824044,
      "loss": 2.2638,
      "step": 1765
    },
    {
      "epoch": 0.3532,
      "grad_norm": 0.024249127134680748,
      "learning_rate": 0.00016347305132022677,
      "loss": 2.2633,
      "step": 1766
    },
    {
      "epoch": 0.3534,
      "grad_norm": 0.07474326342344284,
      "learning_rate": 0.00016341908887621895,
      "loss": 2.2637,
      "step": 1767
    },
    {
      "epoch": 0.3536,
      "grad_norm": 0.09711400419473648,
      "learning_rate": 0.00016336509552251766,
      "loss": 2.264,
      "step": 1768
    },
    {
      "epoch": 0.3538,
      "grad_norm": 0.06761026382446289,
      "learning_rate": 0.00016331107128543857,
      "loss": 2.2621,
      "step": 1769
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.027233947068452835,
      "learning_rate": 0.00016325701619131246,
      "loss": 2.2627,
      "step": 1770
    },
    {
      "epoch": 0.3542,
      "grad_norm": 0.09344633668661118,
      "learning_rate": 0.0001632029302664851,
      "loss": 2.264,
      "step": 1771
    },
    {
      "epoch": 0.3544,
      "grad_norm": 0.12415428459644318,
      "learning_rate": 0.00016314881353731732,
      "loss": 2.2638,
      "step": 1772
    },
    {
      "epoch": 0.3546,
      "grad_norm": 0.11281155049800873,
      "learning_rate": 0.00016309466603018496,
      "loss": 2.2594,
      "step": 1773
    },
    {
      "epoch": 0.3548,
      "grad_norm": 0.06296948343515396,
      "learning_rate": 0.0001630404877714789,
      "loss": 2.2642,
      "step": 1774
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.030796119943261147,
      "learning_rate": 0.00016298627878760487,
      "loss": 2.2622,
      "step": 1775
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.08696480840444565,
      "learning_rate": 0.00016293203910498376,
      "loss": 2.2643,
      "step": 1776
    },
    {
      "epoch": 0.3554,
      "grad_norm": 0.10065480321645737,
      "learning_rate": 0.0001628777687500513,
      "loss": 2.2634,
      "step": 1777
    },
    {
      "epoch": 0.3556,
      "grad_norm": 0.07363198697566986,
      "learning_rate": 0.00016282346774925817,
      "loss": 2.2636,
      "step": 1778
    },
    {
      "epoch": 0.3558,
      "grad_norm": 0.035149361938238144,
      "learning_rate": 0.00016276913612907007,
      "loss": 2.2634,
      "step": 1779
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.027439212426543236,
      "learning_rate": 0.00016271477391596753,
      "loss": 2.2624,
      "step": 1780
    },
    {
      "epoch": 0.3562,
      "grad_norm": 0.05074169486761093,
      "learning_rate": 0.00016266038113644607,
      "loss": 2.2637,
      "step": 1781
    },
    {
      "epoch": 0.3564,
      "grad_norm": 0.06136934459209442,
      "learning_rate": 0.00016260595781701604,
      "loss": 2.2636,
      "step": 1782
    },
    {
      "epoch": 0.3566,
      "grad_norm": 0.04724617302417755,
      "learning_rate": 0.00016255150398420271,
      "loss": 2.2643,
      "step": 1783
    },
    {
      "epoch": 0.3568,
      "grad_norm": 0.02249828912317753,
      "learning_rate": 0.00016249701966454625,
      "loss": 2.2621,
      "step": 1784
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.05729379132390022,
      "learning_rate": 0.00016244250488460158,
      "loss": 2.2635,
      "step": 1785
    },
    {
      "epoch": 0.3572,
      "grad_norm": 0.057126134634017944,
      "learning_rate": 0.00016238795967093864,
      "loss": 2.2624,
      "step": 1786
    },
    {
      "epoch": 0.3574,
      "grad_norm": 0.022226864472031593,
      "learning_rate": 0.00016233338405014202,
      "loss": 2.2619,
      "step": 1787
    },
    {
      "epoch": 0.3576,
      "grad_norm": 0.062124449759721756,
      "learning_rate": 0.00016227877804881127,
      "loss": 2.2626,
      "step": 1788
    },
    {
      "epoch": 0.3578,
      "grad_norm": 0.08396373689174652,
      "learning_rate": 0.00016222414169356065,
      "loss": 2.2644,
      "step": 1789
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.05573159456253052,
      "learning_rate": 0.0001621694750110193,
      "loss": 2.264,
      "step": 1790
    },
    {
      "epoch": 0.3582,
      "grad_norm": 0.02336033806204796,
      "learning_rate": 0.00016211477802783103,
      "loss": 2.2643,
      "step": 1791
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.03461972624063492,
      "learning_rate": 0.00016206005077065458,
      "loss": 2.2646,
      "step": 1792
    },
    {
      "epoch": 0.3586,
      "grad_norm": 0.024460269138216972,
      "learning_rate": 0.00016200529326616328,
      "loss": 2.2608,
      "step": 1793
    },
    {
      "epoch": 0.3588,
      "grad_norm": 0.06802201271057129,
      "learning_rate": 0.00016195050554104528,
      "loss": 2.2628,
      "step": 1794
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.08085691928863525,
      "learning_rate": 0.00016189568762200348,
      "loss": 2.2643,
      "step": 1795
    },
    {
      "epoch": 0.3592,
      "grad_norm": 0.050825610756874084,
      "learning_rate": 0.0001618408395357554,
      "loss": 2.2642,
      "step": 1796
    },
    {
      "epoch": 0.3594,
      "grad_norm": 0.023018311709165573,
      "learning_rate": 0.00016178596130903344,
      "loss": 2.2653,
      "step": 1797
    },
    {
      "epoch": 0.3596,
      "grad_norm": 0.03357858210802078,
      "learning_rate": 0.00016173105296858452,
      "loss": 2.2619,
      "step": 1798
    },
    {
      "epoch": 0.3598,
      "grad_norm": 0.040575385093688965,
      "learning_rate": 0.00016167611454117025,
      "loss": 2.2626,
      "step": 1799
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.03075559437274933,
      "learning_rate": 0.00016162114605356703,
      "loss": 2.266,
      "step": 1800
    },
    {
      "epoch": 0.3602,
      "grad_norm": 0.02585076168179512,
      "learning_rate": 0.0001615661475325658,
      "loss": 2.2688,
      "step": 1801
    },
    {
      "epoch": 0.3604,
      "grad_norm": 0.03208602964878082,
      "learning_rate": 0.00016151111900497225,
      "loss": 2.2637,
      "step": 1802
    },
    {
      "epoch": 0.3606,
      "grad_norm": 0.0755043551325798,
      "learning_rate": 0.00016145606049760644,
      "loss": 2.2643,
      "step": 1803
    },
    {
      "epoch": 0.3608,
      "grad_norm": 0.09882183372974396,
      "learning_rate": 0.00016140097203730337,
      "loss": 2.2625,
      "step": 1804
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.07546348869800568,
      "learning_rate": 0.00016134585365091243,
      "loss": 2.2615,
      "step": 1805
    },
    {
      "epoch": 0.3612,
      "grad_norm": 0.022711938247084618,
      "learning_rate": 0.00016129070536529766,
      "loss": 2.2622,
      "step": 1806
    },
    {
      "epoch": 0.3614,
      "grad_norm": 0.09400403499603271,
      "learning_rate": 0.00016123552720733765,
      "loss": 2.2631,
      "step": 1807
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.13504564762115479,
      "learning_rate": 0.00016118031920392558,
      "loss": 2.2624,
      "step": 1808
    },
    {
      "epoch": 0.3618,
      "grad_norm": 0.09769049286842346,
      "learning_rate": 0.00016112508138196917,
      "loss": 2.2643,
      "step": 1809
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.02213277667760849,
      "learning_rate": 0.00016106981376839066,
      "loss": 2.2624,
      "step": 1810
    },
    {
      "epoch": 0.3622,
      "grad_norm": 0.10809588432312012,
      "learning_rate": 0.0001610145163901268,
      "loss": 2.2619,
      "step": 1811
    },
    {
      "epoch": 0.3624,
      "grad_norm": 0.15427266061306,
      "learning_rate": 0.0001609591892741288,
      "loss": 2.2635,
      "step": 1812
    },
    {
      "epoch": 0.3626,
      "grad_norm": 0.13499905169010162,
      "learning_rate": 0.00016090383244736256,
      "loss": 2.2627,
      "step": 1813
    },
    {
      "epoch": 0.3628,
      "grad_norm": 0.0657903328537941,
      "learning_rate": 0.0001608484459368082,
      "loss": 2.2629,
      "step": 1814
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.031484171748161316,
      "learning_rate": 0.00016079302976946055,
      "loss": 2.2628,
      "step": 1815
    },
    {
      "epoch": 0.3632,
      "grad_norm": 0.09510620683431625,
      "learning_rate": 0.00016073758397232868,
      "loss": 2.2638,
      "step": 1816
    },
    {
      "epoch": 0.3634,
      "grad_norm": 0.12202159315347672,
      "learning_rate": 0.00016068210857243624,
      "loss": 2.2633,
      "step": 1817
    },
    {
      "epoch": 0.3636,
      "grad_norm": 0.10150235146284103,
      "learning_rate": 0.00016062660359682124,
      "loss": 2.2643,
      "step": 1818
    },
    {
      "epoch": 0.3638,
      "grad_norm": 0.04559161514043808,
      "learning_rate": 0.00016057106907253616,
      "loss": 2.2623,
      "step": 1819
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.02437279000878334,
      "learning_rate": 0.0001605155050266478,
      "loss": 2.2633,
      "step": 1820
    },
    {
      "epoch": 0.3642,
      "grad_norm": 0.03431199863553047,
      "learning_rate": 0.0001604599114862375,
      "loss": 2.2648,
      "step": 1821
    },
    {
      "epoch": 0.3644,
      "grad_norm": 0.022553030401468277,
      "learning_rate": 0.00016040428847840079,
      "loss": 2.2619,
      "step": 1822
    },
    {
      "epoch": 0.3646,
      "grad_norm": 0.03018413484096527,
      "learning_rate": 0.00016034863603024767,
      "loss": 2.2608,
      "step": 1823
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.05359705537557602,
      "learning_rate": 0.00016029295416890248,
      "loss": 2.2628,
      "step": 1824
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.08408815413713455,
      "learning_rate": 0.00016023724292150385,
      "loss": 2.2619,
      "step": 1825
    },
    {
      "epoch": 0.3652,
      "grad_norm": 0.10588797926902771,
      "learning_rate": 0.00016018150231520486,
      "loss": 2.2617,
      "step": 1826
    },
    {
      "epoch": 0.3654,
      "grad_norm": 0.09054423123598099,
      "learning_rate": 0.0001601257323771727,
      "loss": 2.2627,
      "step": 1827
    },
    {
      "epoch": 0.3656,
      "grad_norm": 0.04366488754749298,
      "learning_rate": 0.00016006993313458896,
      "loss": 2.2624,
      "step": 1828
    },
    {
      "epoch": 0.3658,
      "grad_norm": 0.04895958676934242,
      "learning_rate": 0.00016001410461464956,
      "loss": 2.2644,
      "step": 1829
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.12263958156108856,
      "learning_rate": 0.00015995824684456465,
      "loss": 2.263,
      "step": 1830
    },
    {
      "epoch": 0.3662,
      "grad_norm": 0.14567363262176514,
      "learning_rate": 0.0001599023598515586,
      "loss": 2.2624,
      "step": 1831
    },
    {
      "epoch": 0.3664,
      "grad_norm": 0.09851444512605667,
      "learning_rate": 0.00015984644366287006,
      "loss": 2.2624,
      "step": 1832
    },
    {
      "epoch": 0.3666,
      "grad_norm": 0.030354123562574387,
      "learning_rate": 0.0001597904983057519,
      "loss": 2.2628,
      "step": 1833
    },
    {
      "epoch": 0.3668,
      "grad_norm": 0.05465419963002205,
      "learning_rate": 0.00015973452380747122,
      "loss": 2.2618,
      "step": 1834
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.10613375157117844,
      "learning_rate": 0.00015967852019530926,
      "loss": 2.2611,
      "step": 1835
    },
    {
      "epoch": 0.3672,
      "grad_norm": 0.1273660808801651,
      "learning_rate": 0.0001596224874965616,
      "loss": 2.2634,
      "step": 1836
    },
    {
      "epoch": 0.3674,
      "grad_norm": 0.10321343690156937,
      "learning_rate": 0.0001595664257385378,
      "loss": 2.2639,
      "step": 1837
    },
    {
      "epoch": 0.3676,
      "grad_norm": 0.04571063444018364,
      "learning_rate": 0.00015951033494856175,
      "loss": 2.2609,
      "step": 1838
    },
    {
      "epoch": 0.3678,
      "grad_norm": 0.05143171548843384,
      "learning_rate": 0.00015945421515397133,
      "loss": 2.2602,
      "step": 1839
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.1192319244146347,
      "learning_rate": 0.0001593980663821187,
      "loss": 2.2629,
      "step": 1840
    },
    {
      "epoch": 0.3682,
      "grad_norm": 0.1402524709701538,
      "learning_rate": 0.00015934188866037016,
      "loss": 2.262,
      "step": 1841
    },
    {
      "epoch": 0.3684,
      "grad_norm": 0.09282241016626358,
      "learning_rate": 0.00015928568201610595,
      "loss": 2.2622,
      "step": 1842
    },
    {
      "epoch": 0.3686,
      "grad_norm": 0.022753525525331497,
      "learning_rate": 0.00015922944647672052,
      "loss": 2.2616,
      "step": 1843
    },
    {
      "epoch": 0.3688,
      "grad_norm": 0.09587835520505905,
      "learning_rate": 0.0001591731820696224,
      "loss": 2.2638,
      "step": 1844
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.11875343322753906,
      "learning_rate": 0.0001591168888222342,
      "loss": 2.2646,
      "step": 1845
    },
    {
      "epoch": 0.3692,
      "grad_norm": 0.07647643238306046,
      "learning_rate": 0.00015906056676199255,
      "loss": 2.2633,
      "step": 1846
    },
    {
      "epoch": 0.3694,
      "grad_norm": 0.022163130342960358,
      "learning_rate": 0.00015900421591634814,
      "loss": 2.2633,
      "step": 1847
    },
    {
      "epoch": 0.3696,
      "grad_norm": 0.0776786059141159,
      "learning_rate": 0.00015894783631276567,
      "loss": 2.2622,
      "step": 1848
    },
    {
      "epoch": 0.3698,
      "grad_norm": 0.11425622552633286,
      "learning_rate": 0.00015889142797872387,
      "loss": 2.2631,
      "step": 1849
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.08238036930561066,
      "learning_rate": 0.00015883499094171554,
      "loss": 2.2631,
      "step": 1850
    },
    {
      "epoch": 0.3702,
      "grad_norm": 0.02644115313887596,
      "learning_rate": 0.00015877852522924732,
      "loss": 2.2658,
      "step": 1851
    },
    {
      "epoch": 0.3704,
      "grad_norm": 0.09848188608884811,
      "learning_rate": 0.00015872203086883996,
      "loss": 2.2618,
      "step": 1852
    },
    {
      "epoch": 0.3706,
      "grad_norm": 0.16159237921237946,
      "learning_rate": 0.00015866550788802813,
      "loss": 2.2666,
      "step": 1853
    },
    {
      "epoch": 0.3708,
      "grad_norm": 0.17647196352481842,
      "learning_rate": 0.00015860895631436043,
      "loss": 2.2617,
      "step": 1854
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.12527470290660858,
      "learning_rate": 0.0001585523761753994,
      "loss": 2.2633,
      "step": 1855
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.034420207142829895,
      "learning_rate": 0.00015849576749872157,
      "loss": 2.2618,
      "step": 1856
    },
    {
      "epoch": 0.3714,
      "grad_norm": 0.07333002239465714,
      "learning_rate": 0.00015843913031191723,
      "loss": 2.2617,
      "step": 1857
    },
    {
      "epoch": 0.3716,
      "grad_norm": 0.12034837156534195,
      "learning_rate": 0.0001583824646425907,
      "loss": 2.2625,
      "step": 1858
    },
    {
      "epoch": 0.3718,
      "grad_norm": 0.10092829912900925,
      "learning_rate": 0.00015832577051836015,
      "loss": 2.2634,
      "step": 1859
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.026041975244879723,
      "learning_rate": 0.00015826904796685762,
      "loss": 2.2627,
      "step": 1860
    },
    {
      "epoch": 0.3722,
      "grad_norm": 0.07175319641828537,
      "learning_rate": 0.00015821229701572896,
      "loss": 2.2649,
      "step": 1861
    },
    {
      "epoch": 0.3724,
      "grad_norm": 0.0931985080242157,
      "learning_rate": 0.00015815551769263387,
      "loss": 2.2643,
      "step": 1862
    },
    {
      "epoch": 0.3726,
      "grad_norm": 0.0800938606262207,
      "learning_rate": 0.000158098710025246,
      "loss": 2.261,
      "step": 1863
    },
    {
      "epoch": 0.3728,
      "grad_norm": 0.06111403554677963,
      "learning_rate": 0.0001580418740412526,
      "loss": 2.2622,
      "step": 1864
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.03417313098907471,
      "learning_rate": 0.00015798500976835493,
      "loss": 2.2632,
      "step": 1865
    },
    {
      "epoch": 0.3732,
      "grad_norm": 0.02204161137342453,
      "learning_rate": 0.0001579281172342679,
      "loss": 2.2616,
      "step": 1866
    },
    {
      "epoch": 0.3734,
      "grad_norm": 0.02794802375137806,
      "learning_rate": 0.00015787119646672025,
      "loss": 2.2598,
      "step": 1867
    },
    {
      "epoch": 0.3736,
      "grad_norm": 0.026285801082849503,
      "learning_rate": 0.00015781424749345446,
      "loss": 2.2612,
      "step": 1868
    },
    {
      "epoch": 0.3738,
      "grad_norm": 0.022365815937519073,
      "learning_rate": 0.00015775727034222675,
      "loss": 2.2615,
      "step": 1869
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.02496866136789322,
      "learning_rate": 0.00015770026504080718,
      "loss": 2.2639,
      "step": 1870
    },
    {
      "epoch": 0.3742,
      "grad_norm": 0.02288416028022766,
      "learning_rate": 0.00015764323161697935,
      "loss": 2.2626,
      "step": 1871
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.02425225079059601,
      "learning_rate": 0.0001575861700985407,
      "loss": 2.2626,
      "step": 1872
    },
    {
      "epoch": 0.3746,
      "grad_norm": 0.026920661330223083,
      "learning_rate": 0.00015752908051330227,
      "loss": 2.2615,
      "step": 1873
    },
    {
      "epoch": 0.3748,
      "grad_norm": 0.024523161351680756,
      "learning_rate": 0.00015747196288908887,
      "loss": 2.2625,
      "step": 1874
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.02314702421426773,
      "learning_rate": 0.000157414817253739,
      "loss": 2.2625,
      "step": 1875
    },
    {
      "epoch": 0.3752,
      "grad_norm": 0.022187361493706703,
      "learning_rate": 0.0001573576436351046,
      "loss": 2.261,
      "step": 1876
    },
    {
      "epoch": 0.3754,
      "grad_norm": 0.022151101380586624,
      "learning_rate": 0.00015730044206105156,
      "loss": 2.259,
      "step": 1877
    },
    {
      "epoch": 0.3756,
      "grad_norm": 0.023417703807353973,
      "learning_rate": 0.0001572432125594591,
      "loss": 2.2615,
      "step": 1878
    },
    {
      "epoch": 0.3758,
      "grad_norm": 0.023830993101000786,
      "learning_rate": 0.00015718595515822027,
      "loss": 2.2608,
      "step": 1879
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.024962378665804863,
      "learning_rate": 0.00015712866988524155,
      "loss": 2.2631,
      "step": 1880
    },
    {
      "epoch": 0.3762,
      "grad_norm": 0.029038487002253532,
      "learning_rate": 0.0001570713567684432,
      "loss": 2.2605,
      "step": 1881
    },
    {
      "epoch": 0.3764,
      "grad_norm": 0.03317686542868614,
      "learning_rate": 0.00015701401583575884,
      "loss": 2.2616,
      "step": 1882
    },
    {
      "epoch": 0.3766,
      "grad_norm": 0.022699251770973206,
      "learning_rate": 0.00015695664711513576,
      "loss": 2.261,
      "step": 1883
    },
    {
      "epoch": 0.3768,
      "grad_norm": 0.03296857699751854,
      "learning_rate": 0.0001568992506345348,
      "loss": 2.26,
      "step": 1884
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.048979803919792175,
      "learning_rate": 0.0001568418264219303,
      "loss": 2.2621,
      "step": 1885
    },
    {
      "epoch": 0.3772,
      "grad_norm": 0.05553796887397766,
      "learning_rate": 0.00015678437450531013,
      "loss": 2.2607,
      "step": 1886
    },
    {
      "epoch": 0.3774,
      "grad_norm": 0.05410953611135483,
      "learning_rate": 0.00015672689491267567,
      "loss": 2.2622,
      "step": 1887
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.037787068635225296,
      "learning_rate": 0.0001566693876720417,
      "loss": 2.2605,
      "step": 1888
    },
    {
      "epoch": 0.3778,
      "grad_norm": 0.03254365548491478,
      "learning_rate": 0.00015661185281143667,
      "loss": 2.2619,
      "step": 1889
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.047884948551654816,
      "learning_rate": 0.0001565542903589023,
      "loss": 2.2622,
      "step": 1890
    },
    {
      "epoch": 0.3782,
      "grad_norm": 0.05890899896621704,
      "learning_rate": 0.0001564967003424938,
      "loss": 2.2617,
      "step": 1891
    },
    {
      "epoch": 0.3784,
      "grad_norm": 0.0513337180018425,
      "learning_rate": 0.00015643908279027992,
      "loss": 2.2606,
      "step": 1892
    },
    {
      "epoch": 0.3786,
      "grad_norm": 0.04706399515271187,
      "learning_rate": 0.00015638143773034267,
      "loss": 2.2628,
      "step": 1893
    },
    {
      "epoch": 0.3788,
      "grad_norm": 0.030410323292016983,
      "learning_rate": 0.00015632376519077767,
      "loss": 2.2622,
      "step": 1894
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.027500614523887634,
      "learning_rate": 0.0001562660651996937,
      "loss": 2.2614,
      "step": 1895
    },
    {
      "epoch": 0.3792,
      "grad_norm": 0.042557891458272934,
      "learning_rate": 0.00015620833778521307,
      "loss": 2.2584,
      "step": 1896
    },
    {
      "epoch": 0.3794,
      "grad_norm": 0.03409736976027489,
      "learning_rate": 0.00015615058297547145,
      "loss": 2.2617,
      "step": 1897
    },
    {
      "epoch": 0.3796,
      "grad_norm": 0.02339145913720131,
      "learning_rate": 0.0001560928007986178,
      "loss": 2.2623,
      "step": 1898
    },
    {
      "epoch": 0.3798,
      "grad_norm": 0.04953969642519951,
      "learning_rate": 0.00015603499128281448,
      "loss": 2.2602,
      "step": 1899
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.07177054136991501,
      "learning_rate": 0.00015597715445623712,
      "loss": 2.2605,
      "step": 1900
    },
    {
      "epoch": 0.3802,
      "grad_norm": 0.07780438661575317,
      "learning_rate": 0.0001559192903470747,
      "loss": 2.2607,
      "step": 1901
    },
    {
      "epoch": 0.3804,
      "grad_norm": 0.06189277023077011,
      "learning_rate": 0.00015586139898352946,
      "loss": 2.2627,
      "step": 1902
    },
    {
      "epoch": 0.3806,
      "grad_norm": 0.04392173886299133,
      "learning_rate": 0.000155803480393817,
      "loss": 2.2637,
      "step": 1903
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.03407058119773865,
      "learning_rate": 0.00015574553460616608,
      "loss": 2.2609,
      "step": 1904
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.030492814257740974,
      "learning_rate": 0.00015568756164881882,
      "loss": 2.262,
      "step": 1905
    },
    {
      "epoch": 0.3812,
      "grad_norm": 0.024938127025961876,
      "learning_rate": 0.0001556295615500305,
      "loss": 2.2633,
      "step": 1906
    },
    {
      "epoch": 0.3814,
      "grad_norm": 0.024051055312156677,
      "learning_rate": 0.00015557153433806966,
      "loss": 2.2625,
      "step": 1907
    },
    {
      "epoch": 0.3816,
      "grad_norm": 0.032745636999607086,
      "learning_rate": 0.0001555134800412181,
      "loss": 2.2627,
      "step": 1908
    },
    {
      "epoch": 0.3818,
      "grad_norm": 0.03817058727145195,
      "learning_rate": 0.00015545539868777074,
      "loss": 2.2614,
      "step": 1909
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.044209614396095276,
      "learning_rate": 0.00015539729030603574,
      "loss": 2.2597,
      "step": 1910
    },
    {
      "epoch": 0.3822,
      "grad_norm": 0.04700690135359764,
      "learning_rate": 0.00015533915492433443,
      "loss": 2.2622,
      "step": 1911
    },
    {
      "epoch": 0.3824,
      "grad_norm": 0.042315252125263214,
      "learning_rate": 0.00015528099257100127,
      "loss": 2.2615,
      "step": 1912
    },
    {
      "epoch": 0.3826,
      "grad_norm": 0.026126787066459656,
      "learning_rate": 0.00015522280327438388,
      "loss": 2.2606,
      "step": 1913
    },
    {
      "epoch": 0.3828,
      "grad_norm": 0.023328278213739395,
      "learning_rate": 0.00015516458706284303,
      "loss": 2.2601,
      "step": 1914
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.028926461935043335,
      "learning_rate": 0.0001551063439647526,
      "loss": 2.2607,
      "step": 1915
    },
    {
      "epoch": 0.3832,
      "grad_norm": 0.04733233153820038,
      "learning_rate": 0.00015504807400849958,
      "loss": 2.2618,
      "step": 1916
    },
    {
      "epoch": 0.3834,
      "grad_norm": 0.06398546695709229,
      "learning_rate": 0.000154989777222484,
      "loss": 2.2589,
      "step": 1917
    },
    {
      "epoch": 0.3836,
      "grad_norm": 0.059509530663490295,
      "learning_rate": 0.000154931453635119,
      "loss": 2.2617,
      "step": 1918
    },
    {
      "epoch": 0.3838,
      "grad_norm": 0.04280368611216545,
      "learning_rate": 0.00015487310327483086,
      "loss": 2.2601,
      "step": 1919
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.025021247565746307,
      "learning_rate": 0.00015481472617005876,
      "loss": 2.2589,
      "step": 1920
    },
    {
      "epoch": 0.3842,
      "grad_norm": 0.024531258270144463,
      "learning_rate": 0.00015475632234925504,
      "loss": 2.2605,
      "step": 1921
    },
    {
      "epoch": 0.3844,
      "grad_norm": 0.0237723421305418,
      "learning_rate": 0.00015469789184088497,
      "loss": 2.2624,
      "step": 1922
    },
    {
      "epoch": 0.3846,
      "grad_norm": 0.024260051548480988,
      "learning_rate": 0.00015463943467342693,
      "loss": 2.2613,
      "step": 1923
    },
    {
      "epoch": 0.3848,
      "grad_norm": 0.023373903706669807,
      "learning_rate": 0.00015458095087537218,
      "loss": 2.2629,
      "step": 1924
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.03690827637910843,
      "learning_rate": 0.00015452244047522502,
      "loss": 2.2593,
      "step": 1925
    },
    {
      "epoch": 0.3852,
      "grad_norm": 0.0759877860546112,
      "learning_rate": 0.00015446390350150273,
      "loss": 2.2614,
      "step": 1926
    },
    {
      "epoch": 0.3854,
      "grad_norm": 0.1043560653924942,
      "learning_rate": 0.00015440533998273547,
      "loss": 2.2628,
      "step": 1927
    },
    {
      "epoch": 0.3856,
      "grad_norm": 0.09826664626598358,
      "learning_rate": 0.0001543467499474665,
      "loss": 2.2616,
      "step": 1928
    },
    {
      "epoch": 0.3858,
      "grad_norm": 0.047135163098573685,
      "learning_rate": 0.00015428813342425177,
      "loss": 2.2612,
      "step": 1929
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.03344644233584404,
      "learning_rate": 0.0001542294904416603,
      "loss": 2.2603,
      "step": 1930
    },
    {
      "epoch": 0.3862,
      "grad_norm": 0.08567211776971817,
      "learning_rate": 0.000154170821028274,
      "loss": 2.2621,
      "step": 1931
    },
    {
      "epoch": 0.3864,
      "grad_norm": 0.11784695833921432,
      "learning_rate": 0.00015411212521268758,
      "loss": 2.261,
      "step": 1932
    },
    {
      "epoch": 0.3866,
      "grad_norm": 0.11946862936019897,
      "learning_rate": 0.00015405340302350871,
      "loss": 2.261,
      "step": 1933
    },
    {
      "epoch": 0.3868,
      "grad_norm": 0.10050247609615326,
      "learning_rate": 0.00015399465448935788,
      "loss": 2.2637,
      "step": 1934
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.07911248505115509,
      "learning_rate": 0.00015393587963886835,
      "loss": 2.2609,
      "step": 1935
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.04306774213910103,
      "learning_rate": 0.0001538770785006863,
      "loss": 2.2623,
      "step": 1936
    },
    {
      "epoch": 0.3874,
      "grad_norm": 0.03850347176194191,
      "learning_rate": 0.00015381825110347074,
      "loss": 2.2625,
      "step": 1937
    },
    {
      "epoch": 0.3876,
      "grad_norm": 0.10403504967689514,
      "learning_rate": 0.00015375939747589335,
      "loss": 2.261,
      "step": 1938
    },
    {
      "epoch": 0.3878,
      "grad_norm": 0.12785737216472626,
      "learning_rate": 0.0001537005176466387,
      "loss": 2.2618,
      "step": 1939
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.08809279650449753,
      "learning_rate": 0.0001536416116444041,
      "loss": 2.2605,
      "step": 1940
    },
    {
      "epoch": 0.3882,
      "grad_norm": 0.02332674339413643,
      "learning_rate": 0.00015358267949789966,
      "loss": 2.2592,
      "step": 1941
    },
    {
      "epoch": 0.3884,
      "grad_norm": 0.09136989712715149,
      "learning_rate": 0.00015352372123584814,
      "loss": 2.2622,
      "step": 1942
    },
    {
      "epoch": 0.3886,
      "grad_norm": 0.13480578362941742,
      "learning_rate": 0.00015346473688698513,
      "loss": 2.262,
      "step": 1943
    },
    {
      "epoch": 0.3888,
      "grad_norm": 0.11357124149799347,
      "learning_rate": 0.00015340572648005888,
      "loss": 2.2612,
      "step": 1944
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.05769458785653114,
      "learning_rate": 0.0001533466900438303,
      "loss": 2.2624,
      "step": 1945
    },
    {
      "epoch": 0.3892,
      "grad_norm": 0.03092828579246998,
      "learning_rate": 0.000153287627607073,
      "loss": 2.2622,
      "step": 1946
    },
    {
      "epoch": 0.3894,
      "grad_norm": 0.08637729287147522,
      "learning_rate": 0.0001532285391985734,
      "loss": 2.2611,
      "step": 1947
    },
    {
      "epoch": 0.3896,
      "grad_norm": 0.11670393496751785,
      "learning_rate": 0.0001531694248471304,
      "loss": 2.26,
      "step": 1948
    },
    {
      "epoch": 0.3898,
      "grad_norm": 0.10425259172916412,
      "learning_rate": 0.00015311028458155567,
      "loss": 2.2604,
      "step": 1949
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.05621179938316345,
      "learning_rate": 0.0001530511184306734,
      "loss": 2.2617,
      "step": 1950
    },
    {
      "epoch": 0.3902,
      "grad_norm": 0.033249616622924805,
      "learning_rate": 0.0001529919264233205,
      "loss": 2.2631,
      "step": 1951
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.08547902852296829,
      "learning_rate": 0.00015293270858834644,
      "loss": 2.2604,
      "step": 1952
    },
    {
      "epoch": 0.3906,
      "grad_norm": 0.11845478415489197,
      "learning_rate": 0.00015287346495461315,
      "loss": 2.2605,
      "step": 1953
    },
    {
      "epoch": 0.3908,
      "grad_norm": 0.115376316010952,
      "learning_rate": 0.00015281419555099546,
      "loss": 2.259,
      "step": 1954
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.0827345922589302,
      "learning_rate": 0.00015275490040638038,
      "loss": 2.2609,
      "step": 1955
    },
    {
      "epoch": 0.3912,
      "grad_norm": 0.03187431022524834,
      "learning_rate": 0.00015269557954966778,
      "loss": 2.2617,
      "step": 1956
    },
    {
      "epoch": 0.3914,
      "grad_norm": 0.04762377217411995,
      "learning_rate": 0.00015263623300976978,
      "loss": 2.261,
      "step": 1957
    },
    {
      "epoch": 0.3916,
      "grad_norm": 0.0922609195113182,
      "learning_rate": 0.00015257686081561134,
      "loss": 2.2614,
      "step": 1958
    },
    {
      "epoch": 0.3918,
      "grad_norm": 0.09407132863998413,
      "learning_rate": 0.0001525174629961296,
      "loss": 2.2606,
      "step": 1959
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.06695912033319473,
      "learning_rate": 0.00015245803958027434,
      "loss": 2.2622,
      "step": 1960
    },
    {
      "epoch": 0.3922,
      "grad_norm": 0.04236817732453346,
      "learning_rate": 0.00015239859059700794,
      "loss": 2.2617,
      "step": 1961
    },
    {
      "epoch": 0.3924,
      "grad_norm": 0.025699716061353683,
      "learning_rate": 0.00015233911607530497,
      "loss": 2.2607,
      "step": 1962
    },
    {
      "epoch": 0.3926,
      "grad_norm": 0.034357428550720215,
      "learning_rate": 0.00015227961604415266,
      "loss": 2.2599,
      "step": 1963
    },
    {
      "epoch": 0.3928,
      "grad_norm": 0.05360817536711693,
      "learning_rate": 0.0001522200905325506,
      "loss": 2.2601,
      "step": 1964
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.057252850383520126,
      "learning_rate": 0.0001521605395695108,
      "loss": 2.2612,
      "step": 1965
    },
    {
      "epoch": 0.3932,
      "grad_norm": 0.05819157138466835,
      "learning_rate": 0.00015210096318405767,
      "loss": 2.2635,
      "step": 1966
    },
    {
      "epoch": 0.3934,
      "grad_norm": 0.054995447397232056,
      "learning_rate": 0.00015204136140522799,
      "loss": 2.2623,
      "step": 1967
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.03879394754767418,
      "learning_rate": 0.00015198173426207094,
      "loss": 2.2612,
      "step": 1968
    },
    {
      "epoch": 0.3938,
      "grad_norm": 0.026291918009519577,
      "learning_rate": 0.00015192208178364816,
      "loss": 2.261,
      "step": 1969
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.02577153965830803,
      "learning_rate": 0.00015186240399903342,
      "loss": 2.2641,
      "step": 1970
    },
    {
      "epoch": 0.3942,
      "grad_norm": 0.034822769463062286,
      "learning_rate": 0.00015180270093731303,
      "loss": 2.2622,
      "step": 1971
    },
    {
      "epoch": 0.3944,
      "grad_norm": 0.036960747092962265,
      "learning_rate": 0.0001517429726275855,
      "loss": 2.2614,
      "step": 1972
    },
    {
      "epoch": 0.3946,
      "grad_norm": 0.02390749566257,
      "learning_rate": 0.00015168321909896172,
      "loss": 2.2608,
      "step": 1973
    },
    {
      "epoch": 0.3948,
      "grad_norm": 0.036738913506269455,
      "learning_rate": 0.00015162344038056476,
      "loss": 2.2616,
      "step": 1974
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.05242566764354706,
      "learning_rate": 0.00015156363650153012,
      "loss": 2.2595,
      "step": 1975
    },
    {
      "epoch": 0.3952,
      "grad_norm": 0.048569004982709885,
      "learning_rate": 0.00015150380749100545,
      "loss": 2.2614,
      "step": 1976
    },
    {
      "epoch": 0.3954,
      "grad_norm": 0.038096025586128235,
      "learning_rate": 0.00015144395337815064,
      "loss": 2.2599,
      "step": 1977
    },
    {
      "epoch": 0.3956,
      "grad_norm": 0.027490440756082535,
      "learning_rate": 0.00015138407419213796,
      "loss": 2.2601,
      "step": 1978
    },
    {
      "epoch": 0.3958,
      "grad_norm": 0.02372065931558609,
      "learning_rate": 0.0001513241699621517,
      "loss": 2.2623,
      "step": 1979
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.02741408906877041,
      "learning_rate": 0.0001512642407173885,
      "loss": 2.2599,
      "step": 1980
    },
    {
      "epoch": 0.3962,
      "grad_norm": 0.027456004172563553,
      "learning_rate": 0.00015120428648705717,
      "loss": 2.263,
      "step": 1981
    },
    {
      "epoch": 0.3964,
      "grad_norm": 0.023829801008105278,
      "learning_rate": 0.0001511443073003786,
      "loss": 2.2601,
      "step": 1982
    },
    {
      "epoch": 0.3966,
      "grad_norm": 0.03645414859056473,
      "learning_rate": 0.000151084303186586,
      "loss": 2.2606,
      "step": 1983
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.051077768206596375,
      "learning_rate": 0.0001510242741749246,
      "loss": 2.2594,
      "step": 1984
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.05351889133453369,
      "learning_rate": 0.00015096422029465178,
      "loss": 2.2618,
      "step": 1985
    },
    {
      "epoch": 0.3972,
      "grad_norm": 0.04252718761563301,
      "learning_rate": 0.00015090414157503714,
      "loss": 2.2614,
      "step": 1986
    },
    {
      "epoch": 0.3974,
      "grad_norm": 0.039059121161699295,
      "learning_rate": 0.0001508440380453623,
      "loss": 2.2629,
      "step": 1987
    },
    {
      "epoch": 0.3976,
      "grad_norm": 0.03787440061569214,
      "learning_rate": 0.00015078390973492092,
      "loss": 2.2593,
      "step": 1988
    },
    {
      "epoch": 0.3978,
      "grad_norm": 0.03608975186944008,
      "learning_rate": 0.00015072375667301893,
      "loss": 2.2616,
      "step": 1989
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.024763844907283783,
      "learning_rate": 0.0001506635788889741,
      "loss": 2.2597,
      "step": 1990
    },
    {
      "epoch": 0.3982,
      "grad_norm": 0.0287175253033638,
      "learning_rate": 0.00015060337641211637,
      "loss": 2.2603,
      "step": 1991
    },
    {
      "epoch": 0.3984,
      "grad_norm": 0.0381942056119442,
      "learning_rate": 0.0001505431492717878,
      "loss": 2.258,
      "step": 1992
    },
    {
      "epoch": 0.3986,
      "grad_norm": 0.029353875666856766,
      "learning_rate": 0.0001504828974973422,
      "loss": 2.2603,
      "step": 1993
    },
    {
      "epoch": 0.3988,
      "grad_norm": 0.02636108547449112,
      "learning_rate": 0.00015042262111814565,
      "loss": 2.2586,
      "step": 1994
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.046515002846717834,
      "learning_rate": 0.0001503623201635761,
      "loss": 2.2619,
      "step": 1995
    },
    {
      "epoch": 0.3992,
      "grad_norm": 0.05892224237322807,
      "learning_rate": 0.00015030199466302353,
      "loss": 2.2602,
      "step": 1996
    },
    {
      "epoch": 0.3994,
      "grad_norm": 0.053582485765218735,
      "learning_rate": 0.00015024164464588982,
      "loss": 2.262,
      "step": 1997
    },
    {
      "epoch": 0.3996,
      "grad_norm": 0.037817731499671936,
      "learning_rate": 0.00015018127014158886,
      "loss": 2.2624,
      "step": 1998
    },
    {
      "epoch": 0.3998,
      "grad_norm": 0.026197368279099464,
      "learning_rate": 0.00015012087117954642,
      "loss": 2.2585,
      "step": 1999
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.027896298095583916,
      "learning_rate": 0.0001500604477892003,
      "loss": 2.2601,
      "step": 2000
    },
    {
      "epoch": 0.4002,
      "grad_norm": 0.04000100865960121,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.2623,
      "step": 2001
    },
    {
      "epoch": 0.4004,
      "grad_norm": 0.05871916934847832,
      "learning_rate": 0.00014993952784140717,
      "loss": 2.2601,
      "step": 2002
    },
    {
      "epoch": 0.4006,
      "grad_norm": 0.0701272040605545,
      "learning_rate": 0.00014987903134289508,
      "loss": 2.2592,
      "step": 2003
    },
    {
      "epoch": 0.4008,
      "grad_norm": 0.07266467809677124,
      "learning_rate": 0.0001498185105339491,
      "loss": 2.2615,
      "step": 2004
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.050111670047044754,
      "learning_rate": 0.00014975796544406625,
      "loss": 2.2599,
      "step": 2005
    },
    {
      "epoch": 0.4012,
      "grad_norm": 0.02476474642753601,
      "learning_rate": 0.00014969739610275556,
      "loss": 2.2604,
      "step": 2006
    },
    {
      "epoch": 0.4014,
      "grad_norm": 0.060178421437740326,
      "learning_rate": 0.0001496368025395377,
      "loss": 2.262,
      "step": 2007
    },
    {
      "epoch": 0.4016,
      "grad_norm": 0.102263905107975,
      "learning_rate": 0.00014957618478394529,
      "loss": 2.2598,
      "step": 2008
    },
    {
      "epoch": 0.4018,
      "grad_norm": 0.13172194361686707,
      "learning_rate": 0.00014951554286552266,
      "loss": 2.2618,
      "step": 2009
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.14223693311214447,
      "learning_rate": 0.00014945487681382598,
      "loss": 2.2595,
      "step": 2010
    },
    {
      "epoch": 0.4022,
      "grad_norm": 0.12872914969921112,
      "learning_rate": 0.0001493941866584231,
      "loss": 2.2606,
      "step": 2011
    },
    {
      "epoch": 0.4024,
      "grad_norm": 0.08791335672140121,
      "learning_rate": 0.0001493334724288937,
      "loss": 2.2585,
      "step": 2012
    },
    {
      "epoch": 0.4026,
      "grad_norm": 0.03992244973778725,
      "learning_rate": 0.00014927273415482915,
      "loss": 2.2604,
      "step": 2013
    },
    {
      "epoch": 0.4028,
      "grad_norm": 0.03268207982182503,
      "learning_rate": 0.00014921197186583255,
      "loss": 2.2594,
      "step": 2014
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.06557431071996689,
      "learning_rate": 0.0001491511855915187,
      "loss": 2.2594,
      "step": 2015
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.08499819785356522,
      "learning_rate": 0.00014909037536151409,
      "loss": 2.2587,
      "step": 2016
    },
    {
      "epoch": 0.4034,
      "grad_norm": 0.08689701557159424,
      "learning_rate": 0.00014902954120545687,
      "loss": 2.2588,
      "step": 2017
    },
    {
      "epoch": 0.4036,
      "grad_norm": 0.07142952084541321,
      "learning_rate": 0.00014896868315299693,
      "loss": 2.2618,
      "step": 2018
    },
    {
      "epoch": 0.4038,
      "grad_norm": 0.04686310887336731,
      "learning_rate": 0.00014890780123379564,
      "loss": 2.2597,
      "step": 2019
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.031153783202171326,
      "learning_rate": 0.0001488468954775262,
      "loss": 2.2614,
      "step": 2020
    },
    {
      "epoch": 0.4042,
      "grad_norm": 0.026057546958327293,
      "learning_rate": 0.0001487859659138733,
      "loss": 2.2601,
      "step": 2021
    },
    {
      "epoch": 0.4044,
      "grad_norm": 0.04768180102109909,
      "learning_rate": 0.00014872501257253323,
      "loss": 2.2588,
      "step": 2022
    },
    {
      "epoch": 0.4046,
      "grad_norm": 0.07035093009471893,
      "learning_rate": 0.000148664035483214,
      "loss": 2.2597,
      "step": 2023
    },
    {
      "epoch": 0.4048,
      "grad_norm": 0.07788503915071487,
      "learning_rate": 0.00014860303467563503,
      "loss": 2.2596,
      "step": 2024
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.07317241281270981,
      "learning_rate": 0.0001485420101795274,
      "loss": 2.2623,
      "step": 2025
    },
    {
      "epoch": 0.4052,
      "grad_norm": 0.05807656794786453,
      "learning_rate": 0.00014848096202463372,
      "loss": 2.2619,
      "step": 2026
    },
    {
      "epoch": 0.4054,
      "grad_norm": 0.04215865582227707,
      "learning_rate": 0.00014841989024070809,
      "loss": 2.2594,
      "step": 2027
    },
    {
      "epoch": 0.4056,
      "grad_norm": 0.02565750479698181,
      "learning_rate": 0.00014835879485751617,
      "loss": 2.2611,
      "step": 2028
    },
    {
      "epoch": 0.4058,
      "grad_norm": 0.03710125386714935,
      "learning_rate": 0.00014829767590483506,
      "loss": 2.2616,
      "step": 2029
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.07095083594322205,
      "learning_rate": 0.00014823653341245353,
      "loss": 2.2618,
      "step": 2030
    },
    {
      "epoch": 0.4062,
      "grad_norm": 0.09825967997312546,
      "learning_rate": 0.00014817536741017152,
      "loss": 2.2595,
      "step": 2031
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.119840607047081,
      "learning_rate": 0.00014811417792780075,
      "loss": 2.2592,
      "step": 2032
    },
    {
      "epoch": 0.4066,
      "grad_norm": 0.11813266575336456,
      "learning_rate": 0.00014805296499516407,
      "loss": 2.2612,
      "step": 2033
    },
    {
      "epoch": 0.4068,
      "grad_norm": 0.08062906563282013,
      "learning_rate": 0.00014799172864209608,
      "loss": 2.2593,
      "step": 2034
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.02741697058081627,
      "learning_rate": 0.0001479304688984425,
      "loss": 2.2621,
      "step": 2035
    },
    {
      "epoch": 0.4072,
      "grad_norm": 0.05347008630633354,
      "learning_rate": 0.0001478691857940607,
      "loss": 2.2622,
      "step": 2036
    },
    {
      "epoch": 0.4074,
      "grad_norm": 0.08572740852832794,
      "learning_rate": 0.00014780787935881923,
      "loss": 2.2592,
      "step": 2037
    },
    {
      "epoch": 0.4076,
      "grad_norm": 0.07644946128129959,
      "learning_rate": 0.00014774654962259812,
      "loss": 2.2621,
      "step": 2038
    },
    {
      "epoch": 0.4078,
      "grad_norm": 0.03928111866116524,
      "learning_rate": 0.0001476851966152888,
      "loss": 2.2596,
      "step": 2039
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.03044668771326542,
      "learning_rate": 0.0001476238203667939,
      "loss": 2.2601,
      "step": 2040
    },
    {
      "epoch": 0.4082,
      "grad_norm": 0.05659167841076851,
      "learning_rate": 0.00014756242090702756,
      "loss": 2.2614,
      "step": 2041
    },
    {
      "epoch": 0.4084,
      "grad_norm": 0.07490435242652893,
      "learning_rate": 0.00014750099826591498,
      "loss": 2.2604,
      "step": 2042
    },
    {
      "epoch": 0.4086,
      "grad_norm": 0.07422783970832825,
      "learning_rate": 0.00014743955247339293,
      "loss": 2.2608,
      "step": 2043
    },
    {
      "epoch": 0.4088,
      "grad_norm": 0.05250106006860733,
      "learning_rate": 0.00014737808355940932,
      "loss": 2.2603,
      "step": 2044
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.024677814915776253,
      "learning_rate": 0.00014731659155392332,
      "loss": 2.2583,
      "step": 2045
    },
    {
      "epoch": 0.4092,
      "grad_norm": 0.057013172656297684,
      "learning_rate": 0.00014725507648690543,
      "loss": 2.261,
      "step": 2046
    },
    {
      "epoch": 0.4094,
      "grad_norm": 0.09172458201646805,
      "learning_rate": 0.0001471935383883373,
      "loss": 2.2591,
      "step": 2047
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.09741536527872086,
      "learning_rate": 0.00014713197728821183,
      "loss": 2.2613,
      "step": 2048
    },
    {
      "epoch": 0.4098,
      "grad_norm": 0.06985251605510712,
      "learning_rate": 0.0001470703932165333,
      "loss": 2.2558,
      "step": 2049
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.027065498754382133,
      "learning_rate": 0.00014700878620331684,
      "loss": 2.2608,
      "step": 2050
    },
    {
      "epoch": 0.4102,
      "grad_norm": 0.04217565432190895,
      "learning_rate": 0.00014694715627858908,
      "loss": 2.2604,
      "step": 2051
    },
    {
      "epoch": 0.4104,
      "grad_norm": 0.06326907873153687,
      "learning_rate": 0.0001468855034723877,
      "loss": 2.2591,
      "step": 2052
    },
    {
      "epoch": 0.4106,
      "grad_norm": 0.06133463978767395,
      "learning_rate": 0.00014682382781476146,
      "loss": 2.2613,
      "step": 2053
    },
    {
      "epoch": 0.4108,
      "grad_norm": 0.04587830603122711,
      "learning_rate": 0.0001467621293357704,
      "loss": 2.2579,
      "step": 2054
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.034749653190374374,
      "learning_rate": 0.00014670040806548555,
      "loss": 2.2596,
      "step": 2055
    },
    {
      "epoch": 0.4112,
      "grad_norm": 0.0333646684885025,
      "learning_rate": 0.00014663866403398913,
      "loss": 2.2607,
      "step": 2056
    },
    {
      "epoch": 0.4114,
      "grad_norm": 0.029165154322981834,
      "learning_rate": 0.00014657689727137443,
      "loss": 2.2595,
      "step": 2057
    },
    {
      "epoch": 0.4116,
      "grad_norm": 0.02500467747449875,
      "learning_rate": 0.00014651510780774583,
      "loss": 2.2572,
      "step": 2058
    },
    {
      "epoch": 0.4118,
      "grad_norm": 0.03474815934896469,
      "learning_rate": 0.0001464532956732188,
      "loss": 2.2601,
      "step": 2059
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.055847518146038055,
      "learning_rate": 0.0001463914608979197,
      "loss": 2.26,
      "step": 2060
    },
    {
      "epoch": 0.4122,
      "grad_norm": 0.060836032032966614,
      "learning_rate": 0.00014632960351198618,
      "loss": 2.2586,
      "step": 2061
    },
    {
      "epoch": 0.4124,
      "grad_norm": 0.04336310550570488,
      "learning_rate": 0.0001462677235455667,
      "loss": 2.2599,
      "step": 2062
    },
    {
      "epoch": 0.4126,
      "grad_norm": 0.03108673356473446,
      "learning_rate": 0.00014620582102882089,
      "loss": 2.2599,
      "step": 2063
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.029055655002593994,
      "learning_rate": 0.00014614389599191917,
      "loss": 2.258,
      "step": 2064
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.031600471585989,
      "learning_rate": 0.0001460819484650431,
      "loss": 2.2587,
      "step": 2065
    },
    {
      "epoch": 0.4132,
      "grad_norm": 0.02936188504099846,
      "learning_rate": 0.00014601997847838518,
      "loss": 2.2598,
      "step": 2066
    },
    {
      "epoch": 0.4134,
      "grad_norm": 0.026178373023867607,
      "learning_rate": 0.00014595798606214882,
      "loss": 2.2617,
      "step": 2067
    },
    {
      "epoch": 0.4136,
      "grad_norm": 0.029225297272205353,
      "learning_rate": 0.00014589597124654833,
      "loss": 2.2602,
      "step": 2068
    },
    {
      "epoch": 0.4138,
      "grad_norm": 0.03574533760547638,
      "learning_rate": 0.000145833934061809,
      "loss": 2.2589,
      "step": 2069
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.03636125847697258,
      "learning_rate": 0.000145771874538167,
      "loss": 2.2577,
      "step": 2070
    },
    {
      "epoch": 0.4142,
      "grad_norm": 0.029872214421629906,
      "learning_rate": 0.00014570979270586945,
      "loss": 2.2602,
      "step": 2071
    },
    {
      "epoch": 0.4144,
      "grad_norm": 0.02558109723031521,
      "learning_rate": 0.00014564768859517418,
      "loss": 2.2586,
      "step": 2072
    },
    {
      "epoch": 0.4146,
      "grad_norm": 0.02614535763859749,
      "learning_rate": 0.00014558556223635003,
      "loss": 2.2598,
      "step": 2073
    },
    {
      "epoch": 0.4148,
      "grad_norm": 0.02863701432943344,
      "learning_rate": 0.00014552341365967658,
      "loss": 2.2607,
      "step": 2074
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.026249319314956665,
      "learning_rate": 0.0001454612428954444,
      "loss": 2.261,
      "step": 2075
    },
    {
      "epoch": 0.4152,
      "grad_norm": 0.0262614656239748,
      "learning_rate": 0.00014539904997395468,
      "loss": 2.2614,
      "step": 2076
    },
    {
      "epoch": 0.4154,
      "grad_norm": 0.026835184544324875,
      "learning_rate": 0.00014533683492551952,
      "loss": 2.2609,
      "step": 2077
    },
    {
      "epoch": 0.4156,
      "grad_norm": 0.026332031935453415,
      "learning_rate": 0.0001452745977804618,
      "loss": 2.2601,
      "step": 2078
    },
    {
      "epoch": 0.4158,
      "grad_norm": 0.05701577290892601,
      "learning_rate": 0.00014521233856911508,
      "loss": 2.2624,
      "step": 2079
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.10317797213792801,
      "learning_rate": 0.00014515005732182383,
      "loss": 2.2593,
      "step": 2080
    },
    {
      "epoch": 0.4162,
      "grad_norm": 0.13751333951950073,
      "learning_rate": 0.00014508775406894307,
      "loss": 2.2592,
      "step": 2081
    },
    {
      "epoch": 0.4164,
      "grad_norm": 0.14162687957286835,
      "learning_rate": 0.00014502542884083875,
      "loss": 2.2618,
      "step": 2082
    },
    {
      "epoch": 0.4166,
      "grad_norm": 0.12804268300533295,
      "learning_rate": 0.0001449630816678874,
      "loss": 2.2583,
      "step": 2083
    },
    {
      "epoch": 0.4168,
      "grad_norm": 0.10165403038263321,
      "learning_rate": 0.00014490071258047623,
      "loss": 2.2622,
      "step": 2084
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.06717651337385178,
      "learning_rate": 0.00014483832160900326,
      "loss": 2.2593,
      "step": 2085
    },
    {
      "epoch": 0.4172,
      "grad_norm": 0.036162205040454865,
      "learning_rate": 0.00014477590878387696,
      "loss": 2.2592,
      "step": 2086
    },
    {
      "epoch": 0.4174,
      "grad_norm": 0.025380820035934448,
      "learning_rate": 0.00014471347413551672,
      "loss": 2.2563,
      "step": 2087
    },
    {
      "epoch": 0.4176,
      "grad_norm": 0.03313155844807625,
      "learning_rate": 0.00014465101769435234,
      "loss": 2.2615,
      "step": 2088
    },
    {
      "epoch": 0.4178,
      "grad_norm": 0.030506109818816185,
      "learning_rate": 0.00014458853949082443,
      "loss": 2.2595,
      "step": 2089
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.02590843476355076,
      "learning_rate": 0.00014452603955538397,
      "loss": 2.2589,
      "step": 2090
    },
    {
      "epoch": 0.4182,
      "grad_norm": 0.025481589138507843,
      "learning_rate": 0.00014446351791849276,
      "loss": 2.2598,
      "step": 2091
    },
    {
      "epoch": 0.4184,
      "grad_norm": 0.026572681963443756,
      "learning_rate": 0.00014440097461062307,
      "loss": 2.2601,
      "step": 2092
    },
    {
      "epoch": 0.4186,
      "grad_norm": 0.02983669564127922,
      "learning_rate": 0.00014433840966225772,
      "loss": 2.2581,
      "step": 2093
    },
    {
      "epoch": 0.4188,
      "grad_norm": 0.03645830228924751,
      "learning_rate": 0.0001442758231038902,
      "loss": 2.2571,
      "step": 2094
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.055296640843153,
      "learning_rate": 0.00014421321496602428,
      "loss": 2.2587,
      "step": 2095
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.05643782019615173,
      "learning_rate": 0.00014415058527917452,
      "loss": 2.2606,
      "step": 2096
    },
    {
      "epoch": 0.4194,
      "grad_norm": 0.046807385981082916,
      "learning_rate": 0.00014408793407386588,
      "loss": 2.258,
      "step": 2097
    },
    {
      "epoch": 0.4196,
      "grad_norm": 0.026586251333355904,
      "learning_rate": 0.00014402526138063373,
      "loss": 2.2574,
      "step": 2098
    },
    {
      "epoch": 0.4198,
      "grad_norm": 0.03826276212930679,
      "learning_rate": 0.000143962567230024,
      "loss": 2.2614,
      "step": 2099
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.0687408447265625,
      "learning_rate": 0.00014389985165259308,
      "loss": 2.2607,
      "step": 2100
    },
    {
      "epoch": 0.4202,
      "grad_norm": 0.10221124440431595,
      "learning_rate": 0.00014383711467890774,
      "loss": 2.2592,
      "step": 2101
    },
    {
      "epoch": 0.4204,
      "grad_norm": 0.1246868371963501,
      "learning_rate": 0.00014377435633954527,
      "loss": 2.2604,
      "step": 2102
    },
    {
      "epoch": 0.4206,
      "grad_norm": 0.1197252869606018,
      "learning_rate": 0.0001437115766650933,
      "loss": 2.2593,
      "step": 2103
    },
    {
      "epoch": 0.4208,
      "grad_norm": 0.09681469947099686,
      "learning_rate": 0.0001436487756861499,
      "loss": 2.2572,
      "step": 2104
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.058221954852342606,
      "learning_rate": 0.00014358595343332342,
      "loss": 2.2605,
      "step": 2105
    },
    {
      "epoch": 0.4212,
      "grad_norm": 0.026096820831298828,
      "learning_rate": 0.00014352310993723277,
      "loss": 2.2576,
      "step": 2106
    },
    {
      "epoch": 0.4214,
      "grad_norm": 0.055660977959632874,
      "learning_rate": 0.00014346024522850703,
      "loss": 2.2591,
      "step": 2107
    },
    {
      "epoch": 0.4216,
      "grad_norm": 0.09871785342693329,
      "learning_rate": 0.00014339735933778576,
      "loss": 2.2589,
      "step": 2108
    },
    {
      "epoch": 0.4218,
      "grad_norm": 0.13600580394268036,
      "learning_rate": 0.00014333445229571873,
      "loss": 2.2615,
      "step": 2109
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.15234515070915222,
      "learning_rate": 0.00014327152413296608,
      "loss": 2.2581,
      "step": 2110
    },
    {
      "epoch": 0.4222,
      "grad_norm": 0.14479351043701172,
      "learning_rate": 0.00014320857488019824,
      "loss": 2.26,
      "step": 2111
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.10252978652715683,
      "learning_rate": 0.0001431456045680959,
      "loss": 2.2606,
      "step": 2112
    },
    {
      "epoch": 0.4226,
      "grad_norm": 0.0494791716337204,
      "learning_rate": 0.00014308261322735005,
      "loss": 2.2606,
      "step": 2113
    },
    {
      "epoch": 0.4228,
      "grad_norm": 0.0277389045804739,
      "learning_rate": 0.00014301960088866186,
      "loss": 2.2589,
      "step": 2114
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.06273595243692398,
      "learning_rate": 0.00014295656758274284,
      "loss": 2.2591,
      "step": 2115
    },
    {
      "epoch": 0.4232,
      "grad_norm": 0.10638439655303955,
      "learning_rate": 0.0001428935133403146,
      "loss": 2.2588,
      "step": 2116
    },
    {
      "epoch": 0.4234,
      "grad_norm": 0.14077924191951752,
      "learning_rate": 0.00014283043819210905,
      "loss": 2.2583,
      "step": 2117
    },
    {
      "epoch": 0.4236,
      "grad_norm": 0.16152983903884888,
      "learning_rate": 0.00014276734216886821,
      "loss": 2.2599,
      "step": 2118
    },
    {
      "epoch": 0.4238,
      "grad_norm": 0.1581408977508545,
      "learning_rate": 0.00014270422530134432,
      "loss": 2.2577,
      "step": 2119
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.12424567341804504,
      "learning_rate": 0.0001426410876202999,
      "loss": 2.258,
      "step": 2120
    },
    {
      "epoch": 0.4242,
      "grad_norm": 0.0574287511408329,
      "learning_rate": 0.00014257792915650728,
      "loss": 2.2609,
      "step": 2121
    },
    {
      "epoch": 0.4244,
      "grad_norm": 0.042300786823034286,
      "learning_rate": 0.00014251474994074928,
      "loss": 2.2586,
      "step": 2122
    },
    {
      "epoch": 0.4246,
      "grad_norm": 0.11969464272260666,
      "learning_rate": 0.0001424515500038186,
      "loss": 2.259,
      "step": 2123
    },
    {
      "epoch": 0.4248,
      "grad_norm": 0.1794278621673584,
      "learning_rate": 0.00014238832937651816,
      "loss": 2.2568,
      "step": 2124
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.19864051043987274,
      "learning_rate": 0.00014232508808966098,
      "loss": 2.2582,
      "step": 2125
    },
    {
      "epoch": 0.4252,
      "grad_norm": 0.1696343868970871,
      "learning_rate": 0.00014226182617406996,
      "loss": 2.2634,
      "step": 2126
    },
    {
      "epoch": 0.4254,
      "grad_norm": 0.11584246158599854,
      "learning_rate": 0.0001421985436605783,
      "loss": 2.2597,
      "step": 2127
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.04813772067427635,
      "learning_rate": 0.0001421352405800291,
      "loss": 2.2581,
      "step": 2128
    },
    {
      "epoch": 0.4258,
      "grad_norm": 0.043385595083236694,
      "learning_rate": 0.00014207191696327548,
      "loss": 2.2597,
      "step": 2129
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.10279418528079987,
      "learning_rate": 0.00014200857284118066,
      "loss": 2.2586,
      "step": 2130
    },
    {
      "epoch": 0.4262,
      "grad_norm": 0.12825998663902283,
      "learning_rate": 0.00014194520824461771,
      "loss": 2.2582,
      "step": 2131
    },
    {
      "epoch": 0.4264,
      "grad_norm": 0.12130255252122879,
      "learning_rate": 0.00014188182320446985,
      "loss": 2.2563,
      "step": 2132
    },
    {
      "epoch": 0.4266,
      "grad_norm": 0.09914086759090424,
      "learning_rate": 0.00014181841775163013,
      "loss": 2.2592,
      "step": 2133
    },
    {
      "epoch": 0.4268,
      "grad_norm": 0.07350296527147293,
      "learning_rate": 0.00014175499191700167,
      "loss": 2.258,
      "step": 2134
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.03388955071568489,
      "learning_rate": 0.00014169154573149737,
      "loss": 2.2595,
      "step": 2135
    },
    {
      "epoch": 0.4272,
      "grad_norm": 0.03861885517835617,
      "learning_rate": 0.00014162807922604012,
      "loss": 2.2594,
      "step": 2136
    },
    {
      "epoch": 0.4274,
      "grad_norm": 0.06287501007318497,
      "learning_rate": 0.0001415645924315628,
      "loss": 2.2596,
      "step": 2137
    },
    {
      "epoch": 0.4276,
      "grad_norm": 0.07868903875350952,
      "learning_rate": 0.00014150108537900805,
      "loss": 2.2592,
      "step": 2138
    },
    {
      "epoch": 0.4278,
      "grad_norm": 0.08550340682268143,
      "learning_rate": 0.00014143755809932845,
      "loss": 2.2621,
      "step": 2139
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.0768226608633995,
      "learning_rate": 0.00014137401062348638,
      "loss": 2.2589,
      "step": 2140
    },
    {
      "epoch": 0.4282,
      "grad_norm": 0.05659716948866844,
      "learning_rate": 0.0001413104429824542,
      "loss": 2.2586,
      "step": 2141
    },
    {
      "epoch": 0.4284,
      "grad_norm": 0.027507396414875984,
      "learning_rate": 0.00014124685520721392,
      "loss": 2.2575,
      "step": 2142
    },
    {
      "epoch": 0.4286,
      "grad_norm": 0.0405757836997509,
      "learning_rate": 0.00014118324732875748,
      "loss": 2.2589,
      "step": 2143
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.06810106337070465,
      "learning_rate": 0.00014111961937808665,
      "loss": 2.258,
      "step": 2144
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.07672035694122314,
      "learning_rate": 0.0001410559713862128,
      "loss": 2.2584,
      "step": 2145
    },
    {
      "epoch": 0.4292,
      "grad_norm": 0.06635713577270508,
      "learning_rate": 0.00014099230338415728,
      "loss": 2.2589,
      "step": 2146
    },
    {
      "epoch": 0.4294,
      "grad_norm": 0.037616729736328125,
      "learning_rate": 0.00014092861540295108,
      "loss": 2.2593,
      "step": 2147
    },
    {
      "epoch": 0.4296,
      "grad_norm": 0.032902639359235764,
      "learning_rate": 0.00014086490747363493,
      "loss": 2.2595,
      "step": 2148
    },
    {
      "epoch": 0.4298,
      "grad_norm": 0.0635729506611824,
      "learning_rate": 0.0001408011796272593,
      "loss": 2.2572,
      "step": 2149
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.07905949652194977,
      "learning_rate": 0.00014073743189488435,
      "loss": 2.2594,
      "step": 2150
    },
    {
      "epoch": 0.4302,
      "grad_norm": 0.07411275058984756,
      "learning_rate": 0.00014067366430758004,
      "loss": 2.2578,
      "step": 2151
    },
    {
      "epoch": 0.4304,
      "grad_norm": 0.05170160159468651,
      "learning_rate": 0.00014060987689642581,
      "loss": 2.2594,
      "step": 2152
    },
    {
      "epoch": 0.4306,
      "grad_norm": 0.030840683728456497,
      "learning_rate": 0.00014054606969251095,
      "loss": 2.2582,
      "step": 2153
    },
    {
      "epoch": 0.4308,
      "grad_norm": 0.0306889396160841,
      "learning_rate": 0.00014048224272693424,
      "loss": 2.2592,
      "step": 2154
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.04626224935054779,
      "learning_rate": 0.00014041839603080422,
      "loss": 2.2572,
      "step": 2155
    },
    {
      "epoch": 0.4312,
      "grad_norm": 0.06176237389445305,
      "learning_rate": 0.00014035452963523902,
      "loss": 2.2586,
      "step": 2156
    },
    {
      "epoch": 0.4314,
      "grad_norm": 0.07059957832098007,
      "learning_rate": 0.00014029064357136628,
      "loss": 2.2575,
      "step": 2157
    },
    {
      "epoch": 0.4316,
      "grad_norm": 0.06624983251094818,
      "learning_rate": 0.00014022673787032332,
      "loss": 2.2618,
      "step": 2158
    },
    {
      "epoch": 0.4318,
      "grad_norm": 0.04265427961945534,
      "learning_rate": 0.000140162812563257,
      "loss": 2.2586,
      "step": 2159
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.026677483692765236,
      "learning_rate": 0.00014009886768132375,
      "loss": 2.2573,
      "step": 2160
    },
    {
      "epoch": 0.4322,
      "grad_norm": 0.05127473175525665,
      "learning_rate": 0.00014003490325568954,
      "loss": 2.2585,
      "step": 2161
    },
    {
      "epoch": 0.4324,
      "grad_norm": 0.08246038854122162,
      "learning_rate": 0.00013997091931752977,
      "loss": 2.2575,
      "step": 2162
    },
    {
      "epoch": 0.4326,
      "grad_norm": 0.09877064824104309,
      "learning_rate": 0.00013990691589802954,
      "loss": 2.2571,
      "step": 2163
    },
    {
      "epoch": 0.4328,
      "grad_norm": 0.07329457998275757,
      "learning_rate": 0.00013984289302838328,
      "loss": 2.2596,
      "step": 2164
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.027480844408273697,
      "learning_rate": 0.000139778850739795,
      "loss": 2.2577,
      "step": 2165
    },
    {
      "epoch": 0.4332,
      "grad_norm": 0.07057588547468185,
      "learning_rate": 0.00013971478906347806,
      "loss": 2.2586,
      "step": 2166
    },
    {
      "epoch": 0.4334,
      "grad_norm": 0.12325457483530045,
      "learning_rate": 0.00013965070803065543,
      "loss": 2.2572,
      "step": 2167
    },
    {
      "epoch": 0.4336,
      "grad_norm": 0.15362600982189178,
      "learning_rate": 0.00013958660767255938,
      "loss": 2.2591,
      "step": 2168
    },
    {
      "epoch": 0.4338,
      "grad_norm": 0.14236006140708923,
      "learning_rate": 0.00013952248802043165,
      "loss": 2.2598,
      "step": 2169
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.09306509792804718,
      "learning_rate": 0.0001394583491055234,
      "loss": 2.2595,
      "step": 2170
    },
    {
      "epoch": 0.4342,
      "grad_norm": 0.027677100151777267,
      "learning_rate": 0.00013939419095909512,
      "loss": 2.2603,
      "step": 2171
    },
    {
      "epoch": 0.4344,
      "grad_norm": 0.08697117865085602,
      "learning_rate": 0.00013933001361241673,
      "loss": 2.2574,
      "step": 2172
    },
    {
      "epoch": 0.4346,
      "grad_norm": 0.1569230556488037,
      "learning_rate": 0.00013926581709676751,
      "loss": 2.2593,
      "step": 2173
    },
    {
      "epoch": 0.4348,
      "grad_norm": 0.19775418937206268,
      "learning_rate": 0.00013920160144343603,
      "loss": 2.2578,
      "step": 2174
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.1955849677324295,
      "learning_rate": 0.00013913736668372026,
      "loss": 2.2604,
      "step": 2175
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.13695724308490753,
      "learning_rate": 0.00013907311284892736,
      "loss": 2.261,
      "step": 2176
    },
    {
      "epoch": 0.4354,
      "grad_norm": 0.04658212885260582,
      "learning_rate": 0.00013900883997037397,
      "loss": 2.2593,
      "step": 2177
    },
    {
      "epoch": 0.4356,
      "grad_norm": 0.06932888180017471,
      "learning_rate": 0.00013894454807938586,
      "loss": 2.26,
      "step": 2178
    },
    {
      "epoch": 0.4358,
      "grad_norm": 0.1360175758600235,
      "learning_rate": 0.0001388802372072981,
      "loss": 2.2599,
      "step": 2179
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.16079944372177124,
      "learning_rate": 0.00013881590738545508,
      "loss": 2.2599,
      "step": 2180
    },
    {
      "epoch": 0.4362,
      "grad_norm": 0.132491335272789,
      "learning_rate": 0.0001387515586452103,
      "loss": 2.2594,
      "step": 2181
    },
    {
      "epoch": 0.4364,
      "grad_norm": 0.06526734679937363,
      "learning_rate": 0.00013868719101792665,
      "loss": 2.2591,
      "step": 2182
    },
    {
      "epoch": 0.4366,
      "grad_norm": 0.040824003517627716,
      "learning_rate": 0.000138622804534976,
      "loss": 2.2592,
      "step": 2183
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.10803896188735962,
      "learning_rate": 0.00013855839922773968,
      "loss": 2.2599,
      "step": 2184
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.15375788509845734,
      "learning_rate": 0.00013849397512760795,
      "loss": 2.2588,
      "step": 2185
    },
    {
      "epoch": 0.4372,
      "grad_norm": 0.14944832026958466,
      "learning_rate": 0.00013842953226598037,
      "loss": 2.2607,
      "step": 2186
    },
    {
      "epoch": 0.4374,
      "grad_norm": 0.09866686165332794,
      "learning_rate": 0.00013836507067426564,
      "loss": 2.261,
      "step": 2187
    },
    {
      "epoch": 0.4376,
      "grad_norm": 0.03125014528632164,
      "learning_rate": 0.0001383005903838815,
      "loss": 2.2576,
      "step": 2188
    },
    {
      "epoch": 0.4378,
      "grad_norm": 0.06501078605651855,
      "learning_rate": 0.0001382360914262549,
      "loss": 2.2585,
      "step": 2189
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.11343173682689667,
      "learning_rate": 0.00013817157383282184,
      "loss": 2.2608,
      "step": 2190
    },
    {
      "epoch": 0.4382,
      "grad_norm": 0.12367937713861465,
      "learning_rate": 0.00013810703763502744,
      "loss": 2.2585,
      "step": 2191
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.105565644800663,
      "learning_rate": 0.00013804248286432578,
      "loss": 2.2597,
      "step": 2192
    },
    {
      "epoch": 0.4386,
      "grad_norm": 0.07272915542125702,
      "learning_rate": 0.00013797790955218014,
      "loss": 2.2573,
      "step": 2193
    },
    {
      "epoch": 0.4388,
      "grad_norm": 0.0447295717895031,
      "learning_rate": 0.0001379133177300627,
      "loss": 2.2598,
      "step": 2194
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.028797926381230354,
      "learning_rate": 0.00013784870742945482,
      "loss": 2.2575,
      "step": 2195
    },
    {
      "epoch": 0.4392,
      "grad_norm": 0.027288835495710373,
      "learning_rate": 0.00013778407868184672,
      "loss": 2.2563,
      "step": 2196
    },
    {
      "epoch": 0.4394,
      "grad_norm": 0.031174026429653168,
      "learning_rate": 0.00013771943151873767,
      "loss": 2.2574,
      "step": 2197
    },
    {
      "epoch": 0.4396,
      "grad_norm": 0.03848635405302048,
      "learning_rate": 0.00013765476597163594,
      "loss": 2.2573,
      "step": 2198
    },
    {
      "epoch": 0.4398,
      "grad_norm": 0.046747419983148575,
      "learning_rate": 0.0001375900820720587,
      "loss": 2.26,
      "step": 2199
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.05338534712791443,
      "learning_rate": 0.0001375253798515321,
      "loss": 2.2573,
      "step": 2200
    },
    {
      "epoch": 0.4402,
      "grad_norm": 0.056610651314258575,
      "learning_rate": 0.00013746065934159123,
      "loss": 2.2575,
      "step": 2201
    },
    {
      "epoch": 0.4404,
      "grad_norm": 0.04106359928846359,
      "learning_rate": 0.00013739592057378003,
      "loss": 2.256,
      "step": 2202
    },
    {
      "epoch": 0.4406,
      "grad_norm": 0.027446581050753593,
      "learning_rate": 0.0001373311635796515,
      "loss": 2.2596,
      "step": 2203
    },
    {
      "epoch": 0.4408,
      "grad_norm": 0.0419372096657753,
      "learning_rate": 0.0001372663883907673,
      "loss": 2.2567,
      "step": 2204
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.0462217777967453,
      "learning_rate": 0.00013720159503869815,
      "loss": 2.2557,
      "step": 2205
    },
    {
      "epoch": 0.4412,
      "grad_norm": 0.04186539351940155,
      "learning_rate": 0.00013713678355502351,
      "loss": 2.2583,
      "step": 2206
    },
    {
      "epoch": 0.4414,
      "grad_norm": 0.03896912559866905,
      "learning_rate": 0.00013707195397133165,
      "loss": 2.2568,
      "step": 2207
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.04099274054169655,
      "learning_rate": 0.00013700710631921984,
      "loss": 2.2594,
      "step": 2208
    },
    {
      "epoch": 0.4418,
      "grad_norm": 0.03512291610240936,
      "learning_rate": 0.00013694224063029396,
      "loss": 2.2565,
      "step": 2209
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.027263741940259933,
      "learning_rate": 0.00013687735693616876,
      "loss": 2.2561,
      "step": 2210
    },
    {
      "epoch": 0.4422,
      "grad_norm": 0.030058296397328377,
      "learning_rate": 0.00013681245526846783,
      "loss": 2.2563,
      "step": 2211
    },
    {
      "epoch": 0.4424,
      "grad_norm": 0.027827739715576172,
      "learning_rate": 0.00013674753565882334,
      "loss": 2.2609,
      "step": 2212
    },
    {
      "epoch": 0.4426,
      "grad_norm": 0.03865962103009224,
      "learning_rate": 0.00013668259813887643,
      "loss": 2.2576,
      "step": 2213
    },
    {
      "epoch": 0.4428,
      "grad_norm": 0.05302705243229866,
      "learning_rate": 0.00013661764274027678,
      "loss": 2.259,
      "step": 2214
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.05075295642018318,
      "learning_rate": 0.0001365526694946829,
      "loss": 2.2576,
      "step": 2215
    },
    {
      "epoch": 0.4432,
      "grad_norm": 0.039108917117118835,
      "learning_rate": 0.00013648767843376196,
      "loss": 2.2571,
      "step": 2216
    },
    {
      "epoch": 0.4434,
      "grad_norm": 0.029007162898778915,
      "learning_rate": 0.00013642266958918984,
      "loss": 2.2603,
      "step": 2217
    },
    {
      "epoch": 0.4436,
      "grad_norm": 0.0471901074051857,
      "learning_rate": 0.000136357642992651,
      "loss": 2.2584,
      "step": 2218
    },
    {
      "epoch": 0.4438,
      "grad_norm": 0.051886338740587234,
      "learning_rate": 0.00013629259867583863,
      "loss": 2.2563,
      "step": 2219
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.03342874348163605,
      "learning_rate": 0.00013622753667045457,
      "loss": 2.257,
      "step": 2220
    },
    {
      "epoch": 0.4442,
      "grad_norm": 0.03618893399834633,
      "learning_rate": 0.00013616245700820922,
      "loss": 2.258,
      "step": 2221
    },
    {
      "epoch": 0.4444,
      "grad_norm": 0.07793637365102768,
      "learning_rate": 0.00013609735972082166,
      "loss": 2.2575,
      "step": 2222
    },
    {
      "epoch": 0.4446,
      "grad_norm": 0.11611457914113998,
      "learning_rate": 0.00013603224484001948,
      "loss": 2.259,
      "step": 2223
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.12703777849674225,
      "learning_rate": 0.0001359671123975389,
      "loss": 2.2579,
      "step": 2224
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.09591969847679138,
      "learning_rate": 0.00013590196242512463,
      "loss": 2.2572,
      "step": 2225
    },
    {
      "epoch": 0.4452,
      "grad_norm": 0.042675893753767014,
      "learning_rate": 0.00013583679495453,
      "loss": 2.2568,
      "step": 2226
    },
    {
      "epoch": 0.4454,
      "grad_norm": 0.03781604766845703,
      "learning_rate": 0.00013577161001751694,
      "loss": 2.2587,
      "step": 2227
    },
    {
      "epoch": 0.4456,
      "grad_norm": 0.07004055380821228,
      "learning_rate": 0.00013570640764585566,
      "loss": 2.257,
      "step": 2228
    },
    {
      "epoch": 0.4458,
      "grad_norm": 0.09123716503381729,
      "learning_rate": 0.00013564118787132506,
      "loss": 2.2581,
      "step": 2229
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.08200162649154663,
      "learning_rate": 0.0001355759507257125,
      "loss": 2.2583,
      "step": 2230
    },
    {
      "epoch": 0.4462,
      "grad_norm": 0.05407446622848511,
      "learning_rate": 0.0001355106962408137,
      "loss": 2.2584,
      "step": 2231
    },
    {
      "epoch": 0.4464,
      "grad_norm": 0.02854936011135578,
      "learning_rate": 0.00013544542444843299,
      "loss": 2.256,
      "step": 2232
    },
    {
      "epoch": 0.4466,
      "grad_norm": 0.07894111424684525,
      "learning_rate": 0.00013538013538038295,
      "loss": 2.2568,
      "step": 2233
    },
    {
      "epoch": 0.4468,
      "grad_norm": 0.13321958482265472,
      "learning_rate": 0.00013531482906848475,
      "loss": 2.2567,
      "step": 2234
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.1313214749097824,
      "learning_rate": 0.00013524950554456784,
      "loss": 2.2586,
      "step": 2235
    },
    {
      "epoch": 0.4472,
      "grad_norm": 0.0780629813671112,
      "learning_rate": 0.00013518416484047018,
      "loss": 2.257,
      "step": 2236
    },
    {
      "epoch": 0.4474,
      "grad_norm": 0.02797723561525345,
      "learning_rate": 0.00013511880698803801,
      "loss": 2.2588,
      "step": 2237
    },
    {
      "epoch": 0.4476,
      "grad_norm": 0.07180168479681015,
      "learning_rate": 0.0001350534320191259,
      "loss": 2.2583,
      "step": 2238
    },
    {
      "epoch": 0.4478,
      "grad_norm": 0.10387291014194489,
      "learning_rate": 0.00013498803996559692,
      "loss": 2.2586,
      "step": 2239
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.09638217836618423,
      "learning_rate": 0.00013492263085932224,
      "loss": 2.2592,
      "step": 2240
    },
    {
      "epoch": 0.4482,
      "grad_norm": 0.05814046040177345,
      "learning_rate": 0.00013485720473218154,
      "loss": 2.2582,
      "step": 2241
    },
    {
      "epoch": 0.4484,
      "grad_norm": 0.02860928513109684,
      "learning_rate": 0.0001347917616160627,
      "loss": 2.2571,
      "step": 2242
    },
    {
      "epoch": 0.4486,
      "grad_norm": 0.06418884545564651,
      "learning_rate": 0.0001347263015428619,
      "loss": 2.259,
      "step": 2243
    },
    {
      "epoch": 0.4488,
      "grad_norm": 0.09886135905981064,
      "learning_rate": 0.00013466082454448362,
      "loss": 2.2566,
      "step": 2244
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.10065021365880966,
      "learning_rate": 0.0001345953306528405,
      "loss": 2.2559,
      "step": 2245
    },
    {
      "epoch": 0.4492,
      "grad_norm": 0.07858720421791077,
      "learning_rate": 0.00013452981989985348,
      "loss": 2.2556,
      "step": 2246
    },
    {
      "epoch": 0.4494,
      "grad_norm": 0.04048420116305351,
      "learning_rate": 0.0001344642923174517,
      "loss": 2.257,
      "step": 2247
    },
    {
      "epoch": 0.4496,
      "grad_norm": 0.030398570001125336,
      "learning_rate": 0.00013439874793757254,
      "loss": 2.2562,
      "step": 2248
    },
    {
      "epoch": 0.4498,
      "grad_norm": 0.049281928688287735,
      "learning_rate": 0.00013433318679216153,
      "loss": 2.2536,
      "step": 2249
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.06432965397834778,
      "learning_rate": 0.00013426760891317236,
      "loss": 2.2606,
      "step": 2250
    },
    {
      "epoch": 0.4502,
      "grad_norm": 0.06477135419845581,
      "learning_rate": 0.00013420201433256689,
      "loss": 2.2564,
      "step": 2251
    },
    {
      "epoch": 0.4504,
      "grad_norm": 0.04786878824234009,
      "learning_rate": 0.0001341364030823151,
      "loss": 2.2561,
      "step": 2252
    },
    {
      "epoch": 0.4506,
      "grad_norm": 0.0303571205586195,
      "learning_rate": 0.0001340707751943952,
      "loss": 2.2585,
      "step": 2253
    },
    {
      "epoch": 0.4508,
      "grad_norm": 0.029651883989572525,
      "learning_rate": 0.0001340051307007933,
      "loss": 2.2585,
      "step": 2254
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.030921634286642075,
      "learning_rate": 0.00013393946963350382,
      "loss": 2.2586,
      "step": 2255
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.03462986648082733,
      "learning_rate": 0.00013387379202452917,
      "loss": 2.2558,
      "step": 2256
    },
    {
      "epoch": 0.4514,
      "grad_norm": 0.03499053046107292,
      "learning_rate": 0.00013380809790587974,
      "loss": 2.2569,
      "step": 2257
    },
    {
      "epoch": 0.4516,
      "grad_norm": 0.03143054619431496,
      "learning_rate": 0.00013374238730957412,
      "loss": 2.2576,
      "step": 2258
    },
    {
      "epoch": 0.4518,
      "grad_norm": 0.029150133952498436,
      "learning_rate": 0.00013367666026763882,
      "loss": 2.258,
      "step": 2259
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.03247435763478279,
      "learning_rate": 0.00013361091681210845,
      "loss": 2.258,
      "step": 2260
    },
    {
      "epoch": 0.4522,
      "grad_norm": 0.0404776893556118,
      "learning_rate": 0.00013354515697502553,
      "loss": 2.259,
      "step": 2261
    },
    {
      "epoch": 0.4524,
      "grad_norm": 0.04394126683473587,
      "learning_rate": 0.0001334793807884406,
      "loss": 2.2603,
      "step": 2262
    },
    {
      "epoch": 0.4526,
      "grad_norm": 0.04055987298488617,
      "learning_rate": 0.00013341358828441218,
      "loss": 2.2607,
      "step": 2263
    },
    {
      "epoch": 0.4528,
      "grad_norm": 0.029747869819402695,
      "learning_rate": 0.00013334777949500673,
      "loss": 2.2597,
      "step": 2264
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.03828541934490204,
      "learning_rate": 0.00013328195445229868,
      "loss": 2.2562,
      "step": 2265
    },
    {
      "epoch": 0.4532,
      "grad_norm": 0.06719018518924713,
      "learning_rate": 0.00013321611318837032,
      "loss": 2.2583,
      "step": 2266
    },
    {
      "epoch": 0.4534,
      "grad_norm": 0.0805705115199089,
      "learning_rate": 0.00013315025573531198,
      "loss": 2.2582,
      "step": 2267
    },
    {
      "epoch": 0.4536,
      "grad_norm": 0.06579513847827911,
      "learning_rate": 0.00013308438212522163,
      "loss": 2.2571,
      "step": 2268
    },
    {
      "epoch": 0.4538,
      "grad_norm": 0.03241194039583206,
      "learning_rate": 0.00013301849239020537,
      "loss": 2.2577,
      "step": 2269
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.04382902756333351,
      "learning_rate": 0.00013295258656237702,
      "loss": 2.2576,
      "step": 2270
    },
    {
      "epoch": 0.4542,
      "grad_norm": 0.07251009345054626,
      "learning_rate": 0.00013288666467385833,
      "loss": 2.2548,
      "step": 2271
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.06464702636003494,
      "learning_rate": 0.0001328207267567788,
      "loss": 2.2575,
      "step": 2272
    },
    {
      "epoch": 0.4546,
      "grad_norm": 0.04058757796883583,
      "learning_rate": 0.0001327547728432757,
      "loss": 2.2583,
      "step": 2273
    },
    {
      "epoch": 0.4548,
      "grad_norm": 0.029362570494413376,
      "learning_rate": 0.00013268880296549425,
      "loss": 2.2571,
      "step": 2274
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.02879798226058483,
      "learning_rate": 0.00013262281715558736,
      "loss": 2.256,
      "step": 2275
    },
    {
      "epoch": 0.4552,
      "grad_norm": 0.029017265886068344,
      "learning_rate": 0.00013255681544571568,
      "loss": 2.2579,
      "step": 2276
    },
    {
      "epoch": 0.4554,
      "grad_norm": 0.030440043658018112,
      "learning_rate": 0.00013249079786804765,
      "loss": 2.2593,
      "step": 2277
    },
    {
      "epoch": 0.4556,
      "grad_norm": 0.0378214567899704,
      "learning_rate": 0.00013242476445475944,
      "loss": 2.2598,
      "step": 2278
    },
    {
      "epoch": 0.4558,
      "grad_norm": 0.04536333680152893,
      "learning_rate": 0.00013235871523803496,
      "loss": 2.2568,
      "step": 2279
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.050095539540052414,
      "learning_rate": 0.00013229265025006576,
      "loss": 2.2587,
      "step": 2280
    },
    {
      "epoch": 0.4562,
      "grad_norm": 0.05214964225888252,
      "learning_rate": 0.00013222656952305113,
      "loss": 2.2571,
      "step": 2281
    },
    {
      "epoch": 0.4564,
      "grad_norm": 0.05856718868017197,
      "learning_rate": 0.00013216047308919798,
      "loss": 2.2576,
      "step": 2282
    },
    {
      "epoch": 0.4566,
      "grad_norm": 0.050109367817640305,
      "learning_rate": 0.00013209436098072095,
      "loss": 2.2534,
      "step": 2283
    },
    {
      "epoch": 0.4568,
      "grad_norm": 0.03582074120640755,
      "learning_rate": 0.00013202823322984228,
      "loss": 2.2582,
      "step": 2284
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.02934819646179676,
      "learning_rate": 0.0001319620898687918,
      "loss": 2.2567,
      "step": 2285
    },
    {
      "epoch": 0.4572,
      "grad_norm": 0.028751472011208534,
      "learning_rate": 0.00013189593092980702,
      "loss": 2.2562,
      "step": 2286
    },
    {
      "epoch": 0.4574,
      "grad_norm": 0.028763195499777794,
      "learning_rate": 0.00013182975644513296,
      "loss": 2.2582,
      "step": 2287
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.03034983016550541,
      "learning_rate": 0.00013176356644702225,
      "loss": 2.2578,
      "step": 2288
    },
    {
      "epoch": 0.4578,
      "grad_norm": 0.02942788228392601,
      "learning_rate": 0.0001316973609677352,
      "loss": 2.2579,
      "step": 2289
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.028587449342012405,
      "learning_rate": 0.0001316311400395394,
      "loss": 2.2563,
      "step": 2290
    },
    {
      "epoch": 0.4582,
      "grad_norm": 0.0324871763586998,
      "learning_rate": 0.00013156490369471027,
      "loss": 2.2577,
      "step": 2291
    },
    {
      "epoch": 0.4584,
      "grad_norm": 0.04134868457913399,
      "learning_rate": 0.0001314986519655305,
      "loss": 2.2565,
      "step": 2292
    },
    {
      "epoch": 0.4586,
      "grad_norm": 0.04932742193341255,
      "learning_rate": 0.0001314323848842904,
      "loss": 2.2567,
      "step": 2293
    },
    {
      "epoch": 0.4588,
      "grad_norm": 0.05531352758407593,
      "learning_rate": 0.00013136610248328778,
      "loss": 2.258,
      "step": 2294
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.06114663928747177,
      "learning_rate": 0.00013129980479482782,
      "loss": 2.2578,
      "step": 2295
    },
    {
      "epoch": 0.4592,
      "grad_norm": 0.05461413040757179,
      "learning_rate": 0.00013123349185122327,
      "loss": 2.2579,
      "step": 2296
    },
    {
      "epoch": 0.4594,
      "grad_norm": 0.0449589341878891,
      "learning_rate": 0.0001311671636847942,
      "loss": 2.2586,
      "step": 2297
    },
    {
      "epoch": 0.4596,
      "grad_norm": 0.03908615559339523,
      "learning_rate": 0.0001311008203278682,
      "loss": 2.2564,
      "step": 2298
    },
    {
      "epoch": 0.4598,
      "grad_norm": 0.04016083478927612,
      "learning_rate": 0.00013103446181278013,
      "loss": 2.2553,
      "step": 2299
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.03403221070766449,
      "learning_rate": 0.00013096808817187242,
      "loss": 2.2588,
      "step": 2300
    },
    {
      "epoch": 0.4602,
      "grad_norm": 0.02924402989447117,
      "learning_rate": 0.00013090169943749476,
      "loss": 2.256,
      "step": 2301
    },
    {
      "epoch": 0.4604,
      "grad_norm": 0.042665034532547,
      "learning_rate": 0.00013083529564200417,
      "loss": 2.2562,
      "step": 2302
    },
    {
      "epoch": 0.4606,
      "grad_norm": 0.07222653180360794,
      "learning_rate": 0.0001307688768177651,
      "loss": 2.2579,
      "step": 2303
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.11704790592193604,
      "learning_rate": 0.0001307024429971492,
      "loss": 2.2564,
      "step": 2304
    },
    {
      "epoch": 0.461,
      "grad_norm": 0.14939701557159424,
      "learning_rate": 0.00013063599421253558,
      "loss": 2.2568,
      "step": 2305
    },
    {
      "epoch": 0.4612,
      "grad_norm": 0.14489616453647614,
      "learning_rate": 0.00013056953049631057,
      "loss": 2.2554,
      "step": 2306
    },
    {
      "epoch": 0.4614,
      "grad_norm": 0.10156918317079544,
      "learning_rate": 0.0001305030518808678,
      "loss": 2.2566,
      "step": 2307
    },
    {
      "epoch": 0.4616,
      "grad_norm": 0.04912867024540901,
      "learning_rate": 0.00013043655839860803,
      "loss": 2.2555,
      "step": 2308
    },
    {
      "epoch": 0.4618,
      "grad_norm": 0.03305302932858467,
      "learning_rate": 0.00013037005008193942,
      "loss": 2.2582,
      "step": 2309
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.06433362513780594,
      "learning_rate": 0.00013030352696327742,
      "loss": 2.2575,
      "step": 2310
    },
    {
      "epoch": 0.4622,
      "grad_norm": 0.0939527079463005,
      "learning_rate": 0.00013023698907504446,
      "loss": 2.2578,
      "step": 2311
    },
    {
      "epoch": 0.4624,
      "grad_norm": 0.0943998247385025,
      "learning_rate": 0.00013017043644967036,
      "loss": 2.2577,
      "step": 2312
    },
    {
      "epoch": 0.4626,
      "grad_norm": 0.06433698534965515,
      "learning_rate": 0.00013010386911959206,
      "loss": 2.2572,
      "step": 2313
    },
    {
      "epoch": 0.4628,
      "grad_norm": 0.02937268279492855,
      "learning_rate": 0.0001300372871172536,
      "loss": 2.254,
      "step": 2314
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.05378207191824913,
      "learning_rate": 0.0001299706904751064,
      "loss": 2.256,
      "step": 2315
    },
    {
      "epoch": 0.4632,
      "grad_norm": 0.08496621996164322,
      "learning_rate": 0.00012990407922560868,
      "loss": 2.2574,
      "step": 2316
    },
    {
      "epoch": 0.4634,
      "grad_norm": 0.1009117141366005,
      "learning_rate": 0.00012983745340122604,
      "loss": 2.2565,
      "step": 2317
    },
    {
      "epoch": 0.4636,
      "grad_norm": 0.10431309044361115,
      "learning_rate": 0.00012977081303443107,
      "loss": 2.2564,
      "step": 2318
    },
    {
      "epoch": 0.4638,
      "grad_norm": 0.08691206574440002,
      "learning_rate": 0.0001297041581577035,
      "loss": 2.2566,
      "step": 2319
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.05689410865306854,
      "learning_rate": 0.00012963748880353011,
      "loss": 2.257,
      "step": 2320
    },
    {
      "epoch": 0.4642,
      "grad_norm": 0.030215349048376083,
      "learning_rate": 0.00012957080500440468,
      "loss": 2.255,
      "step": 2321
    },
    {
      "epoch": 0.4644,
      "grad_norm": 0.04322896525263786,
      "learning_rate": 0.00012950410679282815,
      "loss": 2.2551,
      "step": 2322
    },
    {
      "epoch": 0.4646,
      "grad_norm": 0.07653052359819412,
      "learning_rate": 0.00012943739420130836,
      "loss": 2.2569,
      "step": 2323
    },
    {
      "epoch": 0.4648,
      "grad_norm": 0.11027628928422928,
      "learning_rate": 0.0001293706672623603,
      "loss": 2.2555,
      "step": 2324
    },
    {
      "epoch": 0.465,
      "grad_norm": 0.14229801297187805,
      "learning_rate": 0.00012930392600850573,
      "loss": 2.2565,
      "step": 2325
    },
    {
      "epoch": 0.4652,
      "grad_norm": 0.15588651597499847,
      "learning_rate": 0.00012923717047227368,
      "loss": 2.2586,
      "step": 2326
    },
    {
      "epoch": 0.4654,
      "grad_norm": 0.14854353666305542,
      "learning_rate": 0.0001291704006861999,
      "loss": 2.2582,
      "step": 2327
    },
    {
      "epoch": 0.4656,
      "grad_norm": 0.12022409588098526,
      "learning_rate": 0.00012910361668282719,
      "loss": 2.2582,
      "step": 2328
    },
    {
      "epoch": 0.4658,
      "grad_norm": 0.0743531808257103,
      "learning_rate": 0.00012903681849470527,
      "loss": 2.2567,
      "step": 2329
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.029855944216251373,
      "learning_rate": 0.00012897000615439075,
      "loss": 2.259,
      "step": 2330
    },
    {
      "epoch": 0.4662,
      "grad_norm": 0.06167668476700783,
      "learning_rate": 0.00012890317969444716,
      "loss": 2.2552,
      "step": 2331
    },
    {
      "epoch": 0.4664,
      "grad_norm": 0.09450225532054901,
      "learning_rate": 0.00012883633914744492,
      "loss": 2.2569,
      "step": 2332
    },
    {
      "epoch": 0.4666,
      "grad_norm": 0.10574854165315628,
      "learning_rate": 0.00012876948454596128,
      "loss": 2.2559,
      "step": 2333
    },
    {
      "epoch": 0.4668,
      "grad_norm": 0.10018007457256317,
      "learning_rate": 0.00012870261592258037,
      "loss": 2.2591,
      "step": 2334
    },
    {
      "epoch": 0.467,
      "grad_norm": 0.0771288052201271,
      "learning_rate": 0.00012863573330989313,
      "loss": 2.2575,
      "step": 2335
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.05116257071495056,
      "learning_rate": 0.00012856883674049736,
      "loss": 2.2571,
      "step": 2336
    },
    {
      "epoch": 0.4674,
      "grad_norm": 0.031250353902578354,
      "learning_rate": 0.0001285019262469976,
      "loss": 2.2591,
      "step": 2337
    },
    {
      "epoch": 0.4676,
      "grad_norm": 0.03448353707790375,
      "learning_rate": 0.00012843500186200527,
      "loss": 2.2591,
      "step": 2338
    },
    {
      "epoch": 0.4678,
      "grad_norm": 0.05166180431842804,
      "learning_rate": 0.00012836806361813844,
      "loss": 2.2636,
      "step": 2339
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.05977435037493706,
      "learning_rate": 0.00012830111154802203,
      "loss": 2.2625,
      "step": 2340
    },
    {
      "epoch": 0.4682,
      "grad_norm": 0.0596037432551384,
      "learning_rate": 0.00012823414568428768,
      "loss": 2.2568,
      "step": 2341
    },
    {
      "epoch": 0.4684,
      "grad_norm": 0.040994878858327866,
      "learning_rate": 0.00012816716605957367,
      "loss": 2.256,
      "step": 2342
    },
    {
      "epoch": 0.4686,
      "grad_norm": 0.03166767209768295,
      "learning_rate": 0.00012810017270652513,
      "loss": 2.2578,
      "step": 2343
    },
    {
      "epoch": 0.4688,
      "grad_norm": 0.04643704369664192,
      "learning_rate": 0.00012803316565779377,
      "loss": 2.2572,
      "step": 2344
    },
    {
      "epoch": 0.469,
      "grad_norm": 0.06052107363939285,
      "learning_rate": 0.000127966144946038,
      "loss": 2.256,
      "step": 2345
    },
    {
      "epoch": 0.4692,
      "grad_norm": 0.06363191455602646,
      "learning_rate": 0.00012789911060392294,
      "loss": 2.255,
      "step": 2346
    },
    {
      "epoch": 0.4694,
      "grad_norm": 0.04677058383822441,
      "learning_rate": 0.0001278320626641203,
      "loss": 2.2549,
      "step": 2347
    },
    {
      "epoch": 0.4696,
      "grad_norm": 0.030824856832623482,
      "learning_rate": 0.00012776500115930842,
      "loss": 2.2564,
      "step": 2348
    },
    {
      "epoch": 0.4698,
      "grad_norm": 0.03991983085870743,
      "learning_rate": 0.00012769792612217224,
      "loss": 2.2577,
      "step": 2349
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.05901625379920006,
      "learning_rate": 0.00012763083758540337,
      "loss": 2.2561,
      "step": 2350
    },
    {
      "epoch": 0.4702,
      "grad_norm": 0.06858688592910767,
      "learning_rate": 0.0001275637355816999,
      "loss": 2.2585,
      "step": 2351
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.06172223016619682,
      "learning_rate": 0.00012749662014376663,
      "loss": 2.2549,
      "step": 2352
    },
    {
      "epoch": 0.4706,
      "grad_norm": 0.044460125267505646,
      "learning_rate": 0.00012742949130431467,
      "loss": 2.259,
      "step": 2353
    },
    {
      "epoch": 0.4708,
      "grad_norm": 0.03146405518054962,
      "learning_rate": 0.00012736234909606185,
      "loss": 2.2567,
      "step": 2354
    },
    {
      "epoch": 0.471,
      "grad_norm": 0.048545219004154205,
      "learning_rate": 0.00012729519355173254,
      "loss": 2.2574,
      "step": 2355
    },
    {
      "epoch": 0.4712,
      "grad_norm": 0.06364177912473679,
      "learning_rate": 0.00012722802470405744,
      "loss": 2.2574,
      "step": 2356
    },
    {
      "epoch": 0.4714,
      "grad_norm": 0.061011217534542084,
      "learning_rate": 0.00012716084258577388,
      "loss": 2.2589,
      "step": 2357
    },
    {
      "epoch": 0.4716,
      "grad_norm": 0.0453435480594635,
      "learning_rate": 0.0001270936472296256,
      "loss": 2.2553,
      "step": 2358
    },
    {
      "epoch": 0.4718,
      "grad_norm": 0.030832981690764427,
      "learning_rate": 0.00012702643866836278,
      "loss": 2.2577,
      "step": 2359
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.05026721581816673,
      "learning_rate": 0.00012695921693474212,
      "loss": 2.2573,
      "step": 2360
    },
    {
      "epoch": 0.4722,
      "grad_norm": 0.07925917953252792,
      "learning_rate": 0.00012689198206152657,
      "loss": 2.2558,
      "step": 2361
    },
    {
      "epoch": 0.4724,
      "grad_norm": 0.08880314975976944,
      "learning_rate": 0.0001268247340814857,
      "loss": 2.2555,
      "step": 2362
    },
    {
      "epoch": 0.4726,
      "grad_norm": 0.07040448486804962,
      "learning_rate": 0.00012675747302739527,
      "loss": 2.2559,
      "step": 2363
    },
    {
      "epoch": 0.4728,
      "grad_norm": 0.038574542850255966,
      "learning_rate": 0.00012669019893203759,
      "loss": 2.2571,
      "step": 2364
    },
    {
      "epoch": 0.473,
      "grad_norm": 0.03489730507135391,
      "learning_rate": 0.00012662291182820114,
      "loss": 2.2558,
      "step": 2365
    },
    {
      "epoch": 0.4732,
      "grad_norm": 0.058819908648729324,
      "learning_rate": 0.00012655561174868088,
      "loss": 2.2576,
      "step": 2366
    },
    {
      "epoch": 0.4734,
      "grad_norm": 0.06931158900260925,
      "learning_rate": 0.00012648829872627807,
      "loss": 2.2585,
      "step": 2367
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.06825502961874008,
      "learning_rate": 0.00012642097279380027,
      "loss": 2.2535,
      "step": 2368
    },
    {
      "epoch": 0.4738,
      "grad_norm": 0.047150127589702606,
      "learning_rate": 0.00012635363398406127,
      "loss": 2.2582,
      "step": 2369
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.03340775892138481,
      "learning_rate": 0.0001262862823298812,
      "loss": 2.2594,
      "step": 2370
    },
    {
      "epoch": 0.4742,
      "grad_norm": 0.044783815741539,
      "learning_rate": 0.00012621891786408648,
      "loss": 2.259,
      "step": 2371
    },
    {
      "epoch": 0.4744,
      "grad_norm": 0.07153268903493881,
      "learning_rate": 0.00012615154061950968,
      "loss": 2.2594,
      "step": 2372
    },
    {
      "epoch": 0.4746,
      "grad_norm": 0.09278775751590729,
      "learning_rate": 0.00012608415062898972,
      "loss": 2.2576,
      "step": 2373
    },
    {
      "epoch": 0.4748,
      "grad_norm": 0.10179982334375381,
      "learning_rate": 0.00012601674792537157,
      "loss": 2.2575,
      "step": 2374
    },
    {
      "epoch": 0.475,
      "grad_norm": 0.11558914184570312,
      "learning_rate": 0.00012594933254150655,
      "loss": 2.2564,
      "step": 2375
    },
    {
      "epoch": 0.4752,
      "grad_norm": 0.11516991257667542,
      "learning_rate": 0.00012588190451025207,
      "loss": 2.2567,
      "step": 2376
    },
    {
      "epoch": 0.4754,
      "grad_norm": 0.09298810362815857,
      "learning_rate": 0.0001258144638644718,
      "loss": 2.2572,
      "step": 2377
    },
    {
      "epoch": 0.4756,
      "grad_norm": 0.050350371748209,
      "learning_rate": 0.0001257470106370354,
      "loss": 2.2589,
      "step": 2378
    },
    {
      "epoch": 0.4758,
      "grad_norm": 0.03275075927376747,
      "learning_rate": 0.00012567954486081878,
      "loss": 2.2576,
      "step": 2379
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.05857822299003601,
      "learning_rate": 0.00012561206656870396,
      "loss": 2.2561,
      "step": 2380
    },
    {
      "epoch": 0.4762,
      "grad_norm": 0.07814132422208786,
      "learning_rate": 0.00012554457579357905,
      "loss": 2.258,
      "step": 2381
    },
    {
      "epoch": 0.4764,
      "grad_norm": 0.07371693104505539,
      "learning_rate": 0.00012547707256833823,
      "loss": 2.2554,
      "step": 2382
    },
    {
      "epoch": 0.4766,
      "grad_norm": 0.0501672588288784,
      "learning_rate": 0.00012540955692588173,
      "loss": 2.2585,
      "step": 2383
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.030794374644756317,
      "learning_rate": 0.00012534202889911584,
      "loss": 2.2552,
      "step": 2384
    },
    {
      "epoch": 0.477,
      "grad_norm": 0.04430490732192993,
      "learning_rate": 0.00012527448852095295,
      "loss": 2.2566,
      "step": 2385
    },
    {
      "epoch": 0.4772,
      "grad_norm": 0.056486841291189194,
      "learning_rate": 0.0001252069358243114,
      "loss": 2.2538,
      "step": 2386
    },
    {
      "epoch": 0.4774,
      "grad_norm": 0.058609094470739365,
      "learning_rate": 0.0001251393708421155,
      "loss": 2.2517,
      "step": 2387
    },
    {
      "epoch": 0.4776,
      "grad_norm": 0.056970156729221344,
      "learning_rate": 0.0001250717936072957,
      "loss": 2.2535,
      "step": 2388
    },
    {
      "epoch": 0.4778,
      "grad_norm": 0.0474577397108078,
      "learning_rate": 0.00012500420415278822,
      "loss": 2.2551,
      "step": 2389
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.0314805693924427,
      "learning_rate": 0.0001249366025115354,
      "loss": 2.2557,
      "step": 2390
    },
    {
      "epoch": 0.4782,
      "grad_norm": 0.05976661294698715,
      "learning_rate": 0.0001248689887164855,
      "loss": 2.2536,
      "step": 2391
    },
    {
      "epoch": 0.4784,
      "grad_norm": 0.10040217638015747,
      "learning_rate": 0.00012480136280059256,
      "loss": 2.2566,
      "step": 2392
    },
    {
      "epoch": 0.4786,
      "grad_norm": 0.10876300930976868,
      "learning_rate": 0.00012473372479681672,
      "loss": 2.2551,
      "step": 2393
    },
    {
      "epoch": 0.4788,
      "grad_norm": 0.09260974824428558,
      "learning_rate": 0.00012466607473812387,
      "loss": 2.2561,
      "step": 2394
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.05293447524309158,
      "learning_rate": 0.0001245984126574858,
      "loss": 2.2563,
      "step": 2395
    },
    {
      "epoch": 0.4792,
      "grad_norm": 0.036656249314546585,
      "learning_rate": 0.00012453073858788026,
      "loss": 2.2547,
      "step": 2396
    },
    {
      "epoch": 0.4794,
      "grad_norm": 0.0838983803987503,
      "learning_rate": 0.00012446305256229073,
      "loss": 2.2567,
      "step": 2397
    },
    {
      "epoch": 0.4796,
      "grad_norm": 0.12174234539270401,
      "learning_rate": 0.00012439535461370658,
      "loss": 2.2542,
      "step": 2398
    },
    {
      "epoch": 0.4798,
      "grad_norm": 0.13937489688396454,
      "learning_rate": 0.00012432764477512292,
      "loss": 2.2565,
      "step": 2399
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.11911525577306747,
      "learning_rate": 0.00012425992307954075,
      "loss": 2.2559,
      "step": 2400
    },
    {
      "epoch": 0.4802,
      "grad_norm": 0.07620502263307571,
      "learning_rate": 0.00012419218955996676,
      "loss": 2.2614,
      "step": 2401
    },
    {
      "epoch": 0.4804,
      "grad_norm": 0.03703080862760544,
      "learning_rate": 0.0001241244442494135,
      "loss": 2.2596,
      "step": 2402
    },
    {
      "epoch": 0.4806,
      "grad_norm": 0.04273274913430214,
      "learning_rate": 0.00012405668718089917,
      "loss": 2.2554,
      "step": 2403
    },
    {
      "epoch": 0.4808,
      "grad_norm": 0.08317320793867111,
      "learning_rate": 0.00012398891838744778,
      "loss": 2.2549,
      "step": 2404
    },
    {
      "epoch": 0.481,
      "grad_norm": 0.10633306950330734,
      "learning_rate": 0.00012392113790208895,
      "loss": 2.2556,
      "step": 2405
    },
    {
      "epoch": 0.4812,
      "grad_norm": 0.10683044791221619,
      "learning_rate": 0.0001238533457578581,
      "loss": 2.257,
      "step": 2406
    },
    {
      "epoch": 0.4814,
      "grad_norm": 0.0864560455083847,
      "learning_rate": 0.0001237855419877963,
      "loss": 2.2576,
      "step": 2407
    },
    {
      "epoch": 0.4816,
      "grad_norm": 0.054547443985939026,
      "learning_rate": 0.0001237177266249503,
      "loss": 2.2565,
      "step": 2408
    },
    {
      "epoch": 0.4818,
      "grad_norm": 0.03173983097076416,
      "learning_rate": 0.00012364989970237248,
      "loss": 2.2561,
      "step": 2409
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.048863571137189865,
      "learning_rate": 0.00012358206125312083,
      "loss": 2.2571,
      "step": 2410
    },
    {
      "epoch": 0.4822,
      "grad_norm": 0.08796385675668716,
      "learning_rate": 0.000123514211310259,
      "loss": 2.2559,
      "step": 2411
    },
    {
      "epoch": 0.4824,
      "grad_norm": 0.1066073626279831,
      "learning_rate": 0.00012344634990685624,
      "loss": 2.2589,
      "step": 2412
    },
    {
      "epoch": 0.4826,
      "grad_norm": 0.0993717759847641,
      "learning_rate": 0.00012337847707598738,
      "loss": 2.2592,
      "step": 2413
    },
    {
      "epoch": 0.4828,
      "grad_norm": 0.08186742663383484,
      "learning_rate": 0.0001233105928507328,
      "loss": 2.2568,
      "step": 2414
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.0632479265332222,
      "learning_rate": 0.0001232426972641784,
      "loss": 2.2566,
      "step": 2415
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.034520428627729416,
      "learning_rate": 0.00012317479034941573,
      "loss": 2.2561,
      "step": 2416
    },
    {
      "epoch": 0.4834,
      "grad_norm": 0.04469946399331093,
      "learning_rate": 0.00012310687213954182,
      "loss": 2.2535,
      "step": 2417
    },
    {
      "epoch": 0.4836,
      "grad_norm": 0.08293022960424423,
      "learning_rate": 0.00012303894266765908,
      "loss": 2.2566,
      "step": 2418
    },
    {
      "epoch": 0.4838,
      "grad_norm": 0.11751844733953476,
      "learning_rate": 0.00012297100196687557,
      "loss": 2.2565,
      "step": 2419
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.14407719671726227,
      "learning_rate": 0.00012290305007030478,
      "loss": 2.2564,
      "step": 2420
    },
    {
      "epoch": 0.4842,
      "grad_norm": 0.13432668149471283,
      "learning_rate": 0.00012283508701106557,
      "loss": 2.2582,
      "step": 2421
    },
    {
      "epoch": 0.4844,
      "grad_norm": 0.09094437211751938,
      "learning_rate": 0.0001227671128222824,
      "loss": 2.2543,
      "step": 2422
    },
    {
      "epoch": 0.4846,
      "grad_norm": 0.04020597040653229,
      "learning_rate": 0.00012269912753708502,
      "loss": 2.254,
      "step": 2423
    },
    {
      "epoch": 0.4848,
      "grad_norm": 0.048047278076410294,
      "learning_rate": 0.0001226311311886086,
      "loss": 2.2575,
      "step": 2424
    },
    {
      "epoch": 0.485,
      "grad_norm": 0.09521687030792236,
      "learning_rate": 0.00012256312380999376,
      "loss": 2.2569,
      "step": 2425
    },
    {
      "epoch": 0.4852,
      "grad_norm": 0.11625941842794418,
      "learning_rate": 0.0001224951054343865,
      "loss": 2.2577,
      "step": 2426
    },
    {
      "epoch": 0.4854,
      "grad_norm": 0.09600793570280075,
      "learning_rate": 0.00012242707609493814,
      "loss": 2.2576,
      "step": 2427
    },
    {
      "epoch": 0.4856,
      "grad_norm": 0.05730518326163292,
      "learning_rate": 0.0001223590358248053,
      "loss": 2.2559,
      "step": 2428
    },
    {
      "epoch": 0.4858,
      "grad_norm": 0.0323934331536293,
      "learning_rate": 0.00012229098465715006,
      "loss": 2.2553,
      "step": 2429
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.05184563994407654,
      "learning_rate": 0.00012222292262513965,
      "loss": 2.256,
      "step": 2430
    },
    {
      "epoch": 0.4862,
      "grad_norm": 0.08371777087450027,
      "learning_rate": 0.00012215484976194676,
      "loss": 2.257,
      "step": 2431
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.09266070276498795,
      "learning_rate": 0.00012208676610074917,
      "loss": 2.2568,
      "step": 2432
    },
    {
      "epoch": 0.4866,
      "grad_norm": 0.07080664485692978,
      "learning_rate": 0.00012201867167473015,
      "loss": 2.255,
      "step": 2433
    },
    {
      "epoch": 0.4868,
      "grad_norm": 0.038663677871227264,
      "learning_rate": 0.00012195056651707806,
      "loss": 2.2548,
      "step": 2434
    },
    {
      "epoch": 0.487,
      "grad_norm": 0.03727756440639496,
      "learning_rate": 0.00012188245066098647,
      "loss": 2.255,
      "step": 2435
    },
    {
      "epoch": 0.4872,
      "grad_norm": 0.0466880165040493,
      "learning_rate": 0.00012181432413965428,
      "loss": 2.2558,
      "step": 2436
    },
    {
      "epoch": 0.4874,
      "grad_norm": 0.0407504066824913,
      "learning_rate": 0.00012174618698628549,
      "loss": 2.2557,
      "step": 2437
    },
    {
      "epoch": 0.4876,
      "grad_norm": 0.033740293234586716,
      "learning_rate": 0.00012167803923408934,
      "loss": 2.2532,
      "step": 2438
    },
    {
      "epoch": 0.4878,
      "grad_norm": 0.03173525258898735,
      "learning_rate": 0.00012160988091628022,
      "loss": 2.2539,
      "step": 2439
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.03723864257335663,
      "learning_rate": 0.00012154171206607764,
      "loss": 2.2552,
      "step": 2440
    },
    {
      "epoch": 0.4882,
      "grad_norm": 0.04244433715939522,
      "learning_rate": 0.00012147353271670634,
      "loss": 2.2534,
      "step": 2441
    },
    {
      "epoch": 0.4884,
      "grad_norm": 0.04115131497383118,
      "learning_rate": 0.000121405342901396,
      "loss": 2.2566,
      "step": 2442
    },
    {
      "epoch": 0.4886,
      "grad_norm": 0.03428395837545395,
      "learning_rate": 0.00012133714265338161,
      "loss": 2.2562,
      "step": 2443
    },
    {
      "epoch": 0.4888,
      "grad_norm": 0.03445226326584816,
      "learning_rate": 0.00012126893200590308,
      "loss": 2.2558,
      "step": 2444
    },
    {
      "epoch": 0.489,
      "grad_norm": 0.04038947448134422,
      "learning_rate": 0.00012120071099220549,
      "loss": 2.2526,
      "step": 2445
    },
    {
      "epoch": 0.4892,
      "grad_norm": 0.05185692384839058,
      "learning_rate": 0.00012113247964553888,
      "loss": 2.2559,
      "step": 2446
    },
    {
      "epoch": 0.4894,
      "grad_norm": 0.05518054589629173,
      "learning_rate": 0.0001210642379991584,
      "loss": 2.2529,
      "step": 2447
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.049411047250032425,
      "learning_rate": 0.00012099598608632428,
      "loss": 2.2542,
      "step": 2448
    },
    {
      "epoch": 0.4898,
      "grad_norm": 0.03782451152801514,
      "learning_rate": 0.00012092772394030152,
      "loss": 2.2532,
      "step": 2449
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.03234566003084183,
      "learning_rate": 0.00012085945159436038,
      "loss": 2.2515,
      "step": 2450
    },
    {
      "epoch": 0.4902,
      "grad_norm": 0.03361620381474495,
      "learning_rate": 0.00012079116908177593,
      "loss": 2.2563,
      "step": 2451
    },
    {
      "epoch": 0.4904,
      "grad_norm": 0.03528878092765808,
      "learning_rate": 0.00012072287643582825,
      "loss": 2.2554,
      "step": 2452
    },
    {
      "epoch": 0.4906,
      "grad_norm": 0.0391896516084671,
      "learning_rate": 0.00012065457368980236,
      "loss": 2.2548,
      "step": 2453
    },
    {
      "epoch": 0.4908,
      "grad_norm": 0.040791548788547516,
      "learning_rate": 0.00012058626087698814,
      "loss": 2.2531,
      "step": 2454
    },
    {
      "epoch": 0.491,
      "grad_norm": 0.04204554110765457,
      "learning_rate": 0.00012051793803068046,
      "loss": 2.2534,
      "step": 2455
    },
    {
      "epoch": 0.4912,
      "grad_norm": 0.04115206003189087,
      "learning_rate": 0.00012044960518417903,
      "loss": 2.2529,
      "step": 2456
    },
    {
      "epoch": 0.4914,
      "grad_norm": 0.047343917191028595,
      "learning_rate": 0.0001203812623707885,
      "loss": 2.2567,
      "step": 2457
    },
    {
      "epoch": 0.4916,
      "grad_norm": 0.05638039484620094,
      "learning_rate": 0.00012031290962381823,
      "loss": 2.2564,
      "step": 2458
    },
    {
      "epoch": 0.4918,
      "grad_norm": 0.061982885003089905,
      "learning_rate": 0.00012024454697658261,
      "loss": 2.256,
      "step": 2459
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.05606130138039589,
      "learning_rate": 0.0001201761744624007,
      "loss": 2.2535,
      "step": 2460
    },
    {
      "epoch": 0.4922,
      "grad_norm": 0.04845110699534416,
      "learning_rate": 0.00012010779211459648,
      "loss": 2.2535,
      "step": 2461
    },
    {
      "epoch": 0.4924,
      "grad_norm": 0.03500352427363396,
      "learning_rate": 0.00012003939996649865,
      "loss": 2.2535,
      "step": 2462
    },
    {
      "epoch": 0.4926,
      "grad_norm": 0.03465165197849274,
      "learning_rate": 0.00011997099805144069,
      "loss": 2.2543,
      "step": 2463
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.04334187135100365,
      "learning_rate": 0.00011990258640276094,
      "loss": 2.2558,
      "step": 2464
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.040541186928749084,
      "learning_rate": 0.00011983416505380234,
      "loss": 2.2543,
      "step": 2465
    },
    {
      "epoch": 0.4932,
      "grad_norm": 0.03333282470703125,
      "learning_rate": 0.00011976573403791262,
      "loss": 2.2533,
      "step": 2466
    },
    {
      "epoch": 0.4934,
      "grad_norm": 0.03551569953560829,
      "learning_rate": 0.00011969729338844429,
      "loss": 2.2552,
      "step": 2467
    },
    {
      "epoch": 0.4936,
      "grad_norm": 0.04009980335831642,
      "learning_rate": 0.0001196288431387544,
      "loss": 2.2563,
      "step": 2468
    },
    {
      "epoch": 0.4938,
      "grad_norm": 0.04116707667708397,
      "learning_rate": 0.00011956038332220483,
      "loss": 2.2549,
      "step": 2469
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.03924228996038437,
      "learning_rate": 0.00011949191397216206,
      "loss": 2.2565,
      "step": 2470
    },
    {
      "epoch": 0.4942,
      "grad_norm": 0.03560960665345192,
      "learning_rate": 0.0001194234351219972,
      "loss": 2.2548,
      "step": 2471
    },
    {
      "epoch": 0.4944,
      "grad_norm": 0.03337440267205238,
      "learning_rate": 0.00011935494680508606,
      "loss": 2.2552,
      "step": 2472
    },
    {
      "epoch": 0.4946,
      "grad_norm": 0.03434688597917557,
      "learning_rate": 0.0001192864490548089,
      "loss": 2.2556,
      "step": 2473
    },
    {
      "epoch": 0.4948,
      "grad_norm": 0.037064116448163986,
      "learning_rate": 0.00011921794190455082,
      "loss": 2.254,
      "step": 2474
    },
    {
      "epoch": 0.495,
      "grad_norm": 0.04613760858774185,
      "learning_rate": 0.00011914942538770131,
      "loss": 2.2541,
      "step": 2475
    },
    {
      "epoch": 0.4952,
      "grad_norm": 0.05548027157783508,
      "learning_rate": 0.00011908089953765449,
      "loss": 2.2562,
      "step": 2476
    },
    {
      "epoch": 0.4954,
      "grad_norm": 0.06019706651568413,
      "learning_rate": 0.00011901236438780902,
      "loss": 2.2551,
      "step": 2477
    },
    {
      "epoch": 0.4956,
      "grad_norm": 0.0506066232919693,
      "learning_rate": 0.00011894381997156813,
      "loss": 2.2544,
      "step": 2478
    },
    {
      "epoch": 0.4958,
      "grad_norm": 0.03391333296895027,
      "learning_rate": 0.00011887526632233954,
      "loss": 2.2541,
      "step": 2479
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.04636358842253685,
      "learning_rate": 0.00011880670347353539,
      "loss": 2.2543,
      "step": 2480
    },
    {
      "epoch": 0.4962,
      "grad_norm": 0.0713980570435524,
      "learning_rate": 0.00011873813145857249,
      "loss": 2.2575,
      "step": 2481
    },
    {
      "epoch": 0.4964,
      "grad_norm": 0.1020713597536087,
      "learning_rate": 0.0001186695503108719,
      "loss": 2.2591,
      "step": 2482
    },
    {
      "epoch": 0.4966,
      "grad_norm": 0.13732457160949707,
      "learning_rate": 0.0001186009600638593,
      "loss": 2.2588,
      "step": 2483
    },
    {
      "epoch": 0.4968,
      "grad_norm": 0.16509263217449188,
      "learning_rate": 0.00011853236075096474,
      "loss": 2.2588,
      "step": 2484
    },
    {
      "epoch": 0.497,
      "grad_norm": 0.1714164912700653,
      "learning_rate": 0.0001184637524056227,
      "loss": 2.2581,
      "step": 2485
    },
    {
      "epoch": 0.4972,
      "grad_norm": 0.15014192461967468,
      "learning_rate": 0.00011839513506127203,
      "loss": 2.2583,
      "step": 2486
    },
    {
      "epoch": 0.4974,
      "grad_norm": 0.10826300084590912,
      "learning_rate": 0.00011832650875135598,
      "loss": 2.2533,
      "step": 2487
    },
    {
      "epoch": 0.4976,
      "grad_norm": 0.06272295117378235,
      "learning_rate": 0.00011825787350932222,
      "loss": 2.2542,
      "step": 2488
    },
    {
      "epoch": 0.4978,
      "grad_norm": 0.033373765647411346,
      "learning_rate": 0.00011818922936862269,
      "loss": 2.2546,
      "step": 2489
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.0566544309258461,
      "learning_rate": 0.00011812057636271374,
      "loss": 2.2531,
      "step": 2490
    },
    {
      "epoch": 0.4982,
      "grad_norm": 0.08096616715192795,
      "learning_rate": 0.00011805191452505602,
      "loss": 2.2543,
      "step": 2491
    },
    {
      "epoch": 0.4984,
      "grad_norm": 0.09306299686431885,
      "learning_rate": 0.00011798324388911444,
      "loss": 2.2549,
      "step": 2492
    },
    {
      "epoch": 0.4986,
      "grad_norm": 0.09393053501844406,
      "learning_rate": 0.00011791456448835825,
      "loss": 2.2545,
      "step": 2493
    },
    {
      "epoch": 0.4988,
      "grad_norm": 0.08120952546596527,
      "learning_rate": 0.00011784587635626094,
      "loss": 2.257,
      "step": 2494
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.06491720676422119,
      "learning_rate": 0.00011777717952630031,
      "loss": 2.2554,
      "step": 2495
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.04112532362341881,
      "learning_rate": 0.00011770847403195834,
      "loss": 2.2535,
      "step": 2496
    },
    {
      "epoch": 0.4994,
      "grad_norm": 0.0337531603872776,
      "learning_rate": 0.00011763975990672125,
      "loss": 2.2534,
      "step": 2497
    },
    {
      "epoch": 0.4996,
      "grad_norm": 0.04051388427615166,
      "learning_rate": 0.00011757103718407947,
      "loss": 2.2549,
      "step": 2498
    },
    {
      "epoch": 0.4998,
      "grad_norm": 0.04428940266370773,
      "learning_rate": 0.00011750230589752762,
      "loss": 2.2544,
      "step": 2499
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.04371236637234688,
      "learning_rate": 0.00011743356608056449,
      "loss": 2.2555,
      "step": 2500
    },
    {
      "epoch": 0.5002,
      "grad_norm": 0.03589523956179619,
      "learning_rate": 0.00011736481776669306,
      "loss": 2.2532,
      "step": 2501
    },
    {
      "epoch": 0.5004,
      "grad_norm": 0.03362554684281349,
      "learning_rate": 0.00011729606098942039,
      "loss": 2.258,
      "step": 2502
    },
    {
      "epoch": 0.5006,
      "grad_norm": 0.03870442137122154,
      "learning_rate": 0.00011722729578225769,
      "loss": 2.2574,
      "step": 2503
    },
    {
      "epoch": 0.5008,
      "grad_norm": 0.04325396567583084,
      "learning_rate": 0.0001171585221787203,
      "loss": 2.2538,
      "step": 2504
    },
    {
      "epoch": 0.501,
      "grad_norm": 0.040066029876470566,
      "learning_rate": 0.00011708974021232769,
      "loss": 2.257,
      "step": 2505
    },
    {
      "epoch": 0.5012,
      "grad_norm": 0.033700451254844666,
      "learning_rate": 0.00011702094991660326,
      "loss": 2.2533,
      "step": 2506
    },
    {
      "epoch": 0.5014,
      "grad_norm": 0.051575999706983566,
      "learning_rate": 0.00011695215132507464,
      "loss": 2.2563,
      "step": 2507
    },
    {
      "epoch": 0.5016,
      "grad_norm": 0.08494166284799576,
      "learning_rate": 0.00011688334447127338,
      "loss": 2.2529,
      "step": 2508
    },
    {
      "epoch": 0.5018,
      "grad_norm": 0.1091495230793953,
      "learning_rate": 0.00011681452938873516,
      "loss": 2.2545,
      "step": 2509
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.1132182851433754,
      "learning_rate": 0.00011674570611099955,
      "loss": 2.2533,
      "step": 2510
    },
    {
      "epoch": 0.5022,
      "grad_norm": 0.09938251972198486,
      "learning_rate": 0.00011667687467161024,
      "loss": 2.2553,
      "step": 2511
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.07623118162155151,
      "learning_rate": 0.0001166080351041148,
      "loss": 2.2541,
      "step": 2512
    },
    {
      "epoch": 0.5026,
      "grad_norm": 0.042443469166755676,
      "learning_rate": 0.00011653918744206478,
      "loss": 2.2524,
      "step": 2513
    },
    {
      "epoch": 0.5028,
      "grad_norm": 0.036422424018383026,
      "learning_rate": 0.00011647033171901573,
      "loss": 2.2542,
      "step": 2514
    },
    {
      "epoch": 0.503,
      "grad_norm": 0.057043273001909256,
      "learning_rate": 0.00011640146796852711,
      "loss": 2.2558,
      "step": 2515
    },
    {
      "epoch": 0.5032,
      "grad_norm": 0.08126109838485718,
      "learning_rate": 0.00011633259622416224,
      "loss": 2.2518,
      "step": 2516
    },
    {
      "epoch": 0.5034,
      "grad_norm": 0.10582887381315231,
      "learning_rate": 0.00011626371651948838,
      "loss": 2.2503,
      "step": 2517
    },
    {
      "epoch": 0.5036,
      "grad_norm": 0.11656728386878967,
      "learning_rate": 0.00011619482888807662,
      "loss": 2.2521,
      "step": 2518
    },
    {
      "epoch": 0.5038,
      "grad_norm": 0.11039802432060242,
      "learning_rate": 0.00011612593336350208,
      "loss": 2.248,
      "step": 2519
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.0779513344168663,
      "learning_rate": 0.00011605702997934345,
      "loss": 2.2466,
      "step": 2520
    },
    {
      "epoch": 0.5042,
      "grad_norm": 0.03665342181921005,
      "learning_rate": 0.0001159881187691835,
      "loss": 2.2487,
      "step": 2521
    },
    {
      "epoch": 0.5044,
      "grad_norm": 0.06177924945950508,
      "learning_rate": 0.00011591919976660868,
      "loss": 2.2527,
      "step": 2522
    },
    {
      "epoch": 0.5046,
      "grad_norm": 0.10627396404743195,
      "learning_rate": 0.0001158502730052093,
      "loss": 2.2542,
      "step": 2523
    },
    {
      "epoch": 0.5048,
      "grad_norm": 0.12185534834861755,
      "learning_rate": 0.0001157813385185794,
      "loss": 2.253,
      "step": 2524
    },
    {
      "epoch": 0.505,
      "grad_norm": 0.10972460359334946,
      "learning_rate": 0.00011571239634031679,
      "loss": 2.2539,
      "step": 2525
    },
    {
      "epoch": 0.5052,
      "grad_norm": 0.08169571310281754,
      "learning_rate": 0.0001156434465040231,
      "loss": 2.2558,
      "step": 2526
    },
    {
      "epoch": 0.5054,
      "grad_norm": 0.04006023705005646,
      "learning_rate": 0.00011557448904330362,
      "loss": 2.254,
      "step": 2527
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.04471791163086891,
      "learning_rate": 0.00011550552399176739,
      "loss": 2.2558,
      "step": 2528
    },
    {
      "epoch": 0.5058,
      "grad_norm": 0.0662720799446106,
      "learning_rate": 0.00011543655138302714,
      "loss": 2.2552,
      "step": 2529
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.06277905404567719,
      "learning_rate": 0.00011536757125069923,
      "loss": 2.2555,
      "step": 2530
    },
    {
      "epoch": 0.5062,
      "grad_norm": 0.0528079979121685,
      "learning_rate": 0.00011529858362840382,
      "loss": 2.2549,
      "step": 2531
    },
    {
      "epoch": 0.5064,
      "grad_norm": 0.041077058762311935,
      "learning_rate": 0.00011522958854976458,
      "loss": 2.2528,
      "step": 2532
    },
    {
      "epoch": 0.5066,
      "grad_norm": 0.03458678722381592,
      "learning_rate": 0.00011516058604840891,
      "loss": 2.2546,
      "step": 2533
    },
    {
      "epoch": 0.5068,
      "grad_norm": 0.03408234938979149,
      "learning_rate": 0.00011509157615796776,
      "loss": 2.2537,
      "step": 2534
    },
    {
      "epoch": 0.507,
      "grad_norm": 0.0349009670317173,
      "learning_rate": 0.00011502255891207572,
      "loss": 2.2537,
      "step": 2535
    },
    {
      "epoch": 0.5072,
      "grad_norm": 0.04020395874977112,
      "learning_rate": 0.00011495353434437098,
      "loss": 2.2547,
      "step": 2536
    },
    {
      "epoch": 0.5074,
      "grad_norm": 0.06004386395215988,
      "learning_rate": 0.00011488450248849522,
      "loss": 2.2549,
      "step": 2537
    },
    {
      "epoch": 0.5076,
      "grad_norm": 0.08105111867189407,
      "learning_rate": 0.00011481546337809381,
      "loss": 2.2549,
      "step": 2538
    },
    {
      "epoch": 0.5078,
      "grad_norm": 0.08534441143274307,
      "learning_rate": 0.0001147464170468155,
      "loss": 2.2545,
      "step": 2539
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.06903868913650513,
      "learning_rate": 0.00011467736352831266,
      "loss": 2.2522,
      "step": 2540
    },
    {
      "epoch": 0.5082,
      "grad_norm": 0.046644411981105804,
      "learning_rate": 0.00011460830285624118,
      "loss": 2.2538,
      "step": 2541
    },
    {
      "epoch": 0.5084,
      "grad_norm": 0.03437919542193413,
      "learning_rate": 0.00011453923506426032,
      "loss": 2.2537,
      "step": 2542
    },
    {
      "epoch": 0.5086,
      "grad_norm": 0.05192074179649353,
      "learning_rate": 0.00011447016018603292,
      "loss": 2.2527,
      "step": 2543
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.08308396488428116,
      "learning_rate": 0.00011440107825522521,
      "loss": 2.2537,
      "step": 2544
    },
    {
      "epoch": 0.509,
      "grad_norm": 0.09461339563131332,
      "learning_rate": 0.00011433198930550695,
      "loss": 2.2547,
      "step": 2545
    },
    {
      "epoch": 0.5092,
      "grad_norm": 0.07829435169696808,
      "learning_rate": 0.00011426289337055119,
      "loss": 2.2569,
      "step": 2546
    },
    {
      "epoch": 0.5094,
      "grad_norm": 0.04615357145667076,
      "learning_rate": 0.00011419379048403444,
      "loss": 2.2528,
      "step": 2547
    },
    {
      "epoch": 0.5096,
      "grad_norm": 0.04059170186519623,
      "learning_rate": 0.00011412468067963669,
      "loss": 2.2547,
      "step": 2548
    },
    {
      "epoch": 0.5098,
      "grad_norm": 0.08330275863409042,
      "learning_rate": 0.00011405556399104109,
      "loss": 2.2534,
      "step": 2549
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.11580604314804077,
      "learning_rate": 0.00011398644045193444,
      "loss": 2.2519,
      "step": 2550
    },
    {
      "epoch": 0.5102,
      "grad_norm": 0.1271040141582489,
      "learning_rate": 0.00011391731009600654,
      "loss": 2.2533,
      "step": 2551
    },
    {
      "epoch": 0.5104,
      "grad_norm": 0.11650299280881882,
      "learning_rate": 0.00011384817295695083,
      "loss": 2.2536,
      "step": 2552
    },
    {
      "epoch": 0.5106,
      "grad_norm": 0.10163353383541107,
      "learning_rate": 0.0001137790290684638,
      "loss": 2.2539,
      "step": 2553
    },
    {
      "epoch": 0.5108,
      "grad_norm": 0.08002150058746338,
      "learning_rate": 0.00011370987846424546,
      "loss": 2.2571,
      "step": 2554
    },
    {
      "epoch": 0.511,
      "grad_norm": 0.05028284341096878,
      "learning_rate": 0.00011364072117799885,
      "loss": 2.2509,
      "step": 2555
    },
    {
      "epoch": 0.5112,
      "grad_norm": 0.03570951893925667,
      "learning_rate": 0.00011357155724343045,
      "loss": 2.2525,
      "step": 2556
    },
    {
      "epoch": 0.5114,
      "grad_norm": 0.05799949914216995,
      "learning_rate": 0.00011350238669424993,
      "loss": 2.2528,
      "step": 2557
    },
    {
      "epoch": 0.5116,
      "grad_norm": 0.07568149268627167,
      "learning_rate": 0.00011343320956417014,
      "loss": 2.2554,
      "step": 2558
    },
    {
      "epoch": 0.5118,
      "grad_norm": 0.0769038051366806,
      "learning_rate": 0.00011336402588690726,
      "loss": 2.2552,
      "step": 2559
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.07705814391374588,
      "learning_rate": 0.00011329483569618045,
      "loss": 2.2544,
      "step": 2560
    },
    {
      "epoch": 0.5122,
      "grad_norm": 0.06927946954965591,
      "learning_rate": 0.00011322563902571226,
      "loss": 2.2549,
      "step": 2561
    },
    {
      "epoch": 0.5124,
      "grad_norm": 0.052997589111328125,
      "learning_rate": 0.00011315643590922827,
      "loss": 2.2555,
      "step": 2562
    },
    {
      "epoch": 0.5126,
      "grad_norm": 0.03753871098160744,
      "learning_rate": 0.00011308722638045724,
      "loss": 2.2542,
      "step": 2563
    },
    {
      "epoch": 0.5128,
      "grad_norm": 0.03480939194560051,
      "learning_rate": 0.00011301801047313105,
      "loss": 2.2538,
      "step": 2564
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.03513451665639877,
      "learning_rate": 0.00011294878822098469,
      "loss": 2.2533,
      "step": 2565
    },
    {
      "epoch": 0.5132,
      "grad_norm": 0.03473735973238945,
      "learning_rate": 0.0001128795596577563,
      "loss": 2.2542,
      "step": 2566
    },
    {
      "epoch": 0.5134,
      "grad_norm": 0.03539751470088959,
      "learning_rate": 0.00011281032481718697,
      "loss": 2.2537,
      "step": 2567
    },
    {
      "epoch": 0.5136,
      "grad_norm": 0.03906206041574478,
      "learning_rate": 0.00011274108373302095,
      "loss": 2.2543,
      "step": 2568
    },
    {
      "epoch": 0.5138,
      "grad_norm": 0.04798715561628342,
      "learning_rate": 0.00011267183643900548,
      "loss": 2.2548,
      "step": 2569
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.06310636550188065,
      "learning_rate": 0.00011260258296889086,
      "loss": 2.2555,
      "step": 2570
    },
    {
      "epoch": 0.5142,
      "grad_norm": 0.09093821793794632,
      "learning_rate": 0.00011253332335643043,
      "loss": 2.2536,
      "step": 2571
    },
    {
      "epoch": 0.5144,
      "grad_norm": 0.11366058886051178,
      "learning_rate": 0.00011246405763538046,
      "loss": 2.2533,
      "step": 2572
    },
    {
      "epoch": 0.5146,
      "grad_norm": 0.1186617985367775,
      "learning_rate": 0.00011239478583950018,
      "loss": 2.2529,
      "step": 2573
    },
    {
      "epoch": 0.5148,
      "grad_norm": 0.08597161620855331,
      "learning_rate": 0.00011232550800255188,
      "loss": 2.2567,
      "step": 2574
    },
    {
      "epoch": 0.515,
      "grad_norm": 0.04364580661058426,
      "learning_rate": 0.00011225622415830068,
      "loss": 2.2568,
      "step": 2575
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.03993721306324005,
      "learning_rate": 0.00011218693434051475,
      "loss": 2.2527,
      "step": 2576
    },
    {
      "epoch": 0.5154,
      "grad_norm": 0.053403083235025406,
      "learning_rate": 0.00011211763858296507,
      "loss": 2.2529,
      "step": 2577
    },
    {
      "epoch": 0.5156,
      "grad_norm": 0.06655966490507126,
      "learning_rate": 0.00011204833691942553,
      "loss": 2.2541,
      "step": 2578
    },
    {
      "epoch": 0.5158,
      "grad_norm": 0.06789244711399078,
      "learning_rate": 0.00011197902938367298,
      "loss": 2.2536,
      "step": 2579
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.06631965935230255,
      "learning_rate": 0.00011190971600948699,
      "loss": 2.2528,
      "step": 2580
    },
    {
      "epoch": 0.5162,
      "grad_norm": 0.05350792407989502,
      "learning_rate": 0.00011184039683065013,
      "loss": 2.2539,
      "step": 2581
    },
    {
      "epoch": 0.5164,
      "grad_norm": 0.03826145827770233,
      "learning_rate": 0.00011177107188094764,
      "loss": 2.2564,
      "step": 2582
    },
    {
      "epoch": 0.5166,
      "grad_norm": 0.04122798889875412,
      "learning_rate": 0.00011170174119416776,
      "loss": 2.2553,
      "step": 2583
    },
    {
      "epoch": 0.5168,
      "grad_norm": 0.062357883900403976,
      "learning_rate": 0.00011163240480410135,
      "loss": 2.253,
      "step": 2584
    },
    {
      "epoch": 0.517,
      "grad_norm": 0.08110586553812027,
      "learning_rate": 0.00011156306274454218,
      "loss": 2.2518,
      "step": 2585
    },
    {
      "epoch": 0.5172,
      "grad_norm": 0.08441828191280365,
      "learning_rate": 0.00011149371504928668,
      "loss": 2.2525,
      "step": 2586
    },
    {
      "epoch": 0.5174,
      "grad_norm": 0.07235100865364075,
      "learning_rate": 0.00011142436175213409,
      "loss": 2.2555,
      "step": 2587
    },
    {
      "epoch": 0.5176,
      "grad_norm": 0.04458329826593399,
      "learning_rate": 0.00011135500288688636,
      "loss": 2.2543,
      "step": 2588
    },
    {
      "epoch": 0.5178,
      "grad_norm": 0.03955109044909477,
      "learning_rate": 0.00011128563848734816,
      "loss": 2.2531,
      "step": 2589
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.07615161687135696,
      "learning_rate": 0.0001112162685873269,
      "loss": 2.2516,
      "step": 2590
    },
    {
      "epoch": 0.5182,
      "grad_norm": 0.11719557642936707,
      "learning_rate": 0.00011114689322063255,
      "loss": 2.2563,
      "step": 2591
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.1461007446050644,
      "learning_rate": 0.00011107751242107787,
      "loss": 2.2531,
      "step": 2592
    },
    {
      "epoch": 0.5186,
      "grad_norm": 0.1469339281320572,
      "learning_rate": 0.00011100812622247822,
      "loss": 2.2507,
      "step": 2593
    },
    {
      "epoch": 0.5188,
      "grad_norm": 0.12515299022197723,
      "learning_rate": 0.00011093873465865157,
      "loss": 2.2545,
      "step": 2594
    },
    {
      "epoch": 0.519,
      "grad_norm": 0.07651268690824509,
      "learning_rate": 0.00011086933776341852,
      "loss": 2.2526,
      "step": 2595
    },
    {
      "epoch": 0.5192,
      "grad_norm": 0.03594387322664261,
      "learning_rate": 0.0001107999355706023,
      "loss": 2.2511,
      "step": 2596
    },
    {
      "epoch": 0.5194,
      "grad_norm": 0.05463467910885811,
      "learning_rate": 0.00011073052811402867,
      "loss": 2.2522,
      "step": 2597
    },
    {
      "epoch": 0.5196,
      "grad_norm": 0.073447085916996,
      "learning_rate": 0.000110661115427526,
      "loss": 2.2525,
      "step": 2598
    },
    {
      "epoch": 0.5198,
      "grad_norm": 0.0758962631225586,
      "learning_rate": 0.0001105916975449252,
      "loss": 2.2554,
      "step": 2599
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.06918636709451675,
      "learning_rate": 0.00011052227450005967,
      "loss": 2.2526,
      "step": 2600
    },
    {
      "epoch": 0.5202,
      "grad_norm": 0.06142076849937439,
      "learning_rate": 0.00011045284632676536,
      "loss": 2.2537,
      "step": 2601
    },
    {
      "epoch": 0.5204,
      "grad_norm": 0.04822550714015961,
      "learning_rate": 0.00011038341305888074,
      "loss": 2.2517,
      "step": 2602
    },
    {
      "epoch": 0.5206,
      "grad_norm": 0.03601802513003349,
      "learning_rate": 0.00011031397473024674,
      "loss": 2.255,
      "step": 2603
    },
    {
      "epoch": 0.5208,
      "grad_norm": 0.04205596446990967,
      "learning_rate": 0.00011024453137470677,
      "loss": 2.253,
      "step": 2604
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.05213512107729912,
      "learning_rate": 0.00011017508302610664,
      "loss": 2.2536,
      "step": 2605
    },
    {
      "epoch": 0.5212,
      "grad_norm": 0.052174605429172516,
      "learning_rate": 0.00011010562971829463,
      "loss": 2.2512,
      "step": 2606
    },
    {
      "epoch": 0.5214,
      "grad_norm": 0.04268869757652283,
      "learning_rate": 0.00011003617148512149,
      "loss": 2.2565,
      "step": 2607
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.036170944571495056,
      "learning_rate": 0.0001099667083604403,
      "loss": 2.2591,
      "step": 2608
    },
    {
      "epoch": 0.5218,
      "grad_norm": 0.04359792172908783,
      "learning_rate": 0.00010989724037810652,
      "loss": 2.2536,
      "step": 2609
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.05633258819580078,
      "learning_rate": 0.00010982776757197799,
      "loss": 2.2548,
      "step": 2610
    },
    {
      "epoch": 0.5222,
      "grad_norm": 0.06294973939657211,
      "learning_rate": 0.00010975828997591495,
      "loss": 2.2548,
      "step": 2611
    },
    {
      "epoch": 0.5224,
      "grad_norm": 0.058344192802906036,
      "learning_rate": 0.00010968880762377993,
      "loss": 2.2542,
      "step": 2612
    },
    {
      "epoch": 0.5226,
      "grad_norm": 0.04863126948475838,
      "learning_rate": 0.00010961932054943778,
      "loss": 2.2551,
      "step": 2613
    },
    {
      "epoch": 0.5228,
      "grad_norm": 0.036667075008153915,
      "learning_rate": 0.00010954982878675563,
      "loss": 2.254,
      "step": 2614
    },
    {
      "epoch": 0.523,
      "grad_norm": 0.04322841018438339,
      "learning_rate": 0.00010948033236960294,
      "loss": 2.2543,
      "step": 2615
    },
    {
      "epoch": 0.5232,
      "grad_norm": 0.0615541934967041,
      "learning_rate": 0.00010941083133185146,
      "loss": 2.2515,
      "step": 2616
    },
    {
      "epoch": 0.5234,
      "grad_norm": 0.06733080744743347,
      "learning_rate": 0.00010934132570737507,
      "loss": 2.2537,
      "step": 2617
    },
    {
      "epoch": 0.5236,
      "grad_norm": 0.061912450939416885,
      "learning_rate": 0.00010927181553005002,
      "loss": 2.2532,
      "step": 2618
    },
    {
      "epoch": 0.5238,
      "grad_norm": 0.04114459827542305,
      "learning_rate": 0.00010920230083375473,
      "loss": 2.2547,
      "step": 2619
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.043088965117931366,
      "learning_rate": 0.00010913278165236978,
      "loss": 2.252,
      "step": 2620
    },
    {
      "epoch": 0.5242,
      "grad_norm": 0.08989699184894562,
      "learning_rate": 0.00010906325801977804,
      "loss": 2.2549,
      "step": 2621
    },
    {
      "epoch": 0.5244,
      "grad_norm": 0.14086025953292847,
      "learning_rate": 0.00010899372996986439,
      "loss": 2.2548,
      "step": 2622
    },
    {
      "epoch": 0.5246,
      "grad_norm": 0.17273885011672974,
      "learning_rate": 0.00010892419753651606,
      "loss": 2.2539,
      "step": 2623
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.16789482533931732,
      "learning_rate": 0.00010885466075362223,
      "loss": 2.2541,
      "step": 2624
    },
    {
      "epoch": 0.525,
      "grad_norm": 0.13136136531829834,
      "learning_rate": 0.00010878511965507434,
      "loss": 2.2535,
      "step": 2625
    },
    {
      "epoch": 0.5252,
      "grad_norm": 0.06856507062911987,
      "learning_rate": 0.00010871557427476583,
      "loss": 2.2541,
      "step": 2626
    },
    {
      "epoch": 0.5254,
      "grad_norm": 0.040839288383722305,
      "learning_rate": 0.0001086460246465923,
      "loss": 2.2525,
      "step": 2627
    },
    {
      "epoch": 0.5256,
      "grad_norm": 0.09507431834936142,
      "learning_rate": 0.00010857647080445139,
      "loss": 2.2541,
      "step": 2628
    },
    {
      "epoch": 0.5258,
      "grad_norm": 0.1325189471244812,
      "learning_rate": 0.00010850691278224281,
      "loss": 2.2545,
      "step": 2629
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.14780741930007935,
      "learning_rate": 0.00010843735061386828,
      "loss": 2.2552,
      "step": 2630
    },
    {
      "epoch": 0.5262,
      "grad_norm": 0.12912407517433167,
      "learning_rate": 0.00010836778433323158,
      "loss": 2.2527,
      "step": 2631
    },
    {
      "epoch": 0.5264,
      "grad_norm": 0.08121488988399506,
      "learning_rate": 0.0001082982139742384,
      "loss": 2.254,
      "step": 2632
    },
    {
      "epoch": 0.5266,
      "grad_norm": 0.03598704934120178,
      "learning_rate": 0.00010822863957079656,
      "loss": 2.2536,
      "step": 2633
    },
    {
      "epoch": 0.5268,
      "grad_norm": 0.07612190395593643,
      "learning_rate": 0.00010815906115681578,
      "loss": 2.255,
      "step": 2634
    },
    {
      "epoch": 0.527,
      "grad_norm": 0.10694251209497452,
      "learning_rate": 0.00010808947876620767,
      "loss": 2.2529,
      "step": 2635
    },
    {
      "epoch": 0.5272,
      "grad_norm": 0.09760890156030655,
      "learning_rate": 0.00010801989243288589,
      "loss": 2.2556,
      "step": 2636
    },
    {
      "epoch": 0.5274,
      "grad_norm": 0.061705708503723145,
      "learning_rate": 0.00010795030219076599,
      "loss": 2.255,
      "step": 2637
    },
    {
      "epoch": 0.5276,
      "grad_norm": 0.04161055013537407,
      "learning_rate": 0.00010788070807376536,
      "loss": 2.2563,
      "step": 2638
    },
    {
      "epoch": 0.5278,
      "grad_norm": 0.07024098932743073,
      "learning_rate": 0.00010781111011580336,
      "loss": 2.2534,
      "step": 2639
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.11365289986133575,
      "learning_rate": 0.00010774150835080119,
      "loss": 2.2529,
      "step": 2640
    },
    {
      "epoch": 0.5282,
      "grad_norm": 0.13329482078552246,
      "learning_rate": 0.00010767190281268187,
      "loss": 2.2564,
      "step": 2641
    },
    {
      "epoch": 0.5284,
      "grad_norm": 0.12417105585336685,
      "learning_rate": 0.00010760229353537033,
      "loss": 2.255,
      "step": 2642
    },
    {
      "epoch": 0.5286,
      "grad_norm": 0.08888860046863556,
      "learning_rate": 0.00010753268055279329,
      "loss": 2.2562,
      "step": 2643
    },
    {
      "epoch": 0.5288,
      "grad_norm": 0.04974381998181343,
      "learning_rate": 0.00010746306389887924,
      "loss": 2.2553,
      "step": 2644
    },
    {
      "epoch": 0.529,
      "grad_norm": 0.04523564130067825,
      "learning_rate": 0.00010739344360755852,
      "loss": 2.2554,
      "step": 2645
    },
    {
      "epoch": 0.5292,
      "grad_norm": 0.0916004553437233,
      "learning_rate": 0.00010732381971276318,
      "loss": 2.2526,
      "step": 2646
    },
    {
      "epoch": 0.5294,
      "grad_norm": 0.12194966524839401,
      "learning_rate": 0.0001072541922484271,
      "loss": 2.2549,
      "step": 2647
    },
    {
      "epoch": 0.5296,
      "grad_norm": 0.10657462477684021,
      "learning_rate": 0.00010718456124848583,
      "loss": 2.2534,
      "step": 2648
    },
    {
      "epoch": 0.5298,
      "grad_norm": 0.05975745618343353,
      "learning_rate": 0.00010711492674687671,
      "loss": 2.253,
      "step": 2649
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.03906426578760147,
      "learning_rate": 0.0001070452887775387,
      "loss": 2.2531,
      "step": 2650
    },
    {
      "epoch": 0.5302,
      "grad_norm": 0.07109455019235611,
      "learning_rate": 0.00010697564737441252,
      "loss": 2.2534,
      "step": 2651
    },
    {
      "epoch": 0.5304,
      "grad_norm": 0.09472496062517166,
      "learning_rate": 0.00010690600257144061,
      "loss": 2.2533,
      "step": 2652
    },
    {
      "epoch": 0.5306,
      "grad_norm": 0.09462033212184906,
      "learning_rate": 0.00010683635440256687,
      "loss": 2.2551,
      "step": 2653
    },
    {
      "epoch": 0.5308,
      "grad_norm": 0.08313468843698502,
      "learning_rate": 0.00010676670290173709,
      "loss": 2.2572,
      "step": 2654
    },
    {
      "epoch": 0.531,
      "grad_norm": 0.058747872710227966,
      "learning_rate": 0.00010669704810289851,
      "loss": 2.2529,
      "step": 2655
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.04013830050826073,
      "learning_rate": 0.00010662739004000005,
      "loss": 2.2527,
      "step": 2656
    },
    {
      "epoch": 0.5314,
      "grad_norm": 0.03737836331129074,
      "learning_rate": 0.00010655772874699217,
      "loss": 2.2544,
      "step": 2657
    },
    {
      "epoch": 0.5316,
      "grad_norm": 0.045074306428432465,
      "learning_rate": 0.00010648806425782695,
      "loss": 2.2526,
      "step": 2658
    },
    {
      "epoch": 0.5318,
      "grad_norm": 0.05934572219848633,
      "learning_rate": 0.00010641839660645805,
      "loss": 2.2541,
      "step": 2659
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.07279341667890549,
      "learning_rate": 0.00010634872582684061,
      "loss": 2.2549,
      "step": 2660
    },
    {
      "epoch": 0.5322,
      "grad_norm": 0.08262772858142853,
      "learning_rate": 0.00010627905195293135,
      "loss": 2.2525,
      "step": 2661
    },
    {
      "epoch": 0.5324,
      "grad_norm": 0.07921991497278214,
      "learning_rate": 0.00010620937501868841,
      "loss": 2.2532,
      "step": 2662
    },
    {
      "epoch": 0.5326,
      "grad_norm": 0.060493104159832,
      "learning_rate": 0.00010613969505807156,
      "loss": 2.2523,
      "step": 2663
    },
    {
      "epoch": 0.5328,
      "grad_norm": 0.03690355271100998,
      "learning_rate": 0.00010607001210504191,
      "loss": 2.2546,
      "step": 2664
    },
    {
      "epoch": 0.533,
      "grad_norm": 0.0627632662653923,
      "learning_rate": 0.00010600032619356209,
      "loss": 2.2539,
      "step": 2665
    },
    {
      "epoch": 0.5332,
      "grad_norm": 0.09638942778110504,
      "learning_rate": 0.00010593063735759618,
      "loss": 2.2525,
      "step": 2666
    },
    {
      "epoch": 0.5334,
      "grad_norm": 0.10723550617694855,
      "learning_rate": 0.00010586094563110964,
      "loss": 2.2544,
      "step": 2667
    },
    {
      "epoch": 0.5336,
      "grad_norm": 0.10031808167695999,
      "learning_rate": 0.00010579125104806944,
      "loss": 2.2545,
      "step": 2668
    },
    {
      "epoch": 0.5338,
      "grad_norm": 0.07012747973203659,
      "learning_rate": 0.00010572155364244382,
      "loss": 2.2531,
      "step": 2669
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.04039096459746361,
      "learning_rate": 0.00010565185344820247,
      "loss": 2.2521,
      "step": 2670
    },
    {
      "epoch": 0.5342,
      "grad_norm": 0.04245922341942787,
      "learning_rate": 0.00010558215049931638,
      "loss": 2.2536,
      "step": 2671
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.05528242141008377,
      "learning_rate": 0.00010551244482975798,
      "loss": 2.2541,
      "step": 2672
    },
    {
      "epoch": 0.5346,
      "grad_norm": 0.06062161549925804,
      "learning_rate": 0.00010544273647350092,
      "loss": 2.2537,
      "step": 2673
    },
    {
      "epoch": 0.5348,
      "grad_norm": 0.054264966398477554,
      "learning_rate": 0.00010537302546452022,
      "loss": 2.2529,
      "step": 2674
    },
    {
      "epoch": 0.535,
      "grad_norm": 0.04182315245270729,
      "learning_rate": 0.00010530331183679218,
      "loss": 2.2536,
      "step": 2675
    },
    {
      "epoch": 0.5352,
      "grad_norm": 0.03671992942690849,
      "learning_rate": 0.0001052335956242944,
      "loss": 2.2514,
      "step": 2676
    },
    {
      "epoch": 0.5354,
      "grad_norm": 0.03753357380628586,
      "learning_rate": 0.00010516387686100566,
      "loss": 2.2522,
      "step": 2677
    },
    {
      "epoch": 0.5356,
      "grad_norm": 0.03676513209939003,
      "learning_rate": 0.00010509415558090609,
      "loss": 2.2524,
      "step": 2678
    },
    {
      "epoch": 0.5358,
      "grad_norm": 0.041547976434230804,
      "learning_rate": 0.00010502443181797697,
      "loss": 2.2543,
      "step": 2679
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.044139984995126724,
      "learning_rate": 0.00010495470560620083,
      "loss": 2.2535,
      "step": 2680
    },
    {
      "epoch": 0.5362,
      "grad_norm": 0.04351888597011566,
      "learning_rate": 0.00010488497697956135,
      "loss": 2.2534,
      "step": 2681
    },
    {
      "epoch": 0.5364,
      "grad_norm": 0.03963497281074524,
      "learning_rate": 0.00010481524597204342,
      "loss": 2.2532,
      "step": 2682
    },
    {
      "epoch": 0.5366,
      "grad_norm": 0.037290655076503754,
      "learning_rate": 0.00010474551261763314,
      "loss": 2.254,
      "step": 2683
    },
    {
      "epoch": 0.5368,
      "grad_norm": 0.03712819144129753,
      "learning_rate": 0.00010467577695031762,
      "loss": 2.2518,
      "step": 2684
    },
    {
      "epoch": 0.537,
      "grad_norm": 0.03841441869735718,
      "learning_rate": 0.00010460603900408523,
      "loss": 2.2538,
      "step": 2685
    },
    {
      "epoch": 0.5372,
      "grad_norm": 0.04016151651740074,
      "learning_rate": 0.00010453629881292538,
      "loss": 2.2546,
      "step": 2686
    },
    {
      "epoch": 0.5374,
      "grad_norm": 0.03804207220673561,
      "learning_rate": 0.00010446655641082862,
      "loss": 2.2526,
      "step": 2687
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.037175729870796204,
      "learning_rate": 0.0001043968118317865,
      "loss": 2.2522,
      "step": 2688
    },
    {
      "epoch": 0.5378,
      "grad_norm": 0.037197813391685486,
      "learning_rate": 0.00010432706510979171,
      "loss": 2.2525,
      "step": 2689
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.037842802703380585,
      "learning_rate": 0.00010425731627883797,
      "loss": 2.2533,
      "step": 2690
    },
    {
      "epoch": 0.5382,
      "grad_norm": 0.04005451127886772,
      "learning_rate": 0.00010418756537291996,
      "loss": 2.2543,
      "step": 2691
    },
    {
      "epoch": 0.5384,
      "grad_norm": 0.043113984167575836,
      "learning_rate": 0.00010411781242603352,
      "loss": 2.2527,
      "step": 2692
    },
    {
      "epoch": 0.5386,
      "grad_norm": 0.051554176956415176,
      "learning_rate": 0.00010404805747217526,
      "loss": 2.2527,
      "step": 2693
    },
    {
      "epoch": 0.5388,
      "grad_norm": 0.054528627544641495,
      "learning_rate": 0.000103978300545343,
      "loss": 2.2528,
      "step": 2694
    },
    {
      "epoch": 0.539,
      "grad_norm": 0.047687407582998276,
      "learning_rate": 0.00010390854167953537,
      "loss": 2.2529,
      "step": 2695
    },
    {
      "epoch": 0.5392,
      "grad_norm": 0.03730364516377449,
      "learning_rate": 0.00010383878090875201,
      "loss": 2.2522,
      "step": 2696
    },
    {
      "epoch": 0.5394,
      "grad_norm": 0.061174701899290085,
      "learning_rate": 0.00010376901826699348,
      "loss": 2.2546,
      "step": 2697
    },
    {
      "epoch": 0.5396,
      "grad_norm": 0.0969473198056221,
      "learning_rate": 0.0001036992537882612,
      "loss": 2.2524,
      "step": 2698
    },
    {
      "epoch": 0.5398,
      "grad_norm": 0.11840967833995819,
      "learning_rate": 0.00010362948750655759,
      "loss": 2.253,
      "step": 2699
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.11614800244569778,
      "learning_rate": 0.00010355971945588585,
      "loss": 2.254,
      "step": 2700
    },
    {
      "epoch": 0.5402,
      "grad_norm": 0.08414299786090851,
      "learning_rate": 0.00010348994967025012,
      "loss": 2.2541,
      "step": 2701
    },
    {
      "epoch": 0.5404,
      "grad_norm": 0.05200556293129921,
      "learning_rate": 0.0001034201781836553,
      "loss": 2.2532,
      "step": 2702
    },
    {
      "epoch": 0.5406,
      "grad_norm": 0.037508539855480194,
      "learning_rate": 0.00010335040503010716,
      "loss": 2.2538,
      "step": 2703
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.043445833027362823,
      "learning_rate": 0.00010328063024361232,
      "loss": 2.2536,
      "step": 2704
    },
    {
      "epoch": 0.541,
      "grad_norm": 0.05450231954455376,
      "learning_rate": 0.00010321085385817817,
      "loss": 2.2532,
      "step": 2705
    },
    {
      "epoch": 0.5412,
      "grad_norm": 0.051573846489191055,
      "learning_rate": 0.00010314107590781284,
      "loss": 2.2539,
      "step": 2706
    },
    {
      "epoch": 0.5414,
      "grad_norm": 0.04423009976744652,
      "learning_rate": 0.00010307129642652528,
      "loss": 2.2548,
      "step": 2707
    },
    {
      "epoch": 0.5416,
      "grad_norm": 0.042508143931627274,
      "learning_rate": 0.00010300151544832512,
      "loss": 2.2537,
      "step": 2708
    },
    {
      "epoch": 0.5418,
      "grad_norm": 0.045117609202861786,
      "learning_rate": 0.00010293173300722285,
      "loss": 2.2543,
      "step": 2709
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.04238830506801605,
      "learning_rate": 0.00010286194913722948,
      "loss": 2.2523,
      "step": 2710
    },
    {
      "epoch": 0.5422,
      "grad_norm": 0.03927888348698616,
      "learning_rate": 0.0001027921638723569,
      "loss": 2.2538,
      "step": 2711
    },
    {
      "epoch": 0.5424,
      "grad_norm": 0.03794461116194725,
      "learning_rate": 0.00010272237724661753,
      "loss": 2.2499,
      "step": 2712
    },
    {
      "epoch": 0.5426,
      "grad_norm": 0.037575867027044296,
      "learning_rate": 0.0001026525892940246,
      "loss": 2.2513,
      "step": 2713
    },
    {
      "epoch": 0.5428,
      "grad_norm": 0.03803345188498497,
      "learning_rate": 0.00010258280004859188,
      "loss": 2.2491,
      "step": 2714
    },
    {
      "epoch": 0.543,
      "grad_norm": 0.03853866457939148,
      "learning_rate": 0.00010251300954433376,
      "loss": 2.2498,
      "step": 2715
    },
    {
      "epoch": 0.5432,
      "grad_norm": 0.04372375085949898,
      "learning_rate": 0.00010244321781526533,
      "loss": 2.2534,
      "step": 2716
    },
    {
      "epoch": 0.5434,
      "grad_norm": 0.060124821960926056,
      "learning_rate": 0.0001023734248954022,
      "loss": 2.2523,
      "step": 2717
    },
    {
      "epoch": 0.5436,
      "grad_norm": 0.08681156486272812,
      "learning_rate": 0.00010230363081876064,
      "loss": 2.2511,
      "step": 2718
    },
    {
      "epoch": 0.5438,
      "grad_norm": 0.09887290745973587,
      "learning_rate": 0.00010223383561935738,
      "loss": 2.2518,
      "step": 2719
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.09457878768444061,
      "learning_rate": 0.00010216403933120979,
      "loss": 2.2531,
      "step": 2720
    },
    {
      "epoch": 0.5442,
      "grad_norm": 0.08201874792575836,
      "learning_rate": 0.0001020942419883357,
      "loss": 2.2521,
      "step": 2721
    },
    {
      "epoch": 0.5444,
      "grad_norm": 0.06648856401443481,
      "learning_rate": 0.00010202444362475352,
      "loss": 2.2521,
      "step": 2722
    },
    {
      "epoch": 0.5446,
      "grad_norm": 0.041310884058475494,
      "learning_rate": 0.00010195464427448213,
      "loss": 2.255,
      "step": 2723
    },
    {
      "epoch": 0.5448,
      "grad_norm": 0.05212961509823799,
      "learning_rate": 0.00010188484397154084,
      "loss": 2.251,
      "step": 2724
    },
    {
      "epoch": 0.545,
      "grad_norm": 0.09936342388391495,
      "learning_rate": 0.00010181504274994948,
      "loss": 2.2527,
      "step": 2725
    },
    {
      "epoch": 0.5452,
      "grad_norm": 0.13431304693222046,
      "learning_rate": 0.00010174524064372837,
      "loss": 2.2531,
      "step": 2726
    },
    {
      "epoch": 0.5454,
      "grad_norm": 0.13960151374340057,
      "learning_rate": 0.00010167543768689815,
      "loss": 2.253,
      "step": 2727
    },
    {
      "epoch": 0.5456,
      "grad_norm": 0.10750044137239456,
      "learning_rate": 0.00010160563391347996,
      "loss": 2.2523,
      "step": 2728
    },
    {
      "epoch": 0.5458,
      "grad_norm": 0.05574287474155426,
      "learning_rate": 0.0001015358293574953,
      "loss": 2.2515,
      "step": 2729
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.04406127706170082,
      "learning_rate": 0.00010146602405296607,
      "loss": 2.2531,
      "step": 2730
    },
    {
      "epoch": 0.5462,
      "grad_norm": 0.07081250846385956,
      "learning_rate": 0.00010139621803391455,
      "loss": 2.2551,
      "step": 2731
    },
    {
      "epoch": 0.5464,
      "grad_norm": 0.0856037437915802,
      "learning_rate": 0.0001013264113343633,
      "loss": 2.2522,
      "step": 2732
    },
    {
      "epoch": 0.5466,
      "grad_norm": 0.08208929002285004,
      "learning_rate": 0.00010125660398833528,
      "loss": 2.2529,
      "step": 2733
    },
    {
      "epoch": 0.5468,
      "grad_norm": 0.06212901324033737,
      "learning_rate": 0.00010118679602985373,
      "loss": 2.2511,
      "step": 2734
    },
    {
      "epoch": 0.547,
      "grad_norm": 0.03876010701060295,
      "learning_rate": 0.00010111698749294223,
      "loss": 2.2533,
      "step": 2735
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.052954234182834625,
      "learning_rate": 0.00010104717841162458,
      "loss": 2.2524,
      "step": 2736
    },
    {
      "epoch": 0.5474,
      "grad_norm": 0.08094169944524765,
      "learning_rate": 0.00010097736881992492,
      "loss": 2.2493,
      "step": 2737
    },
    {
      "epoch": 0.5476,
      "grad_norm": 0.08750410377979279,
      "learning_rate": 0.00010090755875186753,
      "loss": 2.2521,
      "step": 2738
    },
    {
      "epoch": 0.5478,
      "grad_norm": 0.07744032889604568,
      "learning_rate": 0.00010083774824147708,
      "loss": 2.251,
      "step": 2739
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.056419625878334045,
      "learning_rate": 0.00010076793732277829,
      "loss": 2.2541,
      "step": 2740
    },
    {
      "epoch": 0.5482,
      "grad_norm": 0.04043165221810341,
      "learning_rate": 0.00010069812602979615,
      "loss": 2.2522,
      "step": 2741
    },
    {
      "epoch": 0.5484,
      "grad_norm": 0.04049070179462433,
      "learning_rate": 0.00010062831439655591,
      "loss": 2.2529,
      "step": 2742
    },
    {
      "epoch": 0.5486,
      "grad_norm": 0.041625965386629105,
      "learning_rate": 0.00010055850245708283,
      "loss": 2.2502,
      "step": 2743
    },
    {
      "epoch": 0.5488,
      "grad_norm": 0.03850638493895531,
      "learning_rate": 0.00010048869024540247,
      "loss": 2.2504,
      "step": 2744
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.04348229989409447,
      "learning_rate": 0.0001004188777955404,
      "loss": 2.2531,
      "step": 2745
    },
    {
      "epoch": 0.5492,
      "grad_norm": 0.05796435847878456,
      "learning_rate": 0.00010034906514152238,
      "loss": 2.2527,
      "step": 2746
    },
    {
      "epoch": 0.5494,
      "grad_norm": 0.07325813174247742,
      "learning_rate": 0.00010027925231737428,
      "loss": 2.2526,
      "step": 2747
    },
    {
      "epoch": 0.5496,
      "grad_norm": 0.07623785734176636,
      "learning_rate": 0.00010020943935712192,
      "loss": 2.249,
      "step": 2748
    },
    {
      "epoch": 0.5498,
      "grad_norm": 0.0675276443362236,
      "learning_rate": 0.00010013962629479146,
      "loss": 2.2527,
      "step": 2749
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.045871444046497345,
      "learning_rate": 0.00010006981316440875,
      "loss": 2.2517,
      "step": 2750
    },
    {
      "epoch": 0.5502,
      "grad_norm": 0.039248786866664886,
      "learning_rate": 0.0001,
      "loss": 2.2505,
      "step": 2751
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.05348920822143555,
      "learning_rate": 9.993018683559126e-05,
      "loss": 2.2517,
      "step": 2752
    },
    {
      "epoch": 0.5506,
      "grad_norm": 0.06278499960899353,
      "learning_rate": 9.986037370520857e-05,
      "loss": 2.2528,
      "step": 2753
    },
    {
      "epoch": 0.5508,
      "grad_norm": 0.060754504054784775,
      "learning_rate": 9.979056064287806e-05,
      "loss": 2.2536,
      "step": 2754
    },
    {
      "epoch": 0.551,
      "grad_norm": 0.051320794969797134,
      "learning_rate": 9.972074768262576e-05,
      "loss": 2.2546,
      "step": 2755
    },
    {
      "epoch": 0.5512,
      "grad_norm": 0.045472584664821625,
      "learning_rate": 9.965093485847767e-05,
      "loss": 2.2484,
      "step": 2756
    },
    {
      "epoch": 0.5514,
      "grad_norm": 0.0406116358935833,
      "learning_rate": 9.958112220445963e-05,
      "loss": 2.2518,
      "step": 2757
    },
    {
      "epoch": 0.5516,
      "grad_norm": 0.03928998485207558,
      "learning_rate": 9.951130975459757e-05,
      "loss": 2.2523,
      "step": 2758
    },
    {
      "epoch": 0.5518,
      "grad_norm": 0.049717340618371964,
      "learning_rate": 9.94414975429172e-05,
      "loss": 2.2505,
      "step": 2759
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.058449577540159225,
      "learning_rate": 9.937168560344412e-05,
      "loss": 2.2511,
      "step": 2760
    },
    {
      "epoch": 0.5522,
      "grad_norm": 0.05971536040306091,
      "learning_rate": 9.930187397020386e-05,
      "loss": 2.2529,
      "step": 2761
    },
    {
      "epoch": 0.5524,
      "grad_norm": 0.05042118951678276,
      "learning_rate": 9.923206267722173e-05,
      "loss": 2.2532,
      "step": 2762
    },
    {
      "epoch": 0.5526,
      "grad_norm": 0.042498987168073654,
      "learning_rate": 9.916225175852293e-05,
      "loss": 2.2499,
      "step": 2763
    },
    {
      "epoch": 0.5528,
      "grad_norm": 0.038722869008779526,
      "learning_rate": 9.909244124813247e-05,
      "loss": 2.2524,
      "step": 2764
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.04063653200864792,
      "learning_rate": 9.902263118007513e-05,
      "loss": 2.2532,
      "step": 2765
    },
    {
      "epoch": 0.5532,
      "grad_norm": 0.0429660938680172,
      "learning_rate": 9.895282158837545e-05,
      "loss": 2.252,
      "step": 2766
    },
    {
      "epoch": 0.5534,
      "grad_norm": 0.040408048778772354,
      "learning_rate": 9.888301250705779e-05,
      "loss": 2.2525,
      "step": 2767
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.03897963836789131,
      "learning_rate": 9.881320397014629e-05,
      "loss": 2.252,
      "step": 2768
    },
    {
      "epoch": 0.5538,
      "grad_norm": 0.04665630683302879,
      "learning_rate": 9.874339601166473e-05,
      "loss": 2.2511,
      "step": 2769
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.053735628724098206,
      "learning_rate": 9.867358866563673e-05,
      "loss": 2.2527,
      "step": 2770
    },
    {
      "epoch": 0.5542,
      "grad_norm": 0.05572671815752983,
      "learning_rate": 9.860378196608549e-05,
      "loss": 2.2553,
      "step": 2771
    },
    {
      "epoch": 0.5544,
      "grad_norm": 0.05171624943614006,
      "learning_rate": 9.853397594703394e-05,
      "loss": 2.2533,
      "step": 2772
    },
    {
      "epoch": 0.5546,
      "grad_norm": 0.04841522499918938,
      "learning_rate": 9.846417064250471e-05,
      "loss": 2.2533,
      "step": 2773
    },
    {
      "epoch": 0.5548,
      "grad_norm": 0.046639762818813324,
      "learning_rate": 9.839436608652007e-05,
      "loss": 2.2519,
      "step": 2774
    },
    {
      "epoch": 0.555,
      "grad_norm": 0.04574999585747719,
      "learning_rate": 9.832456231310189e-05,
      "loss": 2.2539,
      "step": 2775
    },
    {
      "epoch": 0.5552,
      "grad_norm": 0.039051130414009094,
      "learning_rate": 9.825475935627165e-05,
      "loss": 2.2513,
      "step": 2776
    },
    {
      "epoch": 0.5554,
      "grad_norm": 0.045619700103998184,
      "learning_rate": 9.818495725005054e-05,
      "loss": 2.2504,
      "step": 2777
    },
    {
      "epoch": 0.5556,
      "grad_norm": 0.06515227258205414,
      "learning_rate": 9.81151560284592e-05,
      "loss": 2.2513,
      "step": 2778
    },
    {
      "epoch": 0.5558,
      "grad_norm": 0.07768017053604126,
      "learning_rate": 9.804535572551789e-05,
      "loss": 2.2489,
      "step": 2779
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.08016903698444366,
      "learning_rate": 9.797555637524649e-05,
      "loss": 2.2538,
      "step": 2780
    },
    {
      "epoch": 0.5562,
      "grad_norm": 0.060980189591646194,
      "learning_rate": 9.790575801166432e-05,
      "loss": 2.2536,
      "step": 2781
    },
    {
      "epoch": 0.5564,
      "grad_norm": 0.03928131237626076,
      "learning_rate": 9.783596066879022e-05,
      "loss": 2.2528,
      "step": 2782
    },
    {
      "epoch": 0.5566,
      "grad_norm": 0.07196807861328125,
      "learning_rate": 9.776616438064264e-05,
      "loss": 2.2517,
      "step": 2783
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.09770932793617249,
      "learning_rate": 9.76963691812394e-05,
      "loss": 2.2532,
      "step": 2784
    },
    {
      "epoch": 0.557,
      "grad_norm": 0.0997222438454628,
      "learning_rate": 9.762657510459783e-05,
      "loss": 2.2522,
      "step": 2785
    },
    {
      "epoch": 0.5572,
      "grad_norm": 0.08008109778165817,
      "learning_rate": 9.755678218473469e-05,
      "loss": 2.2515,
      "step": 2786
    },
    {
      "epoch": 0.5574,
      "grad_norm": 0.050047680735588074,
      "learning_rate": 9.748699045566626e-05,
      "loss": 2.2518,
      "step": 2787
    },
    {
      "epoch": 0.5576,
      "grad_norm": 0.04326017200946808,
      "learning_rate": 9.741719995140814e-05,
      "loss": 2.2534,
      "step": 2788
    },
    {
      "epoch": 0.5578,
      "grad_norm": 0.07008269429206848,
      "learning_rate": 9.734741070597539e-05,
      "loss": 2.2517,
      "step": 2789
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.0913335457444191,
      "learning_rate": 9.727762275338246e-05,
      "loss": 2.2526,
      "step": 2790
    },
    {
      "epoch": 0.5582,
      "grad_norm": 0.09501408785581589,
      "learning_rate": 9.720783612764314e-05,
      "loss": 2.253,
      "step": 2791
    },
    {
      "epoch": 0.5584,
      "grad_norm": 0.08851312845945358,
      "learning_rate": 9.713805086277054e-05,
      "loss": 2.2525,
      "step": 2792
    },
    {
      "epoch": 0.5586,
      "grad_norm": 0.07215219736099243,
      "learning_rate": 9.706826699277718e-05,
      "loss": 2.2536,
      "step": 2793
    },
    {
      "epoch": 0.5588,
      "grad_norm": 0.04943731427192688,
      "learning_rate": 9.699848455167489e-05,
      "loss": 2.2522,
      "step": 2794
    },
    {
      "epoch": 0.559,
      "grad_norm": 0.042964812368154526,
      "learning_rate": 9.692870357347473e-05,
      "loss": 2.2564,
      "step": 2795
    },
    {
      "epoch": 0.5592,
      "grad_norm": 0.0526539608836174,
      "learning_rate": 9.685892409218717e-05,
      "loss": 2.2546,
      "step": 2796
    },
    {
      "epoch": 0.5594,
      "grad_norm": 0.058619387447834015,
      "learning_rate": 9.678914614182185e-05,
      "loss": 2.2527,
      "step": 2797
    },
    {
      "epoch": 0.5596,
      "grad_norm": 0.05194536969065666,
      "learning_rate": 9.671936975638768e-05,
      "loss": 2.2522,
      "step": 2798
    },
    {
      "epoch": 0.5598,
      "grad_norm": 0.04361266270279884,
      "learning_rate": 9.664959496989285e-05,
      "loss": 2.2501,
      "step": 2799
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.039342910051345825,
      "learning_rate": 9.657982181634475e-05,
      "loss": 2.2521,
      "step": 2800
    },
    {
      "epoch": 0.5602,
      "grad_norm": 0.04465210810303688,
      "learning_rate": 9.651005032974994e-05,
      "loss": 2.2523,
      "step": 2801
    },
    {
      "epoch": 0.5604,
      "grad_norm": 0.06129508092999458,
      "learning_rate": 9.644028054411416e-05,
      "loss": 2.2514,
      "step": 2802
    },
    {
      "epoch": 0.5606,
      "grad_norm": 0.07783286273479462,
      "learning_rate": 9.637051249344243e-05,
      "loss": 2.2503,
      "step": 2803
    },
    {
      "epoch": 0.5608,
      "grad_norm": 0.08646189421415329,
      "learning_rate": 9.630074621173883e-05,
      "loss": 2.2526,
      "step": 2804
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.08335322141647339,
      "learning_rate": 9.623098173300654e-05,
      "loss": 2.2493,
      "step": 2805
    },
    {
      "epoch": 0.5612,
      "grad_norm": 0.07386361807584763,
      "learning_rate": 9.616121909124801e-05,
      "loss": 2.253,
      "step": 2806
    },
    {
      "epoch": 0.5614,
      "grad_norm": 0.05534215271472931,
      "learning_rate": 9.609145832046465e-05,
      "loss": 2.254,
      "step": 2807
    },
    {
      "epoch": 0.5616,
      "grad_norm": 0.042035266757011414,
      "learning_rate": 9.602169945465702e-05,
      "loss": 2.2535,
      "step": 2808
    },
    {
      "epoch": 0.5618,
      "grad_norm": 0.042466603219509125,
      "learning_rate": 9.595194252782477e-05,
      "loss": 2.2523,
      "step": 2809
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.06009724736213684,
      "learning_rate": 9.588218757396655e-05,
      "loss": 2.2516,
      "step": 2810
    },
    {
      "epoch": 0.5622,
      "grad_norm": 0.08319206535816193,
      "learning_rate": 9.581243462708006e-05,
      "loss": 2.2519,
      "step": 2811
    },
    {
      "epoch": 0.5624,
      "grad_norm": 0.10209809988737106,
      "learning_rate": 9.574268372116205e-05,
      "loss": 2.2531,
      "step": 2812
    },
    {
      "epoch": 0.5626,
      "grad_norm": 0.11121319979429245,
      "learning_rate": 9.567293489020831e-05,
      "loss": 2.2527,
      "step": 2813
    },
    {
      "epoch": 0.5628,
      "grad_norm": 0.10277381539344788,
      "learning_rate": 9.560318816821353e-05,
      "loss": 2.254,
      "step": 2814
    },
    {
      "epoch": 0.563,
      "grad_norm": 0.07899581640958786,
      "learning_rate": 9.55334435891714e-05,
      "loss": 2.2515,
      "step": 2815
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.04517989978194237,
      "learning_rate": 9.546370118707463e-05,
      "loss": 2.2525,
      "step": 2816
    },
    {
      "epoch": 0.5634,
      "grad_norm": 0.04874543473124504,
      "learning_rate": 9.539396099591476e-05,
      "loss": 2.2527,
      "step": 2817
    },
    {
      "epoch": 0.5636,
      "grad_norm": 0.08636809885501862,
      "learning_rate": 9.532422304968243e-05,
      "loss": 2.2541,
      "step": 2818
    },
    {
      "epoch": 0.5638,
      "grad_norm": 0.10887370258569717,
      "learning_rate": 9.525448738236691e-05,
      "loss": 2.2541,
      "step": 2819
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.09711309522390366,
      "learning_rate": 9.518475402795661e-05,
      "loss": 2.2522,
      "step": 2820
    },
    {
      "epoch": 0.5642,
      "grad_norm": 0.05867580696940422,
      "learning_rate": 9.511502302043868e-05,
      "loss": 2.2554,
      "step": 2821
    },
    {
      "epoch": 0.5644,
      "grad_norm": 0.051231611520051956,
      "learning_rate": 9.504529439379921e-05,
      "loss": 2.2532,
      "step": 2822
    },
    {
      "epoch": 0.5646,
      "grad_norm": 0.11462923884391785,
      "learning_rate": 9.497556818202306e-05,
      "loss": 2.252,
      "step": 2823
    },
    {
      "epoch": 0.5648,
      "grad_norm": 0.1613958179950714,
      "learning_rate": 9.490584441909392e-05,
      "loss": 2.2529,
      "step": 2824
    },
    {
      "epoch": 0.565,
      "grad_norm": 0.17146477103233337,
      "learning_rate": 9.483612313899435e-05,
      "loss": 2.253,
      "step": 2825
    },
    {
      "epoch": 0.5652,
      "grad_norm": 0.13527818024158478,
      "learning_rate": 9.476640437570562e-05,
      "loss": 2.2512,
      "step": 2826
    },
    {
      "epoch": 0.5654,
      "grad_norm": 0.06879140436649323,
      "learning_rate": 9.469668816320784e-05,
      "loss": 2.251,
      "step": 2827
    },
    {
      "epoch": 0.5656,
      "grad_norm": 0.04924381896853447,
      "learning_rate": 9.462697453547979e-05,
      "loss": 2.2518,
      "step": 2828
    },
    {
      "epoch": 0.5658,
      "grad_norm": 0.1062566488981247,
      "learning_rate": 9.455726352649911e-05,
      "loss": 2.2508,
      "step": 2829
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.14358171820640564,
      "learning_rate": 9.448755517024206e-05,
      "loss": 2.2516,
      "step": 2830
    },
    {
      "epoch": 0.5662,
      "grad_norm": 0.14676614105701447,
      "learning_rate": 9.441784950068362e-05,
      "loss": 2.252,
      "step": 2831
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.13402926921844482,
      "learning_rate": 9.434814655179755e-05,
      "loss": 2.2548,
      "step": 2832
    },
    {
      "epoch": 0.5666,
      "grad_norm": 0.10295087844133377,
      "learning_rate": 9.427844635755619e-05,
      "loss": 2.253,
      "step": 2833
    },
    {
      "epoch": 0.5668,
      "grad_norm": 0.062275927513837814,
      "learning_rate": 9.420874895193056e-05,
      "loss": 2.2491,
      "step": 2834
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.040898263454437256,
      "learning_rate": 9.413905436889035e-05,
      "loss": 2.252,
      "step": 2835
    },
    {
      "epoch": 0.5672,
      "grad_norm": 0.0654892697930336,
      "learning_rate": 9.406936264240386e-05,
      "loss": 2.2492,
      "step": 2836
    },
    {
      "epoch": 0.5674,
      "grad_norm": 0.08466637879610062,
      "learning_rate": 9.399967380643796e-05,
      "loss": 2.2519,
      "step": 2837
    },
    {
      "epoch": 0.5676,
      "grad_norm": 0.08085312694311142,
      "learning_rate": 9.392998789495811e-05,
      "loss": 2.2503,
      "step": 2838
    },
    {
      "epoch": 0.5678,
      "grad_norm": 0.062282249331474304,
      "learning_rate": 9.386030494192846e-05,
      "loss": 2.2506,
      "step": 2839
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.04147384315729141,
      "learning_rate": 9.37906249813116e-05,
      "loss": 2.2515,
      "step": 2840
    },
    {
      "epoch": 0.5682,
      "grad_norm": 0.04905083402991295,
      "learning_rate": 9.372094804706867e-05,
      "loss": 2.252,
      "step": 2841
    },
    {
      "epoch": 0.5684,
      "grad_norm": 0.07711534947156906,
      "learning_rate": 9.36512741731594e-05,
      "loss": 2.2498,
      "step": 2842
    },
    {
      "epoch": 0.5686,
      "grad_norm": 0.09477972984313965,
      "learning_rate": 9.358160339354194e-05,
      "loss": 2.251,
      "step": 2843
    },
    {
      "epoch": 0.5688,
      "grad_norm": 0.08826594799757004,
      "learning_rate": 9.351193574217306e-05,
      "loss": 2.2509,
      "step": 2844
    },
    {
      "epoch": 0.569,
      "grad_norm": 0.06409216672182083,
      "learning_rate": 9.344227125300788e-05,
      "loss": 2.2516,
      "step": 2845
    },
    {
      "epoch": 0.5692,
      "grad_norm": 0.04643873870372772,
      "learning_rate": 9.337260996000002e-05,
      "loss": 2.2509,
      "step": 2846
    },
    {
      "epoch": 0.5694,
      "grad_norm": 0.04016062989830971,
      "learning_rate": 9.330295189710152e-05,
      "loss": 2.2527,
      "step": 2847
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.048114191740751266,
      "learning_rate": 9.323329709826294e-05,
      "loss": 2.253,
      "step": 2848
    },
    {
      "epoch": 0.5698,
      "grad_norm": 0.06702069938182831,
      "learning_rate": 9.316364559743314e-05,
      "loss": 2.2517,
      "step": 2849
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.08167112618684769,
      "learning_rate": 9.309399742855942e-05,
      "loss": 2.2512,
      "step": 2850
    },
    {
      "epoch": 0.5702,
      "grad_norm": 0.08915850520133972,
      "learning_rate": 9.302435262558747e-05,
      "loss": 2.2518,
      "step": 2851
    },
    {
      "epoch": 0.5704,
      "grad_norm": 0.08684986084699631,
      "learning_rate": 9.295471122246131e-05,
      "loss": 2.2528,
      "step": 2852
    },
    {
      "epoch": 0.5706,
      "grad_norm": 0.08279461413621902,
      "learning_rate": 9.288507325312335e-05,
      "loss": 2.2534,
      "step": 2853
    },
    {
      "epoch": 0.5708,
      "grad_norm": 0.06291670352220535,
      "learning_rate": 9.281543875151419e-05,
      "loss": 2.2536,
      "step": 2854
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.04114946350455284,
      "learning_rate": 9.274580775157294e-05,
      "loss": 2.2518,
      "step": 2855
    },
    {
      "epoch": 0.5712,
      "grad_norm": 0.04752233996987343,
      "learning_rate": 9.267618028723686e-05,
      "loss": 2.2506,
      "step": 2856
    },
    {
      "epoch": 0.5714,
      "grad_norm": 0.05268106237053871,
      "learning_rate": 9.26065563924415e-05,
      "loss": 2.2478,
      "step": 2857
    },
    {
      "epoch": 0.5716,
      "grad_norm": 0.04582804813981056,
      "learning_rate": 9.253693610112078e-05,
      "loss": 2.2532,
      "step": 2858
    },
    {
      "epoch": 0.5718,
      "grad_norm": 0.0405990369617939,
      "learning_rate": 9.246731944720675e-05,
      "loss": 2.2539,
      "step": 2859
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.05112968012690544,
      "learning_rate": 9.239770646462968e-05,
      "loss": 2.2522,
      "step": 2860
    },
    {
      "epoch": 0.5722,
      "grad_norm": 0.07285936921834946,
      "learning_rate": 9.232809718731814e-05,
      "loss": 2.2505,
      "step": 2861
    },
    {
      "epoch": 0.5724,
      "grad_norm": 0.0842859074473381,
      "learning_rate": 9.225849164919885e-05,
      "loss": 2.2494,
      "step": 2862
    },
    {
      "epoch": 0.5726,
      "grad_norm": 0.078069768846035,
      "learning_rate": 9.218888988419668e-05,
      "loss": 2.2499,
      "step": 2863
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.05296388640999794,
      "learning_rate": 9.211929192623467e-05,
      "loss": 2.2523,
      "step": 2864
    },
    {
      "epoch": 0.573,
      "grad_norm": 0.04027814045548439,
      "learning_rate": 9.204969780923403e-05,
      "loss": 2.2508,
      "step": 2865
    },
    {
      "epoch": 0.5732,
      "grad_norm": 0.05028754100203514,
      "learning_rate": 9.198010756711412e-05,
      "loss": 2.251,
      "step": 2866
    },
    {
      "epoch": 0.5734,
      "grad_norm": 0.05963124334812164,
      "learning_rate": 9.191052123379234e-05,
      "loss": 2.2506,
      "step": 2867
    },
    {
      "epoch": 0.5736,
      "grad_norm": 0.056063227355480194,
      "learning_rate": 9.184093884318425e-05,
      "loss": 2.2511,
      "step": 2868
    },
    {
      "epoch": 0.5738,
      "grad_norm": 0.05009758472442627,
      "learning_rate": 9.177136042920344e-05,
      "loss": 2.252,
      "step": 2869
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.04504982754588127,
      "learning_rate": 9.17017860257616e-05,
      "loss": 2.2522,
      "step": 2870
    },
    {
      "epoch": 0.5742,
      "grad_norm": 0.04051341488957405,
      "learning_rate": 9.163221566676847e-05,
      "loss": 2.2528,
      "step": 2871
    },
    {
      "epoch": 0.5744,
      "grad_norm": 0.0478254035115242,
      "learning_rate": 9.156264938613174e-05,
      "loss": 2.2521,
      "step": 2872
    },
    {
      "epoch": 0.5746,
      "grad_norm": 0.06253599375486374,
      "learning_rate": 9.14930872177572e-05,
      "loss": 2.2523,
      "step": 2873
    },
    {
      "epoch": 0.5748,
      "grad_norm": 0.06693513691425323,
      "learning_rate": 9.142352919554862e-05,
      "loss": 2.2494,
      "step": 2874
    },
    {
      "epoch": 0.575,
      "grad_norm": 0.0652974545955658,
      "learning_rate": 9.135397535340773e-05,
      "loss": 2.2531,
      "step": 2875
    },
    {
      "epoch": 0.5752,
      "grad_norm": 0.0566631555557251,
      "learning_rate": 9.128442572523417e-05,
      "loss": 2.2529,
      "step": 2876
    },
    {
      "epoch": 0.5754,
      "grad_norm": 0.04655500873923302,
      "learning_rate": 9.121488034492569e-05,
      "loss": 2.2526,
      "step": 2877
    },
    {
      "epoch": 0.5756,
      "grad_norm": 0.04098905622959137,
      "learning_rate": 9.114533924637778e-05,
      "loss": 2.2524,
      "step": 2878
    },
    {
      "epoch": 0.5758,
      "grad_norm": 0.05272220820188522,
      "learning_rate": 9.107580246348395e-05,
      "loss": 2.2492,
      "step": 2879
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.0786268338561058,
      "learning_rate": 9.100627003013562e-05,
      "loss": 2.2529,
      "step": 2880
    },
    {
      "epoch": 0.5762,
      "grad_norm": 0.09645208716392517,
      "learning_rate": 9.093674198022201e-05,
      "loss": 2.2496,
      "step": 2881
    },
    {
      "epoch": 0.5764,
      "grad_norm": 0.09298472851514816,
      "learning_rate": 9.086721834763024e-05,
      "loss": 2.2495,
      "step": 2882
    },
    {
      "epoch": 0.5766,
      "grad_norm": 0.07348407804965973,
      "learning_rate": 9.07976991662453e-05,
      "loss": 2.2529,
      "step": 2883
    },
    {
      "epoch": 0.5768,
      "grad_norm": 0.04999276623129845,
      "learning_rate": 9.072818446994999e-05,
      "loss": 2.2497,
      "step": 2884
    },
    {
      "epoch": 0.577,
      "grad_norm": 0.04455426335334778,
      "learning_rate": 9.065867429262496e-05,
      "loss": 2.2529,
      "step": 2885
    },
    {
      "epoch": 0.5772,
      "grad_norm": 0.07627172768115997,
      "learning_rate": 9.058916866814858e-05,
      "loss": 2.2538,
      "step": 2886
    },
    {
      "epoch": 0.5774,
      "grad_norm": 0.11104367673397064,
      "learning_rate": 9.051966763039707e-05,
      "loss": 2.2531,
      "step": 2887
    },
    {
      "epoch": 0.5776,
      "grad_norm": 0.12235650420188904,
      "learning_rate": 9.045017121324438e-05,
      "loss": 2.2513,
      "step": 2888
    },
    {
      "epoch": 0.5778,
      "grad_norm": 0.09842084348201752,
      "learning_rate": 9.038067945056227e-05,
      "loss": 2.2512,
      "step": 2889
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.051533523947000504,
      "learning_rate": 9.03111923762201e-05,
      "loss": 2.2503,
      "step": 2890
    },
    {
      "epoch": 0.5782,
      "grad_norm": 0.05279460549354553,
      "learning_rate": 9.024171002408506e-05,
      "loss": 2.2513,
      "step": 2891
    },
    {
      "epoch": 0.5784,
      "grad_norm": 0.10426756739616394,
      "learning_rate": 9.017223242802204e-05,
      "loss": 2.25,
      "step": 2892
    },
    {
      "epoch": 0.5786,
      "grad_norm": 0.12989862263202667,
      "learning_rate": 9.01027596218935e-05,
      "loss": 2.2523,
      "step": 2893
    },
    {
      "epoch": 0.5788,
      "grad_norm": 0.11287760734558105,
      "learning_rate": 9.003329163955972e-05,
      "loss": 2.2518,
      "step": 2894
    },
    {
      "epoch": 0.579,
      "grad_norm": 0.0674719288945198,
      "learning_rate": 8.99638285148785e-05,
      "loss": 2.2507,
      "step": 2895
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.041034799069166183,
      "learning_rate": 8.989437028170537e-05,
      "loss": 2.2501,
      "step": 2896
    },
    {
      "epoch": 0.5794,
      "grad_norm": 0.07434824854135513,
      "learning_rate": 8.982491697389338e-05,
      "loss": 2.2538,
      "step": 2897
    },
    {
      "epoch": 0.5796,
      "grad_norm": 0.10091511905193329,
      "learning_rate": 8.975546862529328e-05,
      "loss": 2.2534,
      "step": 2898
    },
    {
      "epoch": 0.5798,
      "grad_norm": 0.09889128059148788,
      "learning_rate": 8.96860252697533e-05,
      "loss": 2.253,
      "step": 2899
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.07014238834381104,
      "learning_rate": 8.961658694111929e-05,
      "loss": 2.2517,
      "step": 2900
    },
    {
      "epoch": 0.5802,
      "grad_norm": 0.0416761077940464,
      "learning_rate": 8.954715367323468e-05,
      "loss": 2.2517,
      "step": 2901
    },
    {
      "epoch": 0.5804,
      "grad_norm": 0.05603697896003723,
      "learning_rate": 8.947772549994035e-05,
      "loss": 2.2525,
      "step": 2902
    },
    {
      "epoch": 0.5806,
      "grad_norm": 0.07404505461454391,
      "learning_rate": 8.940830245507483e-05,
      "loss": 2.2515,
      "step": 2903
    },
    {
      "epoch": 0.5808,
      "grad_norm": 0.07105246931314468,
      "learning_rate": 8.933888457247402e-05,
      "loss": 2.2507,
      "step": 2904
    },
    {
      "epoch": 0.581,
      "grad_norm": 0.04920331388711929,
      "learning_rate": 8.926947188597134e-05,
      "loss": 2.2519,
      "step": 2905
    },
    {
      "epoch": 0.5812,
      "grad_norm": 0.04186251759529114,
      "learning_rate": 8.920006442939772e-05,
      "loss": 2.2504,
      "step": 2906
    },
    {
      "epoch": 0.5814,
      "grad_norm": 0.05709998682141304,
      "learning_rate": 8.913066223658151e-05,
      "loss": 2.2528,
      "step": 2907
    },
    {
      "epoch": 0.5816,
      "grad_norm": 0.06668895483016968,
      "learning_rate": 8.906126534134848e-05,
      "loss": 2.2511,
      "step": 2908
    },
    {
      "epoch": 0.5818,
      "grad_norm": 0.06358695030212402,
      "learning_rate": 8.89918737775218e-05,
      "loss": 2.2517,
      "step": 2909
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.046936824917793274,
      "learning_rate": 8.892248757892214e-05,
      "loss": 2.2514,
      "step": 2910
    },
    {
      "epoch": 0.5822,
      "grad_norm": 0.041901540011167526,
      "learning_rate": 8.885310677936746e-05,
      "loss": 2.2512,
      "step": 2911
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.056413449347019196,
      "learning_rate": 8.878373141267311e-05,
      "loss": 2.2509,
      "step": 2912
    },
    {
      "epoch": 0.5826,
      "grad_norm": 0.06542353332042694,
      "learning_rate": 8.871436151265184e-05,
      "loss": 2.2499,
      "step": 2913
    },
    {
      "epoch": 0.5828,
      "grad_norm": 0.061532896012067795,
      "learning_rate": 8.864499711311362e-05,
      "loss": 2.251,
      "step": 2914
    },
    {
      "epoch": 0.583,
      "grad_norm": 0.04939775913953781,
      "learning_rate": 8.857563824786596e-05,
      "loss": 2.2512,
      "step": 2915
    },
    {
      "epoch": 0.5832,
      "grad_norm": 0.040766261518001556,
      "learning_rate": 8.850628495071336e-05,
      "loss": 2.2505,
      "step": 2916
    },
    {
      "epoch": 0.5834,
      "grad_norm": 0.049392957240343094,
      "learning_rate": 8.843693725545786e-05,
      "loss": 2.2493,
      "step": 2917
    },
    {
      "epoch": 0.5836,
      "grad_norm": 0.05661853775382042,
      "learning_rate": 8.836759519589867e-05,
      "loss": 2.251,
      "step": 2918
    },
    {
      "epoch": 0.5838,
      "grad_norm": 0.05370378866791725,
      "learning_rate": 8.829825880583226e-05,
      "loss": 2.2502,
      "step": 2919
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.044526487588882446,
      "learning_rate": 8.822892811905237e-05,
      "loss": 2.2515,
      "step": 2920
    },
    {
      "epoch": 0.5842,
      "grad_norm": 0.04120916873216629,
      "learning_rate": 8.81596031693499e-05,
      "loss": 2.2511,
      "step": 2921
    },
    {
      "epoch": 0.5844,
      "grad_norm": 0.045657142996788025,
      "learning_rate": 8.809028399051302e-05,
      "loss": 2.2516,
      "step": 2922
    },
    {
      "epoch": 0.5846,
      "grad_norm": 0.0481010340154171,
      "learning_rate": 8.802097061632705e-05,
      "loss": 2.2519,
      "step": 2923
    },
    {
      "epoch": 0.5848,
      "grad_norm": 0.045285459607839584,
      "learning_rate": 8.79516630805745e-05,
      "loss": 2.2528,
      "step": 2924
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.04140923172235489,
      "learning_rate": 8.788236141703498e-05,
      "loss": 2.2522,
      "step": 2925
    },
    {
      "epoch": 0.5852,
      "grad_norm": 0.041382137686014175,
      "learning_rate": 8.781306565948528e-05,
      "loss": 2.2511,
      "step": 2926
    },
    {
      "epoch": 0.5854,
      "grad_norm": 0.04144683852791786,
      "learning_rate": 8.774377584169933e-05,
      "loss": 2.2515,
      "step": 2927
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.041284315288066864,
      "learning_rate": 8.767449199744814e-05,
      "loss": 2.2502,
      "step": 2928
    },
    {
      "epoch": 0.5858,
      "grad_norm": 0.04257747530937195,
      "learning_rate": 8.760521416049983e-05,
      "loss": 2.2534,
      "step": 2929
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.045326922088861465,
      "learning_rate": 8.753594236461957e-05,
      "loss": 2.2524,
      "step": 2930
    },
    {
      "epoch": 0.5862,
      "grad_norm": 0.04534221440553665,
      "learning_rate": 8.746667664356956e-05,
      "loss": 2.2513,
      "step": 2931
    },
    {
      "epoch": 0.5864,
      "grad_norm": 0.04524366185069084,
      "learning_rate": 8.739741703110913e-05,
      "loss": 2.2529,
      "step": 2932
    },
    {
      "epoch": 0.5866,
      "grad_norm": 0.04704603552818298,
      "learning_rate": 8.732816356099456e-05,
      "loss": 2.2514,
      "step": 2933
    },
    {
      "epoch": 0.5868,
      "grad_norm": 0.045590367168188095,
      "learning_rate": 8.72589162669791e-05,
      "loss": 2.253,
      "step": 2934
    },
    {
      "epoch": 0.587,
      "grad_norm": 0.042867377400398254,
      "learning_rate": 8.718967518281307e-05,
      "loss": 2.2508,
      "step": 2935
    },
    {
      "epoch": 0.5872,
      "grad_norm": 0.042393025010824203,
      "learning_rate": 8.712044034224374e-05,
      "loss": 2.2512,
      "step": 2936
    },
    {
      "epoch": 0.5874,
      "grad_norm": 0.04526234418153763,
      "learning_rate": 8.705121177901532e-05,
      "loss": 2.2517,
      "step": 2937
    },
    {
      "epoch": 0.5876,
      "grad_norm": 0.049191612750291824,
      "learning_rate": 8.698198952686896e-05,
      "loss": 2.2529,
      "step": 2938
    },
    {
      "epoch": 0.5878,
      "grad_norm": 0.04890531674027443,
      "learning_rate": 8.691277361954279e-05,
      "loss": 2.2538,
      "step": 2939
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.04170852154493332,
      "learning_rate": 8.684356409077176e-05,
      "loss": 2.2504,
      "step": 2940
    },
    {
      "epoch": 0.5882,
      "grad_norm": 0.04759737849235535,
      "learning_rate": 8.677436097428775e-05,
      "loss": 2.2493,
      "step": 2941
    },
    {
      "epoch": 0.5884,
      "grad_norm": 0.07073279470205307,
      "learning_rate": 8.670516430381958e-05,
      "loss": 2.2522,
      "step": 2942
    },
    {
      "epoch": 0.5886,
      "grad_norm": 0.09056682139635086,
      "learning_rate": 8.663597411309279e-05,
      "loss": 2.2533,
      "step": 2943
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.09070293605327606,
      "learning_rate": 8.656679043582986e-05,
      "loss": 2.2549,
      "step": 2944
    },
    {
      "epoch": 0.589,
      "grad_norm": 0.058459486812353134,
      "learning_rate": 8.64976133057501e-05,
      "loss": 2.2492,
      "step": 2945
    },
    {
      "epoch": 0.5892,
      "grad_norm": 0.04300703853368759,
      "learning_rate": 8.642844275656957e-05,
      "loss": 2.253,
      "step": 2946
    },
    {
      "epoch": 0.5894,
      "grad_norm": 0.05868382751941681,
      "learning_rate": 8.635927882200116e-05,
      "loss": 2.2504,
      "step": 2947
    },
    {
      "epoch": 0.5896,
      "grad_norm": 0.05341348424553871,
      "learning_rate": 8.629012153575458e-05,
      "loss": 2.2506,
      "step": 2948
    },
    {
      "epoch": 0.5898,
      "grad_norm": 0.042828671634197235,
      "learning_rate": 8.62209709315362e-05,
      "loss": 2.2508,
      "step": 2949
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.0423763282597065,
      "learning_rate": 8.615182704304918e-05,
      "loss": 2.248,
      "step": 2950
    },
    {
      "epoch": 0.5902,
      "grad_norm": 0.04478380084037781,
      "learning_rate": 8.608268990399349e-05,
      "loss": 2.2514,
      "step": 2951
    },
    {
      "epoch": 0.5904,
      "grad_norm": 0.046469636261463165,
      "learning_rate": 8.601355954806561e-05,
      "loss": 2.251,
      "step": 2952
    },
    {
      "epoch": 0.5906,
      "grad_norm": 0.04268474876880646,
      "learning_rate": 8.594443600895892e-05,
      "loss": 2.2512,
      "step": 2953
    },
    {
      "epoch": 0.5908,
      "grad_norm": 0.041780997067689896,
      "learning_rate": 8.587531932036335e-05,
      "loss": 2.2508,
      "step": 2954
    },
    {
      "epoch": 0.591,
      "grad_norm": 0.04567885771393776,
      "learning_rate": 8.580620951596557e-05,
      "loss": 2.251,
      "step": 2955
    },
    {
      "epoch": 0.5912,
      "grad_norm": 0.047699835151433945,
      "learning_rate": 8.573710662944885e-05,
      "loss": 2.2513,
      "step": 2956
    },
    {
      "epoch": 0.5914,
      "grad_norm": 0.04823761060833931,
      "learning_rate": 8.566801069449305e-05,
      "loss": 2.2511,
      "step": 2957
    },
    {
      "epoch": 0.5916,
      "grad_norm": 0.05217977613210678,
      "learning_rate": 8.559892174477479e-05,
      "loss": 2.2529,
      "step": 2958
    },
    {
      "epoch": 0.5918,
      "grad_norm": 0.060305070132017136,
      "learning_rate": 8.55298398139671e-05,
      "loss": 2.2522,
      "step": 2959
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.05969778820872307,
      "learning_rate": 8.546076493573972e-05,
      "loss": 2.2504,
      "step": 2960
    },
    {
      "epoch": 0.5922,
      "grad_norm": 0.054478298872709274,
      "learning_rate": 8.539169714375885e-05,
      "loss": 2.2499,
      "step": 2961
    },
    {
      "epoch": 0.5924,
      "grad_norm": 0.04247989133000374,
      "learning_rate": 8.532263647168735e-05,
      "loss": 2.252,
      "step": 2962
    },
    {
      "epoch": 0.5926,
      "grad_norm": 0.049557287245988846,
      "learning_rate": 8.525358295318454e-05,
      "loss": 2.2492,
      "step": 2963
    },
    {
      "epoch": 0.5928,
      "grad_norm": 0.07774637639522552,
      "learning_rate": 8.51845366219062e-05,
      "loss": 2.2483,
      "step": 2964
    },
    {
      "epoch": 0.593,
      "grad_norm": 0.09343428164720535,
      "learning_rate": 8.511549751150479e-05,
      "loss": 2.2508,
      "step": 2965
    },
    {
      "epoch": 0.5932,
      "grad_norm": 0.08043935149908066,
      "learning_rate": 8.504646565562906e-05,
      "loss": 2.2486,
      "step": 2966
    },
    {
      "epoch": 0.5934,
      "grad_norm": 0.048728518187999725,
      "learning_rate": 8.497744108792429e-05,
      "loss": 2.2467,
      "step": 2967
    },
    {
      "epoch": 0.5936,
      "grad_norm": 0.0488530769944191,
      "learning_rate": 8.490842384203225e-05,
      "loss": 2.2477,
      "step": 2968
    },
    {
      "epoch": 0.5938,
      "grad_norm": 0.07723644375801086,
      "learning_rate": 8.483941395159114e-05,
      "loss": 2.2507,
      "step": 2969
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.08616318553686142,
      "learning_rate": 8.477041145023546e-05,
      "loss": 2.2552,
      "step": 2970
    },
    {
      "epoch": 0.5942,
      "grad_norm": 0.07091115415096283,
      "learning_rate": 8.47014163715962e-05,
      "loss": 2.2558,
      "step": 2971
    },
    {
      "epoch": 0.5944,
      "grad_norm": 0.04435138404369354,
      "learning_rate": 8.463242874930079e-05,
      "loss": 2.2488,
      "step": 2972
    },
    {
      "epoch": 0.5946,
      "grad_norm": 0.05390539392828941,
      "learning_rate": 8.456344861697289e-05,
      "loss": 2.2499,
      "step": 2973
    },
    {
      "epoch": 0.5948,
      "grad_norm": 0.07301252335309982,
      "learning_rate": 8.449447600823262e-05,
      "loss": 2.2526,
      "step": 2974
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.06655154377222061,
      "learning_rate": 8.442551095669639e-05,
      "loss": 2.2495,
      "step": 2975
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.05842668563127518,
      "learning_rate": 8.435655349597689e-05,
      "loss": 2.2463,
      "step": 2976
    },
    {
      "epoch": 0.5954,
      "grad_norm": 0.05052134022116661,
      "learning_rate": 8.428760365968326e-05,
      "loss": 2.2493,
      "step": 2977
    },
    {
      "epoch": 0.5956,
      "grad_norm": 0.04401688650250435,
      "learning_rate": 8.421866148142066e-05,
      "loss": 2.251,
      "step": 2978
    },
    {
      "epoch": 0.5958,
      "grad_norm": 0.04375285282731056,
      "learning_rate": 8.414972699479075e-05,
      "loss": 2.2528,
      "step": 2979
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.04975907504558563,
      "learning_rate": 8.408080023339133e-05,
      "loss": 2.2512,
      "step": 2980
    },
    {
      "epoch": 0.5962,
      "grad_norm": 0.04754422977566719,
      "learning_rate": 8.401188123081653e-05,
      "loss": 2.251,
      "step": 2981
    },
    {
      "epoch": 0.5964,
      "grad_norm": 0.04244949296116829,
      "learning_rate": 8.394297002065658e-05,
      "loss": 2.2531,
      "step": 2982
    },
    {
      "epoch": 0.5966,
      "grad_norm": 0.04823329672217369,
      "learning_rate": 8.387406663649795e-05,
      "loss": 2.2506,
      "step": 2983
    },
    {
      "epoch": 0.5968,
      "grad_norm": 0.052687227725982666,
      "learning_rate": 8.380517111192337e-05,
      "loss": 2.2498,
      "step": 2984
    },
    {
      "epoch": 0.597,
      "grad_norm": 0.05638549476861954,
      "learning_rate": 8.373628348051165e-05,
      "loss": 2.2517,
      "step": 2985
    },
    {
      "epoch": 0.5972,
      "grad_norm": 0.061740562319755554,
      "learning_rate": 8.366740377583781e-05,
      "loss": 2.2504,
      "step": 2986
    },
    {
      "epoch": 0.5974,
      "grad_norm": 0.052950408309698105,
      "learning_rate": 8.359853203147291e-05,
      "loss": 2.2496,
      "step": 2987
    },
    {
      "epoch": 0.5976,
      "grad_norm": 0.04283677786588669,
      "learning_rate": 8.352966828098428e-05,
      "loss": 2.2526,
      "step": 2988
    },
    {
      "epoch": 0.5978,
      "grad_norm": 0.04306180775165558,
      "learning_rate": 8.346081255793525e-05,
      "loss": 2.2526,
      "step": 2989
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.04234859719872475,
      "learning_rate": 8.339196489588523e-05,
      "loss": 2.2511,
      "step": 2990
    },
    {
      "epoch": 0.5982,
      "grad_norm": 0.042372506111860275,
      "learning_rate": 8.332312532838978e-05,
      "loss": 2.252,
      "step": 2991
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.04753417149186134,
      "learning_rate": 8.325429388900046e-05,
      "loss": 2.2529,
      "step": 2992
    },
    {
      "epoch": 0.5986,
      "grad_norm": 0.051115550100803375,
      "learning_rate": 8.318547061126485e-05,
      "loss": 2.2518,
      "step": 2993
    },
    {
      "epoch": 0.5988,
      "grad_norm": 0.05116132274270058,
      "learning_rate": 8.311665552872662e-05,
      "loss": 2.251,
      "step": 2994
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.04297903552651405,
      "learning_rate": 8.30478486749254e-05,
      "loss": 2.2507,
      "step": 2995
    },
    {
      "epoch": 0.5992,
      "grad_norm": 0.04933107644319534,
      "learning_rate": 8.297905008339677e-05,
      "loss": 2.2515,
      "step": 2996
    },
    {
      "epoch": 0.5994,
      "grad_norm": 0.0709848627448082,
      "learning_rate": 8.291025978767235e-05,
      "loss": 2.2505,
      "step": 2997
    },
    {
      "epoch": 0.5996,
      "grad_norm": 0.0744006484746933,
      "learning_rate": 8.284147782127971e-05,
      "loss": 2.2522,
      "step": 2998
    },
    {
      "epoch": 0.5998,
      "grad_norm": 0.06430967152118683,
      "learning_rate": 8.277270421774235e-05,
      "loss": 2.2538,
      "step": 2999
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.051336679607629776,
      "learning_rate": 8.270393901057964e-05,
      "loss": 2.2556,
      "step": 3000
    },
    {
      "epoch": 0.6002,
      "grad_norm": 0.04362688586115837,
      "learning_rate": 8.263518223330697e-05,
      "loss": 2.2562,
      "step": 3001
    },
    {
      "epoch": 0.6004,
      "grad_norm": 0.04798604175448418,
      "learning_rate": 8.256643391943551e-05,
      "loss": 2.2537,
      "step": 3002
    },
    {
      "epoch": 0.6006,
      "grad_norm": 0.06502269953489304,
      "learning_rate": 8.249769410247239e-05,
      "loss": 2.2513,
      "step": 3003
    },
    {
      "epoch": 0.6008,
      "grad_norm": 0.08472689986228943,
      "learning_rate": 8.242896281592057e-05,
      "loss": 2.2508,
      "step": 3004
    },
    {
      "epoch": 0.601,
      "grad_norm": 0.09030095487833023,
      "learning_rate": 8.236024009327879e-05,
      "loss": 2.2513,
      "step": 3005
    },
    {
      "epoch": 0.6012,
      "grad_norm": 0.07395274937152863,
      "learning_rate": 8.229152596804168e-05,
      "loss": 2.2524,
      "step": 3006
    },
    {
      "epoch": 0.6014,
      "grad_norm": 0.047006502747535706,
      "learning_rate": 8.222282047369971e-05,
      "loss": 2.2519,
      "step": 3007
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.04778718203306198,
      "learning_rate": 8.215412364373907e-05,
      "loss": 2.2516,
      "step": 3008
    },
    {
      "epoch": 0.6018,
      "grad_norm": 0.06872085481882095,
      "learning_rate": 8.208543551164178e-05,
      "loss": 2.2511,
      "step": 3009
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.07630466669797897,
      "learning_rate": 8.201675611088558e-05,
      "loss": 2.2508,
      "step": 3010
    },
    {
      "epoch": 0.6022,
      "grad_norm": 0.06987916678190231,
      "learning_rate": 8.194808547494401e-05,
      "loss": 2.2496,
      "step": 3011
    },
    {
      "epoch": 0.6024,
      "grad_norm": 0.0494370274245739,
      "learning_rate": 8.187942363728625e-05,
      "loss": 2.2508,
      "step": 3012
    },
    {
      "epoch": 0.6026,
      "grad_norm": 0.042740367352962494,
      "learning_rate": 8.181077063137733e-05,
      "loss": 2.2513,
      "step": 3013
    },
    {
      "epoch": 0.6028,
      "grad_norm": 0.053303271532058716,
      "learning_rate": 8.174212649067781e-05,
      "loss": 2.2508,
      "step": 3014
    },
    {
      "epoch": 0.603,
      "grad_norm": 0.0657331719994545,
      "learning_rate": 8.167349124864405e-05,
      "loss": 2.2504,
      "step": 3015
    },
    {
      "epoch": 0.6032,
      "grad_norm": 0.06830447912216187,
      "learning_rate": 8.160486493872798e-05,
      "loss": 2.2502,
      "step": 3016
    },
    {
      "epoch": 0.6034,
      "grad_norm": 0.05108479782938957,
      "learning_rate": 8.153624759437732e-05,
      "loss": 2.2492,
      "step": 3017
    },
    {
      "epoch": 0.6036,
      "grad_norm": 0.04275594279170036,
      "learning_rate": 8.146763924903527e-05,
      "loss": 2.2488,
      "step": 3018
    },
    {
      "epoch": 0.6038,
      "grad_norm": 0.05909479781985283,
      "learning_rate": 8.139903993614068e-05,
      "loss": 2.2506,
      "step": 3019
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.07084869593381882,
      "learning_rate": 8.133044968912811e-05,
      "loss": 2.2508,
      "step": 3020
    },
    {
      "epoch": 0.6042,
      "grad_norm": 0.06279237568378448,
      "learning_rate": 8.126186854142752e-05,
      "loss": 2.2502,
      "step": 3021
    },
    {
      "epoch": 0.6044,
      "grad_norm": 0.04518973082304001,
      "learning_rate": 8.119329652646463e-05,
      "loss": 2.2531,
      "step": 3022
    },
    {
      "epoch": 0.6046,
      "grad_norm": 0.046796876937150955,
      "learning_rate": 8.11247336776605e-05,
      "loss": 2.252,
      "step": 3023
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.06360036879777908,
      "learning_rate": 8.105618002843189e-05,
      "loss": 2.2505,
      "step": 3024
    },
    {
      "epoch": 0.605,
      "grad_norm": 0.06904488056898117,
      "learning_rate": 8.0987635612191e-05,
      "loss": 2.2484,
      "step": 3025
    },
    {
      "epoch": 0.6052,
      "grad_norm": 0.0638999342918396,
      "learning_rate": 8.091910046234552e-05,
      "loss": 2.2508,
      "step": 3026
    },
    {
      "epoch": 0.6054,
      "grad_norm": 0.047388240694999695,
      "learning_rate": 8.085057461229872e-05,
      "loss": 2.2507,
      "step": 3027
    },
    {
      "epoch": 0.6056,
      "grad_norm": 0.04331745579838753,
      "learning_rate": 8.078205809544917e-05,
      "loss": 2.2513,
      "step": 3028
    },
    {
      "epoch": 0.6058,
      "grad_norm": 0.05352628603577614,
      "learning_rate": 8.071355094519109e-05,
      "loss": 2.2516,
      "step": 3029
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.054953012615442276,
      "learning_rate": 8.064505319491398e-05,
      "loss": 2.2493,
      "step": 3030
    },
    {
      "epoch": 0.6062,
      "grad_norm": 0.04688924923539162,
      "learning_rate": 8.057656487800282e-05,
      "loss": 2.2493,
      "step": 3031
    },
    {
      "epoch": 0.6064,
      "grad_norm": 0.043490417301654816,
      "learning_rate": 8.050808602783795e-05,
      "loss": 2.2496,
      "step": 3032
    },
    {
      "epoch": 0.6066,
      "grad_norm": 0.05730966106057167,
      "learning_rate": 8.04396166777952e-05,
      "loss": 2.2507,
      "step": 3033
    },
    {
      "epoch": 0.6068,
      "grad_norm": 0.07103615254163742,
      "learning_rate": 8.037115686124564e-05,
      "loss": 2.2507,
      "step": 3034
    },
    {
      "epoch": 0.607,
      "grad_norm": 0.07461082935333252,
      "learning_rate": 8.030270661155574e-05,
      "loss": 2.2526,
      "step": 3035
    },
    {
      "epoch": 0.6072,
      "grad_norm": 0.0607343427836895,
      "learning_rate": 8.023426596208739e-05,
      "loss": 2.2523,
      "step": 3036
    },
    {
      "epoch": 0.6074,
      "grad_norm": 0.04248671233654022,
      "learning_rate": 8.016583494619769e-05,
      "loss": 2.2518,
      "step": 3037
    },
    {
      "epoch": 0.6076,
      "grad_norm": 0.054104723036289215,
      "learning_rate": 8.009741359723906e-05,
      "loss": 2.2526,
      "step": 3038
    },
    {
      "epoch": 0.6078,
      "grad_norm": 0.07199501991271973,
      "learning_rate": 8.002900194855932e-05,
      "loss": 2.2478,
      "step": 3039
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.07584595680236816,
      "learning_rate": 7.996060003350139e-05,
      "loss": 2.2513,
      "step": 3040
    },
    {
      "epoch": 0.6082,
      "grad_norm": 0.06834182143211365,
      "learning_rate": 7.989220788540355e-05,
      "loss": 2.2521,
      "step": 3041
    },
    {
      "epoch": 0.6084,
      "grad_norm": 0.05003140866756439,
      "learning_rate": 7.982382553759931e-05,
      "loss": 2.2499,
      "step": 3042
    },
    {
      "epoch": 0.6086,
      "grad_norm": 0.04281144216656685,
      "learning_rate": 7.975545302341743e-05,
      "loss": 2.2515,
      "step": 3043
    },
    {
      "epoch": 0.6088,
      "grad_norm": 0.05624116212129593,
      "learning_rate": 7.96870903761818e-05,
      "loss": 2.2504,
      "step": 3044
    },
    {
      "epoch": 0.609,
      "grad_norm": 0.07195112109184265,
      "learning_rate": 7.961873762921153e-05,
      "loss": 2.2506,
      "step": 3045
    },
    {
      "epoch": 0.6092,
      "grad_norm": 0.07474033534526825,
      "learning_rate": 7.955039481582097e-05,
      "loss": 2.25,
      "step": 3046
    },
    {
      "epoch": 0.6094,
      "grad_norm": 0.05770516023039818,
      "learning_rate": 7.948206196931954e-05,
      "loss": 2.248,
      "step": 3047
    },
    {
      "epoch": 0.6096,
      "grad_norm": 0.0429660864174366,
      "learning_rate": 7.941373912301189e-05,
      "loss": 2.2481,
      "step": 3048
    },
    {
      "epoch": 0.6098,
      "grad_norm": 0.050401341170072556,
      "learning_rate": 7.934542631019768e-05,
      "loss": 2.2513,
      "step": 3049
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.06822075694799423,
      "learning_rate": 7.927712356417176e-05,
      "loss": 2.25,
      "step": 3050
    },
    {
      "epoch": 0.6102,
      "grad_norm": 0.08145581930875778,
      "learning_rate": 7.920883091822408e-05,
      "loss": 2.2505,
      "step": 3051
    },
    {
      "epoch": 0.6104,
      "grad_norm": 0.07435046881437302,
      "learning_rate": 7.914054840563963e-05,
      "loss": 2.249,
      "step": 3052
    },
    {
      "epoch": 0.6106,
      "grad_norm": 0.05960925295948982,
      "learning_rate": 7.907227605969849e-05,
      "loss": 2.251,
      "step": 3053
    },
    {
      "epoch": 0.6108,
      "grad_norm": 0.044836241751909256,
      "learning_rate": 7.900401391367576e-05,
      "loss": 2.249,
      "step": 3054
    },
    {
      "epoch": 0.611,
      "grad_norm": 0.044249046593904495,
      "learning_rate": 7.893576200084159e-05,
      "loss": 2.2515,
      "step": 3055
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.053624074906110764,
      "learning_rate": 7.886752035446114e-05,
      "loss": 2.2501,
      "step": 3056
    },
    {
      "epoch": 0.6114,
      "grad_norm": 0.053928639739751816,
      "learning_rate": 7.879928900779456e-05,
      "loss": 2.2505,
      "step": 3057
    },
    {
      "epoch": 0.6116,
      "grad_norm": 0.04681616276502609,
      "learning_rate": 7.873106799409695e-05,
      "loss": 2.2523,
      "step": 3058
    },
    {
      "epoch": 0.6118,
      "grad_norm": 0.04274589568376541,
      "learning_rate": 7.866285734661841e-05,
      "loss": 2.2504,
      "step": 3059
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.043656378984451294,
      "learning_rate": 7.8594657098604e-05,
      "loss": 2.2495,
      "step": 3060
    },
    {
      "epoch": 0.6122,
      "grad_norm": 0.0476871058344841,
      "learning_rate": 7.852646728329368e-05,
      "loss": 2.2509,
      "step": 3061
    },
    {
      "epoch": 0.6124,
      "grad_norm": 0.05273299664258957,
      "learning_rate": 7.845828793392236e-05,
      "loss": 2.2514,
      "step": 3062
    },
    {
      "epoch": 0.6126,
      "grad_norm": 0.0541733056306839,
      "learning_rate": 7.83901190837198e-05,
      "loss": 2.2519,
      "step": 3063
    },
    {
      "epoch": 0.6128,
      "grad_norm": 0.0438724085688591,
      "learning_rate": 7.832196076591067e-05,
      "loss": 2.2484,
      "step": 3064
    },
    {
      "epoch": 0.613,
      "grad_norm": 0.04677257686853409,
      "learning_rate": 7.825381301371452e-05,
      "loss": 2.251,
      "step": 3065
    },
    {
      "epoch": 0.6132,
      "grad_norm": 0.06204405426979065,
      "learning_rate": 7.818567586034577e-05,
      "loss": 2.2504,
      "step": 3066
    },
    {
      "epoch": 0.6134,
      "grad_norm": 0.0715985894203186,
      "learning_rate": 7.811754933901358e-05,
      "loss": 2.2514,
      "step": 3067
    },
    {
      "epoch": 0.6136,
      "grad_norm": 0.08004756271839142,
      "learning_rate": 7.804943348292197e-05,
      "loss": 2.2497,
      "step": 3068
    },
    {
      "epoch": 0.6138,
      "grad_norm": 0.0840577483177185,
      "learning_rate": 7.798132832526986e-05,
      "loss": 2.2513,
      "step": 3069
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.07353750616312027,
      "learning_rate": 7.791323389925084e-05,
      "loss": 2.2516,
      "step": 3070
    },
    {
      "epoch": 0.6142,
      "grad_norm": 0.04536302387714386,
      "learning_rate": 7.784515023805328e-05,
      "loss": 2.2496,
      "step": 3071
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.05476371571421623,
      "learning_rate": 7.777707737486037e-05,
      "loss": 2.2485,
      "step": 3072
    },
    {
      "epoch": 0.6146,
      "grad_norm": 0.0802014023065567,
      "learning_rate": 7.770901534284995e-05,
      "loss": 2.2517,
      "step": 3073
    },
    {
      "epoch": 0.6148,
      "grad_norm": 0.08346652239561081,
      "learning_rate": 7.76409641751947e-05,
      "loss": 2.2509,
      "step": 3074
    },
    {
      "epoch": 0.615,
      "grad_norm": 0.07306312024593353,
      "learning_rate": 7.75729239050619e-05,
      "loss": 2.2499,
      "step": 3075
    },
    {
      "epoch": 0.6152,
      "grad_norm": 0.05258280783891678,
      "learning_rate": 7.750489456561352e-05,
      "loss": 2.2507,
      "step": 3076
    },
    {
      "epoch": 0.6154,
      "grad_norm": 0.043146055191755295,
      "learning_rate": 7.743687619000626e-05,
      "loss": 2.2504,
      "step": 3077
    },
    {
      "epoch": 0.6156,
      "grad_norm": 0.05423370376229286,
      "learning_rate": 7.736886881139142e-05,
      "loss": 2.2503,
      "step": 3078
    },
    {
      "epoch": 0.6158,
      "grad_norm": 0.053357064723968506,
      "learning_rate": 7.730087246291502e-05,
      "loss": 2.2506,
      "step": 3079
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.04474783316254616,
      "learning_rate": 7.723288717771761e-05,
      "loss": 2.2482,
      "step": 3080
    },
    {
      "epoch": 0.6162,
      "grad_norm": 0.04508621245622635,
      "learning_rate": 7.716491298893442e-05,
      "loss": 2.2536,
      "step": 3081
    },
    {
      "epoch": 0.6164,
      "grad_norm": 0.05410914123058319,
      "learning_rate": 7.709694992969526e-05,
      "loss": 2.2509,
      "step": 3082
    },
    {
      "epoch": 0.6166,
      "grad_norm": 0.06470200419425964,
      "learning_rate": 7.702899803312443e-05,
      "loss": 2.2487,
      "step": 3083
    },
    {
      "epoch": 0.6168,
      "grad_norm": 0.06821881234645844,
      "learning_rate": 7.696105733234098e-05,
      "loss": 2.2516,
      "step": 3084
    },
    {
      "epoch": 0.617,
      "grad_norm": 0.058366384357213974,
      "learning_rate": 7.689312786045823e-05,
      "loss": 2.2491,
      "step": 3085
    },
    {
      "epoch": 0.6172,
      "grad_norm": 0.04320967197418213,
      "learning_rate": 7.682520965058428e-05,
      "loss": 2.2491,
      "step": 3086
    },
    {
      "epoch": 0.6174,
      "grad_norm": 0.05360645800828934,
      "learning_rate": 7.67573027358216e-05,
      "loss": 2.2491,
      "step": 3087
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.07003532350063324,
      "learning_rate": 7.668940714926725e-05,
      "loss": 2.2503,
      "step": 3088
    },
    {
      "epoch": 0.6178,
      "grad_norm": 0.06326913833618164,
      "learning_rate": 7.662152292401264e-05,
      "loss": 2.25,
      "step": 3089
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.04644004628062248,
      "learning_rate": 7.655365009314374e-05,
      "loss": 2.2493,
      "step": 3090
    },
    {
      "epoch": 0.6182,
      "grad_norm": 0.044931117445230484,
      "learning_rate": 7.6485788689741e-05,
      "loss": 2.2497,
      "step": 3091
    },
    {
      "epoch": 0.6184,
      "grad_norm": 0.05099665746092796,
      "learning_rate": 7.641793874687918e-05,
      "loss": 2.2511,
      "step": 3092
    },
    {
      "epoch": 0.6186,
      "grad_norm": 0.05082391947507858,
      "learning_rate": 7.635010029762756e-05,
      "loss": 2.2487,
      "step": 3093
    },
    {
      "epoch": 0.6188,
      "grad_norm": 0.04578569158911705,
      "learning_rate": 7.628227337504972e-05,
      "loss": 2.2516,
      "step": 3094
    },
    {
      "epoch": 0.619,
      "grad_norm": 0.0429779589176178,
      "learning_rate": 7.621445801220371e-05,
      "loss": 2.2498,
      "step": 3095
    },
    {
      "epoch": 0.6192,
      "grad_norm": 0.046841345727443695,
      "learning_rate": 7.614665424214193e-05,
      "loss": 2.2482,
      "step": 3096
    },
    {
      "epoch": 0.6194,
      "grad_norm": 0.05387512966990471,
      "learning_rate": 7.607886209791107e-05,
      "loss": 2.2494,
      "step": 3097
    },
    {
      "epoch": 0.6196,
      "grad_norm": 0.05269436165690422,
      "learning_rate": 7.601108161255226e-05,
      "loss": 2.2492,
      "step": 3098
    },
    {
      "epoch": 0.6198,
      "grad_norm": 0.04480529576539993,
      "learning_rate": 7.594331281910082e-05,
      "loss": 2.2492,
      "step": 3099
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.04326268285512924,
      "learning_rate": 7.587555575058649e-05,
      "loss": 2.2497,
      "step": 3100
    },
    {
      "epoch": 0.6202,
      "grad_norm": 0.04828430339694023,
      "learning_rate": 7.580781044003324e-05,
      "loss": 2.2479,
      "step": 3101
    },
    {
      "epoch": 0.6204,
      "grad_norm": 0.05086647346615791,
      "learning_rate": 7.574007692045928e-05,
      "loss": 2.2507,
      "step": 3102
    },
    {
      "epoch": 0.6206,
      "grad_norm": 0.0488625131547451,
      "learning_rate": 7.567235522487712e-05,
      "loss": 2.2508,
      "step": 3103
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.04372460022568703,
      "learning_rate": 7.560464538629344e-05,
      "loss": 2.2508,
      "step": 3104
    },
    {
      "epoch": 0.621,
      "grad_norm": 0.047008831053972244,
      "learning_rate": 7.553694743770928e-05,
      "loss": 2.249,
      "step": 3105
    },
    {
      "epoch": 0.6212,
      "grad_norm": 0.07408326864242554,
      "learning_rate": 7.546926141211974e-05,
      "loss": 2.2495,
      "step": 3106
    },
    {
      "epoch": 0.6214,
      "grad_norm": 0.09355244785547256,
      "learning_rate": 7.54015873425142e-05,
      "loss": 2.2508,
      "step": 3107
    },
    {
      "epoch": 0.6216,
      "grad_norm": 0.0821254700422287,
      "learning_rate": 7.533392526187617e-05,
      "loss": 2.2486,
      "step": 3108
    },
    {
      "epoch": 0.6218,
      "grad_norm": 0.05177052319049835,
      "learning_rate": 7.526627520318329e-05,
      "loss": 2.2489,
      "step": 3109
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.045630861073732376,
      "learning_rate": 7.519863719940748e-05,
      "loss": 2.2497,
      "step": 3110
    },
    {
      "epoch": 0.6222,
      "grad_norm": 0.0655323714017868,
      "learning_rate": 7.513101128351454e-05,
      "loss": 2.2504,
      "step": 3111
    },
    {
      "epoch": 0.6224,
      "grad_norm": 0.06931561231613159,
      "learning_rate": 7.506339748846461e-05,
      "loss": 2.2485,
      "step": 3112
    },
    {
      "epoch": 0.6226,
      "grad_norm": 0.05760861933231354,
      "learning_rate": 7.499579584721179e-05,
      "loss": 2.2506,
      "step": 3113
    },
    {
      "epoch": 0.6228,
      "grad_norm": 0.04297424852848053,
      "learning_rate": 7.492820639270434e-05,
      "loss": 2.246,
      "step": 3114
    },
    {
      "epoch": 0.623,
      "grad_norm": 0.0551571324467659,
      "learning_rate": 7.486062915788452e-05,
      "loss": 2.2494,
      "step": 3115
    },
    {
      "epoch": 0.6232,
      "grad_norm": 0.07354260236024857,
      "learning_rate": 7.479306417568864e-05,
      "loss": 2.2493,
      "step": 3116
    },
    {
      "epoch": 0.6234,
      "grad_norm": 0.07523467391729355,
      "learning_rate": 7.472551147904708e-05,
      "loss": 2.2486,
      "step": 3117
    },
    {
      "epoch": 0.6236,
      "grad_norm": 0.06053363159298897,
      "learning_rate": 7.465797110088417e-05,
      "loss": 2.2517,
      "step": 3118
    },
    {
      "epoch": 0.6238,
      "grad_norm": 0.04267670214176178,
      "learning_rate": 7.459044307411832e-05,
      "loss": 2.2495,
      "step": 3119
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.06032056361436844,
      "learning_rate": 7.45229274316618e-05,
      "loss": 2.2477,
      "step": 3120
    },
    {
      "epoch": 0.6242,
      "grad_norm": 0.07602176815271378,
      "learning_rate": 7.445542420642097e-05,
      "loss": 2.2483,
      "step": 3121
    },
    {
      "epoch": 0.6244,
      "grad_norm": 0.0635555237531662,
      "learning_rate": 7.438793343129605e-05,
      "loss": 2.2486,
      "step": 3122
    },
    {
      "epoch": 0.6246,
      "grad_norm": 0.04296661168336868,
      "learning_rate": 7.432045513918122e-05,
      "loss": 2.2503,
      "step": 3123
    },
    {
      "epoch": 0.6248,
      "grad_norm": 0.06695044040679932,
      "learning_rate": 7.425298936296463e-05,
      "loss": 2.2528,
      "step": 3124
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.09039997309446335,
      "learning_rate": 7.418553613552825e-05,
      "loss": 2.251,
      "step": 3125
    },
    {
      "epoch": 0.6252,
      "grad_norm": 0.0801958218216896,
      "learning_rate": 7.411809548974792e-05,
      "loss": 2.2504,
      "step": 3126
    },
    {
      "epoch": 0.6254,
      "grad_norm": 0.04458928480744362,
      "learning_rate": 7.405066745849346e-05,
      "loss": 2.2505,
      "step": 3127
    },
    {
      "epoch": 0.6256,
      "grad_norm": 0.07316660135984421,
      "learning_rate": 7.398325207462846e-05,
      "loss": 2.2503,
      "step": 3128
    },
    {
      "epoch": 0.6258,
      "grad_norm": 0.11486947536468506,
      "learning_rate": 7.391584937101033e-05,
      "loss": 2.2474,
      "step": 3129
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.12063782662153244,
      "learning_rate": 7.384845938049031e-05,
      "loss": 2.2511,
      "step": 3130
    },
    {
      "epoch": 0.6262,
      "grad_norm": 0.09432776272296906,
      "learning_rate": 7.378108213591355e-05,
      "loss": 2.2512,
      "step": 3131
    },
    {
      "epoch": 0.6264,
      "grad_norm": 0.049421850591897964,
      "learning_rate": 7.37137176701188e-05,
      "loss": 2.2541,
      "step": 3132
    },
    {
      "epoch": 0.6266,
      "grad_norm": 0.07838404923677444,
      "learning_rate": 7.364636601593875e-05,
      "loss": 2.2537,
      "step": 3133
    },
    {
      "epoch": 0.6268,
      "grad_norm": 0.11114689707756042,
      "learning_rate": 7.357902720619976e-05,
      "loss": 2.2537,
      "step": 3134
    },
    {
      "epoch": 0.627,
      "grad_norm": 0.09753752499818802,
      "learning_rate": 7.351170127372191e-05,
      "loss": 2.2534,
      "step": 3135
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.053893350064754486,
      "learning_rate": 7.344438825131911e-05,
      "loss": 2.253,
      "step": 3136
    },
    {
      "epoch": 0.6274,
      "grad_norm": 0.05685865506529808,
      "learning_rate": 7.33770881717989e-05,
      "loss": 2.2523,
      "step": 3137
    },
    {
      "epoch": 0.6276,
      "grad_norm": 0.08237016946077347,
      "learning_rate": 7.330980106796246e-05,
      "loss": 2.2512,
      "step": 3138
    },
    {
      "epoch": 0.6278,
      "grad_norm": 0.0656891018152237,
      "learning_rate": 7.324252697260474e-05,
      "loss": 2.251,
      "step": 3139
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.04775739833712578,
      "learning_rate": 7.317526591851433e-05,
      "loss": 2.2507,
      "step": 3140
    },
    {
      "epoch": 0.6282,
      "grad_norm": 0.04495927691459656,
      "learning_rate": 7.310801793847344e-05,
      "loss": 2.2512,
      "step": 3141
    },
    {
      "epoch": 0.6284,
      "grad_norm": 0.046536508947610855,
      "learning_rate": 7.30407830652579e-05,
      "loss": 2.2508,
      "step": 3142
    },
    {
      "epoch": 0.6286,
      "grad_norm": 0.04596540331840515,
      "learning_rate": 7.297356133163721e-05,
      "loss": 2.2504,
      "step": 3143
    },
    {
      "epoch": 0.6288,
      "grad_norm": 0.04714493826031685,
      "learning_rate": 7.290635277037442e-05,
      "loss": 2.2505,
      "step": 3144
    },
    {
      "epoch": 0.629,
      "grad_norm": 0.04578232392668724,
      "learning_rate": 7.283915741422612e-05,
      "loss": 2.251,
      "step": 3145
    },
    {
      "epoch": 0.6292,
      "grad_norm": 0.044819340109825134,
      "learning_rate": 7.277197529594257e-05,
      "loss": 2.2497,
      "step": 3146
    },
    {
      "epoch": 0.6294,
      "grad_norm": 0.06300388276576996,
      "learning_rate": 7.270480644826749e-05,
      "loss": 2.2515,
      "step": 3147
    },
    {
      "epoch": 0.6296,
      "grad_norm": 0.07750938832759857,
      "learning_rate": 7.263765090393817e-05,
      "loss": 2.2516,
      "step": 3148
    },
    {
      "epoch": 0.6298,
      "grad_norm": 0.06960883736610413,
      "learning_rate": 7.257050869568535e-05,
      "loss": 2.2515,
      "step": 3149
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.0477108433842659,
      "learning_rate": 7.250337985623342e-05,
      "loss": 2.2512,
      "step": 3150
    },
    {
      "epoch": 0.6302,
      "grad_norm": 0.0501309372484684,
      "learning_rate": 7.243626441830009e-05,
      "loss": 2.2509,
      "step": 3151
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.06642547249794006,
      "learning_rate": 7.236916241459663e-05,
      "loss": 2.251,
      "step": 3152
    },
    {
      "epoch": 0.6306,
      "grad_norm": 0.07384509593248367,
      "learning_rate": 7.230207387782776e-05,
      "loss": 2.2484,
      "step": 3153
    },
    {
      "epoch": 0.6308,
      "grad_norm": 0.06649264693260193,
      "learning_rate": 7.22349988406916e-05,
      "loss": 2.2511,
      "step": 3154
    },
    {
      "epoch": 0.631,
      "grad_norm": 0.04739503934979439,
      "learning_rate": 7.216793733587976e-05,
      "loss": 2.25,
      "step": 3155
    },
    {
      "epoch": 0.6312,
      "grad_norm": 0.05514635518193245,
      "learning_rate": 7.210088939607708e-05,
      "loss": 2.2503,
      "step": 3156
    },
    {
      "epoch": 0.6314,
      "grad_norm": 0.0778113305568695,
      "learning_rate": 7.203385505396203e-05,
      "loss": 2.2501,
      "step": 3157
    },
    {
      "epoch": 0.6316,
      "grad_norm": 0.0687054991722107,
      "learning_rate": 7.196683434220625e-05,
      "loss": 2.2477,
      "step": 3158
    },
    {
      "epoch": 0.6318,
      "grad_norm": 0.04474257677793503,
      "learning_rate": 7.18998272934749e-05,
      "loss": 2.2479,
      "step": 3159
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.06657829880714417,
      "learning_rate": 7.183283394042634e-05,
      "loss": 2.248,
      "step": 3160
    },
    {
      "epoch": 0.6322,
      "grad_norm": 0.07374152541160583,
      "learning_rate": 7.176585431571235e-05,
      "loss": 2.2495,
      "step": 3161
    },
    {
      "epoch": 0.6324,
      "grad_norm": 0.05162114277482033,
      "learning_rate": 7.169888845197798e-05,
      "loss": 2.2499,
      "step": 3162
    },
    {
      "epoch": 0.6326,
      "grad_norm": 0.04876026138663292,
      "learning_rate": 7.163193638186158e-05,
      "loss": 2.2509,
      "step": 3163
    },
    {
      "epoch": 0.6328,
      "grad_norm": 0.06353648751974106,
      "learning_rate": 7.156499813799476e-05,
      "loss": 2.2509,
      "step": 3164
    },
    {
      "epoch": 0.633,
      "grad_norm": 0.05955471098423004,
      "learning_rate": 7.149807375300239e-05,
      "loss": 2.2497,
      "step": 3165
    },
    {
      "epoch": 0.6332,
      "grad_norm": 0.04422984644770622,
      "learning_rate": 7.143116325950265e-05,
      "loss": 2.2482,
      "step": 3166
    },
    {
      "epoch": 0.6334,
      "grad_norm": 0.0541207380592823,
      "learning_rate": 7.136426669010689e-05,
      "loss": 2.2502,
      "step": 3167
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.07235555350780487,
      "learning_rate": 7.129738407741964e-05,
      "loss": 2.2496,
      "step": 3168
    },
    {
      "epoch": 0.6338,
      "grad_norm": 0.06390141695737839,
      "learning_rate": 7.123051545403874e-05,
      "loss": 2.2495,
      "step": 3169
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.04423960670828819,
      "learning_rate": 7.11636608525551e-05,
      "loss": 2.2492,
      "step": 3170
    },
    {
      "epoch": 0.6342,
      "grad_norm": 0.05281616747379303,
      "learning_rate": 7.109682030555283e-05,
      "loss": 2.2491,
      "step": 3171
    },
    {
      "epoch": 0.6344,
      "grad_norm": 0.07105883955955505,
      "learning_rate": 7.102999384560927e-05,
      "loss": 2.25,
      "step": 3172
    },
    {
      "epoch": 0.6346,
      "grad_norm": 0.071198008954525,
      "learning_rate": 7.096318150529477e-05,
      "loss": 2.2484,
      "step": 3173
    },
    {
      "epoch": 0.6348,
      "grad_norm": 0.047796811908483505,
      "learning_rate": 7.089638331717284e-05,
      "loss": 2.2487,
      "step": 3174
    },
    {
      "epoch": 0.635,
      "grad_norm": 0.04550788924098015,
      "learning_rate": 7.082959931380011e-05,
      "loss": 2.2521,
      "step": 3175
    },
    {
      "epoch": 0.6352,
      "grad_norm": 0.05269865691661835,
      "learning_rate": 7.076282952772633e-05,
      "loss": 2.2511,
      "step": 3176
    },
    {
      "epoch": 0.6354,
      "grad_norm": 0.04972384124994278,
      "learning_rate": 7.069607399149428e-05,
      "loss": 2.2492,
      "step": 3177
    },
    {
      "epoch": 0.6356,
      "grad_norm": 0.04470067471265793,
      "learning_rate": 7.062933273763975e-05,
      "loss": 2.2482,
      "step": 3178
    },
    {
      "epoch": 0.6358,
      "grad_norm": 0.0441896952688694,
      "learning_rate": 7.056260579869165e-05,
      "loss": 2.2476,
      "step": 3179
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.051344964653253555,
      "learning_rate": 7.049589320717186e-05,
      "loss": 2.2492,
      "step": 3180
    },
    {
      "epoch": 0.6362,
      "grad_norm": 0.05286978930234909,
      "learning_rate": 7.042919499559537e-05,
      "loss": 2.2473,
      "step": 3181
    },
    {
      "epoch": 0.6364,
      "grad_norm": 0.04645827040076256,
      "learning_rate": 7.036251119646992e-05,
      "loss": 2.2522,
      "step": 3182
    },
    {
      "epoch": 0.6366,
      "grad_norm": 0.04374656081199646,
      "learning_rate": 7.029584184229653e-05,
      "loss": 2.2494,
      "step": 3183
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.043412040919065475,
      "learning_rate": 7.022918696556896e-05,
      "loss": 2.2504,
      "step": 3184
    },
    {
      "epoch": 0.637,
      "grad_norm": 0.04576858505606651,
      "learning_rate": 7.016254659877398e-05,
      "loss": 2.2498,
      "step": 3185
    },
    {
      "epoch": 0.6372,
      "grad_norm": 0.04475827515125275,
      "learning_rate": 7.009592077439134e-05,
      "loss": 2.249,
      "step": 3186
    },
    {
      "epoch": 0.6374,
      "grad_norm": 0.0432325154542923,
      "learning_rate": 7.002930952489362e-05,
      "loss": 2.2501,
      "step": 3187
    },
    {
      "epoch": 0.6376,
      "grad_norm": 0.04668909311294556,
      "learning_rate": 6.996271288274636e-05,
      "loss": 2.2496,
      "step": 3188
    },
    {
      "epoch": 0.6378,
      "grad_norm": 0.047786518931388855,
      "learning_rate": 6.989613088040796e-05,
      "loss": 2.249,
      "step": 3189
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.04407019168138504,
      "learning_rate": 6.982956355032968e-05,
      "loss": 2.2508,
      "step": 3190
    },
    {
      "epoch": 0.6382,
      "grad_norm": 0.043924715369939804,
      "learning_rate": 6.976301092495556e-05,
      "loss": 2.2498,
      "step": 3191
    },
    {
      "epoch": 0.6384,
      "grad_norm": 0.054207686334848404,
      "learning_rate": 6.969647303672262e-05,
      "loss": 2.2499,
      "step": 3192
    },
    {
      "epoch": 0.6386,
      "grad_norm": 0.06268411129713058,
      "learning_rate": 6.962994991806059e-05,
      "loss": 2.2506,
      "step": 3193
    },
    {
      "epoch": 0.6388,
      "grad_norm": 0.057610686868429184,
      "learning_rate": 6.956344160139201e-05,
      "loss": 2.2494,
      "step": 3194
    },
    {
      "epoch": 0.639,
      "grad_norm": 0.04343276098370552,
      "learning_rate": 6.949694811913225e-05,
      "loss": 2.2503,
      "step": 3195
    },
    {
      "epoch": 0.6392,
      "grad_norm": 0.049019359052181244,
      "learning_rate": 6.943046950368944e-05,
      "loss": 2.2498,
      "step": 3196
    },
    {
      "epoch": 0.6394,
      "grad_norm": 0.04800613969564438,
      "learning_rate": 6.93640057874644e-05,
      "loss": 2.2482,
      "step": 3197
    },
    {
      "epoch": 0.6396,
      "grad_norm": 0.04381889849901199,
      "learning_rate": 6.929755700285081e-05,
      "loss": 2.2495,
      "step": 3198
    },
    {
      "epoch": 0.6398,
      "grad_norm": 0.04718461632728577,
      "learning_rate": 6.923112318223496e-05,
      "loss": 2.2487,
      "step": 3199
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.04440631344914436,
      "learning_rate": 6.916470435799587e-05,
      "loss": 2.248,
      "step": 3200
    },
    {
      "epoch": 0.6402,
      "grad_norm": 0.04371071606874466,
      "learning_rate": 6.909830056250527e-05,
      "loss": 2.2509,
      "step": 3201
    },
    {
      "epoch": 0.6404,
      "grad_norm": 0.045811038464307785,
      "learning_rate": 6.90319118281276e-05,
      "loss": 2.2507,
      "step": 3202
    },
    {
      "epoch": 0.6406,
      "grad_norm": 0.04589654505252838,
      "learning_rate": 6.896553818721989e-05,
      "loss": 2.2522,
      "step": 3203
    },
    {
      "epoch": 0.6408,
      "grad_norm": 0.04435838386416435,
      "learning_rate": 6.889917967213185e-05,
      "loss": 2.252,
      "step": 3204
    },
    {
      "epoch": 0.641,
      "grad_norm": 0.04460204765200615,
      "learning_rate": 6.883283631520582e-05,
      "loss": 2.2512,
      "step": 3205
    },
    {
      "epoch": 0.6412,
      "grad_norm": 0.046921148896217346,
      "learning_rate": 6.876650814877674e-05,
      "loss": 2.2544,
      "step": 3206
    },
    {
      "epoch": 0.6414,
      "grad_norm": 0.04416130110621452,
      "learning_rate": 6.870019520517217e-05,
      "loss": 2.2495,
      "step": 3207
    },
    {
      "epoch": 0.6416,
      "grad_norm": 0.043774984776973724,
      "learning_rate": 6.863389751671225e-05,
      "loss": 2.2498,
      "step": 3208
    },
    {
      "epoch": 0.6418,
      "grad_norm": 0.044026799499988556,
      "learning_rate": 6.856761511570963e-05,
      "loss": 2.249,
      "step": 3209
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.04551757499575615,
      "learning_rate": 6.850134803446954e-05,
      "loss": 2.2491,
      "step": 3210
    },
    {
      "epoch": 0.6422,
      "grad_norm": 0.04960237815976143,
      "learning_rate": 6.843509630528977e-05,
      "loss": 2.2511,
      "step": 3211
    },
    {
      "epoch": 0.6424,
      "grad_norm": 0.05215555801987648,
      "learning_rate": 6.836885996046061e-05,
      "loss": 2.2481,
      "step": 3212
    },
    {
      "epoch": 0.6426,
      "grad_norm": 0.050615496933460236,
      "learning_rate": 6.830263903226483e-05,
      "loss": 2.2515,
      "step": 3213
    },
    {
      "epoch": 0.6428,
      "grad_norm": 0.043916333466768265,
      "learning_rate": 6.823643355297773e-05,
      "loss": 2.2493,
      "step": 3214
    },
    {
      "epoch": 0.643,
      "grad_norm": 0.06291504949331284,
      "learning_rate": 6.817024355486706e-05,
      "loss": 2.2477,
      "step": 3215
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.08580249547958374,
      "learning_rate": 6.8104069070193e-05,
      "loss": 2.2503,
      "step": 3216
    },
    {
      "epoch": 0.6434,
      "grad_norm": 0.08165466785430908,
      "learning_rate": 6.803791013120822e-05,
      "loss": 2.2485,
      "step": 3217
    },
    {
      "epoch": 0.6436,
      "grad_norm": 0.055000320076942444,
      "learning_rate": 6.797176677015775e-05,
      "loss": 2.2496,
      "step": 3218
    },
    {
      "epoch": 0.6438,
      "grad_norm": 0.04676193371415138,
      "learning_rate": 6.790563901927907e-05,
      "loss": 2.2507,
      "step": 3219
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.0646531879901886,
      "learning_rate": 6.783952691080203e-05,
      "loss": 2.2517,
      "step": 3220
    },
    {
      "epoch": 0.6442,
      "grad_norm": 0.05741887167096138,
      "learning_rate": 6.77734304769489e-05,
      "loss": 2.2511,
      "step": 3221
    },
    {
      "epoch": 0.6444,
      "grad_norm": 0.045249007642269135,
      "learning_rate": 6.770734974993426e-05,
      "loss": 2.2532,
      "step": 3222
    },
    {
      "epoch": 0.6446,
      "grad_norm": 0.04424303025007248,
      "learning_rate": 6.764128476196505e-05,
      "loss": 2.2501,
      "step": 3223
    },
    {
      "epoch": 0.6448,
      "grad_norm": 0.04427378252148628,
      "learning_rate": 6.757523554524056e-05,
      "loss": 2.2515,
      "step": 3224
    },
    {
      "epoch": 0.645,
      "grad_norm": 0.045767828822135925,
      "learning_rate": 6.750920213195238e-05,
      "loss": 2.2497,
      "step": 3225
    },
    {
      "epoch": 0.6452,
      "grad_norm": 0.0443236343562603,
      "learning_rate": 6.744318455428436e-05,
      "loss": 2.2481,
      "step": 3226
    },
    {
      "epoch": 0.6454,
      "grad_norm": 0.05545181781053543,
      "learning_rate": 6.737718284441267e-05,
      "loss": 2.2495,
      "step": 3227
    },
    {
      "epoch": 0.6456,
      "grad_norm": 0.06708361208438873,
      "learning_rate": 6.731119703450577e-05,
      "loss": 2.2506,
      "step": 3228
    },
    {
      "epoch": 0.6458,
      "grad_norm": 0.054545655846595764,
      "learning_rate": 6.724522715672432e-05,
      "loss": 2.2491,
      "step": 3229
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.04415113851428032,
      "learning_rate": 6.717927324322124e-05,
      "loss": 2.2491,
      "step": 3230
    },
    {
      "epoch": 0.6462,
      "grad_norm": 0.060644544661045074,
      "learning_rate": 6.711333532614168e-05,
      "loss": 2.2505,
      "step": 3231
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.061580855399370193,
      "learning_rate": 6.704741343762295e-05,
      "loss": 2.2509,
      "step": 3232
    },
    {
      "epoch": 0.6466,
      "grad_norm": 0.0459783710539341,
      "learning_rate": 6.698150760979463e-05,
      "loss": 2.2499,
      "step": 3233
    },
    {
      "epoch": 0.6468,
      "grad_norm": 0.05035026744008064,
      "learning_rate": 6.69156178747784e-05,
      "loss": 2.2494,
      "step": 3234
    },
    {
      "epoch": 0.647,
      "grad_norm": 0.057361502200365067,
      "learning_rate": 6.684974426468808e-05,
      "loss": 2.2481,
      "step": 3235
    },
    {
      "epoch": 0.6472,
      "grad_norm": 0.046716537326574326,
      "learning_rate": 6.67838868116297e-05,
      "loss": 2.2504,
      "step": 3236
    },
    {
      "epoch": 0.6474,
      "grad_norm": 0.045477040112018585,
      "learning_rate": 6.671804554770135e-05,
      "loss": 2.2498,
      "step": 3237
    },
    {
      "epoch": 0.6476,
      "grad_norm": 0.050654228776693344,
      "learning_rate": 6.66522205049933e-05,
      "loss": 2.2508,
      "step": 3238
    },
    {
      "epoch": 0.6478,
      "grad_norm": 0.04451736435294151,
      "learning_rate": 6.658641171558785e-05,
      "loss": 2.2506,
      "step": 3239
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.05608830228447914,
      "learning_rate": 6.652061921155943e-05,
      "loss": 2.2502,
      "step": 3240
    },
    {
      "epoch": 0.6482,
      "grad_norm": 0.06159854307770729,
      "learning_rate": 6.64548430249745e-05,
      "loss": 2.249,
      "step": 3241
    },
    {
      "epoch": 0.6484,
      "grad_norm": 0.04819759726524353,
      "learning_rate": 6.638908318789156e-05,
      "loss": 2.2515,
      "step": 3242
    },
    {
      "epoch": 0.6486,
      "grad_norm": 0.044263094663619995,
      "learning_rate": 6.63233397323612e-05,
      "loss": 2.2488,
      "step": 3243
    },
    {
      "epoch": 0.6488,
      "grad_norm": 0.050194066017866135,
      "learning_rate": 6.62576126904259e-05,
      "loss": 2.2483,
      "step": 3244
    },
    {
      "epoch": 0.649,
      "grad_norm": 0.05479355528950691,
      "learning_rate": 6.61919020941203e-05,
      "loss": 2.2506,
      "step": 3245
    },
    {
      "epoch": 0.6492,
      "grad_norm": 0.05415317043662071,
      "learning_rate": 6.612620797547087e-05,
      "loss": 2.2516,
      "step": 3246
    },
    {
      "epoch": 0.6494,
      "grad_norm": 0.05345512181520462,
      "learning_rate": 6.606053036649619e-05,
      "loss": 2.2501,
      "step": 3247
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.04594928398728371,
      "learning_rate": 6.599486929920673e-05,
      "loss": 2.2499,
      "step": 3248
    },
    {
      "epoch": 0.6498,
      "grad_norm": 0.04578505828976631,
      "learning_rate": 6.592922480560483e-05,
      "loss": 2.252,
      "step": 3249
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.057586826384067535,
      "learning_rate": 6.58635969176849e-05,
      "loss": 2.2503,
      "step": 3250
    },
    {
      "epoch": 0.6502,
      "grad_norm": 0.051487188786268234,
      "learning_rate": 6.579798566743314e-05,
      "loss": 2.2494,
      "step": 3251
    },
    {
      "epoch": 0.6504,
      "grad_norm": 0.043290603905916214,
      "learning_rate": 6.573239108682768e-05,
      "loss": 2.2486,
      "step": 3252
    },
    {
      "epoch": 0.6506,
      "grad_norm": 0.04865787923336029,
      "learning_rate": 6.566681320783849e-05,
      "loss": 2.2493,
      "step": 3253
    },
    {
      "epoch": 0.6508,
      "grad_norm": 0.04476117715239525,
      "learning_rate": 6.560125206242746e-05,
      "loss": 2.2503,
      "step": 3254
    },
    {
      "epoch": 0.651,
      "grad_norm": 0.046544305980205536,
      "learning_rate": 6.55357076825483e-05,
      "loss": 2.2495,
      "step": 3255
    },
    {
      "epoch": 0.6512,
      "grad_norm": 0.060339730232954025,
      "learning_rate": 6.547018010014654e-05,
      "loss": 2.2496,
      "step": 3256
    },
    {
      "epoch": 0.6514,
      "grad_norm": 0.05700058117508888,
      "learning_rate": 6.540466934715953e-05,
      "loss": 2.2502,
      "step": 3257
    },
    {
      "epoch": 0.6516,
      "grad_norm": 0.04384414851665497,
      "learning_rate": 6.53391754555164e-05,
      "loss": 2.2482,
      "step": 3258
    },
    {
      "epoch": 0.6518,
      "grad_norm": 0.05448610335588455,
      "learning_rate": 6.52736984571381e-05,
      "loss": 2.2505,
      "step": 3259
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.06743371486663818,
      "learning_rate": 6.520823838393731e-05,
      "loss": 2.2502,
      "step": 3260
    },
    {
      "epoch": 0.6522,
      "grad_norm": 0.056692883372306824,
      "learning_rate": 6.51427952678185e-05,
      "loss": 2.2511,
      "step": 3261
    },
    {
      "epoch": 0.6524,
      "grad_norm": 0.04499679431319237,
      "learning_rate": 6.507736914067781e-05,
      "loss": 2.2491,
      "step": 3262
    },
    {
      "epoch": 0.6526,
      "grad_norm": 0.06407719105482101,
      "learning_rate": 6.501196003440314e-05,
      "loss": 2.2488,
      "step": 3263
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.06613796204328537,
      "learning_rate": 6.494656798087412e-05,
      "loss": 2.2497,
      "step": 3264
    },
    {
      "epoch": 0.653,
      "grad_norm": 0.05021034926176071,
      "learning_rate": 6.488119301196201e-05,
      "loss": 2.2497,
      "step": 3265
    },
    {
      "epoch": 0.6532,
      "grad_norm": 0.045581091195344925,
      "learning_rate": 6.481583515952983e-05,
      "loss": 2.2507,
      "step": 3266
    },
    {
      "epoch": 0.6534,
      "grad_norm": 0.05857613682746887,
      "learning_rate": 6.475049445543215e-05,
      "loss": 2.2509,
      "step": 3267
    },
    {
      "epoch": 0.6536,
      "grad_norm": 0.05452260375022888,
      "learning_rate": 6.468517093151525e-05,
      "loss": 2.2515,
      "step": 3268
    },
    {
      "epoch": 0.6538,
      "grad_norm": 0.04927503690123558,
      "learning_rate": 6.461986461961706e-05,
      "loss": 2.2513,
      "step": 3269
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.04438156262040138,
      "learning_rate": 6.455457555156705e-05,
      "loss": 2.2502,
      "step": 3270
    },
    {
      "epoch": 0.6542,
      "grad_norm": 0.053239453583955765,
      "learning_rate": 6.448930375918631e-05,
      "loss": 2.248,
      "step": 3271
    },
    {
      "epoch": 0.6544,
      "grad_norm": 0.07205071300268173,
      "learning_rate": 6.442404927428751e-05,
      "loss": 2.2508,
      "step": 3272
    },
    {
      "epoch": 0.6546,
      "grad_norm": 0.06147437170147896,
      "learning_rate": 6.435881212867493e-05,
      "loss": 2.2493,
      "step": 3273
    },
    {
      "epoch": 0.6548,
      "grad_norm": 0.043817128986120224,
      "learning_rate": 6.429359235414436e-05,
      "loss": 2.2515,
      "step": 3274
    },
    {
      "epoch": 0.655,
      "grad_norm": 0.0628380998969078,
      "learning_rate": 6.422838998248307e-05,
      "loss": 2.2504,
      "step": 3275
    },
    {
      "epoch": 0.6552,
      "grad_norm": 0.07362399250268936,
      "learning_rate": 6.416320504546997e-05,
      "loss": 2.2472,
      "step": 3276
    },
    {
      "epoch": 0.6554,
      "grad_norm": 0.06137099862098694,
      "learning_rate": 6.409803757487538e-05,
      "loss": 2.2502,
      "step": 3277
    },
    {
      "epoch": 0.6556,
      "grad_norm": 0.04429709538817406,
      "learning_rate": 6.403288760246112e-05,
      "loss": 2.2489,
      "step": 3278
    },
    {
      "epoch": 0.6558,
      "grad_norm": 0.046913936734199524,
      "learning_rate": 6.396775515998055e-05,
      "loss": 2.2488,
      "step": 3279
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.05013040080666542,
      "learning_rate": 6.390264027917836e-05,
      "loss": 2.2509,
      "step": 3280
    },
    {
      "epoch": 0.6562,
      "grad_norm": 0.04851338639855385,
      "learning_rate": 6.383754299179079e-05,
      "loss": 2.2486,
      "step": 3281
    },
    {
      "epoch": 0.6564,
      "grad_norm": 0.04479034245014191,
      "learning_rate": 6.377246332954544e-05,
      "loss": 2.2511,
      "step": 3282
    },
    {
      "epoch": 0.6566,
      "grad_norm": 0.04493977501988411,
      "learning_rate": 6.370740132416138e-05,
      "loss": 2.2532,
      "step": 3283
    },
    {
      "epoch": 0.6568,
      "grad_norm": 0.04614586383104324,
      "learning_rate": 6.364235700734903e-05,
      "loss": 2.2529,
      "step": 3284
    },
    {
      "epoch": 0.657,
      "grad_norm": 0.04734201729297638,
      "learning_rate": 6.357733041081018e-05,
      "loss": 2.2532,
      "step": 3285
    },
    {
      "epoch": 0.6572,
      "grad_norm": 0.05409156531095505,
      "learning_rate": 6.351232156623803e-05,
      "loss": 2.2535,
      "step": 3286
    },
    {
      "epoch": 0.6574,
      "grad_norm": 0.08752614259719849,
      "learning_rate": 6.344733050531713e-05,
      "loss": 2.2512,
      "step": 3287
    },
    {
      "epoch": 0.6576,
      "grad_norm": 0.09066589921712875,
      "learning_rate": 6.338235725972325e-05,
      "loss": 2.2481,
      "step": 3288
    },
    {
      "epoch": 0.6578,
      "grad_norm": 0.048206936568021774,
      "learning_rate": 6.33174018611236e-05,
      "loss": 2.2491,
      "step": 3289
    },
    {
      "epoch": 0.658,
      "grad_norm": 0.0663878470659256,
      "learning_rate": 6.325246434117668e-05,
      "loss": 2.2501,
      "step": 3290
    },
    {
      "epoch": 0.6582,
      "grad_norm": 0.08912371844053268,
      "learning_rate": 6.318754473153221e-05,
      "loss": 2.2502,
      "step": 3291
    },
    {
      "epoch": 0.6584,
      "grad_norm": 0.06764467805624008,
      "learning_rate": 6.312264306383124e-05,
      "loss": 2.2508,
      "step": 3292
    },
    {
      "epoch": 0.6586,
      "grad_norm": 0.04383425414562225,
      "learning_rate": 6.305775936970606e-05,
      "loss": 2.2482,
      "step": 3293
    },
    {
      "epoch": 0.6588,
      "grad_norm": 0.06308526545763016,
      "learning_rate": 6.299289368078016e-05,
      "loss": 2.2501,
      "step": 3294
    },
    {
      "epoch": 0.659,
      "grad_norm": 0.06720224767923355,
      "learning_rate": 6.292804602866834e-05,
      "loss": 2.248,
      "step": 3295
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.047222308814525604,
      "learning_rate": 6.286321644497655e-05,
      "loss": 2.2483,
      "step": 3296
    },
    {
      "epoch": 0.6594,
      "grad_norm": 0.053008388727903366,
      "learning_rate": 6.27984049613019e-05,
      "loss": 2.249,
      "step": 3297
    },
    {
      "epoch": 0.6596,
      "grad_norm": 0.06790782511234283,
      "learning_rate": 6.27336116092327e-05,
      "loss": 2.2496,
      "step": 3298
    },
    {
      "epoch": 0.6598,
      "grad_norm": 0.055292923003435135,
      "learning_rate": 6.266883642034853e-05,
      "loss": 2.2484,
      "step": 3299
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.04340741038322449,
      "learning_rate": 6.260407942621998e-05,
      "loss": 2.2473,
      "step": 3300
    },
    {
      "epoch": 0.6602,
      "grad_norm": 0.05498648062348366,
      "learning_rate": 6.25393406584088e-05,
      "loss": 2.2489,
      "step": 3301
    },
    {
      "epoch": 0.6604,
      "grad_norm": 0.05948701500892639,
      "learning_rate": 6.247462014846792e-05,
      "loss": 2.247,
      "step": 3302
    },
    {
      "epoch": 0.6606,
      "grad_norm": 0.04443967714905739,
      "learning_rate": 6.240991792794133e-05,
      "loss": 2.2472,
      "step": 3303
    },
    {
      "epoch": 0.6608,
      "grad_norm": 0.05590837076306343,
      "learning_rate": 6.234523402836407e-05,
      "loss": 2.2486,
      "step": 3304
    },
    {
      "epoch": 0.661,
      "grad_norm": 0.07550916075706482,
      "learning_rate": 6.228056848126236e-05,
      "loss": 2.2515,
      "step": 3305
    },
    {
      "epoch": 0.6612,
      "grad_norm": 0.06288622319698334,
      "learning_rate": 6.22159213181533e-05,
      "loss": 2.2493,
      "step": 3306
    },
    {
      "epoch": 0.6614,
      "grad_norm": 0.04341036081314087,
      "learning_rate": 6.215129257054522e-05,
      "loss": 2.2488,
      "step": 3307
    },
    {
      "epoch": 0.6616,
      "grad_norm": 0.06184759363532066,
      "learning_rate": 6.20866822699373e-05,
      "loss": 2.2503,
      "step": 3308
    },
    {
      "epoch": 0.6618,
      "grad_norm": 0.06485436111688614,
      "learning_rate": 6.20220904478199e-05,
      "loss": 2.2487,
      "step": 3309
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.049354664981365204,
      "learning_rate": 6.195751713567427e-05,
      "loss": 2.2493,
      "step": 3310
    },
    {
      "epoch": 0.6622,
      "grad_norm": 0.05169341713190079,
      "learning_rate": 6.18929623649726e-05,
      "loss": 2.2499,
      "step": 3311
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.08293081074953079,
      "learning_rate": 6.182842616717817e-05,
      "loss": 2.25,
      "step": 3312
    },
    {
      "epoch": 0.6626,
      "grad_norm": 0.08938521146774292,
      "learning_rate": 6.176390857374507e-05,
      "loss": 2.2503,
      "step": 3313
    },
    {
      "epoch": 0.6628,
      "grad_norm": 0.0517149493098259,
      "learning_rate": 6.169940961611854e-05,
      "loss": 2.2504,
      "step": 3314
    },
    {
      "epoch": 0.663,
      "grad_norm": 0.06070300564169884,
      "learning_rate": 6.163492932573438e-05,
      "loss": 2.2491,
      "step": 3315
    },
    {
      "epoch": 0.6632,
      "grad_norm": 0.09089697152376175,
      "learning_rate": 6.157046773401964e-05,
      "loss": 2.2476,
      "step": 3316
    },
    {
      "epoch": 0.6634,
      "grad_norm": 0.06924617290496826,
      "learning_rate": 6.150602487239206e-05,
      "loss": 2.2504,
      "step": 3317
    },
    {
      "epoch": 0.6636,
      "grad_norm": 0.050874292850494385,
      "learning_rate": 6.144160077226036e-05,
      "loss": 2.2494,
      "step": 3318
    },
    {
      "epoch": 0.6638,
      "grad_norm": 0.10592562705278397,
      "learning_rate": 6.1377195465024e-05,
      "loss": 2.2491,
      "step": 3319
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.09892558306455612,
      "learning_rate": 6.131280898207339e-05,
      "loss": 2.2476,
      "step": 3320
    },
    {
      "epoch": 0.6642,
      "grad_norm": 0.04362037405371666,
      "learning_rate": 6.12484413547897e-05,
      "loss": 2.2504,
      "step": 3321
    },
    {
      "epoch": 0.6644,
      "grad_norm": 0.09730537980794907,
      "learning_rate": 6.118409261454494e-05,
      "loss": 2.2495,
      "step": 3322
    },
    {
      "epoch": 0.6646,
      "grad_norm": 0.10823437571525574,
      "learning_rate": 6.111976279270192e-05,
      "loss": 2.2489,
      "step": 3323
    },
    {
      "epoch": 0.6648,
      "grad_norm": 0.04920780286192894,
      "learning_rate": 6.105545192061416e-05,
      "loss": 2.2504,
      "step": 3324
    },
    {
      "epoch": 0.665,
      "grad_norm": 0.0838194265961647,
      "learning_rate": 6.099116002962604e-05,
      "loss": 2.2478,
      "step": 3325
    },
    {
      "epoch": 0.6652,
      "grad_norm": 0.1053294986486435,
      "learning_rate": 6.092688715107264e-05,
      "loss": 2.2491,
      "step": 3326
    },
    {
      "epoch": 0.6654,
      "grad_norm": 0.05858572944998741,
      "learning_rate": 6.086263331627976e-05,
      "loss": 2.25,
      "step": 3327
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.06837553530931473,
      "learning_rate": 6.0798398556563976e-05,
      "loss": 2.2499,
      "step": 3328
    },
    {
      "epoch": 0.6658,
      "grad_norm": 0.10028330981731415,
      "learning_rate": 6.073418290323251e-05,
      "loss": 2.2471,
      "step": 3329
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.0657181590795517,
      "learning_rate": 6.066998638758326e-05,
      "loss": 2.2488,
      "step": 3330
    },
    {
      "epoch": 0.6662,
      "grad_norm": 0.0539432168006897,
      "learning_rate": 6.0605809040904894e-05,
      "loss": 2.2486,
      "step": 3331
    },
    {
      "epoch": 0.6664,
      "grad_norm": 0.09486139565706253,
      "learning_rate": 6.054165089447663e-05,
      "loss": 2.2492,
      "step": 3332
    },
    {
      "epoch": 0.6666,
      "grad_norm": 0.08214554190635681,
      "learning_rate": 6.047751197956838e-05,
      "loss": 2.2537,
      "step": 3333
    },
    {
      "epoch": 0.6668,
      "grad_norm": 0.04541086032986641,
      "learning_rate": 6.0413392327440635e-05,
      "loss": 2.2509,
      "step": 3334
    },
    {
      "epoch": 0.667,
      "grad_norm": 0.10966705530881882,
      "learning_rate": 6.034929196934459e-05,
      "loss": 2.2469,
      "step": 3335
    },
    {
      "epoch": 0.6672,
      "grad_norm": 0.11283647269010544,
      "learning_rate": 6.0285210936521955e-05,
      "loss": 2.2497,
      "step": 3336
    },
    {
      "epoch": 0.6674,
      "grad_norm": 0.044089023023843765,
      "learning_rate": 6.022114926020504e-05,
      "loss": 2.2479,
      "step": 3337
    },
    {
      "epoch": 0.6676,
      "grad_norm": 0.09828078001737595,
      "learning_rate": 6.015710697161674e-05,
      "loss": 2.2502,
      "step": 3338
    },
    {
      "epoch": 0.6678,
      "grad_norm": 0.09593603760004044,
      "learning_rate": 6.009308410197047e-05,
      "loss": 2.2486,
      "step": 3339
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.04558422788977623,
      "learning_rate": 6.002908068247024e-05,
      "loss": 2.2482,
      "step": 3340
    },
    {
      "epoch": 0.6682,
      "grad_norm": 0.10640701651573181,
      "learning_rate": 5.9965096744310526e-05,
      "loss": 2.2491,
      "step": 3341
    },
    {
      "epoch": 0.6684,
      "grad_norm": 0.09842637181282043,
      "learning_rate": 5.990113231867629e-05,
      "loss": 2.2477,
      "step": 3342
    },
    {
      "epoch": 0.6686,
      "grad_norm": 0.04350535571575165,
      "learning_rate": 5.9837187436743016e-05,
      "loss": 2.2472,
      "step": 3343
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.08064314723014832,
      "learning_rate": 5.977326212967671e-05,
      "loss": 2.2485,
      "step": 3344
    },
    {
      "epoch": 0.669,
      "grad_norm": 0.08069884777069092,
      "learning_rate": 5.9709356428633746e-05,
      "loss": 2.2477,
      "step": 3345
    },
    {
      "epoch": 0.6692,
      "grad_norm": 0.04409792274236679,
      "learning_rate": 5.964547036476099e-05,
      "loss": 2.2471,
      "step": 3346
    },
    {
      "epoch": 0.6694,
      "grad_norm": 0.059183165431022644,
      "learning_rate": 5.9581603969195766e-05,
      "loss": 2.2498,
      "step": 3347
    },
    {
      "epoch": 0.6696,
      "grad_norm": 0.054731398820877075,
      "learning_rate": 5.951775727306577e-05,
      "loss": 2.2488,
      "step": 3348
    },
    {
      "epoch": 0.6698,
      "grad_norm": 0.044970735907554626,
      "learning_rate": 5.94539303074891e-05,
      "loss": 2.248,
      "step": 3349
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.059369444847106934,
      "learning_rate": 5.9390123103574215e-05,
      "loss": 2.2458,
      "step": 3350
    },
    {
      "epoch": 0.6702,
      "grad_norm": 0.05663641542196274,
      "learning_rate": 5.9326335692419995e-05,
      "loss": 2.2477,
      "step": 3351
    },
    {
      "epoch": 0.6704,
      "grad_norm": 0.04386400803923607,
      "learning_rate": 5.9262568105115654e-05,
      "loss": 2.25,
      "step": 3352
    },
    {
      "epoch": 0.6706,
      "grad_norm": 0.047565072774887085,
      "learning_rate": 5.9198820372740726e-05,
      "loss": 2.2483,
      "step": 3353
    },
    {
      "epoch": 0.6708,
      "grad_norm": 0.04537694901227951,
      "learning_rate": 5.913509252636511e-05,
      "loss": 2.2499,
      "step": 3354
    },
    {
      "epoch": 0.671,
      "grad_norm": 0.046887267380952835,
      "learning_rate": 5.907138459704895e-05,
      "loss": 2.2507,
      "step": 3355
    },
    {
      "epoch": 0.6712,
      "grad_norm": 0.051190122961997986,
      "learning_rate": 5.900769661584272e-05,
      "loss": 2.2482,
      "step": 3356
    },
    {
      "epoch": 0.6714,
      "grad_norm": 0.044390883296728134,
      "learning_rate": 5.8944028613787206e-05,
      "loss": 2.2499,
      "step": 3357
    },
    {
      "epoch": 0.6716,
      "grad_norm": 0.04516726732254028,
      "learning_rate": 5.88803806219134e-05,
      "loss": 2.2474,
      "step": 3358
    },
    {
      "epoch": 0.6718,
      "grad_norm": 0.04462460055947304,
      "learning_rate": 5.881675267124254e-05,
      "loss": 2.2501,
      "step": 3359
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.04402879253029823,
      "learning_rate": 5.8753144792786096e-05,
      "loss": 2.2481,
      "step": 3360
    },
    {
      "epoch": 0.6722,
      "grad_norm": 0.051826346665620804,
      "learning_rate": 5.868955701754584e-05,
      "loss": 2.2475,
      "step": 3361
    },
    {
      "epoch": 0.6724,
      "grad_norm": 0.048659298568964005,
      "learning_rate": 5.862598937651365e-05,
      "loss": 2.2465,
      "step": 3362
    },
    {
      "epoch": 0.6726,
      "grad_norm": 0.04378709942102432,
      "learning_rate": 5.856244190067159e-05,
      "loss": 2.2498,
      "step": 3363
    },
    {
      "epoch": 0.6728,
      "grad_norm": 0.05332536622881889,
      "learning_rate": 5.849891462099198e-05,
      "loss": 2.2492,
      "step": 3364
    },
    {
      "epoch": 0.673,
      "grad_norm": 0.049874916672706604,
      "learning_rate": 5.8435407568437216e-05,
      "loss": 2.2489,
      "step": 3365
    },
    {
      "epoch": 0.6732,
      "grad_norm": 0.044616635888814926,
      "learning_rate": 5.83719207739599e-05,
      "loss": 2.2486,
      "step": 3366
    },
    {
      "epoch": 0.6734,
      "grad_norm": 0.06065170839428902,
      "learning_rate": 5.8308454268502675e-05,
      "loss": 2.2462,
      "step": 3367
    },
    {
      "epoch": 0.6736,
      "grad_norm": 0.054568078368902206,
      "learning_rate": 5.8245008082998364e-05,
      "loss": 2.2474,
      "step": 3368
    },
    {
      "epoch": 0.6738,
      "grad_norm": 0.04331440106034279,
      "learning_rate": 5.818158224836987e-05,
      "loss": 2.2469,
      "step": 3369
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.053836844861507416,
      "learning_rate": 5.8118176795530176e-05,
      "loss": 2.2492,
      "step": 3370
    },
    {
      "epoch": 0.6742,
      "grad_norm": 0.04553777351975441,
      "learning_rate": 5.805479175538229e-05,
      "loss": 2.2501,
      "step": 3371
    },
    {
      "epoch": 0.6744,
      "grad_norm": 0.050529513508081436,
      "learning_rate": 5.799142715881938e-05,
      "loss": 2.2485,
      "step": 3372
    },
    {
      "epoch": 0.6746,
      "grad_norm": 0.06091774255037308,
      "learning_rate": 5.792808303672454e-05,
      "loss": 2.246,
      "step": 3373
    },
    {
      "epoch": 0.6748,
      "grad_norm": 0.047246053814888,
      "learning_rate": 5.786475941997094e-05,
      "loss": 2.2484,
      "step": 3374
    },
    {
      "epoch": 0.675,
      "grad_norm": 0.04800136387348175,
      "learning_rate": 5.780145633942173e-05,
      "loss": 2.2492,
      "step": 3375
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.05492372810840607,
      "learning_rate": 5.773817382593008e-05,
      "loss": 2.2484,
      "step": 3376
    },
    {
      "epoch": 0.6754,
      "grad_norm": 0.0435781292617321,
      "learning_rate": 5.7674911910339094e-05,
      "loss": 2.2475,
      "step": 3377
    },
    {
      "epoch": 0.6756,
      "grad_norm": 0.05371342971920967,
      "learning_rate": 5.7611670623481864e-05,
      "loss": 2.2488,
      "step": 3378
    },
    {
      "epoch": 0.6758,
      "grad_norm": 0.055878881365060806,
      "learning_rate": 5.754844999618144e-05,
      "loss": 2.2495,
      "step": 3379
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.044444456696510315,
      "learning_rate": 5.748525005925074e-05,
      "loss": 2.2506,
      "step": 3380
    },
    {
      "epoch": 0.6762,
      "grad_norm": 0.05631497874855995,
      "learning_rate": 5.7422070843492734e-05,
      "loss": 2.2537,
      "step": 3381
    },
    {
      "epoch": 0.6764,
      "grad_norm": 0.07325559854507446,
      "learning_rate": 5.735891237970015e-05,
      "loss": 2.2488,
      "step": 3382
    },
    {
      "epoch": 0.6766,
      "grad_norm": 0.053775932639837265,
      "learning_rate": 5.7295774698655655e-05,
      "loss": 2.2479,
      "step": 3383
    },
    {
      "epoch": 0.6768,
      "grad_norm": 0.04815937951207161,
      "learning_rate": 5.723265783113181e-05,
      "loss": 2.2488,
      "step": 3384
    },
    {
      "epoch": 0.677,
      "grad_norm": 0.060982637107372284,
      "learning_rate": 5.716956180789098e-05,
      "loss": 2.249,
      "step": 3385
    },
    {
      "epoch": 0.6772,
      "grad_norm": 0.04529368132352829,
      "learning_rate": 5.710648665968543e-05,
      "loss": 2.2498,
      "step": 3386
    },
    {
      "epoch": 0.6774,
      "grad_norm": 0.05031241849064827,
      "learning_rate": 5.704343241725719e-05,
      "loss": 2.2503,
      "step": 3387
    },
    {
      "epoch": 0.6776,
      "grad_norm": 0.050562385469675064,
      "learning_rate": 5.6980399111338156e-05,
      "loss": 2.2483,
      "step": 3388
    },
    {
      "epoch": 0.6778,
      "grad_norm": 0.04376978427171707,
      "learning_rate": 5.691738677265e-05,
      "loss": 2.2476,
      "step": 3389
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.050853535532951355,
      "learning_rate": 5.6854395431904094e-05,
      "loss": 2.2485,
      "step": 3390
    },
    {
      "epoch": 0.6782,
      "grad_norm": 0.04622604325413704,
      "learning_rate": 5.679142511980175e-05,
      "loss": 2.2508,
      "step": 3391
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.046668823808431625,
      "learning_rate": 5.6728475867033925e-05,
      "loss": 2.2476,
      "step": 3392
    },
    {
      "epoch": 0.6786,
      "grad_norm": 0.049916960299015045,
      "learning_rate": 5.666554770428129e-05,
      "loss": 2.2465,
      "step": 3393
    },
    {
      "epoch": 0.6788,
      "grad_norm": 0.04445553198456764,
      "learning_rate": 5.6602640662214256e-05,
      "loss": 2.2486,
      "step": 3394
    },
    {
      "epoch": 0.679,
      "grad_norm": 0.04497900232672691,
      "learning_rate": 5.653975477149298e-05,
      "loss": 2.2475,
      "step": 3395
    },
    {
      "epoch": 0.6792,
      "grad_norm": 0.045793335884809494,
      "learning_rate": 5.647689006276726e-05,
      "loss": 2.2462,
      "step": 3396
    },
    {
      "epoch": 0.6794,
      "grad_norm": 0.043948255479335785,
      "learning_rate": 5.641404656667661e-05,
      "loss": 2.2481,
      "step": 3397
    },
    {
      "epoch": 0.6796,
      "grad_norm": 0.04316512867808342,
      "learning_rate": 5.635122431385016e-05,
      "loss": 2.2479,
      "step": 3398
    },
    {
      "epoch": 0.6798,
      "grad_norm": 0.04469825699925423,
      "learning_rate": 5.6288423334906735e-05,
      "loss": 2.2484,
      "step": 3399
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.04584405943751335,
      "learning_rate": 5.622564366045472e-05,
      "loss": 2.2487,
      "step": 3400
    },
    {
      "epoch": 0.6802,
      "grad_norm": 0.04428691044449806,
      "learning_rate": 5.616288532109225e-05,
      "loss": 2.2482,
      "step": 3401
    },
    {
      "epoch": 0.6804,
      "grad_norm": 0.04405900463461876,
      "learning_rate": 5.610014834740693e-05,
      "loss": 2.2511,
      "step": 3402
    },
    {
      "epoch": 0.6806,
      "grad_norm": 0.044760219752788544,
      "learning_rate": 5.6037432769976064e-05,
      "loss": 2.2476,
      "step": 3403
    },
    {
      "epoch": 0.6808,
      "grad_norm": 0.04377727955579758,
      "learning_rate": 5.5974738619366295e-05,
      "loss": 2.2492,
      "step": 3404
    },
    {
      "epoch": 0.681,
      "grad_norm": 0.04535974562168121,
      "learning_rate": 5.591206592613416e-05,
      "loss": 2.2483,
      "step": 3405
    },
    {
      "epoch": 0.6812,
      "grad_norm": 0.04837596043944359,
      "learning_rate": 5.584941472082549e-05,
      "loss": 2.2506,
      "step": 3406
    },
    {
      "epoch": 0.6814,
      "grad_norm": 0.04384641721844673,
      "learning_rate": 5.578678503397574e-05,
      "loss": 2.2477,
      "step": 3407
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.04534120485186577,
      "learning_rate": 5.572417689610987e-05,
      "loss": 2.25,
      "step": 3408
    },
    {
      "epoch": 0.6818,
      "grad_norm": 0.04585394263267517,
      "learning_rate": 5.566159033774225e-05,
      "loss": 2.2493,
      "step": 3409
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.04343205317854881,
      "learning_rate": 5.5599025389376935e-05,
      "loss": 2.2476,
      "step": 3410
    },
    {
      "epoch": 0.6822,
      "grad_norm": 0.0439441055059433,
      "learning_rate": 5.553648208150728e-05,
      "loss": 2.2474,
      "step": 3411
    },
    {
      "epoch": 0.6824,
      "grad_norm": 0.04326486960053444,
      "learning_rate": 5.547396044461608e-05,
      "loss": 2.2468,
      "step": 3412
    },
    {
      "epoch": 0.6826,
      "grad_norm": 0.04355590417981148,
      "learning_rate": 5.541146050917561e-05,
      "loss": 2.2474,
      "step": 3413
    },
    {
      "epoch": 0.6828,
      "grad_norm": 0.043559104204177856,
      "learning_rate": 5.5348982305647643e-05,
      "loss": 2.2494,
      "step": 3414
    },
    {
      "epoch": 0.683,
      "grad_norm": 0.04606525972485542,
      "learning_rate": 5.5286525864483285e-05,
      "loss": 2.2482,
      "step": 3415
    },
    {
      "epoch": 0.6832,
      "grad_norm": 0.04669791832566261,
      "learning_rate": 5.522409121612304e-05,
      "loss": 2.2498,
      "step": 3416
    },
    {
      "epoch": 0.6834,
      "grad_norm": 0.04368447884917259,
      "learning_rate": 5.5161678390996796e-05,
      "loss": 2.248,
      "step": 3417
    },
    {
      "epoch": 0.6836,
      "grad_norm": 0.04937957227230072,
      "learning_rate": 5.509928741952379e-05,
      "loss": 2.2487,
      "step": 3418
    },
    {
      "epoch": 0.6838,
      "grad_norm": 0.049119334667921066,
      "learning_rate": 5.50369183321126e-05,
      "loss": 2.2504,
      "step": 3419
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.04353886470198631,
      "learning_rate": 5.497457115916127e-05,
      "loss": 2.2479,
      "step": 3420
    },
    {
      "epoch": 0.6842,
      "grad_norm": 0.056724727153778076,
      "learning_rate": 5.491224593105695e-05,
      "loss": 2.25,
      "step": 3421
    },
    {
      "epoch": 0.6844,
      "grad_norm": 0.05408460646867752,
      "learning_rate": 5.484994267817624e-05,
      "loss": 2.248,
      "step": 3422
    },
    {
      "epoch": 0.6846,
      "grad_norm": 0.04350957274436951,
      "learning_rate": 5.478766143088492e-05,
      "loss": 2.248,
      "step": 3423
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.05226632580161095,
      "learning_rate": 5.4725402219538236e-05,
      "loss": 2.246,
      "step": 3424
    },
    {
      "epoch": 0.685,
      "grad_norm": 0.046637751162052155,
      "learning_rate": 5.466316507448048e-05,
      "loss": 2.2482,
      "step": 3425
    },
    {
      "epoch": 0.6852,
      "grad_norm": 0.045481644570827484,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 2.2476,
      "step": 3426
    },
    {
      "epoch": 0.6854,
      "grad_norm": 0.05509251356124878,
      "learning_rate": 5.4538757104555615e-05,
      "loss": 2.2468,
      "step": 3427
    },
    {
      "epoch": 0.6856,
      "grad_norm": 0.04688287526369095,
      "learning_rate": 5.447658634032338e-05,
      "loss": 2.2466,
      "step": 3428
    },
    {
      "epoch": 0.6858,
      "grad_norm": 0.04428037628531456,
      "learning_rate": 5.441443776365003e-05,
      "loss": 2.2479,
      "step": 3429
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.04724736884236336,
      "learning_rate": 5.435231140482587e-05,
      "loss": 2.2462,
      "step": 3430
    },
    {
      "epoch": 0.6862,
      "grad_norm": 0.043834883719682693,
      "learning_rate": 5.4290207294130615e-05,
      "loss": 2.2482,
      "step": 3431
    },
    {
      "epoch": 0.6864,
      "grad_norm": 0.04364657774567604,
      "learning_rate": 5.4228125461833024e-05,
      "loss": 2.2484,
      "step": 3432
    },
    {
      "epoch": 0.6866,
      "grad_norm": 0.043223120272159576,
      "learning_rate": 5.416606593819101e-05,
      "loss": 2.2462,
      "step": 3433
    },
    {
      "epoch": 0.6868,
      "grad_norm": 0.04467546194791794,
      "learning_rate": 5.4104028753451696e-05,
      "loss": 2.2491,
      "step": 3434
    },
    {
      "epoch": 0.687,
      "grad_norm": 0.043835416436195374,
      "learning_rate": 5.404201393785122e-05,
      "loss": 2.2474,
      "step": 3435
    },
    {
      "epoch": 0.6872,
      "grad_norm": 0.04377090185880661,
      "learning_rate": 5.398002152161484e-05,
      "loss": 2.2469,
      "step": 3436
    },
    {
      "epoch": 0.6874,
      "grad_norm": 0.04649663716554642,
      "learning_rate": 5.3918051534956926e-05,
      "loss": 2.2479,
      "step": 3437
    },
    {
      "epoch": 0.6876,
      "grad_norm": 0.04425317049026489,
      "learning_rate": 5.3856104008080876e-05,
      "loss": 2.2476,
      "step": 3438
    },
    {
      "epoch": 0.6878,
      "grad_norm": 0.04412775859236717,
      "learning_rate": 5.3794178971179165e-05,
      "loss": 2.2486,
      "step": 3439
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.046897172927856445,
      "learning_rate": 5.373227645443332e-05,
      "loss": 2.2468,
      "step": 3440
    },
    {
      "epoch": 0.6882,
      "grad_norm": 0.04520728439092636,
      "learning_rate": 5.3670396488013854e-05,
      "loss": 2.2482,
      "step": 3441
    },
    {
      "epoch": 0.6884,
      "grad_norm": 0.04363364353775978,
      "learning_rate": 5.360853910208028e-05,
      "loss": 2.248,
      "step": 3442
    },
    {
      "epoch": 0.6886,
      "grad_norm": 0.04995275288820267,
      "learning_rate": 5.3546704326781236e-05,
      "loss": 2.2471,
      "step": 3443
    },
    {
      "epoch": 0.6888,
      "grad_norm": 0.05970809981226921,
      "learning_rate": 5.348489219225416e-05,
      "loss": 2.2488,
      "step": 3444
    },
    {
      "epoch": 0.689,
      "grad_norm": 0.04932196065783501,
      "learning_rate": 5.3423102728625574e-05,
      "loss": 2.248,
      "step": 3445
    },
    {
      "epoch": 0.6892,
      "grad_norm": 0.04493122547864914,
      "learning_rate": 5.33613359660109e-05,
      "loss": 2.2501,
      "step": 3446
    },
    {
      "epoch": 0.6894,
      "grad_norm": 0.050814077258110046,
      "learning_rate": 5.329959193451448e-05,
      "loss": 2.248,
      "step": 3447
    },
    {
      "epoch": 0.6896,
      "grad_norm": 0.04682989791035652,
      "learning_rate": 5.3237870664229636e-05,
      "loss": 2.2497,
      "step": 3448
    },
    {
      "epoch": 0.6898,
      "grad_norm": 0.043656181544065475,
      "learning_rate": 5.317617218523856e-05,
      "loss": 2.2498,
      "step": 3449
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.04809102788567543,
      "learning_rate": 5.311449652761235e-05,
      "loss": 2.2468,
      "step": 3450
    },
    {
      "epoch": 0.6902,
      "grad_norm": 0.04499198868870735,
      "learning_rate": 5.305284372141095e-05,
      "loss": 2.2485,
      "step": 3451
    },
    {
      "epoch": 0.6904,
      "grad_norm": 0.045435696840286255,
      "learning_rate": 5.299121379668316e-05,
      "loss": 2.2489,
      "step": 3452
    },
    {
      "epoch": 0.6906,
      "grad_norm": 0.04695555195212364,
      "learning_rate": 5.292960678346675e-05,
      "loss": 2.2492,
      "step": 3453
    },
    {
      "epoch": 0.6908,
      "grad_norm": 0.043885063380002975,
      "learning_rate": 5.286802271178815e-05,
      "loss": 2.2498,
      "step": 3454
    },
    {
      "epoch": 0.691,
      "grad_norm": 0.04398662969470024,
      "learning_rate": 5.2806461611662735e-05,
      "loss": 2.2493,
      "step": 3455
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.04524889960885048,
      "learning_rate": 5.274492351309461e-05,
      "loss": 2.2473,
      "step": 3456
    },
    {
      "epoch": 0.6914,
      "grad_norm": 0.053079161792993546,
      "learning_rate": 5.26834084460767e-05,
      "loss": 2.2531,
      "step": 3457
    },
    {
      "epoch": 0.6916,
      "grad_norm": 0.06262307614088058,
      "learning_rate": 5.2621916440590715e-05,
      "loss": 2.2526,
      "step": 3458
    },
    {
      "epoch": 0.6918,
      "grad_norm": 0.04580818489193916,
      "learning_rate": 5.256044752660709e-05,
      "loss": 2.2493,
      "step": 3459
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.05982528254389763,
      "learning_rate": 5.2499001734085044e-05,
      "loss": 2.2491,
      "step": 3460
    },
    {
      "epoch": 0.6922,
      "grad_norm": 0.06118384376168251,
      "learning_rate": 5.243757909297247e-05,
      "loss": 2.2471,
      "step": 3461
    },
    {
      "epoch": 0.6924,
      "grad_norm": 0.0446353554725647,
      "learning_rate": 5.237617963320608e-05,
      "loss": 2.2518,
      "step": 3462
    },
    {
      "epoch": 0.6926,
      "grad_norm": 0.050736766308546066,
      "learning_rate": 5.2314803384711195e-05,
      "loss": 2.2493,
      "step": 3463
    },
    {
      "epoch": 0.6928,
      "grad_norm": 0.04627757892012596,
      "learning_rate": 5.225345037740186e-05,
      "loss": 2.2471,
      "step": 3464
    },
    {
      "epoch": 0.693,
      "grad_norm": 0.04590899124741554,
      "learning_rate": 5.2192120641180786e-05,
      "loss": 2.2482,
      "step": 3465
    },
    {
      "epoch": 0.6932,
      "grad_norm": 0.05088716372847557,
      "learning_rate": 5.213081420593933e-05,
      "loss": 2.2474,
      "step": 3466
    },
    {
      "epoch": 0.6934,
      "grad_norm": 0.04390627518296242,
      "learning_rate": 5.20695311015575e-05,
      "loss": 2.2468,
      "step": 3467
    },
    {
      "epoch": 0.6936,
      "grad_norm": 0.04864330217242241,
      "learning_rate": 5.200827135790396e-05,
      "loss": 2.2505,
      "step": 3468
    },
    {
      "epoch": 0.6938,
      "grad_norm": 0.04600329324603081,
      "learning_rate": 5.194703500483593e-05,
      "loss": 2.2493,
      "step": 3469
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.044377490878105164,
      "learning_rate": 5.188582207219931e-05,
      "loss": 2.2488,
      "step": 3470
    },
    {
      "epoch": 0.6942,
      "grad_norm": 0.04719150811433792,
      "learning_rate": 5.182463258982846e-05,
      "loss": 2.2492,
      "step": 3471
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.04448991268873215,
      "learning_rate": 5.1763466587546485e-05,
      "loss": 2.2502,
      "step": 3472
    },
    {
      "epoch": 0.6946,
      "grad_norm": 0.043776772916316986,
      "learning_rate": 5.170232409516496e-05,
      "loss": 2.2484,
      "step": 3473
    },
    {
      "epoch": 0.6948,
      "grad_norm": 0.04361778125166893,
      "learning_rate": 5.1641205142483894e-05,
      "loss": 2.248,
      "step": 3474
    },
    {
      "epoch": 0.695,
      "grad_norm": 0.04531606659293175,
      "learning_rate": 5.158010975929193e-05,
      "loss": 2.2495,
      "step": 3475
    },
    {
      "epoch": 0.6952,
      "grad_norm": 0.0448589026927948,
      "learning_rate": 5.15190379753663e-05,
      "loss": 2.248,
      "step": 3476
    },
    {
      "epoch": 0.6954,
      "grad_norm": 0.043422844260931015,
      "learning_rate": 5.145798982047261e-05,
      "loss": 2.2479,
      "step": 3477
    },
    {
      "epoch": 0.6956,
      "grad_norm": 0.04527623951435089,
      "learning_rate": 5.1396965324364986e-05,
      "loss": 2.2486,
      "step": 3478
    },
    {
      "epoch": 0.6958,
      "grad_norm": 0.0444105826318264,
      "learning_rate": 5.133596451678603e-05,
      "loss": 2.2516,
      "step": 3479
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.04380786418914795,
      "learning_rate": 5.127498742746675e-05,
      "loss": 2.2479,
      "step": 3480
    },
    {
      "epoch": 0.6962,
      "grad_norm": 0.04489584267139435,
      "learning_rate": 5.121403408612672e-05,
      "loss": 2.2483,
      "step": 3481
    },
    {
      "epoch": 0.6964,
      "grad_norm": 0.043661873787641525,
      "learning_rate": 5.115310452247386e-05,
      "loss": 2.2495,
      "step": 3482
    },
    {
      "epoch": 0.6966,
      "grad_norm": 0.046714868396520615,
      "learning_rate": 5.1092198766204415e-05,
      "loss": 2.2492,
      "step": 3483
    },
    {
      "epoch": 0.6968,
      "grad_norm": 0.045469049364328384,
      "learning_rate": 5.103131684700314e-05,
      "loss": 2.2503,
      "step": 3484
    },
    {
      "epoch": 0.697,
      "grad_norm": 0.048064231872558594,
      "learning_rate": 5.097045879454313e-05,
      "loss": 2.2489,
      "step": 3485
    },
    {
      "epoch": 0.6972,
      "grad_norm": 0.04721171036362648,
      "learning_rate": 5.090962463848592e-05,
      "loss": 2.248,
      "step": 3486
    },
    {
      "epoch": 0.6974,
      "grad_norm": 0.04461055248975754,
      "learning_rate": 5.0848814408481305e-05,
      "loss": 2.2488,
      "step": 3487
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.057256605476140976,
      "learning_rate": 5.0788028134167456e-05,
      "loss": 2.2481,
      "step": 3488
    },
    {
      "epoch": 0.6978,
      "grad_norm": 0.057538293302059174,
      "learning_rate": 5.072726584517086e-05,
      "loss": 2.2475,
      "step": 3489
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.043430134654045105,
      "learning_rate": 5.066652757110628e-05,
      "loss": 2.2468,
      "step": 3490
    },
    {
      "epoch": 0.6982,
      "grad_norm": 0.05868206545710564,
      "learning_rate": 5.0605813341576924e-05,
      "loss": 2.249,
      "step": 3491
    },
    {
      "epoch": 0.6984,
      "grad_norm": 0.0562085285782814,
      "learning_rate": 5.054512318617406e-05,
      "loss": 2.2477,
      "step": 3492
    },
    {
      "epoch": 0.6986,
      "grad_norm": 0.04367415979504585,
      "learning_rate": 5.048445713447738e-05,
      "loss": 2.2482,
      "step": 3493
    },
    {
      "epoch": 0.6988,
      "grad_norm": 0.053822606801986694,
      "learning_rate": 5.0423815216054724e-05,
      "loss": 2.2494,
      "step": 3494
    },
    {
      "epoch": 0.699,
      "grad_norm": 0.046439029276371,
      "learning_rate": 5.036319746046232e-05,
      "loss": 2.2482,
      "step": 3495
    },
    {
      "epoch": 0.6992,
      "grad_norm": 0.04509136453270912,
      "learning_rate": 5.0302603897244474e-05,
      "loss": 2.2485,
      "step": 3496
    },
    {
      "epoch": 0.6994,
      "grad_norm": 0.04795043170452118,
      "learning_rate": 5.024203455593375e-05,
      "loss": 2.2486,
      "step": 3497
    },
    {
      "epoch": 0.6996,
      "grad_norm": 0.04343557730317116,
      "learning_rate": 5.018148946605092e-05,
      "loss": 2.2464,
      "step": 3498
    },
    {
      "epoch": 0.6998,
      "grad_norm": 0.04713596776127815,
      "learning_rate": 5.012096865710494e-05,
      "loss": 2.2479,
      "step": 3499
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.05017903447151184,
      "learning_rate": 5.006047215859289e-05,
      "loss": 2.2491,
      "step": 3500
    },
    {
      "epoch": 0.7002,
      "grad_norm": 0.043422892689704895,
      "learning_rate": 5.000000000000002e-05,
      "loss": 2.2488,
      "step": 3501
    },
    {
      "epoch": 0.7004,
      "grad_norm": 0.0480036623775959,
      "learning_rate": 4.9939552210799755e-05,
      "loss": 2.2464,
      "step": 3502
    },
    {
      "epoch": 0.7006,
      "grad_norm": 0.04788430780172348,
      "learning_rate": 4.98791288204536e-05,
      "loss": 2.2495,
      "step": 3503
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.04427356645464897,
      "learning_rate": 4.981872985841115e-05,
      "loss": 2.2477,
      "step": 3504
    },
    {
      "epoch": 0.701,
      "grad_norm": 0.05334961786866188,
      "learning_rate": 4.97583553541102e-05,
      "loss": 2.249,
      "step": 3505
    },
    {
      "epoch": 0.7012,
      "grad_norm": 0.04849403351545334,
      "learning_rate": 4.969800533697649e-05,
      "loss": 2.2462,
      "step": 3506
    },
    {
      "epoch": 0.7014,
      "grad_norm": 0.048140574246644974,
      "learning_rate": 4.9637679836423924e-05,
      "loss": 2.2405,
      "step": 3507
    },
    {
      "epoch": 0.7016,
      "grad_norm": 0.06544103473424911,
      "learning_rate": 4.957737888185439e-05,
      "loss": 2.2436,
      "step": 3508
    },
    {
      "epoch": 0.7018,
      "grad_norm": 0.0529162622988224,
      "learning_rate": 4.9517102502657845e-05,
      "loss": 2.2474,
      "step": 3509
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.046926092356443405,
      "learning_rate": 4.945685072821227e-05,
      "loss": 2.2475,
      "step": 3510
    },
    {
      "epoch": 0.7022,
      "grad_norm": 0.054402850568294525,
      "learning_rate": 4.939662358788364e-05,
      "loss": 2.2485,
      "step": 3511
    },
    {
      "epoch": 0.7024,
      "grad_norm": 0.044252630323171616,
      "learning_rate": 4.933642111102594e-05,
      "loss": 2.2482,
      "step": 3512
    },
    {
      "epoch": 0.7026,
      "grad_norm": 0.04883267357945442,
      "learning_rate": 4.927624332698109e-05,
      "loss": 2.2494,
      "step": 3513
    },
    {
      "epoch": 0.7028,
      "grad_norm": 0.04798687621951103,
      "learning_rate": 4.921609026507907e-05,
      "loss": 2.2482,
      "step": 3514
    },
    {
      "epoch": 0.703,
      "grad_norm": 0.0438782274723053,
      "learning_rate": 4.915596195463773e-05,
      "loss": 2.2487,
      "step": 3515
    },
    {
      "epoch": 0.7032,
      "grad_norm": 0.05514473095536232,
      "learning_rate": 4.909585842496287e-05,
      "loss": 2.2481,
      "step": 3516
    },
    {
      "epoch": 0.7034,
      "grad_norm": 0.052665505558252335,
      "learning_rate": 4.9035779705348226e-05,
      "loss": 2.2483,
      "step": 3517
    },
    {
      "epoch": 0.7036,
      "grad_norm": 0.043842218816280365,
      "learning_rate": 4.8975725825075435e-05,
      "loss": 2.2503,
      "step": 3518
    },
    {
      "epoch": 0.7038,
      "grad_norm": 0.05288993939757347,
      "learning_rate": 4.8915696813414026e-05,
      "loss": 2.2498,
      "step": 3519
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.04788551479578018,
      "learning_rate": 4.885569269962142e-05,
      "loss": 2.2501,
      "step": 3520
    },
    {
      "epoch": 0.7042,
      "grad_norm": 0.04579389840364456,
      "learning_rate": 4.8795713512942865e-05,
      "loss": 2.2475,
      "step": 3521
    },
    {
      "epoch": 0.7044,
      "grad_norm": 0.05787007138133049,
      "learning_rate": 4.8735759282611516e-05,
      "loss": 2.2497,
      "step": 3522
    },
    {
      "epoch": 0.7046,
      "grad_norm": 0.04911714047193527,
      "learning_rate": 4.867583003784829e-05,
      "loss": 2.2492,
      "step": 3523
    },
    {
      "epoch": 0.7048,
      "grad_norm": 0.04608847573399544,
      "learning_rate": 4.861592580786205e-05,
      "loss": 2.2449,
      "step": 3524
    },
    {
      "epoch": 0.705,
      "grad_norm": 0.053234223276376724,
      "learning_rate": 4.8556046621849346e-05,
      "loss": 2.2496,
      "step": 3525
    },
    {
      "epoch": 0.7052,
      "grad_norm": 0.04437782242894173,
      "learning_rate": 4.8496192508994576e-05,
      "loss": 2.251,
      "step": 3526
    },
    {
      "epoch": 0.7054,
      "grad_norm": 0.051470015197992325,
      "learning_rate": 4.8436363498469906e-05,
      "loss": 2.2537,
      "step": 3527
    },
    {
      "epoch": 0.7056,
      "grad_norm": 0.0587628073990345,
      "learning_rate": 4.837655961943526e-05,
      "loss": 2.2525,
      "step": 3528
    },
    {
      "epoch": 0.7058,
      "grad_norm": 0.04439152032136917,
      "learning_rate": 4.8316780901038314e-05,
      "loss": 2.2459,
      "step": 3529
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.050223130732774734,
      "learning_rate": 4.825702737241452e-05,
      "loss": 2.2486,
      "step": 3530
    },
    {
      "epoch": 0.7062,
      "grad_norm": 0.04903262108564377,
      "learning_rate": 4.8197299062686995e-05,
      "loss": 2.2452,
      "step": 3531
    },
    {
      "epoch": 0.7064,
      "grad_norm": 0.04701049253344536,
      "learning_rate": 4.8137596000966614e-05,
      "loss": 2.2479,
      "step": 3532
    },
    {
      "epoch": 0.7066,
      "grad_norm": 0.06338362395763397,
      "learning_rate": 4.807791821635186e-05,
      "loss": 2.2487,
      "step": 3533
    },
    {
      "epoch": 0.7068,
      "grad_norm": 0.05234332010149956,
      "learning_rate": 4.8018265737929044e-05,
      "loss": 2.2455,
      "step": 3534
    },
    {
      "epoch": 0.707,
      "grad_norm": 0.050354063510894775,
      "learning_rate": 4.7958638594772064e-05,
      "loss": 2.2483,
      "step": 3535
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.0668630450963974,
      "learning_rate": 4.78990368159424e-05,
      "loss": 2.2497,
      "step": 3536
    },
    {
      "epoch": 0.7074,
      "grad_norm": 0.045763954520225525,
      "learning_rate": 4.783946043048923e-05,
      "loss": 2.2485,
      "step": 3537
    },
    {
      "epoch": 0.7076,
      "grad_norm": 0.057782646268606186,
      "learning_rate": 4.7779909467449414e-05,
      "loss": 2.2501,
      "step": 3538
    },
    {
      "epoch": 0.7078,
      "grad_norm": 0.05471006780862808,
      "learning_rate": 4.7720383955847345e-05,
      "loss": 2.2492,
      "step": 3539
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.04825177416205406,
      "learning_rate": 4.766088392469506e-05,
      "loss": 2.2494,
      "step": 3540
    },
    {
      "epoch": 0.7082,
      "grad_norm": 0.06014271080493927,
      "learning_rate": 4.7601409402992106e-05,
      "loss": 2.2494,
      "step": 3541
    },
    {
      "epoch": 0.7084,
      "grad_norm": 0.048671383410692215,
      "learning_rate": 4.7541960419725626e-05,
      "loss": 2.2504,
      "step": 3542
    },
    {
      "epoch": 0.7086,
      "grad_norm": 0.05505266785621643,
      "learning_rate": 4.748253700387042e-05,
      "loss": 2.2483,
      "step": 3543
    },
    {
      "epoch": 0.7088,
      "grad_norm": 0.0617169588804245,
      "learning_rate": 4.742313918438872e-05,
      "loss": 2.2472,
      "step": 3544
    },
    {
      "epoch": 0.709,
      "grad_norm": 0.043992217630147934,
      "learning_rate": 4.736376699023023e-05,
      "loss": 2.2478,
      "step": 3545
    },
    {
      "epoch": 0.7092,
      "grad_norm": 0.06278485059738159,
      "learning_rate": 4.7304420450332244e-05,
      "loss": 2.2475,
      "step": 3546
    },
    {
      "epoch": 0.7094,
      "grad_norm": 0.04960522800683975,
      "learning_rate": 4.724509959361961e-05,
      "loss": 2.2505,
      "step": 3547
    },
    {
      "epoch": 0.7096,
      "grad_norm": 0.053978946059942245,
      "learning_rate": 4.7185804449004565e-05,
      "loss": 2.2481,
      "step": 3548
    },
    {
      "epoch": 0.7098,
      "grad_norm": 0.056778013706207275,
      "learning_rate": 4.712653504538683e-05,
      "loss": 2.2493,
      "step": 3549
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.045877501368522644,
      "learning_rate": 4.706729141165361e-05,
      "loss": 2.2503,
      "step": 3550
    },
    {
      "epoch": 0.7102,
      "grad_norm": 0.06248707324266434,
      "learning_rate": 4.700807357667952e-05,
      "loss": 2.2489,
      "step": 3551
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.04495299980044365,
      "learning_rate": 4.694888156932658e-05,
      "loss": 2.2465,
      "step": 3552
    },
    {
      "epoch": 0.7106,
      "grad_norm": 0.053209543228149414,
      "learning_rate": 4.688971541844436e-05,
      "loss": 2.249,
      "step": 3553
    },
    {
      "epoch": 0.7108,
      "grad_norm": 0.0537678562104702,
      "learning_rate": 4.6830575152869616e-05,
      "loss": 2.2464,
      "step": 3554
    },
    {
      "epoch": 0.711,
      "grad_norm": 0.04389020800590515,
      "learning_rate": 4.6771460801426635e-05,
      "loss": 2.248,
      "step": 3555
    },
    {
      "epoch": 0.7112,
      "grad_norm": 0.050975650548934937,
      "learning_rate": 4.6712372392927e-05,
      "loss": 2.2468,
      "step": 3556
    },
    {
      "epoch": 0.7114,
      "grad_norm": 0.04619363322854042,
      "learning_rate": 4.665330995616974e-05,
      "loss": 2.2453,
      "step": 3557
    },
    {
      "epoch": 0.7116,
      "grad_norm": 0.048186756670475006,
      "learning_rate": 4.6594273519941154e-05,
      "loss": 2.2515,
      "step": 3558
    },
    {
      "epoch": 0.7118,
      "grad_norm": 0.054643336683511734,
      "learning_rate": 4.653526311301488e-05,
      "loss": 2.2501,
      "step": 3559
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.044812604784965515,
      "learning_rate": 4.647627876415186e-05,
      "loss": 2.248,
      "step": 3560
    },
    {
      "epoch": 0.7122,
      "grad_norm": 0.04733280837535858,
      "learning_rate": 4.6417320502100316e-05,
      "loss": 2.2466,
      "step": 3561
    },
    {
      "epoch": 0.7124,
      "grad_norm": 0.044027235358953476,
      "learning_rate": 4.6358388355595904e-05,
      "loss": 2.2489,
      "step": 3562
    },
    {
      "epoch": 0.7126,
      "grad_norm": 0.04806281998753548,
      "learning_rate": 4.629948235336133e-05,
      "loss": 2.247,
      "step": 3563
    },
    {
      "epoch": 0.7128,
      "grad_norm": 0.04913327097892761,
      "learning_rate": 4.62406025241067e-05,
      "loss": 2.246,
      "step": 3564
    },
    {
      "epoch": 0.713,
      "grad_norm": 0.04424583539366722,
      "learning_rate": 4.6181748896529273e-05,
      "loss": 2.2481,
      "step": 3565
    },
    {
      "epoch": 0.7132,
      "grad_norm": 0.04952290654182434,
      "learning_rate": 4.612292149931369e-05,
      "loss": 2.2475,
      "step": 3566
    },
    {
      "epoch": 0.7134,
      "grad_norm": 0.04396660998463631,
      "learning_rate": 4.6064120361131656e-05,
      "loss": 2.2501,
      "step": 3567
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.053504690527915955,
      "learning_rate": 4.600534551064215e-05,
      "loss": 2.2459,
      "step": 3568
    },
    {
      "epoch": 0.7138,
      "grad_norm": 0.04564760625362396,
      "learning_rate": 4.5946596976491295e-05,
      "loss": 2.2484,
      "step": 3569
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.05741512030363083,
      "learning_rate": 4.588787478731242e-05,
      "loss": 2.2483,
      "step": 3570
    },
    {
      "epoch": 0.7142,
      "grad_norm": 0.0637587457895279,
      "learning_rate": 4.582917897172603e-05,
      "loss": 2.2488,
      "step": 3571
    },
    {
      "epoch": 0.7144,
      "grad_norm": 0.04367498680949211,
      "learning_rate": 4.577050955833973e-05,
      "loss": 2.2484,
      "step": 3572
    },
    {
      "epoch": 0.7146,
      "grad_norm": 0.059962231665849686,
      "learning_rate": 4.5711866575748276e-05,
      "loss": 2.2478,
      "step": 3573
    },
    {
      "epoch": 0.7148,
      "grad_norm": 0.04809006676077843,
      "learning_rate": 4.565325005253356e-05,
      "loss": 2.2479,
      "step": 3574
    },
    {
      "epoch": 0.715,
      "grad_norm": 0.050963059067726135,
      "learning_rate": 4.559466001726451e-05,
      "loss": 2.2468,
      "step": 3575
    },
    {
      "epoch": 0.7152,
      "grad_norm": 0.05730389431118965,
      "learning_rate": 4.5536096498497295e-05,
      "loss": 2.2478,
      "step": 3576
    },
    {
      "epoch": 0.7154,
      "grad_norm": 0.043295718729496,
      "learning_rate": 4.5477559524774994e-05,
      "loss": 2.2471,
      "step": 3577
    },
    {
      "epoch": 0.7156,
      "grad_norm": 0.05685611069202423,
      "learning_rate": 4.541904912462784e-05,
      "loss": 2.2465,
      "step": 3578
    },
    {
      "epoch": 0.7158,
      "grad_norm": 0.05222656577825546,
      "learning_rate": 4.5360565326573104e-05,
      "loss": 2.2476,
      "step": 3579
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.045770976692438126,
      "learning_rate": 4.530210815911504e-05,
      "loss": 2.2475,
      "step": 3580
    },
    {
      "epoch": 0.7162,
      "grad_norm": 0.061393484473228455,
      "learning_rate": 4.524367765074499e-05,
      "loss": 2.2475,
      "step": 3581
    },
    {
      "epoch": 0.7164,
      "grad_norm": 0.043727412819862366,
      "learning_rate": 4.518527382994127e-05,
      "loss": 2.246,
      "step": 3582
    },
    {
      "epoch": 0.7166,
      "grad_norm": 0.06079760938882828,
      "learning_rate": 4.512689672516918e-05,
      "loss": 2.2494,
      "step": 3583
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.058676257729530334,
      "learning_rate": 4.506854636488103e-05,
      "loss": 2.2454,
      "step": 3584
    },
    {
      "epoch": 0.717,
      "grad_norm": 0.04390779882669449,
      "learning_rate": 4.501022277751602e-05,
      "loss": 2.2475,
      "step": 3585
    },
    {
      "epoch": 0.7172,
      "grad_norm": 0.06036575511097908,
      "learning_rate": 4.495192599150044e-05,
      "loss": 2.2475,
      "step": 3586
    },
    {
      "epoch": 0.7174,
      "grad_norm": 0.049313150346279144,
      "learning_rate": 4.48936560352474e-05,
      "loss": 2.2453,
      "step": 3587
    },
    {
      "epoch": 0.7176,
      "grad_norm": 0.049007829278707504,
      "learning_rate": 4.483541293715698e-05,
      "loss": 2.2457,
      "step": 3588
    },
    {
      "epoch": 0.7178,
      "grad_norm": 0.054551467299461365,
      "learning_rate": 4.4777196725616146e-05,
      "loss": 2.2467,
      "step": 3589
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.04323901981115341,
      "learning_rate": 4.471900742899876e-05,
      "loss": 2.2463,
      "step": 3590
    },
    {
      "epoch": 0.7182,
      "grad_norm": 0.05348814278841019,
      "learning_rate": 4.46608450756656e-05,
      "loss": 2.2458,
      "step": 3591
    },
    {
      "epoch": 0.7184,
      "grad_norm": 0.048908691853284836,
      "learning_rate": 4.460270969396429e-05,
      "loss": 2.2474,
      "step": 3592
    },
    {
      "epoch": 0.7186,
      "grad_norm": 0.04417140781879425,
      "learning_rate": 4.45446013122293e-05,
      "loss": 2.2465,
      "step": 3593
    },
    {
      "epoch": 0.7188,
      "grad_norm": 0.0485621877014637,
      "learning_rate": 4.44865199587819e-05,
      "loss": 2.2482,
      "step": 3594
    },
    {
      "epoch": 0.719,
      "grad_norm": 0.04518631845712662,
      "learning_rate": 4.4428465661930343e-05,
      "loss": 2.2478,
      "step": 3595
    },
    {
      "epoch": 0.7192,
      "grad_norm": 0.0530904084444046,
      "learning_rate": 4.437043844996952e-05,
      "loss": 2.2495,
      "step": 3596
    },
    {
      "epoch": 0.7194,
      "grad_norm": 0.04286698251962662,
      "learning_rate": 4.431243835118124e-05,
      "loss": 2.247,
      "step": 3597
    },
    {
      "epoch": 0.7196,
      "grad_norm": 0.046682532876729965,
      "learning_rate": 4.425446539383393e-05,
      "loss": 2.2483,
      "step": 3598
    },
    {
      "epoch": 0.7198,
      "grad_norm": 0.04658105969429016,
      "learning_rate": 4.419651960618302e-05,
      "loss": 2.2479,
      "step": 3599
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.058978933840990067,
      "learning_rate": 4.413860101647055e-05,
      "loss": 2.2479,
      "step": 3600
    },
    {
      "epoch": 0.7202,
      "grad_norm": 0.04543992131948471,
      "learning_rate": 4.4080709652925336e-05,
      "loss": 2.2525,
      "step": 3601
    },
    {
      "epoch": 0.7204,
      "grad_norm": 0.05337828770279884,
      "learning_rate": 4.4022845543762915e-05,
      "loss": 2.2524,
      "step": 3602
    },
    {
      "epoch": 0.7206,
      "grad_norm": 0.05029558762907982,
      "learning_rate": 4.396500871718555e-05,
      "loss": 2.2513,
      "step": 3603
    },
    {
      "epoch": 0.7208,
      "grad_norm": 0.045208945870399475,
      "learning_rate": 4.39071992013822e-05,
      "loss": 2.2472,
      "step": 3604
    },
    {
      "epoch": 0.721,
      "grad_norm": 0.05401340126991272,
      "learning_rate": 4.3849417024528564e-05,
      "loss": 2.2465,
      "step": 3605
    },
    {
      "epoch": 0.7212,
      "grad_norm": 0.04380156099796295,
      "learning_rate": 4.379166221478697e-05,
      "loss": 2.25,
      "step": 3606
    },
    {
      "epoch": 0.7214,
      "grad_norm": 0.05205322802066803,
      "learning_rate": 4.3733934800306366e-05,
      "loss": 2.2493,
      "step": 3607
    },
    {
      "epoch": 0.7216,
      "grad_norm": 0.048920467495918274,
      "learning_rate": 4.367623480922236e-05,
      "loss": 2.2491,
      "step": 3608
    },
    {
      "epoch": 0.7218,
      "grad_norm": 0.045232780277729034,
      "learning_rate": 4.361856226965733e-05,
      "loss": 2.2486,
      "step": 3609
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.04844867065548897,
      "learning_rate": 4.356091720972011e-05,
      "loss": 2.2495,
      "step": 3610
    },
    {
      "epoch": 0.7222,
      "grad_norm": 0.04397585615515709,
      "learning_rate": 4.350329965750621e-05,
      "loss": 2.2486,
      "step": 3611
    },
    {
      "epoch": 0.7224,
      "grad_norm": 0.043845485895872116,
      "learning_rate": 4.3445709641097745e-05,
      "loss": 2.2464,
      "step": 3612
    },
    {
      "epoch": 0.7226,
      "grad_norm": 0.043143630027770996,
      "learning_rate": 4.3388147188563325e-05,
      "loss": 2.2461,
      "step": 3613
    },
    {
      "epoch": 0.7228,
      "grad_norm": 0.043425627052783966,
      "learning_rate": 4.333061232795826e-05,
      "loss": 2.2447,
      "step": 3614
    },
    {
      "epoch": 0.723,
      "grad_norm": 0.04606479033827782,
      "learning_rate": 4.327310508732437e-05,
      "loss": 2.2492,
      "step": 3615
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.044024281203746796,
      "learning_rate": 4.32156254946899e-05,
      "loss": 2.2481,
      "step": 3616
    },
    {
      "epoch": 0.7234,
      "grad_norm": 0.04332940652966499,
      "learning_rate": 4.315817357806974e-05,
      "loss": 2.2502,
      "step": 3617
    },
    {
      "epoch": 0.7236,
      "grad_norm": 0.043427590280771255,
      "learning_rate": 4.310074936546521e-05,
      "loss": 2.247,
      "step": 3618
    },
    {
      "epoch": 0.7238,
      "grad_norm": 0.04367993399500847,
      "learning_rate": 4.304335288486426e-05,
      "loss": 2.2474,
      "step": 3619
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.044296037405729294,
      "learning_rate": 4.29859841642412e-05,
      "loss": 2.2473,
      "step": 3620
    },
    {
      "epoch": 0.7242,
      "grad_norm": 0.047012560069561005,
      "learning_rate": 4.2928643231556844e-05,
      "loss": 2.2429,
      "step": 3621
    },
    {
      "epoch": 0.7244,
      "grad_norm": 0.0438995324075222,
      "learning_rate": 4.287133011475847e-05,
      "loss": 2.2471,
      "step": 3622
    },
    {
      "epoch": 0.7246,
      "grad_norm": 0.04712672531604767,
      "learning_rate": 4.2814044841779745e-05,
      "loss": 2.2481,
      "step": 3623
    },
    {
      "epoch": 0.7248,
      "grad_norm": 0.048081472516059875,
      "learning_rate": 4.2756787440540936e-05,
      "loss": 2.2477,
      "step": 3624
    },
    {
      "epoch": 0.725,
      "grad_norm": 0.04498900845646858,
      "learning_rate": 4.26995579389485e-05,
      "loss": 2.2481,
      "step": 3625
    },
    {
      "epoch": 0.7252,
      "grad_norm": 0.055248841643333435,
      "learning_rate": 4.264235636489542e-05,
      "loss": 2.2462,
      "step": 3626
    },
    {
      "epoch": 0.7254,
      "grad_norm": 0.04619801789522171,
      "learning_rate": 4.2585182746261035e-05,
      "loss": 2.2519,
      "step": 3627
    },
    {
      "epoch": 0.7256,
      "grad_norm": 0.05548837035894394,
      "learning_rate": 4.252803711091112e-05,
      "loss": 2.2507,
      "step": 3628
    },
    {
      "epoch": 0.7258,
      "grad_norm": 0.04703717306256294,
      "learning_rate": 4.2470919486697744e-05,
      "loss": 2.2469,
      "step": 3629
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.04705357551574707,
      "learning_rate": 4.2413829901459344e-05,
      "loss": 2.2487,
      "step": 3630
    },
    {
      "epoch": 0.7262,
      "grad_norm": 0.04878963157534599,
      "learning_rate": 4.235676838302068e-05,
      "loss": 2.2487,
      "step": 3631
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.04336884617805481,
      "learning_rate": 4.229973495919286e-05,
      "loss": 2.2468,
      "step": 3632
    },
    {
      "epoch": 0.7266,
      "grad_norm": 0.04583846032619476,
      "learning_rate": 4.224272965777326e-05,
      "loss": 2.2477,
      "step": 3633
    },
    {
      "epoch": 0.7268,
      "grad_norm": 0.043528661131858826,
      "learning_rate": 4.2185752506545585e-05,
      "loss": 2.2479,
      "step": 3634
    },
    {
      "epoch": 0.727,
      "grad_norm": 0.04734785854816437,
      "learning_rate": 4.212880353327979e-05,
      "loss": 2.249,
      "step": 3635
    },
    {
      "epoch": 0.7272,
      "grad_norm": 0.05467696860432625,
      "learning_rate": 4.207188276573214e-05,
      "loss": 2.2476,
      "step": 3636
    },
    {
      "epoch": 0.7274,
      "grad_norm": 0.04711592197418213,
      "learning_rate": 4.201499023164508e-05,
      "loss": 2.2469,
      "step": 3637
    },
    {
      "epoch": 0.7276,
      "grad_norm": 0.04849999397993088,
      "learning_rate": 4.195812595874739e-05,
      "loss": 2.2489,
      "step": 3638
    },
    {
      "epoch": 0.7278,
      "grad_norm": 0.051871176809072495,
      "learning_rate": 4.190128997475402e-05,
      "loss": 2.2475,
      "step": 3639
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.0431196503341198,
      "learning_rate": 4.184448230736613e-05,
      "loss": 2.2488,
      "step": 3640
    },
    {
      "epoch": 0.7282,
      "grad_norm": 0.06125929579138756,
      "learning_rate": 4.1787702984271074e-05,
      "loss": 2.249,
      "step": 3641
    },
    {
      "epoch": 0.7284,
      "grad_norm": 0.05971783399581909,
      "learning_rate": 4.173095203314241e-05,
      "loss": 2.2468,
      "step": 3642
    },
    {
      "epoch": 0.7286,
      "grad_norm": 0.04364870861172676,
      "learning_rate": 4.167422948163986e-05,
      "loss": 2.2497,
      "step": 3643
    },
    {
      "epoch": 0.7288,
      "grad_norm": 0.05424895137548447,
      "learning_rate": 4.161753535740932e-05,
      "loss": 2.249,
      "step": 3644
    },
    {
      "epoch": 0.729,
      "grad_norm": 0.04458721727132797,
      "learning_rate": 4.15608696880828e-05,
      "loss": 2.2495,
      "step": 3645
    },
    {
      "epoch": 0.7292,
      "grad_norm": 0.05307374894618988,
      "learning_rate": 4.150423250127845e-05,
      "loss": 2.2485,
      "step": 3646
    },
    {
      "epoch": 0.7294,
      "grad_norm": 0.05517127737402916,
      "learning_rate": 4.144762382460059e-05,
      "loss": 2.2486,
      "step": 3647
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.046195898205041885,
      "learning_rate": 4.1391043685639576e-05,
      "loss": 2.2468,
      "step": 3648
    },
    {
      "epoch": 0.7298,
      "grad_norm": 0.065590500831604,
      "learning_rate": 4.133449211197188e-05,
      "loss": 2.2476,
      "step": 3649
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.047206420451402664,
      "learning_rate": 4.1277969131160045e-05,
      "loss": 2.2477,
      "step": 3650
    },
    {
      "epoch": 0.7302,
      "grad_norm": 0.05116748809814453,
      "learning_rate": 4.12214747707527e-05,
      "loss": 2.2475,
      "step": 3651
    },
    {
      "epoch": 0.7304,
      "grad_norm": 0.0490836501121521,
      "learning_rate": 4.11650090582845e-05,
      "loss": 2.2462,
      "step": 3652
    },
    {
      "epoch": 0.7306,
      "grad_norm": 0.044095251709222794,
      "learning_rate": 4.110857202127615e-05,
      "loss": 2.2479,
      "step": 3653
    },
    {
      "epoch": 0.7308,
      "grad_norm": 0.049163125455379486,
      "learning_rate": 4.1052163687234366e-05,
      "loss": 2.2479,
      "step": 3654
    },
    {
      "epoch": 0.731,
      "grad_norm": 0.04446861520409584,
      "learning_rate": 4.099578408365191e-05,
      "loss": 2.2489,
      "step": 3655
    },
    {
      "epoch": 0.7312,
      "grad_norm": 0.04552249610424042,
      "learning_rate": 4.093943323800745e-05,
      "loss": 2.2464,
      "step": 3656
    },
    {
      "epoch": 0.7314,
      "grad_norm": 0.04398418962955475,
      "learning_rate": 4.08831111777658e-05,
      "loss": 2.2497,
      "step": 3657
    },
    {
      "epoch": 0.7316,
      "grad_norm": 0.04614340513944626,
      "learning_rate": 4.08268179303776e-05,
      "loss": 2.2492,
      "step": 3658
    },
    {
      "epoch": 0.7318,
      "grad_norm": 0.04563388228416443,
      "learning_rate": 4.077055352327953e-05,
      "loss": 2.2464,
      "step": 3659
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.04353368654847145,
      "learning_rate": 4.071431798389408e-05,
      "loss": 2.2471,
      "step": 3660
    },
    {
      "epoch": 0.7322,
      "grad_norm": 0.045966148376464844,
      "learning_rate": 4.065811133962987e-05,
      "loss": 2.2455,
      "step": 3661
    },
    {
      "epoch": 0.7324,
      "grad_norm": 0.04351382330060005,
      "learning_rate": 4.0601933617881294e-05,
      "loss": 2.2485,
      "step": 3662
    },
    {
      "epoch": 0.7326,
      "grad_norm": 0.04356320947408676,
      "learning_rate": 4.05457848460287e-05,
      "loss": 2.2475,
      "step": 3663
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.0430820994079113,
      "learning_rate": 4.048966505143831e-05,
      "loss": 2.2483,
      "step": 3664
    },
    {
      "epoch": 0.733,
      "grad_norm": 0.04423351213335991,
      "learning_rate": 4.0433574261462206e-05,
      "loss": 2.2487,
      "step": 3665
    },
    {
      "epoch": 0.7332,
      "grad_norm": 0.04399638622999191,
      "learning_rate": 4.037751250343841e-05,
      "loss": 2.2477,
      "step": 3666
    },
    {
      "epoch": 0.7334,
      "grad_norm": 0.04738782346248627,
      "learning_rate": 4.032147980469072e-05,
      "loss": 2.2485,
      "step": 3667
    },
    {
      "epoch": 0.7336,
      "grad_norm": 0.04390334710478783,
      "learning_rate": 4.026547619252883e-05,
      "loss": 2.2472,
      "step": 3668
    },
    {
      "epoch": 0.7338,
      "grad_norm": 0.04457469284534454,
      "learning_rate": 4.020950169424815e-05,
      "loss": 2.249,
      "step": 3669
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.0429653562605381,
      "learning_rate": 4.015355633712996e-05,
      "loss": 2.2477,
      "step": 3670
    },
    {
      "epoch": 0.7342,
      "grad_norm": 0.048050541430711746,
      "learning_rate": 4.009764014844143e-05,
      "loss": 2.2488,
      "step": 3671
    },
    {
      "epoch": 0.7344,
      "grad_norm": 0.045114681124687195,
      "learning_rate": 4.004175315543538e-05,
      "loss": 2.2489,
      "step": 3672
    },
    {
      "epoch": 0.7346,
      "grad_norm": 0.04430629312992096,
      "learning_rate": 3.9985895385350456e-05,
      "loss": 2.2465,
      "step": 3673
    },
    {
      "epoch": 0.7348,
      "grad_norm": 0.043964091688394547,
      "learning_rate": 3.9930066865411075e-05,
      "loss": 2.2467,
      "step": 3674
    },
    {
      "epoch": 0.735,
      "grad_norm": 0.04294174909591675,
      "learning_rate": 3.987426762282733e-05,
      "loss": 2.2463,
      "step": 3675
    },
    {
      "epoch": 0.7352,
      "grad_norm": 0.04341582953929901,
      "learning_rate": 3.981849768479517e-05,
      "loss": 2.2488,
      "step": 3676
    },
    {
      "epoch": 0.7354,
      "grad_norm": 0.04400061070919037,
      "learning_rate": 3.976275707849616e-05,
      "loss": 2.247,
      "step": 3677
    },
    {
      "epoch": 0.7356,
      "grad_norm": 0.04347170516848564,
      "learning_rate": 3.9707045831097555e-05,
      "loss": 2.2475,
      "step": 3678
    },
    {
      "epoch": 0.7358,
      "grad_norm": 0.043140362948179245,
      "learning_rate": 3.9651363969752344e-05,
      "loss": 2.2496,
      "step": 3679
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.043325845152139664,
      "learning_rate": 3.9595711521599224e-05,
      "loss": 2.247,
      "step": 3680
    },
    {
      "epoch": 0.7362,
      "grad_norm": 0.04358040541410446,
      "learning_rate": 3.954008851376252e-05,
      "loss": 2.2432,
      "step": 3681
    },
    {
      "epoch": 0.7364,
      "grad_norm": 0.04347475618124008,
      "learning_rate": 3.948449497335219e-05,
      "loss": 2.2459,
      "step": 3682
    },
    {
      "epoch": 0.7366,
      "grad_norm": 0.04674066603183746,
      "learning_rate": 3.942893092746387e-05,
      "loss": 2.2479,
      "step": 3683
    },
    {
      "epoch": 0.7368,
      "grad_norm": 0.049238257110118866,
      "learning_rate": 3.9373396403178786e-05,
      "loss": 2.247,
      "step": 3684
    },
    {
      "epoch": 0.737,
      "grad_norm": 0.04327504709362984,
      "learning_rate": 3.931789142756377e-05,
      "loss": 2.2468,
      "step": 3685
    },
    {
      "epoch": 0.7372,
      "grad_norm": 0.050731807947158813,
      "learning_rate": 3.9262416027671356e-05,
      "loss": 2.2497,
      "step": 3686
    },
    {
      "epoch": 0.7374,
      "grad_norm": 0.045587535947561264,
      "learning_rate": 3.920697023053949e-05,
      "loss": 2.2486,
      "step": 3687
    },
    {
      "epoch": 0.7376,
      "grad_norm": 0.04399876669049263,
      "learning_rate": 3.915155406319181e-05,
      "loss": 2.2484,
      "step": 3688
    },
    {
      "epoch": 0.7378,
      "grad_norm": 0.04599481821060181,
      "learning_rate": 3.9096167552637454e-05,
      "loss": 2.2461,
      "step": 3689
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.04315338656306267,
      "learning_rate": 3.904081072587119e-05,
      "loss": 2.2458,
      "step": 3690
    },
    {
      "epoch": 0.7382,
      "grad_norm": 0.045589614659547806,
      "learning_rate": 3.8985483609873244e-05,
      "loss": 2.2464,
      "step": 3691
    },
    {
      "epoch": 0.7384,
      "grad_norm": 0.04326820373535156,
      "learning_rate": 3.893018623160938e-05,
      "loss": 2.2473,
      "step": 3692
    },
    {
      "epoch": 0.7386,
      "grad_norm": 0.04417683929204941,
      "learning_rate": 3.887491861803085e-05,
      "loss": 2.2473,
      "step": 3693
    },
    {
      "epoch": 0.7388,
      "grad_norm": 0.04362903907895088,
      "learning_rate": 3.88196807960744e-05,
      "loss": 2.2461,
      "step": 3694
    },
    {
      "epoch": 0.739,
      "grad_norm": 0.04433811828494072,
      "learning_rate": 3.876447279266238e-05,
      "loss": 2.2476,
      "step": 3695
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.049436699599027634,
      "learning_rate": 3.8709294634702376e-05,
      "loss": 2.2492,
      "step": 3696
    },
    {
      "epoch": 0.7394,
      "grad_norm": 0.045295245945453644,
      "learning_rate": 3.8654146349087606e-05,
      "loss": 2.248,
      "step": 3697
    },
    {
      "epoch": 0.7396,
      "grad_norm": 0.04552413523197174,
      "learning_rate": 3.859902796269663e-05,
      "loss": 2.2468,
      "step": 3698
    },
    {
      "epoch": 0.7398,
      "grad_norm": 0.0455162413418293,
      "learning_rate": 3.854393950239355e-05,
      "loss": 2.2495,
      "step": 3699
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.05190965160727501,
      "learning_rate": 3.8488880995027786e-05,
      "loss": 2.2466,
      "step": 3700
    },
    {
      "epoch": 0.7402,
      "grad_norm": 0.04755708575248718,
      "learning_rate": 3.843385246743417e-05,
      "loss": 2.2481,
      "step": 3701
    },
    {
      "epoch": 0.7404,
      "grad_norm": 0.046713944524526596,
      "learning_rate": 3.837885394643296e-05,
      "loss": 2.2467,
      "step": 3702
    },
    {
      "epoch": 0.7406,
      "grad_norm": 0.0491618812084198,
      "learning_rate": 3.832388545882976e-05,
      "loss": 2.2478,
      "step": 3703
    },
    {
      "epoch": 0.7408,
      "grad_norm": 0.043353836983442307,
      "learning_rate": 3.8268947031415514e-05,
      "loss": 2.2471,
      "step": 3704
    },
    {
      "epoch": 0.741,
      "grad_norm": 0.045286476612091064,
      "learning_rate": 3.821403869096658e-05,
      "loss": 2.2482,
      "step": 3705
    },
    {
      "epoch": 0.7412,
      "grad_norm": 0.043359577655792236,
      "learning_rate": 3.8159160464244606e-05,
      "loss": 2.245,
      "step": 3706
    },
    {
      "epoch": 0.7414,
      "grad_norm": 0.04638367518782616,
      "learning_rate": 3.8104312377996564e-05,
      "loss": 2.2479,
      "step": 3707
    },
    {
      "epoch": 0.7416,
      "grad_norm": 0.043197859078645706,
      "learning_rate": 3.8049494458954725e-05,
      "loss": 2.2456,
      "step": 3708
    },
    {
      "epoch": 0.7418,
      "grad_norm": 0.04570884630084038,
      "learning_rate": 3.799470673383674e-05,
      "loss": 2.2473,
      "step": 3709
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.04482327774167061,
      "learning_rate": 3.793994922934544e-05,
      "loss": 2.2503,
      "step": 3710
    },
    {
      "epoch": 0.7422,
      "grad_norm": 0.04396402835845947,
      "learning_rate": 3.788522197216897e-05,
      "loss": 2.2465,
      "step": 3711
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.04321090877056122,
      "learning_rate": 3.783052498898073e-05,
      "loss": 2.248,
      "step": 3712
    },
    {
      "epoch": 0.7426,
      "grad_norm": 0.04404432326555252,
      "learning_rate": 3.777585830643937e-05,
      "loss": 2.2505,
      "step": 3713
    },
    {
      "epoch": 0.7428,
      "grad_norm": 0.05103129893541336,
      "learning_rate": 3.7721221951188765e-05,
      "loss": 2.2474,
      "step": 3714
    },
    {
      "epoch": 0.743,
      "grad_norm": 0.05218076333403587,
      "learning_rate": 3.766661594985801e-05,
      "loss": 2.2463,
      "step": 3715
    },
    {
      "epoch": 0.7432,
      "grad_norm": 0.04713495448231697,
      "learning_rate": 3.7612040329061405e-05,
      "loss": 2.2478,
      "step": 3716
    },
    {
      "epoch": 0.7434,
      "grad_norm": 0.052727509289979935,
      "learning_rate": 3.755749511539845e-05,
      "loss": 2.2474,
      "step": 3717
    },
    {
      "epoch": 0.7436,
      "grad_norm": 0.04376762732863426,
      "learning_rate": 3.7502980335453774e-05,
      "loss": 2.248,
      "step": 3718
    },
    {
      "epoch": 0.7438,
      "grad_norm": 0.04622356593608856,
      "learning_rate": 3.7448496015797295e-05,
      "loss": 2.2471,
      "step": 3719
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.044147830456495285,
      "learning_rate": 3.739404218298398e-05,
      "loss": 2.2496,
      "step": 3720
    },
    {
      "epoch": 0.7442,
      "grad_norm": 0.044175393879413605,
      "learning_rate": 3.733961886355398e-05,
      "loss": 2.2518,
      "step": 3721
    },
    {
      "epoch": 0.7444,
      "grad_norm": 0.04526311531662941,
      "learning_rate": 3.728522608403249e-05,
      "loss": 2.2488,
      "step": 3722
    },
    {
      "epoch": 0.7446,
      "grad_norm": 0.04371926560997963,
      "learning_rate": 3.7230863870929964e-05,
      "loss": 2.2492,
      "step": 3723
    },
    {
      "epoch": 0.7448,
      "grad_norm": 0.04314769431948662,
      "learning_rate": 3.717653225074186e-05,
      "loss": 2.2466,
      "step": 3724
    },
    {
      "epoch": 0.745,
      "grad_norm": 0.04311686381697655,
      "learning_rate": 3.712223124994875e-05,
      "loss": 2.2461,
      "step": 3725
    },
    {
      "epoch": 0.7452,
      "grad_norm": 0.04570721462368965,
      "learning_rate": 3.7067960895016275e-05,
      "loss": 2.2496,
      "step": 3726
    },
    {
      "epoch": 0.7454,
      "grad_norm": 0.042942460626363754,
      "learning_rate": 3.701372121239512e-05,
      "loss": 2.2467,
      "step": 3727
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.043917812407016754,
      "learning_rate": 3.695951222852112e-05,
      "loss": 2.2459,
      "step": 3728
    },
    {
      "epoch": 0.7458,
      "grad_norm": 0.043362025171518326,
      "learning_rate": 3.690533396981504e-05,
      "loss": 2.2459,
      "step": 3729
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.04659676551818848,
      "learning_rate": 3.685118646268272e-05,
      "loss": 2.2479,
      "step": 3730
    },
    {
      "epoch": 0.7462,
      "grad_norm": 0.04457828775048256,
      "learning_rate": 3.679706973351491e-05,
      "loss": 2.2497,
      "step": 3731
    },
    {
      "epoch": 0.7464,
      "grad_norm": 0.046061549335718155,
      "learning_rate": 3.674298380868756e-05,
      "loss": 2.2483,
      "step": 3732
    },
    {
      "epoch": 0.7466,
      "grad_norm": 0.044102419167757034,
      "learning_rate": 3.668892871456144e-05,
      "loss": 2.2472,
      "step": 3733
    },
    {
      "epoch": 0.7468,
      "grad_norm": 0.04528801515698433,
      "learning_rate": 3.6634904477482354e-05,
      "loss": 2.2476,
      "step": 3734
    },
    {
      "epoch": 0.747,
      "grad_norm": 0.04598249867558479,
      "learning_rate": 3.6580911123781056e-05,
      "loss": 2.2491,
      "step": 3735
    },
    {
      "epoch": 0.7472,
      "grad_norm": 0.044027067720890045,
      "learning_rate": 3.6526948679773257e-05,
      "loss": 2.2481,
      "step": 3736
    },
    {
      "epoch": 0.7474,
      "grad_norm": 0.044715240597724915,
      "learning_rate": 3.647301717175956e-05,
      "loss": 2.2468,
      "step": 3737
    },
    {
      "epoch": 0.7476,
      "grad_norm": 0.04344501718878746,
      "learning_rate": 3.641911662602559e-05,
      "loss": 2.2469,
      "step": 3738
    },
    {
      "epoch": 0.7478,
      "grad_norm": 0.04402906447649002,
      "learning_rate": 3.6365247068841814e-05,
      "loss": 2.2482,
      "step": 3739
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.044257309287786484,
      "learning_rate": 3.631140852646355e-05,
      "loss": 2.2497,
      "step": 3740
    },
    {
      "epoch": 0.7482,
      "grad_norm": 0.0438835583627224,
      "learning_rate": 3.6257601025131026e-05,
      "loss": 2.2466,
      "step": 3741
    },
    {
      "epoch": 0.7484,
      "grad_norm": 0.04407086223363876,
      "learning_rate": 3.6203824591069456e-05,
      "loss": 2.2481,
      "step": 3742
    },
    {
      "epoch": 0.7486,
      "grad_norm": 0.043893713504076004,
      "learning_rate": 3.615007925048878e-05,
      "loss": 2.2496,
      "step": 3743
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.044255420565605164,
      "learning_rate": 3.60963650295838e-05,
      "loss": 2.2496,
      "step": 3744
    },
    {
      "epoch": 0.749,
      "grad_norm": 0.043120428919792175,
      "learning_rate": 3.604268195453421e-05,
      "loss": 2.2468,
      "step": 3745
    },
    {
      "epoch": 0.7492,
      "grad_norm": 0.04452433064579964,
      "learning_rate": 3.5989030051504434e-05,
      "loss": 2.2489,
      "step": 3746
    },
    {
      "epoch": 0.7494,
      "grad_norm": 0.04361582547426224,
      "learning_rate": 3.5935409346643835e-05,
      "loss": 2.2466,
      "step": 3747
    },
    {
      "epoch": 0.7496,
      "grad_norm": 0.045242439955472946,
      "learning_rate": 3.5881819866086484e-05,
      "loss": 2.2447,
      "step": 3748
    },
    {
      "epoch": 0.7498,
      "grad_norm": 0.04305049404501915,
      "learning_rate": 3.582826163595119e-05,
      "loss": 2.2466,
      "step": 3749
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.044138334691524506,
      "learning_rate": 3.577473468234156e-05,
      "loss": 2.247,
      "step": 3750
    },
    {
      "epoch": 0.7502,
      "grad_norm": 0.04382839798927307,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 2.2474,
      "step": 3751
    },
    {
      "epoch": 0.7504,
      "grad_norm": 0.04434545710682869,
      "learning_rate": 3.56677747090378e-05,
      "loss": 2.2475,
      "step": 3752
    },
    {
      "epoch": 0.7506,
      "grad_norm": 0.04364481568336487,
      "learning_rate": 3.561434174147463e-05,
      "loss": 2.2471,
      "step": 3753
    },
    {
      "epoch": 0.7508,
      "grad_norm": 0.04604643955826759,
      "learning_rate": 3.556094015469913e-05,
      "loss": 2.2478,
      "step": 3754
    },
    {
      "epoch": 0.751,
      "grad_norm": 0.04304558411240578,
      "learning_rate": 3.5507569974738574e-05,
      "loss": 2.2486,
      "step": 3755
    },
    {
      "epoch": 0.7512,
      "grad_norm": 0.04460751265287399,
      "learning_rate": 3.545423122760493e-05,
      "loss": 2.2464,
      "step": 3756
    },
    {
      "epoch": 0.7514,
      "grad_norm": 0.04298251122236252,
      "learning_rate": 3.540092393929494e-05,
      "loss": 2.2468,
      "step": 3757
    },
    {
      "epoch": 0.7516,
      "grad_norm": 0.04316344112157822,
      "learning_rate": 3.534764813578982e-05,
      "loss": 2.2475,
      "step": 3758
    },
    {
      "epoch": 0.7518,
      "grad_norm": 0.04334551841020584,
      "learning_rate": 3.52944038430556e-05,
      "loss": 2.2469,
      "step": 3759
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.04349389299750328,
      "learning_rate": 3.524119108704286e-05,
      "loss": 2.2499,
      "step": 3760
    },
    {
      "epoch": 0.7522,
      "grad_norm": 0.042937010526657104,
      "learning_rate": 3.518800989368691e-05,
      "loss": 2.2444,
      "step": 3761
    },
    {
      "epoch": 0.7524,
      "grad_norm": 0.042970653623342514,
      "learning_rate": 3.51348602889076e-05,
      "loss": 2.2466,
      "step": 3762
    },
    {
      "epoch": 0.7526,
      "grad_norm": 0.04297403246164322,
      "learning_rate": 3.50817422986094e-05,
      "loss": 2.2459,
      "step": 3763
    },
    {
      "epoch": 0.7528,
      "grad_norm": 0.04364338144659996,
      "learning_rate": 3.5028655948681355e-05,
      "loss": 2.248,
      "step": 3764
    },
    {
      "epoch": 0.753,
      "grad_norm": 0.044651098549366,
      "learning_rate": 3.497560126499709e-05,
      "loss": 2.2473,
      "step": 3765
    },
    {
      "epoch": 0.7532,
      "grad_norm": 0.04299464821815491,
      "learning_rate": 3.492257827341492e-05,
      "loss": 2.2482,
      "step": 3766
    },
    {
      "epoch": 0.7534,
      "grad_norm": 0.043776825070381165,
      "learning_rate": 3.4869586999777495e-05,
      "loss": 2.2472,
      "step": 3767
    },
    {
      "epoch": 0.7536,
      "grad_norm": 0.04348917677998543,
      "learning_rate": 3.481662746991214e-05,
      "loss": 2.2491,
      "step": 3768
    },
    {
      "epoch": 0.7538,
      "grad_norm": 0.044562239199876785,
      "learning_rate": 3.4763699709630716e-05,
      "loss": 2.2461,
      "step": 3769
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.04305689036846161,
      "learning_rate": 3.4710803744729515e-05,
      "loss": 2.248,
      "step": 3770
    },
    {
      "epoch": 0.7542,
      "grad_norm": 0.044059623032808304,
      "learning_rate": 3.465793960098945e-05,
      "loss": 2.2447,
      "step": 3771
    },
    {
      "epoch": 0.7544,
      "grad_norm": 0.04669122025370598,
      "learning_rate": 3.460510730417585e-05,
      "loss": 2.2501,
      "step": 3772
    },
    {
      "epoch": 0.7546,
      "grad_norm": 0.04602985456585884,
      "learning_rate": 3.455230688003852e-05,
      "loss": 2.2475,
      "step": 3773
    },
    {
      "epoch": 0.7548,
      "grad_norm": 0.042857762426137924,
      "learning_rate": 3.4499538354311755e-05,
      "loss": 2.2471,
      "step": 3774
    },
    {
      "epoch": 0.755,
      "grad_norm": 0.04445553570985794,
      "learning_rate": 3.444680175271428e-05,
      "loss": 2.2486,
      "step": 3775
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.0431755892932415,
      "learning_rate": 3.439409710094929e-05,
      "loss": 2.2495,
      "step": 3776
    },
    {
      "epoch": 0.7554,
      "grad_norm": 0.04644105210900307,
      "learning_rate": 3.4341424424704375e-05,
      "loss": 2.2496,
      "step": 3777
    },
    {
      "epoch": 0.7556,
      "grad_norm": 0.045591846108436584,
      "learning_rate": 3.4288783749651564e-05,
      "loss": 2.2469,
      "step": 3778
    },
    {
      "epoch": 0.7558,
      "grad_norm": 0.04371289163827896,
      "learning_rate": 3.423617510144727e-05,
      "loss": 2.2497,
      "step": 3779
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.04374853894114494,
      "learning_rate": 3.418359850573234e-05,
      "loss": 2.2462,
      "step": 3780
    },
    {
      "epoch": 0.7562,
      "grad_norm": 0.043018266558647156,
      "learning_rate": 3.413105398813195e-05,
      "loss": 2.2455,
      "step": 3781
    },
    {
      "epoch": 0.7564,
      "grad_norm": 0.043668802827596664,
      "learning_rate": 3.4078541574255664e-05,
      "loss": 2.2469,
      "step": 3782
    },
    {
      "epoch": 0.7566,
      "grad_norm": 0.04369719326496124,
      "learning_rate": 3.40260612896974e-05,
      "loss": 2.2488,
      "step": 3783
    },
    {
      "epoch": 0.7568,
      "grad_norm": 0.043997857719659805,
      "learning_rate": 3.397361316003539e-05,
      "loss": 2.2484,
      "step": 3784
    },
    {
      "epoch": 0.757,
      "grad_norm": 0.04330883547663689,
      "learning_rate": 3.3921197210832235e-05,
      "loss": 2.2453,
      "step": 3785
    },
    {
      "epoch": 0.7572,
      "grad_norm": 0.04889485985040665,
      "learning_rate": 3.386881346763483e-05,
      "loss": 2.2461,
      "step": 3786
    },
    {
      "epoch": 0.7574,
      "grad_norm": 0.04477841779589653,
      "learning_rate": 3.3816461955974365e-05,
      "loss": 2.2476,
      "step": 3787
    },
    {
      "epoch": 0.7576,
      "grad_norm": 0.04380331188440323,
      "learning_rate": 3.376414270136633e-05,
      "loss": 2.2449,
      "step": 3788
    },
    {
      "epoch": 0.7578,
      "grad_norm": 0.04387950897216797,
      "learning_rate": 3.371185572931048e-05,
      "loss": 2.2487,
      "step": 3789
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.043615542352199554,
      "learning_rate": 3.3659601065290893e-05,
      "loss": 2.2474,
      "step": 3790
    },
    {
      "epoch": 0.7582,
      "grad_norm": 0.042777154594659805,
      "learning_rate": 3.360737873477584e-05,
      "loss": 2.2475,
      "step": 3791
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.04402711242437363,
      "learning_rate": 3.355518876321787e-05,
      "loss": 2.2484,
      "step": 3792
    },
    {
      "epoch": 0.7586,
      "grad_norm": 0.04332396015524864,
      "learning_rate": 3.3503031176053656e-05,
      "loss": 2.2472,
      "step": 3793
    },
    {
      "epoch": 0.7588,
      "grad_norm": 0.04718505218625069,
      "learning_rate": 3.3450905998704275e-05,
      "loss": 2.2478,
      "step": 3794
    },
    {
      "epoch": 0.759,
      "grad_norm": 0.04334387183189392,
      "learning_rate": 3.339881325657484e-05,
      "loss": 2.2454,
      "step": 3795
    },
    {
      "epoch": 0.7592,
      "grad_norm": 0.044340409338474274,
      "learning_rate": 3.334675297505476e-05,
      "loss": 2.2473,
      "step": 3796
    },
    {
      "epoch": 0.7594,
      "grad_norm": 0.04430157691240311,
      "learning_rate": 3.3294725179517574e-05,
      "loss": 2.2465,
      "step": 3797
    },
    {
      "epoch": 0.7596,
      "grad_norm": 0.043489303439855576,
      "learning_rate": 3.3242729895320946e-05,
      "loss": 2.2491,
      "step": 3798
    },
    {
      "epoch": 0.7598,
      "grad_norm": 0.04471897706389427,
      "learning_rate": 3.319076714780682e-05,
      "loss": 2.2475,
      "step": 3799
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.04296937957406044,
      "learning_rate": 3.313883696230119e-05,
      "loss": 2.2482,
      "step": 3800
    },
    {
      "epoch": 0.7602,
      "grad_norm": 0.0435057207942009,
      "learning_rate": 3.308693936411421e-05,
      "loss": 2.2499,
      "step": 3801
    },
    {
      "epoch": 0.7604,
      "grad_norm": 0.04641132056713104,
      "learning_rate": 3.303507437854009e-05,
      "loss": 2.2444,
      "step": 3802
    },
    {
      "epoch": 0.7606,
      "grad_norm": 0.042793817818164825,
      "learning_rate": 3.2983242030857174e-05,
      "loss": 2.2481,
      "step": 3803
    },
    {
      "epoch": 0.7608,
      "grad_norm": 0.0526215024292469,
      "learning_rate": 3.2931442346328004e-05,
      "loss": 2.2475,
      "step": 3804
    },
    {
      "epoch": 0.761,
      "grad_norm": 0.04802757501602173,
      "learning_rate": 3.287967535019908e-05,
      "loss": 2.2488,
      "step": 3805
    },
    {
      "epoch": 0.7612,
      "grad_norm": 0.04561171680688858,
      "learning_rate": 3.2827941067700996e-05,
      "loss": 2.2473,
      "step": 3806
    },
    {
      "epoch": 0.7614,
      "grad_norm": 0.0461902916431427,
      "learning_rate": 3.277623952404842e-05,
      "loss": 2.2464,
      "step": 3807
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.043275829404592514,
      "learning_rate": 3.272457074444003e-05,
      "loss": 2.2485,
      "step": 3808
    },
    {
      "epoch": 0.7618,
      "grad_norm": 0.04559134319424629,
      "learning_rate": 3.2672934754058616e-05,
      "loss": 2.2489,
      "step": 3809
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.0432986244559288,
      "learning_rate": 3.2621331578070934e-05,
      "loss": 2.2471,
      "step": 3810
    },
    {
      "epoch": 0.7622,
      "grad_norm": 0.04367348924279213,
      "learning_rate": 3.2569761241627696e-05,
      "loss": 2.247,
      "step": 3811
    },
    {
      "epoch": 0.7624,
      "grad_norm": 0.04330682009458542,
      "learning_rate": 3.251822376986363e-05,
      "loss": 2.2472,
      "step": 3812
    },
    {
      "epoch": 0.7626,
      "grad_norm": 0.04343635216355324,
      "learning_rate": 3.246671918789755e-05,
      "loss": 2.2488,
      "step": 3813
    },
    {
      "epoch": 0.7628,
      "grad_norm": 0.043006617575883865,
      "learning_rate": 3.2415247520832146e-05,
      "loss": 2.2456,
      "step": 3814
    },
    {
      "epoch": 0.763,
      "grad_norm": 0.04473123699426651,
      "learning_rate": 3.236380879375408e-05,
      "loss": 2.2495,
      "step": 3815
    },
    {
      "epoch": 0.7632,
      "grad_norm": 0.04526223987340927,
      "learning_rate": 3.231240303173394e-05,
      "loss": 2.2486,
      "step": 3816
    },
    {
      "epoch": 0.7634,
      "grad_norm": 0.04338522255420685,
      "learning_rate": 3.226103025982628e-05,
      "loss": 2.2482,
      "step": 3817
    },
    {
      "epoch": 0.7636,
      "grad_norm": 0.044016748666763306,
      "learning_rate": 3.220969050306955e-05,
      "loss": 2.248,
      "step": 3818
    },
    {
      "epoch": 0.7638,
      "grad_norm": 0.04315854236483574,
      "learning_rate": 3.21583837864862e-05,
      "loss": 2.2467,
      "step": 3819
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.04318187013268471,
      "learning_rate": 3.210711013508242e-05,
      "loss": 2.2489,
      "step": 3820
    },
    {
      "epoch": 0.7642,
      "grad_norm": 0.043124351650476456,
      "learning_rate": 3.205586957384838e-05,
      "loss": 2.249,
      "step": 3821
    },
    {
      "epoch": 0.7644,
      "grad_norm": 0.0429903045296669,
      "learning_rate": 3.200466212775808e-05,
      "loss": 2.2467,
      "step": 3822
    },
    {
      "epoch": 0.7646,
      "grad_norm": 0.043964650481939316,
      "learning_rate": 3.195348782176948e-05,
      "loss": 2.2468,
      "step": 3823
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.044888418167829514,
      "learning_rate": 3.190234668082427e-05,
      "loss": 2.2476,
      "step": 3824
    },
    {
      "epoch": 0.765,
      "grad_norm": 0.044221024960279465,
      "learning_rate": 3.1851238729848034e-05,
      "loss": 2.2468,
      "step": 3825
    },
    {
      "epoch": 0.7652,
      "grad_norm": 0.04354993253946304,
      "learning_rate": 3.1800163993750166e-05,
      "loss": 2.248,
      "step": 3826
    },
    {
      "epoch": 0.7654,
      "grad_norm": 0.045575350522994995,
      "learning_rate": 3.174912249742382e-05,
      "loss": 2.2486,
      "step": 3827
    },
    {
      "epoch": 0.7656,
      "grad_norm": 0.0428493358194828,
      "learning_rate": 3.1698114265746124e-05,
      "loss": 2.2473,
      "step": 3828
    },
    {
      "epoch": 0.7658,
      "grad_norm": 0.04470749571919441,
      "learning_rate": 3.164713932357776e-05,
      "loss": 2.2439,
      "step": 3829
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.04421084374189377,
      "learning_rate": 3.159619769576333e-05,
      "loss": 2.2466,
      "step": 3830
    },
    {
      "epoch": 0.7662,
      "grad_norm": 0.043442774564027786,
      "learning_rate": 3.154528940713113e-05,
      "loss": 2.2462,
      "step": 3831
    },
    {
      "epoch": 0.7664,
      "grad_norm": 0.046266328543424606,
      "learning_rate": 3.1494414482493304e-05,
      "loss": 2.2497,
      "step": 3832
    },
    {
      "epoch": 0.7666,
      "grad_norm": 0.06259813904762268,
      "learning_rate": 3.144357294664565e-05,
      "loss": 2.246,
      "step": 3833
    },
    {
      "epoch": 0.7668,
      "grad_norm": 0.04493393748998642,
      "learning_rate": 3.1392764824367704e-05,
      "loss": 2.2486,
      "step": 3834
    },
    {
      "epoch": 0.767,
      "grad_norm": 0.061224814504384995,
      "learning_rate": 3.134199014042274e-05,
      "loss": 2.2465,
      "step": 3835
    },
    {
      "epoch": 0.7672,
      "grad_norm": 0.04586179181933403,
      "learning_rate": 3.129124891955771e-05,
      "loss": 2.2463,
      "step": 3836
    },
    {
      "epoch": 0.7674,
      "grad_norm": 0.061906199902296066,
      "learning_rate": 3.124054118650327e-05,
      "loss": 2.247,
      "step": 3837
    },
    {
      "epoch": 0.7676,
      "grad_norm": 0.04915378615260124,
      "learning_rate": 3.1189866965973766e-05,
      "loss": 2.2451,
      "step": 3838
    },
    {
      "epoch": 0.7678,
      "grad_norm": 0.05077783390879631,
      "learning_rate": 3.113922628266718e-05,
      "loss": 2.2461,
      "step": 3839
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.04416922479867935,
      "learning_rate": 3.108861916126518e-05,
      "loss": 2.2473,
      "step": 3840
    },
    {
      "epoch": 0.7682,
      "grad_norm": 0.054836809635162354,
      "learning_rate": 3.103804562643302e-05,
      "loss": 2.2467,
      "step": 3841
    },
    {
      "epoch": 0.7684,
      "grad_norm": 0.04373233765363693,
      "learning_rate": 3.098750570281969e-05,
      "loss": 2.2477,
      "step": 3842
    },
    {
      "epoch": 0.7686,
      "grad_norm": 0.056486826390028,
      "learning_rate": 3.093699941505771e-05,
      "loss": 2.2465,
      "step": 3843
    },
    {
      "epoch": 0.7688,
      "grad_norm": 0.04563102871179581,
      "learning_rate": 3.0886526787763234e-05,
      "loss": 2.2465,
      "step": 3844
    },
    {
      "epoch": 0.769,
      "grad_norm": 0.0554550401866436,
      "learning_rate": 3.0836087845536e-05,
      "loss": 2.2468,
      "step": 3845
    },
    {
      "epoch": 0.7692,
      "grad_norm": 0.047156114131212234,
      "learning_rate": 3.078568261295933e-05,
      "loss": 2.2422,
      "step": 3846
    },
    {
      "epoch": 0.7694,
      "grad_norm": 0.05189557746052742,
      "learning_rate": 3.073531111460013e-05,
      "loss": 2.2431,
      "step": 3847
    },
    {
      "epoch": 0.7696,
      "grad_norm": 0.04938704892992973,
      "learning_rate": 3.068497337500886e-05,
      "loss": 2.2452,
      "step": 3848
    },
    {
      "epoch": 0.7698,
      "grad_norm": 0.049502670764923096,
      "learning_rate": 3.063466941871952e-05,
      "loss": 2.241,
      "step": 3849
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.047218501567840576,
      "learning_rate": 3.058439927024962e-05,
      "loss": 2.2459,
      "step": 3850
    },
    {
      "epoch": 0.7702,
      "grad_norm": 0.0470452718436718,
      "learning_rate": 3.053416295410026e-05,
      "loss": 2.2465,
      "step": 3851
    },
    {
      "epoch": 0.7704,
      "grad_norm": 0.044752392917871475,
      "learning_rate": 3.0483960494756016e-05,
      "loss": 2.2481,
      "step": 3852
    },
    {
      "epoch": 0.7706,
      "grad_norm": 0.04909840226173401,
      "learning_rate": 3.0433791916684916e-05,
      "loss": 2.2455,
      "step": 3853
    },
    {
      "epoch": 0.7708,
      "grad_norm": 0.045548055320978165,
      "learning_rate": 3.038365724433858e-05,
      "loss": 2.2476,
      "step": 3854
    },
    {
      "epoch": 0.771,
      "grad_norm": 0.046611323952674866,
      "learning_rate": 3.0333556502151926e-05,
      "loss": 2.2488,
      "step": 3855
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.04449962079524994,
      "learning_rate": 3.0283489714543556e-05,
      "loss": 2.2488,
      "step": 3856
    },
    {
      "epoch": 0.7714,
      "grad_norm": 0.04706541448831558,
      "learning_rate": 3.023345690591537e-05,
      "loss": 2.2481,
      "step": 3857
    },
    {
      "epoch": 0.7716,
      "grad_norm": 0.04434105381369591,
      "learning_rate": 3.018345810065275e-05,
      "loss": 2.2436,
      "step": 3858
    },
    {
      "epoch": 0.7718,
      "grad_norm": 0.04628472775220871,
      "learning_rate": 3.0133493323124505e-05,
      "loss": 2.2481,
      "step": 3859
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.045269034802913666,
      "learning_rate": 3.0083562597682847e-05,
      "loss": 2.2467,
      "step": 3860
    },
    {
      "epoch": 0.7722,
      "grad_norm": 0.044568054378032684,
      "learning_rate": 3.0033665948663448e-05,
      "loss": 2.2483,
      "step": 3861
    },
    {
      "epoch": 0.7724,
      "grad_norm": 0.04321078583598137,
      "learning_rate": 2.9983803400385312e-05,
      "loss": 2.2477,
      "step": 3862
    },
    {
      "epoch": 0.7726,
      "grad_norm": 0.04742109775543213,
      "learning_rate": 2.993397497715086e-05,
      "loss": 2.2461,
      "step": 3863
    },
    {
      "epoch": 0.7728,
      "grad_norm": 0.04427780583500862,
      "learning_rate": 2.9884180703245767e-05,
      "loss": 2.2469,
      "step": 3864
    },
    {
      "epoch": 0.773,
      "grad_norm": 0.04497529938817024,
      "learning_rate": 2.9834420602939263e-05,
      "loss": 2.247,
      "step": 3865
    },
    {
      "epoch": 0.7732,
      "grad_norm": 0.042855288833379745,
      "learning_rate": 2.9784694700483762e-05,
      "loss": 2.246,
      "step": 3866
    },
    {
      "epoch": 0.7734,
      "grad_norm": 0.04731624573469162,
      "learning_rate": 2.9735003020115092e-05,
      "loss": 2.245,
      "step": 3867
    },
    {
      "epoch": 0.7736,
      "grad_norm": 0.042967788875103,
      "learning_rate": 2.968534558605236e-05,
      "loss": 2.2465,
      "step": 3868
    },
    {
      "epoch": 0.7738,
      "grad_norm": 0.04857281595468521,
      "learning_rate": 2.9635722422497993e-05,
      "loss": 2.2496,
      "step": 3869
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.04264584183692932,
      "learning_rate": 2.9586133553637683e-05,
      "loss": 2.2461,
      "step": 3870
    },
    {
      "epoch": 0.7742,
      "grad_norm": 0.05224639177322388,
      "learning_rate": 2.953657900364053e-05,
      "loss": 2.2468,
      "step": 3871
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.04384768009185791,
      "learning_rate": 2.9487058796658783e-05,
      "loss": 2.2478,
      "step": 3872
    },
    {
      "epoch": 0.7746,
      "grad_norm": 0.05025855451822281,
      "learning_rate": 2.9437572956827964e-05,
      "loss": 2.247,
      "step": 3873
    },
    {
      "epoch": 0.7748,
      "grad_norm": 0.043290890753269196,
      "learning_rate": 2.938812150826684e-05,
      "loss": 2.2467,
      "step": 3874
    },
    {
      "epoch": 0.775,
      "grad_norm": 0.04596670717000961,
      "learning_rate": 2.933870447507753e-05,
      "loss": 2.2459,
      "step": 3875
    },
    {
      "epoch": 0.7752,
      "grad_norm": 0.042704757302999496,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 2.2468,
      "step": 3876
    },
    {
      "epoch": 0.7754,
      "grad_norm": 0.047898903489112854,
      "learning_rate": 2.9239973751138495e-05,
      "loss": 2.2484,
      "step": 3877
    },
    {
      "epoch": 0.7756,
      "grad_norm": 0.042811889201402664,
      "learning_rate": 2.9190660108508917e-05,
      "loss": 2.2447,
      "step": 3878
    },
    {
      "epoch": 0.7758,
      "grad_norm": 0.04456919804215431,
      "learning_rate": 2.9141380977491373e-05,
      "loss": 2.2464,
      "step": 3879
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.04288101941347122,
      "learning_rate": 2.9092136382103973e-05,
      "loss": 2.2491,
      "step": 3880
    },
    {
      "epoch": 0.7762,
      "grad_norm": 0.04714420065283775,
      "learning_rate": 2.904292634634793e-05,
      "loss": 2.2464,
      "step": 3881
    },
    {
      "epoch": 0.7764,
      "grad_norm": 0.04280823841691017,
      "learning_rate": 2.899375089420756e-05,
      "loss": 2.2451,
      "step": 3882
    },
    {
      "epoch": 0.7766,
      "grad_norm": 0.047531403601169586,
      "learning_rate": 2.894461004965038e-05,
      "loss": 2.2455,
      "step": 3883
    },
    {
      "epoch": 0.7768,
      "grad_norm": 0.042738817632198334,
      "learning_rate": 2.8895503836627103e-05,
      "loss": 2.245,
      "step": 3884
    },
    {
      "epoch": 0.777,
      "grad_norm": 0.047206584364175797,
      "learning_rate": 2.8846432279071467e-05,
      "loss": 2.2452,
      "step": 3885
    },
    {
      "epoch": 0.7772,
      "grad_norm": 0.044425349682569504,
      "learning_rate": 2.879739540090036e-05,
      "loss": 2.2515,
      "step": 3886
    },
    {
      "epoch": 0.7774,
      "grad_norm": 0.04445197060704231,
      "learning_rate": 2.874839322601375e-05,
      "loss": 2.2458,
      "step": 3887
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.04291234910488129,
      "learning_rate": 2.869942577829471e-05,
      "loss": 2.2458,
      "step": 3888
    },
    {
      "epoch": 0.7778,
      "grad_norm": 0.04402870312333107,
      "learning_rate": 2.865049308160934e-05,
      "loss": 2.2494,
      "step": 3889
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.04498887062072754,
      "learning_rate": 2.860159515980695e-05,
      "loss": 2.2474,
      "step": 3890
    },
    {
      "epoch": 0.7782,
      "grad_norm": 0.04669259116053581,
      "learning_rate": 2.8552732036719687e-05,
      "loss": 2.2473,
      "step": 3891
    },
    {
      "epoch": 0.7784,
      "grad_norm": 0.04409320652484894,
      "learning_rate": 2.8503903736162875e-05,
      "loss": 2.2461,
      "step": 3892
    },
    {
      "epoch": 0.7786,
      "grad_norm": 0.04731551930308342,
      "learning_rate": 2.8455110281934803e-05,
      "loss": 2.2464,
      "step": 3893
    },
    {
      "epoch": 0.7788,
      "grad_norm": 0.045735206454992294,
      "learning_rate": 2.840635169781688e-05,
      "loss": 2.2459,
      "step": 3894
    },
    {
      "epoch": 0.779,
      "grad_norm": 0.0517541728913784,
      "learning_rate": 2.835762800757341e-05,
      "loss": 2.2477,
      "step": 3895
    },
    {
      "epoch": 0.7792,
      "grad_norm": 0.044531092047691345,
      "learning_rate": 2.8308939234951726e-05,
      "loss": 2.2489,
      "step": 3896
    },
    {
      "epoch": 0.7794,
      "grad_norm": 0.05354612320661545,
      "learning_rate": 2.826028540368215e-05,
      "loss": 2.2481,
      "step": 3897
    },
    {
      "epoch": 0.7796,
      "grad_norm": 0.0429520383477211,
      "learning_rate": 2.8211666537477933e-05,
      "loss": 2.2492,
      "step": 3898
    },
    {
      "epoch": 0.7798,
      "grad_norm": 0.04554592818021774,
      "learning_rate": 2.816308266003541e-05,
      "loss": 2.2503,
      "step": 3899
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.04423745721578598,
      "learning_rate": 2.8114533795033683e-05,
      "loss": 2.248,
      "step": 3900
    },
    {
      "epoch": 0.7802,
      "grad_norm": 0.044597748667001724,
      "learning_rate": 2.8066019966134904e-05,
      "loss": 2.2493,
      "step": 3901
    },
    {
      "epoch": 0.7804,
      "grad_norm": 0.0462055541574955,
      "learning_rate": 2.8017541196984142e-05,
      "loss": 2.2478,
      "step": 3902
    },
    {
      "epoch": 0.7806,
      "grad_norm": 0.044962164014577866,
      "learning_rate": 2.7969097511209308e-05,
      "loss": 2.247,
      "step": 3903
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.045565709471702576,
      "learning_rate": 2.7920688932421335e-05,
      "loss": 2.2498,
      "step": 3904
    },
    {
      "epoch": 0.781,
      "grad_norm": 0.04606812819838524,
      "learning_rate": 2.7872315484213952e-05,
      "loss": 2.2506,
      "step": 3905
    },
    {
      "epoch": 0.7812,
      "grad_norm": 0.04424980282783508,
      "learning_rate": 2.7823977190163786e-05,
      "loss": 2.25,
      "step": 3906
    },
    {
      "epoch": 0.7814,
      "grad_norm": 0.042877472937107086,
      "learning_rate": 2.7775674073830337e-05,
      "loss": 2.2466,
      "step": 3907
    },
    {
      "epoch": 0.7816,
      "grad_norm": 0.044136226177215576,
      "learning_rate": 2.772740615875594e-05,
      "loss": 2.2491,
      "step": 3908
    },
    {
      "epoch": 0.7818,
      "grad_norm": 0.04560239613056183,
      "learning_rate": 2.7679173468465812e-05,
      "loss": 2.2471,
      "step": 3909
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.04292147979140282,
      "learning_rate": 2.7630976026467968e-05,
      "loss": 2.2463,
      "step": 3910
    },
    {
      "epoch": 0.7822,
      "grad_norm": 0.04362970590591431,
      "learning_rate": 2.7582813856253275e-05,
      "loss": 2.2474,
      "step": 3911
    },
    {
      "epoch": 0.7824,
      "grad_norm": 0.0426667295396328,
      "learning_rate": 2.753468698129533e-05,
      "loss": 2.2466,
      "step": 3912
    },
    {
      "epoch": 0.7826,
      "grad_norm": 0.04313972592353821,
      "learning_rate": 2.7486595425050665e-05,
      "loss": 2.2473,
      "step": 3913
    },
    {
      "epoch": 0.7828,
      "grad_norm": 0.043289896100759506,
      "learning_rate": 2.743853921095848e-05,
      "loss": 2.2481,
      "step": 3914
    },
    {
      "epoch": 0.783,
      "grad_norm": 0.04473060742020607,
      "learning_rate": 2.7390518362440808e-05,
      "loss": 2.2481,
      "step": 3915
    },
    {
      "epoch": 0.7832,
      "grad_norm": 0.04871669039130211,
      "learning_rate": 2.734253290290242e-05,
      "loss": 2.2512,
      "step": 3916
    },
    {
      "epoch": 0.7834,
      "grad_norm": 0.05653618276119232,
      "learning_rate": 2.7294582855730832e-05,
      "loss": 2.2484,
      "step": 3917
    },
    {
      "epoch": 0.7836,
      "grad_norm": 0.042974893003702164,
      "learning_rate": 2.7246668244296323e-05,
      "loss": 2.2474,
      "step": 3918
    },
    {
      "epoch": 0.7838,
      "grad_norm": 0.059263624250888824,
      "learning_rate": 2.7198789091951902e-05,
      "loss": 2.2472,
      "step": 3919
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.04299822822213173,
      "learning_rate": 2.715094542203327e-05,
      "loss": 2.248,
      "step": 3920
    },
    {
      "epoch": 0.7842,
      "grad_norm": 0.0606023333966732,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 2.2476,
      "step": 3921
    },
    {
      "epoch": 0.7844,
      "grad_norm": 0.04380832985043526,
      "learning_rate": 2.7055364622729773e-05,
      "loss": 2.2484,
      "step": 3922
    },
    {
      "epoch": 0.7846,
      "grad_norm": 0.06024012342095375,
      "learning_rate": 2.700762753992985e-05,
      "loss": 2.2457,
      "step": 3923
    },
    {
      "epoch": 0.7848,
      "grad_norm": 0.04397067055106163,
      "learning_rate": 2.6959926032725535e-05,
      "loss": 2.2477,
      "step": 3924
    },
    {
      "epoch": 0.785,
      "grad_norm": 0.0545315258204937,
      "learning_rate": 2.6912260124366006e-05,
      "loss": 2.2479,
      "step": 3925
    },
    {
      "epoch": 0.7852,
      "grad_norm": 0.047849953174591064,
      "learning_rate": 2.6864629838082956e-05,
      "loss": 2.2485,
      "step": 3926
    },
    {
      "epoch": 0.7854,
      "grad_norm": 0.052340298891067505,
      "learning_rate": 2.681703519709089e-05,
      "loss": 2.2453,
      "step": 3927
    },
    {
      "epoch": 0.7856,
      "grad_norm": 0.04801769554615021,
      "learning_rate": 2.676947622458683e-05,
      "loss": 2.2476,
      "step": 3928
    },
    {
      "epoch": 0.7858,
      "grad_norm": 0.052130959928035736,
      "learning_rate": 2.672195294375045e-05,
      "loss": 2.2485,
      "step": 3929
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.044636406004428864,
      "learning_rate": 2.6674465377744017e-05,
      "loss": 2.2466,
      "step": 3930
    },
    {
      "epoch": 0.7862,
      "grad_norm": 0.04428806155920029,
      "learning_rate": 2.6627013549712355e-05,
      "loss": 2.2449,
      "step": 3931
    },
    {
      "epoch": 0.7864,
      "grad_norm": 0.045109450817108154,
      "learning_rate": 2.657959748278297e-05,
      "loss": 2.2464,
      "step": 3932
    },
    {
      "epoch": 0.7866,
      "grad_norm": 0.0430612787604332,
      "learning_rate": 2.6532217200065858e-05,
      "loss": 2.2453,
      "step": 3933
    },
    {
      "epoch": 0.7868,
      "grad_norm": 0.04692564532160759,
      "learning_rate": 2.6484872724653608e-05,
      "loss": 2.2472,
      "step": 3934
    },
    {
      "epoch": 0.787,
      "grad_norm": 0.042583268135786057,
      "learning_rate": 2.6437564079621267e-05,
      "loss": 2.2486,
      "step": 3935
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.04442555457353592,
      "learning_rate": 2.639029128802657e-05,
      "loss": 2.2471,
      "step": 3936
    },
    {
      "epoch": 0.7874,
      "grad_norm": 0.04385305568575859,
      "learning_rate": 2.634305437290968e-05,
      "loss": 2.2472,
      "step": 3937
    },
    {
      "epoch": 0.7876,
      "grad_norm": 0.0464387983083725,
      "learning_rate": 2.6295853357293298e-05,
      "loss": 2.246,
      "step": 3938
    },
    {
      "epoch": 0.7878,
      "grad_norm": 0.04309503734111786,
      "learning_rate": 2.624868826418262e-05,
      "loss": 2.2475,
      "step": 3939
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.04442499950528145,
      "learning_rate": 2.6201559116565345e-05,
      "loss": 2.2474,
      "step": 3940
    },
    {
      "epoch": 0.7882,
      "grad_norm": 0.04312480241060257,
      "learning_rate": 2.615446593741161e-05,
      "loss": 2.2474,
      "step": 3941
    },
    {
      "epoch": 0.7884,
      "grad_norm": 0.04663326218724251,
      "learning_rate": 2.6107408749674122e-05,
      "loss": 2.2476,
      "step": 3942
    },
    {
      "epoch": 0.7886,
      "grad_norm": 0.04295874387025833,
      "learning_rate": 2.606038757628798e-05,
      "loss": 2.247,
      "step": 3943
    },
    {
      "epoch": 0.7888,
      "grad_norm": 0.04256637766957283,
      "learning_rate": 2.6013402440170676e-05,
      "loss": 2.2473,
      "step": 3944
    },
    {
      "epoch": 0.789,
      "grad_norm": 0.042664509266614914,
      "learning_rate": 2.5966453364222186e-05,
      "loss": 2.2462,
      "step": 3945
    },
    {
      "epoch": 0.7892,
      "grad_norm": 0.042628396302461624,
      "learning_rate": 2.5919540371325e-05,
      "loss": 2.2459,
      "step": 3946
    },
    {
      "epoch": 0.7894,
      "grad_norm": 0.042927682399749756,
      "learning_rate": 2.5872663484343884e-05,
      "loss": 2.2462,
      "step": 3947
    },
    {
      "epoch": 0.7896,
      "grad_norm": 0.04375307634472847,
      "learning_rate": 2.582582272612609e-05,
      "loss": 2.2451,
      "step": 3948
    },
    {
      "epoch": 0.7898,
      "grad_norm": 0.04292015731334686,
      "learning_rate": 2.5779018119501208e-05,
      "loss": 2.2481,
      "step": 3949
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.04329948127269745,
      "learning_rate": 2.573224968728123e-05,
      "loss": 2.2473,
      "step": 3950
    },
    {
      "epoch": 0.7902,
      "grad_norm": 0.04426323249936104,
      "learning_rate": 2.5685517452260567e-05,
      "loss": 2.2498,
      "step": 3951
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.043475158512592316,
      "learning_rate": 2.5638821437215944e-05,
      "loss": 2.2464,
      "step": 3952
    },
    {
      "epoch": 0.7906,
      "grad_norm": 0.04318339750170708,
      "learning_rate": 2.5592161664906368e-05,
      "loss": 2.247,
      "step": 3953
    },
    {
      "epoch": 0.7908,
      "grad_norm": 0.04472862929105759,
      "learning_rate": 2.554553815807328e-05,
      "loss": 2.2485,
      "step": 3954
    },
    {
      "epoch": 0.791,
      "grad_norm": 0.04282104969024658,
      "learning_rate": 2.549895093944039e-05,
      "loss": 2.2487,
      "step": 3955
    },
    {
      "epoch": 0.7912,
      "grad_norm": 0.04433750733733177,
      "learning_rate": 2.5452400031713785e-05,
      "loss": 2.2475,
      "step": 3956
    },
    {
      "epoch": 0.7914,
      "grad_norm": 0.04253934696316719,
      "learning_rate": 2.540588545758179e-05,
      "loss": 2.2486,
      "step": 3957
    },
    {
      "epoch": 0.7916,
      "grad_norm": 0.0442885085940361,
      "learning_rate": 2.535940723971505e-05,
      "loss": 2.2473,
      "step": 3958
    },
    {
      "epoch": 0.7918,
      "grad_norm": 0.04221853241324425,
      "learning_rate": 2.5312965400766474e-05,
      "loss": 2.2463,
      "step": 3959
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.042581215500831604,
      "learning_rate": 2.5266559963371216e-05,
      "loss": 2.2465,
      "step": 3960
    },
    {
      "epoch": 0.7922,
      "grad_norm": 0.044596217572689056,
      "learning_rate": 2.5220190950146827e-05,
      "loss": 2.2465,
      "step": 3961
    },
    {
      "epoch": 0.7924,
      "grad_norm": 0.04375210031867027,
      "learning_rate": 2.5173858383692906e-05,
      "loss": 2.2457,
      "step": 3962
    },
    {
      "epoch": 0.7926,
      "grad_norm": 0.042877428233623505,
      "learning_rate": 2.512756228659141e-05,
      "loss": 2.2475,
      "step": 3963
    },
    {
      "epoch": 0.7928,
      "grad_norm": 0.042822882533073425,
      "learning_rate": 2.508130268140646e-05,
      "loss": 2.2451,
      "step": 3964
    },
    {
      "epoch": 0.793,
      "grad_norm": 0.04383544251322746,
      "learning_rate": 2.5035079590684497e-05,
      "loss": 2.2465,
      "step": 3965
    },
    {
      "epoch": 0.7932,
      "grad_norm": 0.04352884739637375,
      "learning_rate": 2.4988893036954043e-05,
      "loss": 2.2481,
      "step": 3966
    },
    {
      "epoch": 0.7934,
      "grad_norm": 0.043424539268016815,
      "learning_rate": 2.4942743042725892e-05,
      "loss": 2.2459,
      "step": 3967
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.043184686452150345,
      "learning_rate": 2.4896629630492973e-05,
      "loss": 2.2472,
      "step": 3968
    },
    {
      "epoch": 0.7938,
      "grad_norm": 0.043667569756507874,
      "learning_rate": 2.48505528227304e-05,
      "loss": 2.2472,
      "step": 3969
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.04350176826119423,
      "learning_rate": 2.480451264189546e-05,
      "loss": 2.2491,
      "step": 3970
    },
    {
      "epoch": 0.7942,
      "grad_norm": 0.0430961512029171,
      "learning_rate": 2.4758509110427575e-05,
      "loss": 2.247,
      "step": 3971
    },
    {
      "epoch": 0.7944,
      "grad_norm": 0.04327191784977913,
      "learning_rate": 2.4712542250748304e-05,
      "loss": 2.2474,
      "step": 3972
    },
    {
      "epoch": 0.7946,
      "grad_norm": 0.04260185360908508,
      "learning_rate": 2.4666612085261342e-05,
      "loss": 2.2482,
      "step": 3973
    },
    {
      "epoch": 0.7948,
      "grad_norm": 0.044274505227804184,
      "learning_rate": 2.4620718636352457e-05,
      "loss": 2.2487,
      "step": 3974
    },
    {
      "epoch": 0.795,
      "grad_norm": 0.04274379462003708,
      "learning_rate": 2.4574861926389615e-05,
      "loss": 2.2477,
      "step": 3975
    },
    {
      "epoch": 0.7952,
      "grad_norm": 0.043024104088544846,
      "learning_rate": 2.45290419777228e-05,
      "loss": 2.2477,
      "step": 3976
    },
    {
      "epoch": 0.7954,
      "grad_norm": 0.04304726794362068,
      "learning_rate": 2.4483258812684096e-05,
      "loss": 2.2479,
      "step": 3977
    },
    {
      "epoch": 0.7956,
      "grad_norm": 0.043114155530929565,
      "learning_rate": 2.443751245358765e-05,
      "loss": 2.2467,
      "step": 3978
    },
    {
      "epoch": 0.7958,
      "grad_norm": 0.043726179748773575,
      "learning_rate": 2.43918029227297e-05,
      "loss": 2.2474,
      "step": 3979
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.04247238114476204,
      "learning_rate": 2.43461302423885e-05,
      "loss": 2.2482,
      "step": 3980
    },
    {
      "epoch": 0.7962,
      "grad_norm": 0.04250417649745941,
      "learning_rate": 2.4300494434824373e-05,
      "loss": 2.2476,
      "step": 3981
    },
    {
      "epoch": 0.7964,
      "grad_norm": 0.04460495710372925,
      "learning_rate": 2.425489552227964e-05,
      "loss": 2.2501,
      "step": 3982
    },
    {
      "epoch": 0.7966,
      "grad_norm": 0.04452277347445488,
      "learning_rate": 2.420933352697865e-05,
      "loss": 2.247,
      "step": 3983
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.04303939640522003,
      "learning_rate": 2.4163808471127812e-05,
      "loss": 2.2478,
      "step": 3984
    },
    {
      "epoch": 0.797,
      "grad_norm": 0.04302297160029411,
      "learning_rate": 2.411832037691545e-05,
      "loss": 2.2495,
      "step": 3985
    },
    {
      "epoch": 0.7972,
      "grad_norm": 0.043988119810819626,
      "learning_rate": 2.407286926651192e-05,
      "loss": 2.248,
      "step": 3986
    },
    {
      "epoch": 0.7974,
      "grad_norm": 0.04238108918070793,
      "learning_rate": 2.4027455162069567e-05,
      "loss": 2.2469,
      "step": 3987
    },
    {
      "epoch": 0.7976,
      "grad_norm": 0.046329401433467865,
      "learning_rate": 2.3982078085722582e-05,
      "loss": 2.2502,
      "step": 3988
    },
    {
      "epoch": 0.7978,
      "grad_norm": 0.04461927339434624,
      "learning_rate": 2.3936738059587282e-05,
      "loss": 2.2488,
      "step": 3989
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.042583003640174866,
      "learning_rate": 2.389143510576184e-05,
      "loss": 2.2479,
      "step": 3990
    },
    {
      "epoch": 0.7982,
      "grad_norm": 0.04643017798662186,
      "learning_rate": 2.3846169246326343e-05,
      "loss": 2.2495,
      "step": 3991
    },
    {
      "epoch": 0.7984,
      "grad_norm": 0.04295787587761879,
      "learning_rate": 2.3800940503342828e-05,
      "loss": 2.2471,
      "step": 3992
    },
    {
      "epoch": 0.7986,
      "grad_norm": 0.04383156821131706,
      "learning_rate": 2.37557488988552e-05,
      "loss": 2.2466,
      "step": 3993
    },
    {
      "epoch": 0.7988,
      "grad_norm": 0.04297374188899994,
      "learning_rate": 2.3710594454889378e-05,
      "loss": 2.2482,
      "step": 3994
    },
    {
      "epoch": 0.799,
      "grad_norm": 0.042747337371110916,
      "learning_rate": 2.3665477193453034e-05,
      "loss": 2.2462,
      "step": 3995
    },
    {
      "epoch": 0.7992,
      "grad_norm": 0.04249376803636551,
      "learning_rate": 2.362039713653581e-05,
      "loss": 2.2451,
      "step": 3996
    },
    {
      "epoch": 0.7994,
      "grad_norm": 0.04275410622358322,
      "learning_rate": 2.35753543061091e-05,
      "loss": 2.246,
      "step": 3997
    },
    {
      "epoch": 0.7996,
      "grad_norm": 0.04246923699975014,
      "learning_rate": 2.3530348724126307e-05,
      "loss": 2.2472,
      "step": 3998
    },
    {
      "epoch": 0.7998,
      "grad_norm": 0.042986128479242325,
      "learning_rate": 2.3485380412522585e-05,
      "loss": 2.2479,
      "step": 3999
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.04235870763659477,
      "learning_rate": 2.3440449393214948e-05,
      "loss": 2.2459,
      "step": 4000
    },
    {
      "epoch": 0.8002,
      "grad_norm": 0.04347938299179077,
      "learning_rate": 2.339555568810221e-05,
      "loss": 2.248,
      "step": 4001
    },
    {
      "epoch": 0.8004,
      "grad_norm": 0.042568884789943695,
      "learning_rate": 2.3350699319065026e-05,
      "loss": 2.2461,
      "step": 4002
    },
    {
      "epoch": 0.8006,
      "grad_norm": 0.04285457730293274,
      "learning_rate": 2.3305880307965833e-05,
      "loss": 2.2465,
      "step": 4003
    },
    {
      "epoch": 0.8008,
      "grad_norm": 0.04247830808162689,
      "learning_rate": 2.3261098676648907e-05,
      "loss": 2.2461,
      "step": 4004
    },
    {
      "epoch": 0.801,
      "grad_norm": 0.043211258947849274,
      "learning_rate": 2.321635444694028e-05,
      "loss": 2.248,
      "step": 4005
    },
    {
      "epoch": 0.8012,
      "grad_norm": 0.04414999857544899,
      "learning_rate": 2.3171647640647687e-05,
      "loss": 2.247,
      "step": 4006
    },
    {
      "epoch": 0.8014,
      "grad_norm": 0.043235257267951965,
      "learning_rate": 2.3126978279560684e-05,
      "loss": 2.248,
      "step": 4007
    },
    {
      "epoch": 0.8016,
      "grad_norm": 0.043725159019231796,
      "learning_rate": 2.3082346385450637e-05,
      "loss": 2.2494,
      "step": 4008
    },
    {
      "epoch": 0.8018,
      "grad_norm": 0.04388371855020523,
      "learning_rate": 2.3037751980070555e-05,
      "loss": 2.2466,
      "step": 4009
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.04448014870285988,
      "learning_rate": 2.2993195085155205e-05,
      "loss": 2.2493,
      "step": 4010
    },
    {
      "epoch": 0.8022,
      "grad_norm": 0.04307398572564125,
      "learning_rate": 2.2948675722421086e-05,
      "loss": 2.2487,
      "step": 4011
    },
    {
      "epoch": 0.8024,
      "grad_norm": 0.042765650898218155,
      "learning_rate": 2.2904193913566363e-05,
      "loss": 2.2482,
      "step": 4012
    },
    {
      "epoch": 0.8026,
      "grad_norm": 0.04370567202568054,
      "learning_rate": 2.2859749680270982e-05,
      "loss": 2.2466,
      "step": 4013
    },
    {
      "epoch": 0.8028,
      "grad_norm": 0.04259917140007019,
      "learning_rate": 2.2815343044196525e-05,
      "loss": 2.2466,
      "step": 4014
    },
    {
      "epoch": 0.803,
      "grad_norm": 0.043378595262765884,
      "learning_rate": 2.277097402698619e-05,
      "loss": 2.2458,
      "step": 4015
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.04252930358052254,
      "learning_rate": 2.2726642650264895e-05,
      "loss": 2.2461,
      "step": 4016
    },
    {
      "epoch": 0.8034,
      "grad_norm": 0.04256908595561981,
      "learning_rate": 2.2682348935639274e-05,
      "loss": 2.2469,
      "step": 4017
    },
    {
      "epoch": 0.8036,
      "grad_norm": 0.042589783668518066,
      "learning_rate": 2.2638092904697518e-05,
      "loss": 2.2487,
      "step": 4018
    },
    {
      "epoch": 0.8038,
      "grad_norm": 0.04265584424138069,
      "learning_rate": 2.259387457900948e-05,
      "loss": 2.2464,
      "step": 4019
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.043072059750556946,
      "learning_rate": 2.254969398012663e-05,
      "loss": 2.2495,
      "step": 4020
    },
    {
      "epoch": 0.8042,
      "grad_norm": 0.044470589607954025,
      "learning_rate": 2.2505551129582047e-05,
      "loss": 2.2473,
      "step": 4021
    },
    {
      "epoch": 0.8044,
      "grad_norm": 0.043554678559303284,
      "learning_rate": 2.246144604889042e-05,
      "loss": 2.2469,
      "step": 4022
    },
    {
      "epoch": 0.8046,
      "grad_norm": 0.04373255372047424,
      "learning_rate": 2.241737875954808e-05,
      "loss": 2.2487,
      "step": 4023
    },
    {
      "epoch": 0.8048,
      "grad_norm": 0.04315250739455223,
      "learning_rate": 2.237334928303283e-05,
      "loss": 2.2473,
      "step": 4024
    },
    {
      "epoch": 0.805,
      "grad_norm": 0.0432584248483181,
      "learning_rate": 2.2329357640804117e-05,
      "loss": 2.246,
      "step": 4025
    },
    {
      "epoch": 0.8052,
      "grad_norm": 0.04322300851345062,
      "learning_rate": 2.2285403854302912e-05,
      "loss": 2.2471,
      "step": 4026
    },
    {
      "epoch": 0.8054,
      "grad_norm": 0.04293752834200859,
      "learning_rate": 2.22414879449518e-05,
      "loss": 2.2471,
      "step": 4027
    },
    {
      "epoch": 0.8056,
      "grad_norm": 0.046400390565395355,
      "learning_rate": 2.219760993415485e-05,
      "loss": 2.2488,
      "step": 4028
    },
    {
      "epoch": 0.8058,
      "grad_norm": 0.04366804659366608,
      "learning_rate": 2.2153769843297667e-05,
      "loss": 2.2459,
      "step": 4029
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.04672745615243912,
      "learning_rate": 2.210996769374737e-05,
      "loss": 2.2467,
      "step": 4030
    },
    {
      "epoch": 0.8062,
      "grad_norm": 0.04254091903567314,
      "learning_rate": 2.2066203506852566e-05,
      "loss": 2.247,
      "step": 4031
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.04798227548599243,
      "learning_rate": 2.202247730394349e-05,
      "loss": 2.2473,
      "step": 4032
    },
    {
      "epoch": 0.8066,
      "grad_norm": 0.04277210682630539,
      "learning_rate": 2.1978789106331665e-05,
      "loss": 2.2471,
      "step": 4033
    },
    {
      "epoch": 0.8068,
      "grad_norm": 0.048542458564043045,
      "learning_rate": 2.1935138935310206e-05,
      "loss": 2.2462,
      "step": 4034
    },
    {
      "epoch": 0.807,
      "grad_norm": 0.043250247836112976,
      "learning_rate": 2.1891526812153672e-05,
      "loss": 2.2461,
      "step": 4035
    },
    {
      "epoch": 0.8072,
      "grad_norm": 0.04641755297780037,
      "learning_rate": 2.1847952758118117e-05,
      "loss": 2.2466,
      "step": 4036
    },
    {
      "epoch": 0.8074,
      "grad_norm": 0.04354877397418022,
      "learning_rate": 2.1804416794440995e-05,
      "loss": 2.2465,
      "step": 4037
    },
    {
      "epoch": 0.8076,
      "grad_norm": 0.046106815338134766,
      "learning_rate": 2.1760918942341192e-05,
      "loss": 2.2492,
      "step": 4038
    },
    {
      "epoch": 0.8078,
      "grad_norm": 0.044260423630476,
      "learning_rate": 2.1717459223019033e-05,
      "loss": 2.2478,
      "step": 4039
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.043805815279483795,
      "learning_rate": 2.1674037657656266e-05,
      "loss": 2.247,
      "step": 4040
    },
    {
      "epoch": 0.8082,
      "grad_norm": 0.0448303185403347,
      "learning_rate": 2.163065426741603e-05,
      "loss": 2.2478,
      "step": 4041
    },
    {
      "epoch": 0.8084,
      "grad_norm": 0.04384468495845795,
      "learning_rate": 2.1587309073442863e-05,
      "loss": 2.2444,
      "step": 4042
    },
    {
      "epoch": 0.8086,
      "grad_norm": 0.04288894310593605,
      "learning_rate": 2.154400209686268e-05,
      "loss": 2.2476,
      "step": 4043
    },
    {
      "epoch": 0.8088,
      "grad_norm": 0.04263492301106453,
      "learning_rate": 2.1500733358782786e-05,
      "loss": 2.2467,
      "step": 4044
    },
    {
      "epoch": 0.809,
      "grad_norm": 0.04538034275174141,
      "learning_rate": 2.1457502880291812e-05,
      "loss": 2.2473,
      "step": 4045
    },
    {
      "epoch": 0.8092,
      "grad_norm": 0.04270567372441292,
      "learning_rate": 2.1414310682459802e-05,
      "loss": 2.2461,
      "step": 4046
    },
    {
      "epoch": 0.8094,
      "grad_norm": 0.04441223666071892,
      "learning_rate": 2.137115678633811e-05,
      "loss": 2.2466,
      "step": 4047
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.042989157140254974,
      "learning_rate": 2.1328041212959403e-05,
      "loss": 2.244,
      "step": 4048
    },
    {
      "epoch": 0.8098,
      "grad_norm": 0.04329875856637955,
      "learning_rate": 2.128496398333768e-05,
      "loss": 2.2467,
      "step": 4049
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.043924447149038315,
      "learning_rate": 2.1241925118468287e-05,
      "loss": 2.248,
      "step": 4050
    },
    {
      "epoch": 0.8102,
      "grad_norm": 0.04250532016158104,
      "learning_rate": 2.119892463932781e-05,
      "loss": 2.2467,
      "step": 4051
    },
    {
      "epoch": 0.8104,
      "grad_norm": 0.044043559581041336,
      "learning_rate": 2.115596256687419e-05,
      "loss": 2.2461,
      "step": 4052
    },
    {
      "epoch": 0.8106,
      "grad_norm": 0.0424211248755455,
      "learning_rate": 2.1113038922046602e-05,
      "loss": 2.2464,
      "step": 4053
    },
    {
      "epoch": 0.8108,
      "grad_norm": 0.0437840074300766,
      "learning_rate": 2.107015372576552e-05,
      "loss": 2.2466,
      "step": 4054
    },
    {
      "epoch": 0.811,
      "grad_norm": 0.04227857664227486,
      "learning_rate": 2.1027306998932627e-05,
      "loss": 2.2459,
      "step": 4055
    },
    {
      "epoch": 0.8112,
      "grad_norm": 0.043394237756729126,
      "learning_rate": 2.098449876243096e-05,
      "loss": 2.2487,
      "step": 4056
    },
    {
      "epoch": 0.8114,
      "grad_norm": 0.04231567680835724,
      "learning_rate": 2.09417290371247e-05,
      "loss": 2.2469,
      "step": 4057
    },
    {
      "epoch": 0.8116,
      "grad_norm": 0.042540282011032104,
      "learning_rate": 2.0898997843859334e-05,
      "loss": 2.2459,
      "step": 4058
    },
    {
      "epoch": 0.8118,
      "grad_norm": 0.04288550838828087,
      "learning_rate": 2.0856305203461436e-05,
      "loss": 2.2478,
      "step": 4059
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.04239378869533539,
      "learning_rate": 2.0813651136738954e-05,
      "loss": 2.2463,
      "step": 4060
    },
    {
      "epoch": 0.8122,
      "grad_norm": 0.04255305603146553,
      "learning_rate": 2.0771035664480942e-05,
      "loss": 2.2466,
      "step": 4061
    },
    {
      "epoch": 0.8124,
      "grad_norm": 0.0423002615571022,
      "learning_rate": 2.0728458807457662e-05,
      "loss": 2.2464,
      "step": 4062
    },
    {
      "epoch": 0.8126,
      "grad_norm": 0.0433778278529644,
      "learning_rate": 2.068592058642056e-05,
      "loss": 2.2473,
      "step": 4063
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.043289683759212494,
      "learning_rate": 2.0643421022102217e-05,
      "loss": 2.2463,
      "step": 4064
    },
    {
      "epoch": 0.813,
      "grad_norm": 0.04294027388095856,
      "learning_rate": 2.0600960135216462e-05,
      "loss": 2.2484,
      "step": 4065
    },
    {
      "epoch": 0.8132,
      "grad_norm": 0.04230961576104164,
      "learning_rate": 2.0558537946458177e-05,
      "loss": 2.2457,
      "step": 4066
    },
    {
      "epoch": 0.8134,
      "grad_norm": 0.04385892674326897,
      "learning_rate": 2.051615447650347e-05,
      "loss": 2.2506,
      "step": 4067
    },
    {
      "epoch": 0.8136,
      "grad_norm": 0.04346568509936333,
      "learning_rate": 2.0473809746009444e-05,
      "loss": 2.2461,
      "step": 4068
    },
    {
      "epoch": 0.8138,
      "grad_norm": 0.044096872210502625,
      "learning_rate": 2.0431503775614457e-05,
      "loss": 2.2466,
      "step": 4069
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.042609333992004395,
      "learning_rate": 2.0389236585937945e-05,
      "loss": 2.2467,
      "step": 4070
    },
    {
      "epoch": 0.8142,
      "grad_norm": 0.042398031800985336,
      "learning_rate": 2.0347008197580374e-05,
      "loss": 2.248,
      "step": 4071
    },
    {
      "epoch": 0.8144,
      "grad_norm": 0.0461578406393528,
      "learning_rate": 2.030481863112339e-05,
      "loss": 2.2462,
      "step": 4072
    },
    {
      "epoch": 0.8146,
      "grad_norm": 0.04302477464079857,
      "learning_rate": 2.026266790712965e-05,
      "loss": 2.2462,
      "step": 4073
    },
    {
      "epoch": 0.8148,
      "grad_norm": 0.0446154922246933,
      "learning_rate": 2.0220556046142893e-05,
      "loss": 2.2462,
      "step": 4074
    },
    {
      "epoch": 0.815,
      "grad_norm": 0.04244913160800934,
      "learning_rate": 2.017848306868797e-05,
      "loss": 2.2465,
      "step": 4075
    },
    {
      "epoch": 0.8152,
      "grad_norm": 0.04385795071721077,
      "learning_rate": 2.013644899527074e-05,
      "loss": 2.2478,
      "step": 4076
    },
    {
      "epoch": 0.8154,
      "grad_norm": 0.043010469526052475,
      "learning_rate": 2.009445384637805e-05,
      "loss": 2.2489,
      "step": 4077
    },
    {
      "epoch": 0.8156,
      "grad_norm": 0.042541664093732834,
      "learning_rate": 2.005249764247783e-05,
      "loss": 2.2484,
      "step": 4078
    },
    {
      "epoch": 0.8158,
      "grad_norm": 0.043181996792554855,
      "learning_rate": 2.0010580404019062e-05,
      "loss": 2.2468,
      "step": 4079
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.042761217802762985,
      "learning_rate": 1.9968702151431696e-05,
      "loss": 2.2469,
      "step": 4080
    },
    {
      "epoch": 0.8162,
      "grad_norm": 0.04299420863389969,
      "learning_rate": 1.9926862905126665e-05,
      "loss": 2.2469,
      "step": 4081
    },
    {
      "epoch": 0.8164,
      "grad_norm": 0.04251562058925629,
      "learning_rate": 1.9885062685495904e-05,
      "loss": 2.2459,
      "step": 4082
    },
    {
      "epoch": 0.8166,
      "grad_norm": 0.043009836226701736,
      "learning_rate": 1.9843301512912327e-05,
      "loss": 2.2481,
      "step": 4083
    },
    {
      "epoch": 0.8168,
      "grad_norm": 0.0426003523170948,
      "learning_rate": 1.9801579407729866e-05,
      "loss": 2.2478,
      "step": 4084
    },
    {
      "epoch": 0.817,
      "grad_norm": 0.04233984276652336,
      "learning_rate": 1.9759896390283362e-05,
      "loss": 2.2465,
      "step": 4085
    },
    {
      "epoch": 0.8172,
      "grad_norm": 0.04299848899245262,
      "learning_rate": 1.9718252480888566e-05,
      "loss": 2.2475,
      "step": 4086
    },
    {
      "epoch": 0.8174,
      "grad_norm": 0.042751602828502655,
      "learning_rate": 1.9676647699842242e-05,
      "loss": 2.2467,
      "step": 4087
    },
    {
      "epoch": 0.8176,
      "grad_norm": 0.04464944452047348,
      "learning_rate": 1.963508206742202e-05,
      "loss": 2.2464,
      "step": 4088
    },
    {
      "epoch": 0.8178,
      "grad_norm": 0.04280589148402214,
      "learning_rate": 1.9593555603886538e-05,
      "loss": 2.246,
      "step": 4089
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.0425100214779377,
      "learning_rate": 1.955206832947526e-05,
      "loss": 2.2477,
      "step": 4090
    },
    {
      "epoch": 0.8182,
      "grad_norm": 0.043543897569179535,
      "learning_rate": 1.9510620264408596e-05,
      "loss": 2.2491,
      "step": 4091
    },
    {
      "epoch": 0.8184,
      "grad_norm": 0.04231572896242142,
      "learning_rate": 1.946921142888781e-05,
      "loss": 2.2464,
      "step": 4092
    },
    {
      "epoch": 0.8186,
      "grad_norm": 0.04404304176568985,
      "learning_rate": 1.9427841843095063e-05,
      "loss": 2.2445,
      "step": 4093
    },
    {
      "epoch": 0.8188,
      "grad_norm": 0.04212678596377373,
      "learning_rate": 1.938651152719344e-05,
      "loss": 2.2462,
      "step": 4094
    },
    {
      "epoch": 0.819,
      "grad_norm": 0.04308634623885155,
      "learning_rate": 1.9345220501326777e-05,
      "loss": 2.2485,
      "step": 4095
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.04264509677886963,
      "learning_rate": 1.930396878561983e-05,
      "loss": 2.247,
      "step": 4096
    },
    {
      "epoch": 0.8194,
      "grad_norm": 0.042206522077322006,
      "learning_rate": 1.9262756400178162e-05,
      "loss": 2.2473,
      "step": 4097
    },
    {
      "epoch": 0.8196,
      "grad_norm": 0.0430840328335762,
      "learning_rate": 1.922158336508825e-05,
      "loss": 2.2463,
      "step": 4098
    },
    {
      "epoch": 0.8198,
      "grad_norm": 0.04235004261136055,
      "learning_rate": 1.918044970041729e-05,
      "loss": 2.2466,
      "step": 4099
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.04289722070097923,
      "learning_rate": 1.9139355426213347e-05,
      "loss": 2.2464,
      "step": 4100
    },
    {
      "epoch": 0.8202,
      "grad_norm": 0.04241228476166725,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 2.2472,
      "step": 4101
    },
    {
      "epoch": 0.8204,
      "grad_norm": 0.042470160871744156,
      "learning_rate": 1.9057285129302683e-05,
      "loss": 2.2469,
      "step": 4102
    },
    {
      "epoch": 0.8206,
      "grad_norm": 0.042375870048999786,
      "learning_rate": 1.9016309146596023e-05,
      "loss": 2.2453,
      "step": 4103
    },
    {
      "epoch": 0.8208,
      "grad_norm": 0.04221430420875549,
      "learning_rate": 1.897537263435648e-05,
      "loss": 2.2453,
      "step": 4104
    },
    {
      "epoch": 0.821,
      "grad_norm": 0.04274389520287514,
      "learning_rate": 1.8934475612536017e-05,
      "loss": 2.2477,
      "step": 4105
    },
    {
      "epoch": 0.8212,
      "grad_norm": 0.04231968894600868,
      "learning_rate": 1.8893618101067355e-05,
      "loss": 2.2462,
      "step": 4106
    },
    {
      "epoch": 0.8214,
      "grad_norm": 0.04255298897624016,
      "learning_rate": 1.885280011986391e-05,
      "loss": 2.2464,
      "step": 4107
    },
    {
      "epoch": 0.8216,
      "grad_norm": 0.042457208037376404,
      "learning_rate": 1.8812021688819915e-05,
      "loss": 2.2478,
      "step": 4108
    },
    {
      "epoch": 0.8218,
      "grad_norm": 0.04276324436068535,
      "learning_rate": 1.877128282781028e-05,
      "loss": 2.2471,
      "step": 4109
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.042404044419527054,
      "learning_rate": 1.8730583556690605e-05,
      "loss": 2.2463,
      "step": 4110
    },
    {
      "epoch": 0.8222,
      "grad_norm": 0.042531151324510574,
      "learning_rate": 1.8689923895297245e-05,
      "loss": 2.2464,
      "step": 4111
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.04276575148105621,
      "learning_rate": 1.8649303863447198e-05,
      "loss": 2.2476,
      "step": 4112
    },
    {
      "epoch": 0.8226,
      "grad_norm": 0.04256764054298401,
      "learning_rate": 1.8608723480938206e-05,
      "loss": 2.2467,
      "step": 4113
    },
    {
      "epoch": 0.8228,
      "grad_norm": 0.04237718507647514,
      "learning_rate": 1.8568182767548626e-05,
      "loss": 2.2485,
      "step": 4114
    },
    {
      "epoch": 0.823,
      "grad_norm": 0.04382926598191261,
      "learning_rate": 1.852768174303752e-05,
      "loss": 2.2445,
      "step": 4115
    },
    {
      "epoch": 0.8232,
      "grad_norm": 0.042583707720041275,
      "learning_rate": 1.848722042714457e-05,
      "loss": 2.2465,
      "step": 4116
    },
    {
      "epoch": 0.8234,
      "grad_norm": 0.04483901709318161,
      "learning_rate": 1.8446798839590186e-05,
      "loss": 2.2491,
      "step": 4117
    },
    {
      "epoch": 0.8236,
      "grad_norm": 0.04263860732316971,
      "learning_rate": 1.8406417000075325e-05,
      "loss": 2.2461,
      "step": 4118
    },
    {
      "epoch": 0.8238,
      "grad_norm": 0.04373256117105484,
      "learning_rate": 1.8366074928281607e-05,
      "loss": 2.2472,
      "step": 4119
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.04261251166462898,
      "learning_rate": 1.8325772643871265e-05,
      "loss": 2.2494,
      "step": 4120
    },
    {
      "epoch": 0.8242,
      "grad_norm": 0.04324120655655861,
      "learning_rate": 1.8285510166487152e-05,
      "loss": 2.2492,
      "step": 4121
    },
    {
      "epoch": 0.8244,
      "grad_norm": 0.042822934687137604,
      "learning_rate": 1.8245287515752708e-05,
      "loss": 2.246,
      "step": 4122
    },
    {
      "epoch": 0.8246,
      "grad_norm": 0.042503613978624344,
      "learning_rate": 1.8205104711271957e-05,
      "loss": 2.2473,
      "step": 4123
    },
    {
      "epoch": 0.8248,
      "grad_norm": 0.043856311589479446,
      "learning_rate": 1.816496177262952e-05,
      "loss": 2.2463,
      "step": 4124
    },
    {
      "epoch": 0.825,
      "grad_norm": 0.0423063263297081,
      "learning_rate": 1.812485871939056e-05,
      "loss": 2.2462,
      "step": 4125
    },
    {
      "epoch": 0.8252,
      "grad_norm": 0.04351063072681427,
      "learning_rate": 1.808479557110081e-05,
      "loss": 2.2473,
      "step": 4126
    },
    {
      "epoch": 0.8254,
      "grad_norm": 0.04410536587238312,
      "learning_rate": 1.804477234728661e-05,
      "loss": 2.2456,
      "step": 4127
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.04235951974987984,
      "learning_rate": 1.8004789067454764e-05,
      "loss": 2.245,
      "step": 4128
    },
    {
      "epoch": 0.8258,
      "grad_norm": 0.04518783837556839,
      "learning_rate": 1.7964845751092664e-05,
      "loss": 2.246,
      "step": 4129
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.04301745817065239,
      "learning_rate": 1.792494241766811e-05,
      "loss": 2.2471,
      "step": 4130
    },
    {
      "epoch": 0.8262,
      "grad_norm": 0.044412195682525635,
      "learning_rate": 1.78850790866296e-05,
      "loss": 2.2475,
      "step": 4131
    },
    {
      "epoch": 0.8264,
      "grad_norm": 0.04477781429886818,
      "learning_rate": 1.7845255777406e-05,
      "loss": 2.2469,
      "step": 4132
    },
    {
      "epoch": 0.8266,
      "grad_norm": 0.04253169521689415,
      "learning_rate": 1.7805472509406696e-05,
      "loss": 2.249,
      "step": 4133
    },
    {
      "epoch": 0.8268,
      "grad_norm": 0.044635310769081116,
      "learning_rate": 1.7765729302021596e-05,
      "loss": 2.2456,
      "step": 4134
    },
    {
      "epoch": 0.827,
      "grad_norm": 0.04260510578751564,
      "learning_rate": 1.7726026174621003e-05,
      "loss": 2.2467,
      "step": 4135
    },
    {
      "epoch": 0.8272,
      "grad_norm": 0.04461664333939552,
      "learning_rate": 1.7686363146555805e-05,
      "loss": 2.244,
      "step": 4136
    },
    {
      "epoch": 0.8274,
      "grad_norm": 0.042576923966407776,
      "learning_rate": 1.7646740237157256e-05,
      "loss": 2.2474,
      "step": 4137
    },
    {
      "epoch": 0.8276,
      "grad_norm": 0.042816367000341415,
      "learning_rate": 1.760715746573709e-05,
      "loss": 2.2449,
      "step": 4138
    },
    {
      "epoch": 0.8278,
      "grad_norm": 0.042964182794094086,
      "learning_rate": 1.7567614851587443e-05,
      "loss": 2.2443,
      "step": 4139
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.04329131171107292,
      "learning_rate": 1.752811241398089e-05,
      "loss": 2.2463,
      "step": 4140
    },
    {
      "epoch": 0.8282,
      "grad_norm": 0.0424012690782547,
      "learning_rate": 1.7488650172170496e-05,
      "loss": 2.2482,
      "step": 4141
    },
    {
      "epoch": 0.8284,
      "grad_norm": 0.04382287338376045,
      "learning_rate": 1.744922814538964e-05,
      "loss": 2.2497,
      "step": 4142
    },
    {
      "epoch": 0.8286,
      "grad_norm": 0.04487747699022293,
      "learning_rate": 1.7409846352852143e-05,
      "loss": 2.2476,
      "step": 4143
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.04301789775490761,
      "learning_rate": 1.737050481375223e-05,
      "loss": 2.2496,
      "step": 4144
    },
    {
      "epoch": 0.829,
      "grad_norm": 0.042713407427072525,
      "learning_rate": 1.7331203547264453e-05,
      "loss": 2.2434,
      "step": 4145
    },
    {
      "epoch": 0.8292,
      "grad_norm": 0.043304041028022766,
      "learning_rate": 1.7291942572543807e-05,
      "loss": 2.2468,
      "step": 4146
    },
    {
      "epoch": 0.8294,
      "grad_norm": 0.042175717651844025,
      "learning_rate": 1.7252721908725632e-05,
      "loss": 2.2453,
      "step": 4147
    },
    {
      "epoch": 0.8296,
      "grad_norm": 0.04388795047998428,
      "learning_rate": 1.721354157492555e-05,
      "loss": 2.2461,
      "step": 4148
    },
    {
      "epoch": 0.8298,
      "grad_norm": 0.043051574379205704,
      "learning_rate": 1.7174401590239587e-05,
      "loss": 2.2466,
      "step": 4149
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.04273408278822899,
      "learning_rate": 1.7135301973744124e-05,
      "loss": 2.2449,
      "step": 4150
    },
    {
      "epoch": 0.8302,
      "grad_norm": 0.04314368590712547,
      "learning_rate": 1.7096242744495837e-05,
      "loss": 2.2453,
      "step": 4151
    },
    {
      "epoch": 0.8304,
      "grad_norm": 0.04245135560631752,
      "learning_rate": 1.7057223921531707e-05,
      "loss": 2.2468,
      "step": 4152
    },
    {
      "epoch": 0.8306,
      "grad_norm": 0.04321559518575668,
      "learning_rate": 1.7018245523869036e-05,
      "loss": 2.2481,
      "step": 4153
    },
    {
      "epoch": 0.8308,
      "grad_norm": 0.042642563581466675,
      "learning_rate": 1.697930757050542e-05,
      "loss": 2.2475,
      "step": 4154
    },
    {
      "epoch": 0.831,
      "grad_norm": 0.042736101895570755,
      "learning_rate": 1.6940410080418723e-05,
      "loss": 2.2489,
      "step": 4155
    },
    {
      "epoch": 0.8312,
      "grad_norm": 0.04319055378437042,
      "learning_rate": 1.690155307256719e-05,
      "loss": 2.2472,
      "step": 4156
    },
    {
      "epoch": 0.8314,
      "grad_norm": 0.042257774621248245,
      "learning_rate": 1.686273656588917e-05,
      "loss": 2.2466,
      "step": 4157
    },
    {
      "epoch": 0.8316,
      "grad_norm": 0.04346176236867905,
      "learning_rate": 1.6823960579303377e-05,
      "loss": 2.2476,
      "step": 4158
    },
    {
      "epoch": 0.8318,
      "grad_norm": 0.0422145314514637,
      "learning_rate": 1.678522513170875e-05,
      "loss": 2.2458,
      "step": 4159
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.04230855777859688,
      "learning_rate": 1.6746530241984504e-05,
      "loss": 2.2473,
      "step": 4160
    },
    {
      "epoch": 0.8322,
      "grad_norm": 0.04299319535493851,
      "learning_rate": 1.6707875928990058e-05,
      "loss": 2.2473,
      "step": 4161
    },
    {
      "epoch": 0.8324,
      "grad_norm": 0.04215241223573685,
      "learning_rate": 1.6669262211565028e-05,
      "loss": 2.2456,
      "step": 4162
    },
    {
      "epoch": 0.8326,
      "grad_norm": 0.04289764538407326,
      "learning_rate": 1.6630689108529284e-05,
      "loss": 2.2446,
      "step": 4163
    },
    {
      "epoch": 0.8328,
      "grad_norm": 0.04286947846412659,
      "learning_rate": 1.6592156638682886e-05,
      "loss": 2.2463,
      "step": 4164
    },
    {
      "epoch": 0.833,
      "grad_norm": 0.04263506457209587,
      "learning_rate": 1.6553664820806102e-05,
      "loss": 2.2451,
      "step": 4165
    },
    {
      "epoch": 0.8332,
      "grad_norm": 0.042450301349163055,
      "learning_rate": 1.6515213673659357e-05,
      "loss": 2.2471,
      "step": 4166
    },
    {
      "epoch": 0.8334,
      "grad_norm": 0.042483944445848465,
      "learning_rate": 1.6476803215983294e-05,
      "loss": 2.2476,
      "step": 4167
    },
    {
      "epoch": 0.8336,
      "grad_norm": 0.043153222650289536,
      "learning_rate": 1.643843346649866e-05,
      "loss": 2.2478,
      "step": 4168
    },
    {
      "epoch": 0.8338,
      "grad_norm": 0.04267827421426773,
      "learning_rate": 1.640010444390646e-05,
      "loss": 2.2465,
      "step": 4169
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.04219396412372589,
      "learning_rate": 1.6361816166887768e-05,
      "loss": 2.2476,
      "step": 4170
    },
    {
      "epoch": 0.8342,
      "grad_norm": 0.04227854311466217,
      "learning_rate": 1.632356865410384e-05,
      "loss": 2.2449,
      "step": 4171
    },
    {
      "epoch": 0.8344,
      "grad_norm": 0.042713165283203125,
      "learning_rate": 1.628536192419603e-05,
      "loss": 2.2461,
      "step": 4172
    },
    {
      "epoch": 0.8346,
      "grad_norm": 0.042815450578927994,
      "learning_rate": 1.6247195995785837e-05,
      "loss": 2.2479,
      "step": 4173
    },
    {
      "epoch": 0.8348,
      "grad_norm": 0.04348187521100044,
      "learning_rate": 1.6209070887474876e-05,
      "loss": 2.2462,
      "step": 4174
    },
    {
      "epoch": 0.835,
      "grad_norm": 0.042716093361377716,
      "learning_rate": 1.6170986617844863e-05,
      "loss": 2.2462,
      "step": 4175
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.04257708042860031,
      "learning_rate": 1.6132943205457606e-05,
      "loss": 2.2479,
      "step": 4176
    },
    {
      "epoch": 0.8354,
      "grad_norm": 0.04261988773941994,
      "learning_rate": 1.6094940668855006e-05,
      "loss": 2.247,
      "step": 4177
    },
    {
      "epoch": 0.8356,
      "grad_norm": 0.04239582270383835,
      "learning_rate": 1.6056979026559004e-05,
      "loss": 2.2486,
      "step": 4178
    },
    {
      "epoch": 0.8358,
      "grad_norm": 0.04266057908535004,
      "learning_rate": 1.601905829707171e-05,
      "loss": 2.246,
      "step": 4179
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.04235038906335831,
      "learning_rate": 1.598117849887518e-05,
      "loss": 2.2468,
      "step": 4180
    },
    {
      "epoch": 0.8362,
      "grad_norm": 0.04310173541307449,
      "learning_rate": 1.5943339650431576e-05,
      "loss": 2.2471,
      "step": 4181
    },
    {
      "epoch": 0.8364,
      "grad_norm": 0.04216710478067398,
      "learning_rate": 1.5905541770183096e-05,
      "loss": 2.2466,
      "step": 4182
    },
    {
      "epoch": 0.8366,
      "grad_norm": 0.042886003851890564,
      "learning_rate": 1.586778487655197e-05,
      "loss": 2.2462,
      "step": 4183
    },
    {
      "epoch": 0.8368,
      "grad_norm": 0.04275131598114967,
      "learning_rate": 1.5830068987940438e-05,
      "loss": 2.2466,
      "step": 4184
    },
    {
      "epoch": 0.837,
      "grad_norm": 0.04345928132534027,
      "learning_rate": 1.579239412273078e-05,
      "loss": 2.2479,
      "step": 4185
    },
    {
      "epoch": 0.8372,
      "grad_norm": 0.04242853820323944,
      "learning_rate": 1.5754760299285252e-05,
      "loss": 2.2462,
      "step": 4186
    },
    {
      "epoch": 0.8374,
      "grad_norm": 0.042384013533592224,
      "learning_rate": 1.5717167535946144e-05,
      "loss": 2.2473,
      "step": 4187
    },
    {
      "epoch": 0.8376,
      "grad_norm": 0.04252569004893303,
      "learning_rate": 1.567961585103567e-05,
      "loss": 2.2481,
      "step": 4188
    },
    {
      "epoch": 0.8378,
      "grad_norm": 0.042520277202129364,
      "learning_rate": 1.564210526285612e-05,
      "loss": 2.2484,
      "step": 4189
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.04225335270166397,
      "learning_rate": 1.560463578968967e-05,
      "loss": 2.248,
      "step": 4190
    },
    {
      "epoch": 0.8382,
      "grad_norm": 0.043613746762275696,
      "learning_rate": 1.5567207449798515e-05,
      "loss": 2.2474,
      "step": 4191
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.04241429269313812,
      "learning_rate": 1.5529820261424698e-05,
      "loss": 2.2477,
      "step": 4192
    },
    {
      "epoch": 0.8386,
      "grad_norm": 0.04318554699420929,
      "learning_rate": 1.5492474242790366e-05,
      "loss": 2.2463,
      "step": 4193
    },
    {
      "epoch": 0.8388,
      "grad_norm": 0.04219919815659523,
      "learning_rate": 1.545516941209747e-05,
      "loss": 2.2475,
      "step": 4194
    },
    {
      "epoch": 0.839,
      "grad_norm": 0.043487682938575745,
      "learning_rate": 1.541790578752794e-05,
      "loss": 2.2465,
      "step": 4195
    },
    {
      "epoch": 0.8392,
      "grad_norm": 0.04231768473982811,
      "learning_rate": 1.538068338724361e-05,
      "loss": 2.2474,
      "step": 4196
    },
    {
      "epoch": 0.8394,
      "grad_norm": 0.04375532642006874,
      "learning_rate": 1.5343502229386207e-05,
      "loss": 2.2452,
      "step": 4197
    },
    {
      "epoch": 0.8396,
      "grad_norm": 0.043390631675720215,
      "learning_rate": 1.530636233207743e-05,
      "loss": 2.2456,
      "step": 4198
    },
    {
      "epoch": 0.8398,
      "grad_norm": 0.042602453380823135,
      "learning_rate": 1.526926371341878e-05,
      "loss": 2.2479,
      "step": 4199
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.04549301788210869,
      "learning_rate": 1.5232206391491699e-05,
      "loss": 2.2475,
      "step": 4200
    },
    {
      "epoch": 0.8402,
      "grad_norm": 0.04282153770327568,
      "learning_rate": 1.5195190384357404e-05,
      "loss": 2.2489,
      "step": 4201
    },
    {
      "epoch": 0.8404,
      "grad_norm": 0.04389774054288864,
      "learning_rate": 1.5158215710057123e-05,
      "loss": 2.2463,
      "step": 4202
    },
    {
      "epoch": 0.8406,
      "grad_norm": 0.04213562607765198,
      "learning_rate": 1.5121282386611824e-05,
      "loss": 2.2436,
      "step": 4203
    },
    {
      "epoch": 0.8408,
      "grad_norm": 0.0423583947122097,
      "learning_rate": 1.5084390432022377e-05,
      "loss": 2.2457,
      "step": 4204
    },
    {
      "epoch": 0.841,
      "grad_norm": 0.042128391563892365,
      "learning_rate": 1.5047539864269478e-05,
      "loss": 2.2429,
      "step": 4205
    },
    {
      "epoch": 0.8412,
      "grad_norm": 0.04236996918916702,
      "learning_rate": 1.5010730701313625e-05,
      "loss": 2.2456,
      "step": 4206
    },
    {
      "epoch": 0.8414,
      "grad_norm": 0.042877353727817535,
      "learning_rate": 1.4973962961095134e-05,
      "loss": 2.2479,
      "step": 4207
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.042507484555244446,
      "learning_rate": 1.4937236661534226e-05,
      "loss": 2.2486,
      "step": 4208
    },
    {
      "epoch": 0.8418,
      "grad_norm": 0.043403513729572296,
      "learning_rate": 1.4900551820530828e-05,
      "loss": 2.2482,
      "step": 4209
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.043889425694942474,
      "learning_rate": 1.486390845596466e-05,
      "loss": 2.2477,
      "step": 4210
    },
    {
      "epoch": 0.8422,
      "grad_norm": 0.04249398037791252,
      "learning_rate": 1.4827306585695234e-05,
      "loss": 2.2458,
      "step": 4211
    },
    {
      "epoch": 0.8424,
      "grad_norm": 0.04434576258063316,
      "learning_rate": 1.4790746227561925e-05,
      "loss": 2.2463,
      "step": 4212
    },
    {
      "epoch": 0.8426,
      "grad_norm": 0.04334045574069023,
      "learning_rate": 1.4754227399383757e-05,
      "loss": 2.2485,
      "step": 4213
    },
    {
      "epoch": 0.8428,
      "grad_norm": 0.042224712669849396,
      "learning_rate": 1.4717750118959584e-05,
      "loss": 2.2468,
      "step": 4214
    },
    {
      "epoch": 0.843,
      "grad_norm": 0.04244210943579674,
      "learning_rate": 1.468131440406798e-05,
      "loss": 2.2465,
      "step": 4215
    },
    {
      "epoch": 0.8432,
      "grad_norm": 0.04266824945807457,
      "learning_rate": 1.4644920272467244e-05,
      "loss": 2.2475,
      "step": 4216
    },
    {
      "epoch": 0.8434,
      "grad_norm": 0.04361007735133171,
      "learning_rate": 1.4608567741895495e-05,
      "loss": 2.2485,
      "step": 4217
    },
    {
      "epoch": 0.8436,
      "grad_norm": 0.04247981309890747,
      "learning_rate": 1.4572256830070497e-05,
      "loss": 2.2445,
      "step": 4218
    },
    {
      "epoch": 0.8438,
      "grad_norm": 0.04304089769721031,
      "learning_rate": 1.4535987554689712e-05,
      "loss": 2.2468,
      "step": 4219
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.043053317815065384,
      "learning_rate": 1.4499759933430346e-05,
      "loss": 2.2468,
      "step": 4220
    },
    {
      "epoch": 0.8442,
      "grad_norm": 0.042784061282873154,
      "learning_rate": 1.4463573983949341e-05,
      "loss": 2.2462,
      "step": 4221
    },
    {
      "epoch": 0.8444,
      "grad_norm": 0.04231737554073334,
      "learning_rate": 1.4427429723883257e-05,
      "loss": 2.2457,
      "step": 4222
    },
    {
      "epoch": 0.8446,
      "grad_norm": 0.04297446832060814,
      "learning_rate": 1.439132717084839e-05,
      "loss": 2.2476,
      "step": 4223
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.04239623248577118,
      "learning_rate": 1.4355266342440677e-05,
      "loss": 2.2472,
      "step": 4224
    },
    {
      "epoch": 0.845,
      "grad_norm": 0.04224582388997078,
      "learning_rate": 1.4319247256235714e-05,
      "loss": 2.2484,
      "step": 4225
    },
    {
      "epoch": 0.8452,
      "grad_norm": 0.042304422706365585,
      "learning_rate": 1.4283269929788779e-05,
      "loss": 2.2455,
      "step": 4226
    },
    {
      "epoch": 0.8454,
      "grad_norm": 0.04252997785806656,
      "learning_rate": 1.4247334380634792e-05,
      "loss": 2.2449,
      "step": 4227
    },
    {
      "epoch": 0.8456,
      "grad_norm": 0.04215705394744873,
      "learning_rate": 1.4211440626288286e-05,
      "loss": 2.2471,
      "step": 4228
    },
    {
      "epoch": 0.8458,
      "grad_norm": 0.04232439398765564,
      "learning_rate": 1.4175588684243446e-05,
      "loss": 2.248,
      "step": 4229
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.04222467541694641,
      "learning_rate": 1.4139778571974049e-05,
      "loss": 2.245,
      "step": 4230
    },
    {
      "epoch": 0.8462,
      "grad_norm": 0.042594898492097855,
      "learning_rate": 1.4104010306933557e-05,
      "loss": 2.2482,
      "step": 4231
    },
    {
      "epoch": 0.8464,
      "grad_norm": 0.04244188591837883,
      "learning_rate": 1.406828390655497e-05,
      "loss": 2.2464,
      "step": 4232
    },
    {
      "epoch": 0.8466,
      "grad_norm": 0.04213394969701767,
      "learning_rate": 1.40325993882509e-05,
      "loss": 2.2433,
      "step": 4233
    },
    {
      "epoch": 0.8468,
      "grad_norm": 0.04303458705544472,
      "learning_rate": 1.3996956769413538e-05,
      "loss": 2.2445,
      "step": 4234
    },
    {
      "epoch": 0.847,
      "grad_norm": 0.04262850433588028,
      "learning_rate": 1.3961356067414666e-05,
      "loss": 2.2459,
      "step": 4235
    },
    {
      "epoch": 0.8472,
      "grad_norm": 0.04216993972659111,
      "learning_rate": 1.3925797299605647e-05,
      "loss": 2.2464,
      "step": 4236
    },
    {
      "epoch": 0.8474,
      "grad_norm": 0.04266618192195892,
      "learning_rate": 1.3890280483317374e-05,
      "loss": 2.2463,
      "step": 4237
    },
    {
      "epoch": 0.8476,
      "grad_norm": 0.04257845878601074,
      "learning_rate": 1.3854805635860336e-05,
      "loss": 2.2443,
      "step": 4238
    },
    {
      "epoch": 0.8478,
      "grad_norm": 0.04258664324879646,
      "learning_rate": 1.3819372774524508e-05,
      "loss": 2.2473,
      "step": 4239
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.04416991025209427,
      "learning_rate": 1.3783981916579446e-05,
      "loss": 2.2469,
      "step": 4240
    },
    {
      "epoch": 0.8482,
      "grad_norm": 0.042294975370168686,
      "learning_rate": 1.3748633079274253e-05,
      "loss": 2.2458,
      "step": 4241
    },
    {
      "epoch": 0.8484,
      "grad_norm": 0.04278896003961563,
      "learning_rate": 1.3713326279837501e-05,
      "loss": 2.2468,
      "step": 4242
    },
    {
      "epoch": 0.8486,
      "grad_norm": 0.04292275756597519,
      "learning_rate": 1.3678061535477304e-05,
      "loss": 2.2467,
      "step": 4243
    },
    {
      "epoch": 0.8488,
      "grad_norm": 0.0423789843916893,
      "learning_rate": 1.3642838863381257e-05,
      "loss": 2.2482,
      "step": 4244
    },
    {
      "epoch": 0.849,
      "grad_norm": 0.042969442903995514,
      "learning_rate": 1.3607658280716473e-05,
      "loss": 2.2472,
      "step": 4245
    },
    {
      "epoch": 0.8492,
      "grad_norm": 0.04219621792435646,
      "learning_rate": 1.3572519804629536e-05,
      "loss": 2.2461,
      "step": 4246
    },
    {
      "epoch": 0.8494,
      "grad_norm": 0.042385075241327286,
      "learning_rate": 1.3537423452246523e-05,
      "loss": 2.2458,
      "step": 4247
    },
    {
      "epoch": 0.8496,
      "grad_norm": 0.04332932084798813,
      "learning_rate": 1.3502369240672941e-05,
      "loss": 2.2457,
      "step": 4248
    },
    {
      "epoch": 0.8498,
      "grad_norm": 0.04199465736746788,
      "learning_rate": 1.34673571869938e-05,
      "loss": 2.2444,
      "step": 4249
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.042348068207502365,
      "learning_rate": 1.3432387308273575e-05,
      "loss": 2.2466,
      "step": 4250
    },
    {
      "epoch": 0.8502,
      "grad_norm": 0.04198942333459854,
      "learning_rate": 1.339745962155613e-05,
      "loss": 2.246,
      "step": 4251
    },
    {
      "epoch": 0.8504,
      "grad_norm": 0.042475197464227676,
      "learning_rate": 1.3362574143864814e-05,
      "loss": 2.2441,
      "step": 4252
    },
    {
      "epoch": 0.8506,
      "grad_norm": 0.04255584627389908,
      "learning_rate": 1.3327730892202383e-05,
      "loss": 2.2486,
      "step": 4253
    },
    {
      "epoch": 0.8508,
      "grad_norm": 0.04229597747325897,
      "learning_rate": 1.3292929883550998e-05,
      "loss": 2.2483,
      "step": 4254
    },
    {
      "epoch": 0.851,
      "grad_norm": 0.04210527241230011,
      "learning_rate": 1.3258171134872265e-05,
      "loss": 2.2421,
      "step": 4255
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.04183517023921013,
      "learning_rate": 1.3223454663107172e-05,
      "loss": 2.245,
      "step": 4256
    },
    {
      "epoch": 0.8514,
      "grad_norm": 0.04224444180727005,
      "learning_rate": 1.3188780485176088e-05,
      "loss": 2.2452,
      "step": 4257
    },
    {
      "epoch": 0.8516,
      "grad_norm": 0.042254868894815445,
      "learning_rate": 1.3154148617978812e-05,
      "loss": 2.2488,
      "step": 4258
    },
    {
      "epoch": 0.8518,
      "grad_norm": 0.04228084906935692,
      "learning_rate": 1.3119559078394461e-05,
      "loss": 2.2465,
      "step": 4259
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.04202711954712868,
      "learning_rate": 1.3085011883281606e-05,
      "loss": 2.2427,
      "step": 4260
    },
    {
      "epoch": 0.8522,
      "grad_norm": 0.04264817386865616,
      "learning_rate": 1.30505070494781e-05,
      "loss": 2.2447,
      "step": 4261
    },
    {
      "epoch": 0.8524,
      "grad_norm": 0.042129699140787125,
      "learning_rate": 1.3016044593801202e-05,
      "loss": 2.2455,
      "step": 4262
    },
    {
      "epoch": 0.8526,
      "grad_norm": 0.04322460666298866,
      "learning_rate": 1.2981624533047432e-05,
      "loss": 2.2468,
      "step": 4263
    },
    {
      "epoch": 0.8528,
      "grad_norm": 0.04263778403401375,
      "learning_rate": 1.294724688399278e-05,
      "loss": 2.2459,
      "step": 4264
    },
    {
      "epoch": 0.853,
      "grad_norm": 0.04227285087108612,
      "learning_rate": 1.2912911663392469e-05,
      "loss": 2.2472,
      "step": 4265
    },
    {
      "epoch": 0.8532,
      "grad_norm": 0.042583491653203964,
      "learning_rate": 1.2878618887981064e-05,
      "loss": 2.2468,
      "step": 4266
    },
    {
      "epoch": 0.8534,
      "grad_norm": 0.04437197372317314,
      "learning_rate": 1.2844368574472454e-05,
      "loss": 2.2481,
      "step": 4267
    },
    {
      "epoch": 0.8536,
      "grad_norm": 0.04373043030500412,
      "learning_rate": 1.2810160739559796e-05,
      "loss": 2.2494,
      "step": 4268
    },
    {
      "epoch": 0.8538,
      "grad_norm": 0.04275861755013466,
      "learning_rate": 1.2775995399915631e-05,
      "loss": 2.2477,
      "step": 4269
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.04516648128628731,
      "learning_rate": 1.2741872572191682e-05,
      "loss": 2.2474,
      "step": 4270
    },
    {
      "epoch": 0.8542,
      "grad_norm": 0.04273242875933647,
      "learning_rate": 1.2707792273019048e-05,
      "loss": 2.2468,
      "step": 4271
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.04344552382826805,
      "learning_rate": 1.2673754519008008e-05,
      "loss": 2.2479,
      "step": 4272
    },
    {
      "epoch": 0.8546,
      "grad_norm": 0.04501458629965782,
      "learning_rate": 1.2639759326748135e-05,
      "loss": 2.2448,
      "step": 4273
    },
    {
      "epoch": 0.8548,
      "grad_norm": 0.04227891564369202,
      "learning_rate": 1.260580671280832e-05,
      "loss": 2.2448,
      "step": 4274
    },
    {
      "epoch": 0.855,
      "grad_norm": 0.04527784511446953,
      "learning_rate": 1.257189669373664e-05,
      "loss": 2.2478,
      "step": 4275
    },
    {
      "epoch": 0.8552,
      "grad_norm": 0.04445177689194679,
      "learning_rate": 1.2538029286060426e-05,
      "loss": 2.2489,
      "step": 4276
    },
    {
      "epoch": 0.8554,
      "grad_norm": 0.042956508696079254,
      "learning_rate": 1.2504204506286243e-05,
      "loss": 2.2459,
      "step": 4277
    },
    {
      "epoch": 0.8556,
      "grad_norm": 0.04287036508321762,
      "learning_rate": 1.2470422370899838e-05,
      "loss": 2.248,
      "step": 4278
    },
    {
      "epoch": 0.8558,
      "grad_norm": 0.0421048142015934,
      "learning_rate": 1.243668289636628e-05,
      "loss": 2.2477,
      "step": 4279
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.042257245630025864,
      "learning_rate": 1.2402986099129765e-05,
      "loss": 2.247,
      "step": 4280
    },
    {
      "epoch": 0.8562,
      "grad_norm": 0.04357968643307686,
      "learning_rate": 1.2369331995613665e-05,
      "loss": 2.2444,
      "step": 4281
    },
    {
      "epoch": 0.8564,
      "grad_norm": 0.04216937720775604,
      "learning_rate": 1.2335720602220569e-05,
      "loss": 2.247,
      "step": 4282
    },
    {
      "epoch": 0.8566,
      "grad_norm": 0.04211278632283211,
      "learning_rate": 1.2302151935332329e-05,
      "loss": 2.246,
      "step": 4283
    },
    {
      "epoch": 0.8568,
      "grad_norm": 0.04276847839355469,
      "learning_rate": 1.2268626011309858e-05,
      "loss": 2.2463,
      "step": 4284
    },
    {
      "epoch": 0.857,
      "grad_norm": 0.041919413954019547,
      "learning_rate": 1.2235142846493308e-05,
      "loss": 2.2453,
      "step": 4285
    },
    {
      "epoch": 0.8572,
      "grad_norm": 0.04185524210333824,
      "learning_rate": 1.2201702457201947e-05,
      "loss": 2.246,
      "step": 4286
    },
    {
      "epoch": 0.8574,
      "grad_norm": 0.04256486892700195,
      "learning_rate": 1.2168304859734226e-05,
      "loss": 2.2453,
      "step": 4287
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.04226496443152428,
      "learning_rate": 1.2134950070367723e-05,
      "loss": 2.2447,
      "step": 4288
    },
    {
      "epoch": 0.8578,
      "grad_norm": 0.04205084219574928,
      "learning_rate": 1.2101638105359169e-05,
      "loss": 2.2471,
      "step": 4289
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.04247293248772621,
      "learning_rate": 1.206836898094439e-05,
      "loss": 2.2459,
      "step": 4290
    },
    {
      "epoch": 0.8582,
      "grad_norm": 0.04207956790924072,
      "learning_rate": 1.2035142713338366e-05,
      "loss": 2.2472,
      "step": 4291
    },
    {
      "epoch": 0.8584,
      "grad_norm": 0.04198762774467468,
      "learning_rate": 1.2001959318735156e-05,
      "loss": 2.246,
      "step": 4292
    },
    {
      "epoch": 0.8586,
      "grad_norm": 0.04207804799079895,
      "learning_rate": 1.196881881330798e-05,
      "loss": 2.2467,
      "step": 4293
    },
    {
      "epoch": 0.8588,
      "grad_norm": 0.0423450693488121,
      "learning_rate": 1.1935721213209105e-05,
      "loss": 2.2484,
      "step": 4294
    },
    {
      "epoch": 0.859,
      "grad_norm": 0.042577315121889114,
      "learning_rate": 1.1902666534569883e-05,
      "loss": 2.249,
      "step": 4295
    },
    {
      "epoch": 0.8592,
      "grad_norm": 0.04242873564362526,
      "learning_rate": 1.1869654793500784e-05,
      "loss": 2.2465,
      "step": 4296
    },
    {
      "epoch": 0.8594,
      "grad_norm": 0.04240456968545914,
      "learning_rate": 1.1836686006091313e-05,
      "loss": 2.2463,
      "step": 4297
    },
    {
      "epoch": 0.8596,
      "grad_norm": 0.0427541509270668,
      "learning_rate": 1.1803760188410073e-05,
      "loss": 2.2466,
      "step": 4298
    },
    {
      "epoch": 0.8598,
      "grad_norm": 0.04344424232840538,
      "learning_rate": 1.1770877356504683e-05,
      "loss": 2.2489,
      "step": 4299
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.04369115084409714,
      "learning_rate": 1.1738037526401857e-05,
      "loss": 2.2458,
      "step": 4300
    },
    {
      "epoch": 0.8602,
      "grad_norm": 0.04223785921931267,
      "learning_rate": 1.1705240714107302e-05,
      "loss": 2.2466,
      "step": 4301
    },
    {
      "epoch": 0.8604,
      "grad_norm": 0.043701160699129105,
      "learning_rate": 1.1672486935605831e-05,
      "loss": 2.2468,
      "step": 4302
    },
    {
      "epoch": 0.8606,
      "grad_norm": 0.043343834578990936,
      "learning_rate": 1.1639776206861196e-05,
      "loss": 2.247,
      "step": 4303
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.0429883636534214,
      "learning_rate": 1.1607108543816248e-05,
      "loss": 2.2457,
      "step": 4304
    },
    {
      "epoch": 0.861,
      "grad_norm": 0.04488733783364296,
      "learning_rate": 1.1574483962392767e-05,
      "loss": 2.2478,
      "step": 4305
    },
    {
      "epoch": 0.8612,
      "grad_norm": 0.042108744382858276,
      "learning_rate": 1.1541902478491606e-05,
      "loss": 2.2469,
      "step": 4306
    },
    {
      "epoch": 0.8614,
      "grad_norm": 0.043609619140625,
      "learning_rate": 1.1509364107992583e-05,
      "loss": 2.2447,
      "step": 4307
    },
    {
      "epoch": 0.8616,
      "grad_norm": 0.04334346204996109,
      "learning_rate": 1.1476868866754486e-05,
      "loss": 2.245,
      "step": 4308
    },
    {
      "epoch": 0.8618,
      "grad_norm": 0.0423419289290905,
      "learning_rate": 1.1444416770615118e-05,
      "loss": 2.2471,
      "step": 4309
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.04345468431711197,
      "learning_rate": 1.1412007835391236e-05,
      "loss": 2.2502,
      "step": 4310
    },
    {
      "epoch": 0.8622,
      "grad_norm": 0.04621249437332153,
      "learning_rate": 1.1379642076878527e-05,
      "loss": 2.246,
      "step": 4311
    },
    {
      "epoch": 0.8624,
      "grad_norm": 0.04245927184820175,
      "learning_rate": 1.1347319510851717e-05,
      "loss": 2.2477,
      "step": 4312
    },
    {
      "epoch": 0.8626,
      "grad_norm": 0.04398512467741966,
      "learning_rate": 1.1315040153064416e-05,
      "loss": 2.2472,
      "step": 4313
    },
    {
      "epoch": 0.8628,
      "grad_norm": 0.04238060489296913,
      "learning_rate": 1.1282804019249182e-05,
      "loss": 2.2468,
      "step": 4314
    },
    {
      "epoch": 0.863,
      "grad_norm": 0.042352356016635895,
      "learning_rate": 1.1250611125117527e-05,
      "loss": 2.2487,
      "step": 4315
    },
    {
      "epoch": 0.8632,
      "grad_norm": 0.04311046004295349,
      "learning_rate": 1.1218461486359877e-05,
      "loss": 2.2497,
      "step": 4316
    },
    {
      "epoch": 0.8634,
      "grad_norm": 0.04327397421002388,
      "learning_rate": 1.1186355118645554e-05,
      "loss": 2.2476,
      "step": 4317
    },
    {
      "epoch": 0.8636,
      "grad_norm": 0.042662594467401505,
      "learning_rate": 1.1154292037622838e-05,
      "loss": 2.2476,
      "step": 4318
    },
    {
      "epoch": 0.8638,
      "grad_norm": 0.0420435406267643,
      "learning_rate": 1.1122272258918865e-05,
      "loss": 2.2466,
      "step": 4319
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.04314056411385536,
      "learning_rate": 1.1090295798139672e-05,
      "loss": 2.2464,
      "step": 4320
    },
    {
      "epoch": 0.8642,
      "grad_norm": 0.042234547436237335,
      "learning_rate": 1.1058362670870249e-05,
      "loss": 2.2469,
      "step": 4321
    },
    {
      "epoch": 0.8644,
      "grad_norm": 0.04251323267817497,
      "learning_rate": 1.1026472892674378e-05,
      "loss": 2.248,
      "step": 4322
    },
    {
      "epoch": 0.8646,
      "grad_norm": 0.04258756339550018,
      "learning_rate": 1.099462647909475e-05,
      "loss": 2.2457,
      "step": 4323
    },
    {
      "epoch": 0.8648,
      "grad_norm": 0.04266500845551491,
      "learning_rate": 1.0962823445652959e-05,
      "loss": 2.2457,
      "step": 4324
    },
    {
      "epoch": 0.865,
      "grad_norm": 0.04204858839511871,
      "learning_rate": 1.093106380784934e-05,
      "loss": 2.2473,
      "step": 4325
    },
    {
      "epoch": 0.8652,
      "grad_norm": 0.04213659092783928,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 2.2461,
      "step": 4326
    },
    {
      "epoch": 0.8654,
      "grad_norm": 0.0428352877497673,
      "learning_rate": 1.0867674781052684e-05,
      "loss": 2.247,
      "step": 4327
    },
    {
      "epoch": 0.8656,
      "grad_norm": 0.04251393675804138,
      "learning_rate": 1.0836045422954666e-05,
      "loss": 2.2469,
      "step": 4328
    },
    {
      "epoch": 0.8658,
      "grad_norm": 0.04402275010943413,
      "learning_rate": 1.0804459522284926e-05,
      "loss": 2.2461,
      "step": 4329
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.04297014698386192,
      "learning_rate": 1.0772917094438051e-05,
      "loss": 2.2495,
      "step": 4330
    },
    {
      "epoch": 0.8662,
      "grad_norm": 0.04289510101079941,
      "learning_rate": 1.0741418154787442e-05,
      "loss": 2.2475,
      "step": 4331
    },
    {
      "epoch": 0.8664,
      "grad_norm": 0.042321715503931046,
      "learning_rate": 1.0709962718685318e-05,
      "loss": 2.2456,
      "step": 4332
    },
    {
      "epoch": 0.8666,
      "grad_norm": 0.042740024626255035,
      "learning_rate": 1.067855080146266e-05,
      "loss": 2.2467,
      "step": 4333
    },
    {
      "epoch": 0.8668,
      "grad_norm": 0.04219790920615196,
      "learning_rate": 1.0647182418429225e-05,
      "loss": 2.2478,
      "step": 4334
    },
    {
      "epoch": 0.867,
      "grad_norm": 0.04201864078640938,
      "learning_rate": 1.0615857584873623e-05,
      "loss": 2.2477,
      "step": 4335
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.0419863760471344,
      "learning_rate": 1.0584576316063188e-05,
      "loss": 2.2469,
      "step": 4336
    },
    {
      "epoch": 0.8674,
      "grad_norm": 0.04216728359460831,
      "learning_rate": 1.0553338627244025e-05,
      "loss": 2.246,
      "step": 4337
    },
    {
      "epoch": 0.8676,
      "grad_norm": 0.0419631265103817,
      "learning_rate": 1.0522144533640998e-05,
      "loss": 2.2465,
      "step": 4338
    },
    {
      "epoch": 0.8678,
      "grad_norm": 0.04219980165362358,
      "learning_rate": 1.0490994050457748e-05,
      "loss": 2.246,
      "step": 4339
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.04209089279174805,
      "learning_rate": 1.0459887192876594e-05,
      "loss": 2.2476,
      "step": 4340
    },
    {
      "epoch": 0.8682,
      "grad_norm": 0.04212191328406334,
      "learning_rate": 1.042882397605871e-05,
      "loss": 2.2446,
      "step": 4341
    },
    {
      "epoch": 0.8684,
      "grad_norm": 0.042805153876543045,
      "learning_rate": 1.0397804415143909e-05,
      "loss": 2.2463,
      "step": 4342
    },
    {
      "epoch": 0.8686,
      "grad_norm": 0.04247391223907471,
      "learning_rate": 1.0366828525250726e-05,
      "loss": 2.2456,
      "step": 4343
    },
    {
      "epoch": 0.8688,
      "grad_norm": 0.0434224009513855,
      "learning_rate": 1.0335896321476413e-05,
      "loss": 2.247,
      "step": 4344
    },
    {
      "epoch": 0.869,
      "grad_norm": 0.04229902848601341,
      "learning_rate": 1.0305007818897006e-05,
      "loss": 2.2456,
      "step": 4345
    },
    {
      "epoch": 0.8692,
      "grad_norm": 0.04203388839960098,
      "learning_rate": 1.0274163032567163e-05,
      "loss": 2.247,
      "step": 4346
    },
    {
      "epoch": 0.8694,
      "grad_norm": 0.043223317712545395,
      "learning_rate": 1.0243361977520249e-05,
      "loss": 2.2458,
      "step": 4347
    },
    {
      "epoch": 0.8696,
      "grad_norm": 0.042722027748823166,
      "learning_rate": 1.0212604668768343e-05,
      "loss": 2.2464,
      "step": 4348
    },
    {
      "epoch": 0.8698,
      "grad_norm": 0.04287485405802727,
      "learning_rate": 1.0181891121302145e-05,
      "loss": 2.2473,
      "step": 4349
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.04229862615466118,
      "learning_rate": 1.0151221350091134e-05,
      "loss": 2.2478,
      "step": 4350
    },
    {
      "epoch": 0.8702,
      "grad_norm": 0.043793145567178726,
      "learning_rate": 1.0120595370083318e-05,
      "loss": 2.2486,
      "step": 4351
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.0418410524725914,
      "learning_rate": 1.009001319620545e-05,
      "loss": 2.2427,
      "step": 4352
    },
    {
      "epoch": 0.8706,
      "grad_norm": 0.04252767562866211,
      "learning_rate": 1.0059474843362892e-05,
      "loss": 2.244,
      "step": 4353
    },
    {
      "epoch": 0.8708,
      "grad_norm": 0.043481532484292984,
      "learning_rate": 1.0028980326439707e-05,
      "loss": 2.2464,
      "step": 4354
    },
    {
      "epoch": 0.871,
      "grad_norm": 0.04215213283896446,
      "learning_rate": 9.998529660298539e-06,
      "loss": 2.2471,
      "step": 4355
    },
    {
      "epoch": 0.8712,
      "grad_norm": 0.043759722262620926,
      "learning_rate": 9.968122859780648e-06,
      "loss": 2.2472,
      "step": 4356
    },
    {
      "epoch": 0.8714,
      "grad_norm": 0.0428805947303772,
      "learning_rate": 9.937759939705971e-06,
      "loss": 2.2472,
      "step": 4357
    },
    {
      "epoch": 0.8716,
      "grad_norm": 0.04227194935083389,
      "learning_rate": 9.907440914873e-06,
      "loss": 2.2469,
      "step": 4358
    },
    {
      "epoch": 0.8718,
      "grad_norm": 0.04188724234700203,
      "learning_rate": 9.877165800058874e-06,
      "loss": 2.2457,
      "step": 4359
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.0425703227519989,
      "learning_rate": 9.84693461001932e-06,
      "loss": 2.2469,
      "step": 4360
    },
    {
      "epoch": 0.8722,
      "grad_norm": 0.04223932698369026,
      "learning_rate": 9.816747359488632e-06,
      "loss": 2.2474,
      "step": 4361
    },
    {
      "epoch": 0.8724,
      "grad_norm": 0.04225975647568703,
      "learning_rate": 9.786604063179728e-06,
      "loss": 2.2469,
      "step": 4362
    },
    {
      "epoch": 0.8726,
      "grad_norm": 0.04255088046193123,
      "learning_rate": 9.756504735784067e-06,
      "loss": 2.2471,
      "step": 4363
    },
    {
      "epoch": 0.8728,
      "grad_norm": 0.04379403963685036,
      "learning_rate": 9.726449391971714e-06,
      "loss": 2.2478,
      "step": 4364
    },
    {
      "epoch": 0.873,
      "grad_norm": 0.04268002510070801,
      "learning_rate": 9.696438046391288e-06,
      "loss": 2.2486,
      "step": 4365
    },
    {
      "epoch": 0.8732,
      "grad_norm": 0.04312638193368912,
      "learning_rate": 9.666470713669918e-06,
      "loss": 2.2461,
      "step": 4366
    },
    {
      "epoch": 0.8734,
      "grad_norm": 0.04338795319199562,
      "learning_rate": 9.636547408413355e-06,
      "loss": 2.2469,
      "step": 4367
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.042302656918764114,
      "learning_rate": 9.606668145205833e-06,
      "loss": 2.2457,
      "step": 4368
    },
    {
      "epoch": 0.8738,
      "grad_norm": 0.04206240549683571,
      "learning_rate": 9.576832938610137e-06,
      "loss": 2.248,
      "step": 4369
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.04277682676911354,
      "learning_rate": 9.5470418031676e-06,
      "loss": 2.2473,
      "step": 4370
    },
    {
      "epoch": 0.8742,
      "grad_norm": 0.04336588457226753,
      "learning_rate": 9.517294753398064e-06,
      "loss": 2.2475,
      "step": 4371
    },
    {
      "epoch": 0.8744,
      "grad_norm": 0.04201910272240639,
      "learning_rate": 9.487591803799856e-06,
      "loss": 2.247,
      "step": 4372
    },
    {
      "epoch": 0.8746,
      "grad_norm": 0.04232001677155495,
      "learning_rate": 9.457932968849825e-06,
      "loss": 2.2446,
      "step": 4373
    },
    {
      "epoch": 0.8748,
      "grad_norm": 0.04355714097619057,
      "learning_rate": 9.428318263003378e-06,
      "loss": 2.2442,
      "step": 4374
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.044539324939250946,
      "learning_rate": 9.398747700694322e-06,
      "loss": 2.2491,
      "step": 4375
    },
    {
      "epoch": 0.8752,
      "grad_norm": 0.04255610704421997,
      "learning_rate": 9.369221296335006e-06,
      "loss": 2.2477,
      "step": 4376
    },
    {
      "epoch": 0.8754,
      "grad_norm": 0.04290711507201195,
      "learning_rate": 9.339739064316233e-06,
      "loss": 2.2457,
      "step": 4377
    },
    {
      "epoch": 0.8756,
      "grad_norm": 0.04927946999669075,
      "learning_rate": 9.310301019007285e-06,
      "loss": 2.2485,
      "step": 4378
    },
    {
      "epoch": 0.8758,
      "grad_norm": 0.04475207254290581,
      "learning_rate": 9.280907174755915e-06,
      "loss": 2.2467,
      "step": 4379
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.045855239033699036,
      "learning_rate": 9.251557545888312e-06,
      "loss": 2.2477,
      "step": 4380
    },
    {
      "epoch": 0.8762,
      "grad_norm": 0.04230930656194687,
      "learning_rate": 9.222252146709142e-06,
      "loss": 2.2463,
      "step": 4381
    },
    {
      "epoch": 0.8764,
      "grad_norm": 0.042546357959508896,
      "learning_rate": 9.192990991501482e-06,
      "loss": 2.2473,
      "step": 4382
    },
    {
      "epoch": 0.8766,
      "grad_norm": 0.04518045485019684,
      "learning_rate": 9.163774094526889e-06,
      "loss": 2.248,
      "step": 4383
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.042531393468379974,
      "learning_rate": 9.134601470025306e-06,
      "loss": 2.2457,
      "step": 4384
    },
    {
      "epoch": 0.877,
      "grad_norm": 0.042162779718637466,
      "learning_rate": 9.105473132215125e-06,
      "loss": 2.2472,
      "step": 4385
    },
    {
      "epoch": 0.8772,
      "grad_norm": 0.044599127024412155,
      "learning_rate": 9.076389095293148e-06,
      "loss": 2.2471,
      "step": 4386
    },
    {
      "epoch": 0.8774,
      "grad_norm": 0.043389011174440384,
      "learning_rate": 9.047349373434566e-06,
      "loss": 2.2496,
      "step": 4387
    },
    {
      "epoch": 0.8776,
      "grad_norm": 0.04217509925365448,
      "learning_rate": 9.018353980792993e-06,
      "loss": 2.2468,
      "step": 4388
    },
    {
      "epoch": 0.8778,
      "grad_norm": 0.043369390070438385,
      "learning_rate": 8.989402931500434e-06,
      "loss": 2.2463,
      "step": 4389
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.043551500886678696,
      "learning_rate": 8.960496239667282e-06,
      "loss": 2.2469,
      "step": 4390
    },
    {
      "epoch": 0.8782,
      "grad_norm": 0.0420420877635479,
      "learning_rate": 8.931633919382298e-06,
      "loss": 2.2471,
      "step": 4391
    },
    {
      "epoch": 0.8784,
      "grad_norm": 0.04445742443203926,
      "learning_rate": 8.90281598471262e-06,
      "loss": 2.2475,
      "step": 4392
    },
    {
      "epoch": 0.8786,
      "grad_norm": 0.04208650067448616,
      "learning_rate": 8.87404244970378e-06,
      "loss": 2.2465,
      "step": 4393
    },
    {
      "epoch": 0.8788,
      "grad_norm": 0.04355565831065178,
      "learning_rate": 8.845313328379634e-06,
      "loss": 2.2473,
      "step": 4394
    },
    {
      "epoch": 0.879,
      "grad_norm": 0.04239115118980408,
      "learning_rate": 8.816628634742441e-06,
      "loss": 2.2483,
      "step": 4395
    },
    {
      "epoch": 0.8792,
      "grad_norm": 0.04239444434642792,
      "learning_rate": 8.787988382772705e-06,
      "loss": 2.2471,
      "step": 4396
    },
    {
      "epoch": 0.8794,
      "grad_norm": 0.042485978454351425,
      "learning_rate": 8.759392586429393e-06,
      "loss": 2.2463,
      "step": 4397
    },
    {
      "epoch": 0.8796,
      "grad_norm": 0.04317314550280571,
      "learning_rate": 8.730841259649725e-06,
      "loss": 2.2472,
      "step": 4398
    },
    {
      "epoch": 0.8798,
      "grad_norm": 0.04287562146782875,
      "learning_rate": 8.702334416349278e-06,
      "loss": 2.2477,
      "step": 4399
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.0451127253472805,
      "learning_rate": 8.67387207042194e-06,
      "loss": 2.2465,
      "step": 4400
    },
    {
      "epoch": 0.8802,
      "grad_norm": 0.04232543334364891,
      "learning_rate": 8.645454235739903e-06,
      "loss": 2.2473,
      "step": 4401
    },
    {
      "epoch": 0.8804,
      "grad_norm": 0.042325183749198914,
      "learning_rate": 8.617080926153698e-06,
      "loss": 2.2479,
      "step": 4402
    },
    {
      "epoch": 0.8806,
      "grad_norm": 0.042301952838897705,
      "learning_rate": 8.588752155492119e-06,
      "loss": 2.2473,
      "step": 4403
    },
    {
      "epoch": 0.8808,
      "grad_norm": 0.04233676940202713,
      "learning_rate": 8.560467937562277e-06,
      "loss": 2.2469,
      "step": 4404
    },
    {
      "epoch": 0.881,
      "grad_norm": 0.042364202439785004,
      "learning_rate": 8.532228286149501e-06,
      "loss": 2.2465,
      "step": 4405
    },
    {
      "epoch": 0.8812,
      "grad_norm": 0.042219460010528564,
      "learning_rate": 8.504033215017527e-06,
      "loss": 2.2458,
      "step": 4406
    },
    {
      "epoch": 0.8814,
      "grad_norm": 0.042322177439928055,
      "learning_rate": 8.475882737908248e-06,
      "loss": 2.2467,
      "step": 4407
    },
    {
      "epoch": 0.8816,
      "grad_norm": 0.042210210114717484,
      "learning_rate": 8.447776868541879e-06,
      "loss": 2.2459,
      "step": 4408
    },
    {
      "epoch": 0.8818,
      "grad_norm": 0.0424596406519413,
      "learning_rate": 8.419715620616874e-06,
      "loss": 2.2459,
      "step": 4409
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.04220402240753174,
      "learning_rate": 8.39169900780995e-06,
      "loss": 2.2454,
      "step": 4410
    },
    {
      "epoch": 0.8822,
      "grad_norm": 0.042293161153793335,
      "learning_rate": 8.363727043776038e-06,
      "loss": 2.2459,
      "step": 4411
    },
    {
      "epoch": 0.8824,
      "grad_norm": 0.04241516813635826,
      "learning_rate": 8.335799742148387e-06,
      "loss": 2.2475,
      "step": 4412
    },
    {
      "epoch": 0.8826,
      "grad_norm": 0.0425039604306221,
      "learning_rate": 8.307917116538378e-06,
      "loss": 2.2494,
      "step": 4413
    },
    {
      "epoch": 0.8828,
      "grad_norm": 0.04222685471177101,
      "learning_rate": 8.280079180535672e-06,
      "loss": 2.247,
      "step": 4414
    },
    {
      "epoch": 0.883,
      "grad_norm": 0.04232260212302208,
      "learning_rate": 8.252285947708139e-06,
      "loss": 2.2445,
      "step": 4415
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.04204900190234184,
      "learning_rate": 8.224537431601886e-06,
      "loss": 2.2458,
      "step": 4416
    },
    {
      "epoch": 0.8834,
      "grad_norm": 0.043279897421598434,
      "learning_rate": 8.196833645741186e-06,
      "loss": 2.2482,
      "step": 4417
    },
    {
      "epoch": 0.8836,
      "grad_norm": 0.04349076747894287,
      "learning_rate": 8.169174603628538e-06,
      "loss": 2.2474,
      "step": 4418
    },
    {
      "epoch": 0.8838,
      "grad_norm": 0.04247501865029335,
      "learning_rate": 8.1415603187446e-06,
      "loss": 2.2474,
      "step": 4419
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.04316752031445503,
      "learning_rate": 8.113990804548244e-06,
      "loss": 2.2501,
      "step": 4420
    },
    {
      "epoch": 0.8842,
      "grad_norm": 0.04218990355730057,
      "learning_rate": 8.086466074476563e-06,
      "loss": 2.2453,
      "step": 4421
    },
    {
      "epoch": 0.8844,
      "grad_norm": 0.04221469908952713,
      "learning_rate": 8.058986141944724e-06,
      "loss": 2.249,
      "step": 4422
    },
    {
      "epoch": 0.8846,
      "grad_norm": 0.04245307296514511,
      "learning_rate": 8.031551020346128e-06,
      "loss": 2.2492,
      "step": 4423
    },
    {
      "epoch": 0.8848,
      "grad_norm": 0.042556032538414,
      "learning_rate": 8.004160723052312e-06,
      "loss": 2.2481,
      "step": 4424
    },
    {
      "epoch": 0.885,
      "grad_norm": 0.04212749004364014,
      "learning_rate": 7.976815263412963e-06,
      "loss": 2.2467,
      "step": 4425
    },
    {
      "epoch": 0.8852,
      "grad_norm": 0.04221435636281967,
      "learning_rate": 7.949514654755962e-06,
      "loss": 2.2484,
      "step": 4426
    },
    {
      "epoch": 0.8854,
      "grad_norm": 0.04235000163316727,
      "learning_rate": 7.922258910387282e-06,
      "loss": 2.2484,
      "step": 4427
    },
    {
      "epoch": 0.8856,
      "grad_norm": 0.04425071179866791,
      "learning_rate": 7.895048043591036e-06,
      "loss": 2.2499,
      "step": 4428
    },
    {
      "epoch": 0.8858,
      "grad_norm": 0.042118821293115616,
      "learning_rate": 7.867882067629472e-06,
      "loss": 2.248,
      "step": 4429
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.04240143299102783,
      "learning_rate": 7.840760995742946e-06,
      "loss": 2.2494,
      "step": 4430
    },
    {
      "epoch": 0.8862,
      "grad_norm": 0.042363397777080536,
      "learning_rate": 7.81368484114996e-06,
      "loss": 2.2471,
      "step": 4431
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.042064767330884933,
      "learning_rate": 7.786653617047079e-06,
      "loss": 2.2481,
      "step": 4432
    },
    {
      "epoch": 0.8866,
      "grad_norm": 0.04227539896965027,
      "learning_rate": 7.75966733660901e-06,
      "loss": 2.2487,
      "step": 4433
    },
    {
      "epoch": 0.8868,
      "grad_norm": 0.04236982762813568,
      "learning_rate": 7.73272601298851e-06,
      "loss": 2.2462,
      "step": 4434
    },
    {
      "epoch": 0.887,
      "grad_norm": 0.042364560067653656,
      "learning_rate": 7.7058296593165e-06,
      "loss": 2.2485,
      "step": 4435
    },
    {
      "epoch": 0.8872,
      "grad_norm": 0.04216094687581062,
      "learning_rate": 7.67897828870191e-06,
      "loss": 2.2477,
      "step": 4436
    },
    {
      "epoch": 0.8874,
      "grad_norm": 0.04256565868854523,
      "learning_rate": 7.652171914231776e-06,
      "loss": 2.2465,
      "step": 4437
    },
    {
      "epoch": 0.8876,
      "grad_norm": 0.04260299727320671,
      "learning_rate": 7.625410548971191e-06,
      "loss": 2.2455,
      "step": 4438
    },
    {
      "epoch": 0.8878,
      "grad_norm": 0.042191945016384125,
      "learning_rate": 7.59869420596333e-06,
      "loss": 2.2453,
      "step": 4439
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.042240168899297714,
      "learning_rate": 7.572022898229403e-06,
      "loss": 2.2471,
      "step": 4440
    },
    {
      "epoch": 0.8882,
      "grad_norm": 0.042641639709472656,
      "learning_rate": 7.545396638768698e-06,
      "loss": 2.2478,
      "step": 4441
    },
    {
      "epoch": 0.8884,
      "grad_norm": 0.04250340536236763,
      "learning_rate": 7.518815440558513e-06,
      "loss": 2.2481,
      "step": 4442
    },
    {
      "epoch": 0.8886,
      "grad_norm": 0.042032379657030106,
      "learning_rate": 7.492279316554207e-06,
      "loss": 2.2469,
      "step": 4443
    },
    {
      "epoch": 0.8888,
      "grad_norm": 0.04243841767311096,
      "learning_rate": 7.465788279689156e-06,
      "loss": 2.2478,
      "step": 4444
    },
    {
      "epoch": 0.889,
      "grad_norm": 0.042077645659446716,
      "learning_rate": 7.439342342874789e-06,
      "loss": 2.2474,
      "step": 4445
    },
    {
      "epoch": 0.8892,
      "grad_norm": 0.042653512209653854,
      "learning_rate": 7.412941519000527e-06,
      "loss": 2.2461,
      "step": 4446
    },
    {
      "epoch": 0.8894,
      "grad_norm": 0.04197916388511658,
      "learning_rate": 7.386585820933811e-06,
      "loss": 2.2483,
      "step": 4447
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.042191192507743835,
      "learning_rate": 7.360275261520078e-06,
      "loss": 2.2483,
      "step": 4448
    },
    {
      "epoch": 0.8898,
      "grad_norm": 0.04223394766449928,
      "learning_rate": 7.3340098535827905e-06,
      "loss": 2.2476,
      "step": 4449
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.04209490865468979,
      "learning_rate": 7.3077896099233765e-06,
      "loss": 2.2461,
      "step": 4450
    },
    {
      "epoch": 0.8902,
      "grad_norm": 0.04170761629939079,
      "learning_rate": 7.281614543321269e-06,
      "loss": 2.2465,
      "step": 4451
    },
    {
      "epoch": 0.8904,
      "grad_norm": 0.04217234253883362,
      "learning_rate": 7.255484666533874e-06,
      "loss": 2.2463,
      "step": 4452
    },
    {
      "epoch": 0.8906,
      "grad_norm": 0.042304374277591705,
      "learning_rate": 7.2293999922965705e-06,
      "loss": 2.2466,
      "step": 4453
    },
    {
      "epoch": 0.8908,
      "grad_norm": 0.042026586830616,
      "learning_rate": 7.203360533322734e-06,
      "loss": 2.2469,
      "step": 4454
    },
    {
      "epoch": 0.891,
      "grad_norm": 0.04254571348428726,
      "learning_rate": 7.177366302303667e-06,
      "loss": 2.2486,
      "step": 4455
    },
    {
      "epoch": 0.8912,
      "grad_norm": 0.04215645417571068,
      "learning_rate": 7.151417311908648e-06,
      "loss": 2.2454,
      "step": 4456
    },
    {
      "epoch": 0.8914,
      "grad_norm": 0.04204672947525978,
      "learning_rate": 7.125513574784903e-06,
      "loss": 2.2463,
      "step": 4457
    },
    {
      "epoch": 0.8916,
      "grad_norm": 0.04237568378448486,
      "learning_rate": 7.099655103557556e-06,
      "loss": 2.2454,
      "step": 4458
    },
    {
      "epoch": 0.8918,
      "grad_norm": 0.04390484094619751,
      "learning_rate": 7.07384191082977e-06,
      "loss": 2.2502,
      "step": 4459
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.042452748864889145,
      "learning_rate": 7.048074009182548e-06,
      "loss": 2.2475,
      "step": 4460
    },
    {
      "epoch": 0.8922,
      "grad_norm": 0.04220503941178322,
      "learning_rate": 7.022351411174866e-06,
      "loss": 2.2477,
      "step": 4461
    },
    {
      "epoch": 0.8924,
      "grad_norm": 0.04223506525158882,
      "learning_rate": 6.996674129343606e-06,
      "loss": 2.2478,
      "step": 4462
    },
    {
      "epoch": 0.8926,
      "grad_norm": 0.042983610183000565,
      "learning_rate": 6.971042176203535e-06,
      "loss": 2.2467,
      "step": 4463
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.04207136482000351,
      "learning_rate": 6.945455564247394e-06,
      "loss": 2.2456,
      "step": 4464
    },
    {
      "epoch": 0.893,
      "grad_norm": 0.04220470413565636,
      "learning_rate": 6.919914305945774e-06,
      "loss": 2.247,
      "step": 4465
    },
    {
      "epoch": 0.8932,
      "grad_norm": 0.042150549590587616,
      "learning_rate": 6.894418413747183e-06,
      "loss": 2.248,
      "step": 4466
    },
    {
      "epoch": 0.8934,
      "grad_norm": 0.04276266694068909,
      "learning_rate": 6.868967900077972e-06,
      "loss": 2.2456,
      "step": 4467
    },
    {
      "epoch": 0.8936,
      "grad_norm": 0.04273210093379021,
      "learning_rate": 6.8435627773424495e-06,
      "loss": 2.2473,
      "step": 4468
    },
    {
      "epoch": 0.8938,
      "grad_norm": 0.043414805084466934,
      "learning_rate": 6.818203057922757e-06,
      "loss": 2.2492,
      "step": 4469
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.04313512146472931,
      "learning_rate": 6.7928887541789055e-06,
      "loss": 2.2472,
      "step": 4470
    },
    {
      "epoch": 0.8942,
      "grad_norm": 0.042200446128845215,
      "learning_rate": 6.767619878448783e-06,
      "loss": 2.2458,
      "step": 4471
    },
    {
      "epoch": 0.8944,
      "grad_norm": 0.04300621896982193,
      "learning_rate": 6.742396443048138e-06,
      "loss": 2.2491,
      "step": 4472
    },
    {
      "epoch": 0.8946,
      "grad_norm": 0.04263417050242424,
      "learning_rate": 6.717218460270536e-06,
      "loss": 2.2473,
      "step": 4473
    },
    {
      "epoch": 0.8948,
      "grad_norm": 0.04213889315724373,
      "learning_rate": 6.692085942387483e-06,
      "loss": 2.2465,
      "step": 4474
    },
    {
      "epoch": 0.895,
      "grad_norm": 0.042169149965047836,
      "learning_rate": 6.666998901648203e-06,
      "loss": 2.2475,
      "step": 4475
    },
    {
      "epoch": 0.8952,
      "grad_norm": 0.04312710091471672,
      "learning_rate": 6.6419573502798374e-06,
      "loss": 2.2485,
      "step": 4476
    },
    {
      "epoch": 0.8954,
      "grad_norm": 0.044529203325510025,
      "learning_rate": 6.616961300487324e-06,
      "loss": 2.2448,
      "step": 4477
    },
    {
      "epoch": 0.8956,
      "grad_norm": 0.042799048125743866,
      "learning_rate": 6.592010764453449e-06,
      "loss": 2.2462,
      "step": 4478
    },
    {
      "epoch": 0.8958,
      "grad_norm": 0.04328519478440285,
      "learning_rate": 6.5671057543387985e-06,
      "loss": 2.247,
      "step": 4479
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.04273047670722008,
      "learning_rate": 6.542246282281772e-06,
      "loss": 2.2457,
      "step": 4480
    },
    {
      "epoch": 0.8962,
      "grad_norm": 0.04265209659934044,
      "learning_rate": 6.517432360398556e-06,
      "loss": 2.2458,
      "step": 4481
    },
    {
      "epoch": 0.8964,
      "grad_norm": 0.04236316680908203,
      "learning_rate": 6.492664000783166e-06,
      "loss": 2.2474,
      "step": 4482
    },
    {
      "epoch": 0.8966,
      "grad_norm": 0.042268864810466766,
      "learning_rate": 6.467941215507433e-06,
      "loss": 2.2448,
      "step": 4483
    },
    {
      "epoch": 0.8968,
      "grad_norm": 0.041839729994535446,
      "learning_rate": 6.443264016620887e-06,
      "loss": 2.2445,
      "step": 4484
    },
    {
      "epoch": 0.897,
      "grad_norm": 0.04193761199712753,
      "learning_rate": 6.418632416150927e-06,
      "loss": 2.2457,
      "step": 4485
    },
    {
      "epoch": 0.8972,
      "grad_norm": 0.04223021864891052,
      "learning_rate": 6.394046426102674e-06,
      "loss": 2.2456,
      "step": 4486
    },
    {
      "epoch": 0.8974,
      "grad_norm": 0.042054276913404465,
      "learning_rate": 6.3695060584590625e-06,
      "loss": 2.2471,
      "step": 4487
    },
    {
      "epoch": 0.8976,
      "grad_norm": 0.041923586279153824,
      "learning_rate": 6.345011325180772e-06,
      "loss": 2.2456,
      "step": 4488
    },
    {
      "epoch": 0.8978,
      "grad_norm": 0.04198106750845909,
      "learning_rate": 6.320562238206218e-06,
      "loss": 2.2461,
      "step": 4489
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.04186665266752243,
      "learning_rate": 6.296158809451602e-06,
      "loss": 2.2462,
      "step": 4490
    },
    {
      "epoch": 0.8982,
      "grad_norm": 0.04208570718765259,
      "learning_rate": 6.2718010508108545e-06,
      "loss": 2.2467,
      "step": 4491
    },
    {
      "epoch": 0.8984,
      "grad_norm": 0.04302738606929779,
      "learning_rate": 6.2474889741556575e-06,
      "loss": 2.2492,
      "step": 4492
    },
    {
      "epoch": 0.8986,
      "grad_norm": 0.043631941080093384,
      "learning_rate": 6.223222591335409e-06,
      "loss": 2.2517,
      "step": 4493
    },
    {
      "epoch": 0.8988,
      "grad_norm": 0.04313802719116211,
      "learning_rate": 6.1990019141772605e-06,
      "loss": 2.2506,
      "step": 4494
    },
    {
      "epoch": 0.899,
      "grad_norm": 0.04332403838634491,
      "learning_rate": 6.174826954486068e-06,
      "loss": 2.2491,
      "step": 4495
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.041906777769327164,
      "learning_rate": 6.1506977240444074e-06,
      "loss": 2.2454,
      "step": 4496
    },
    {
      "epoch": 0.8994,
      "grad_norm": 0.04195789620280266,
      "learning_rate": 6.126614234612593e-06,
      "loss": 2.2462,
      "step": 4497
    },
    {
      "epoch": 0.8996,
      "grad_norm": 0.04295383766293526,
      "learning_rate": 6.102576497928614e-06,
      "loss": 2.2467,
      "step": 4498
    },
    {
      "epoch": 0.8998,
      "grad_norm": 0.042315393686294556,
      "learning_rate": 6.078584525708176e-06,
      "loss": 2.2465,
      "step": 4499
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.04225185140967369,
      "learning_rate": 6.054638329644657e-06,
      "loss": 2.2457,
      "step": 4500
    },
    {
      "epoch": 0.9002,
      "grad_norm": 0.04241446033120155,
      "learning_rate": 6.030737921409169e-06,
      "loss": 2.2461,
      "step": 4501
    },
    {
      "epoch": 0.9004,
      "grad_norm": 0.0421096533536911,
      "learning_rate": 6.006883312650457e-06,
      "loss": 2.2443,
      "step": 4502
    },
    {
      "epoch": 0.9006,
      "grad_norm": 0.04315692558884621,
      "learning_rate": 5.98307451499498e-06,
      "loss": 2.2456,
      "step": 4503
    },
    {
      "epoch": 0.9008,
      "grad_norm": 0.04203221946954727,
      "learning_rate": 5.9593115400468636e-06,
      "loss": 2.2456,
      "step": 4504
    },
    {
      "epoch": 0.901,
      "grad_norm": 0.04302586242556572,
      "learning_rate": 5.935594399387856e-06,
      "loss": 2.2486,
      "step": 4505
    },
    {
      "epoch": 0.9012,
      "grad_norm": 0.042068496346473694,
      "learning_rate": 5.911923104577455e-06,
      "loss": 2.2456,
      "step": 4506
    },
    {
      "epoch": 0.9014,
      "grad_norm": 0.04253055155277252,
      "learning_rate": 5.888297667152731e-06,
      "loss": 2.2474,
      "step": 4507
    },
    {
      "epoch": 0.9016,
      "grad_norm": 0.0424773246049881,
      "learning_rate": 5.864718098628441e-06,
      "loss": 2.2454,
      "step": 4508
    },
    {
      "epoch": 0.9018,
      "grad_norm": 0.042437031865119934,
      "learning_rate": 5.8411844104969916e-06,
      "loss": 2.2444,
      "step": 4509
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.04224766045808792,
      "learning_rate": 5.8176966142283965e-06,
      "loss": 2.245,
      "step": 4510
    },
    {
      "epoch": 0.9022,
      "grad_norm": 0.042413968592882156,
      "learning_rate": 5.7942547212703315e-06,
      "loss": 2.246,
      "step": 4511
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.04311200603842735,
      "learning_rate": 5.770858743048091e-06,
      "loss": 2.2468,
      "step": 4512
    },
    {
      "epoch": 0.9026,
      "grad_norm": 0.041987717151641846,
      "learning_rate": 5.747508690964598e-06,
      "loss": 2.2448,
      "step": 4513
    },
    {
      "epoch": 0.9028,
      "grad_norm": 0.042389921844005585,
      "learning_rate": 5.724204576400371e-06,
      "loss": 2.2434,
      "step": 4514
    },
    {
      "epoch": 0.903,
      "grad_norm": 0.042077936232089996,
      "learning_rate": 5.700946410713548e-06,
      "loss": 2.2458,
      "step": 4515
    },
    {
      "epoch": 0.9032,
      "grad_norm": 0.047447700053453445,
      "learning_rate": 5.6777342052399045e-06,
      "loss": 2.2467,
      "step": 4516
    },
    {
      "epoch": 0.9034,
      "grad_norm": 0.05228515341877937,
      "learning_rate": 5.654567971292757e-06,
      "loss": 2.2478,
      "step": 4517
    },
    {
      "epoch": 0.9036,
      "grad_norm": 0.046592287719249725,
      "learning_rate": 5.6314477201630745e-06,
      "loss": 2.2461,
      "step": 4518
    },
    {
      "epoch": 0.9038,
      "grad_norm": 0.04925387352705002,
      "learning_rate": 5.608373463119354e-06,
      "loss": 2.249,
      "step": 4519
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.044598959386348724,
      "learning_rate": 5.585345211407733e-06,
      "loss": 2.2453,
      "step": 4520
    },
    {
      "epoch": 0.9042,
      "grad_norm": 0.042134642601013184,
      "learning_rate": 5.562362976251901e-06,
      "loss": 2.246,
      "step": 4521
    },
    {
      "epoch": 0.9044,
      "grad_norm": 0.043870288878679276,
      "learning_rate": 5.5394267688531066e-06,
      "loss": 2.2482,
      "step": 4522
    },
    {
      "epoch": 0.9046,
      "grad_norm": 0.04796696826815605,
      "learning_rate": 5.516536600390188e-06,
      "loss": 2.2495,
      "step": 4523
    },
    {
      "epoch": 0.9048,
      "grad_norm": 0.046737585216760635,
      "learning_rate": 5.49369248201953e-06,
      "loss": 2.249,
      "step": 4524
    },
    {
      "epoch": 0.905,
      "grad_norm": 0.042963895946741104,
      "learning_rate": 5.470894424875062e-06,
      "loss": 2.2481,
      "step": 4525
    },
    {
      "epoch": 0.9052,
      "grad_norm": 0.04557519406080246,
      "learning_rate": 5.448142440068316e-06,
      "loss": 2.2474,
      "step": 4526
    },
    {
      "epoch": 0.9054,
      "grad_norm": 0.05313529446721077,
      "learning_rate": 5.425436538688322e-06,
      "loss": 2.2477,
      "step": 4527
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.045883920043706894,
      "learning_rate": 5.402776731801662e-06,
      "loss": 2.2464,
      "step": 4528
    },
    {
      "epoch": 0.9058,
      "grad_norm": 0.04268309846520424,
      "learning_rate": 5.380163030452412e-06,
      "loss": 2.248,
      "step": 4529
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.04267093166708946,
      "learning_rate": 5.357595445662267e-06,
      "loss": 2.2477,
      "step": 4530
    },
    {
      "epoch": 0.9062,
      "grad_norm": 0.044431302696466446,
      "learning_rate": 5.335073988430372e-06,
      "loss": 2.2492,
      "step": 4531
    },
    {
      "epoch": 0.9064,
      "grad_norm": 0.04724140465259552,
      "learning_rate": 5.312598669733404e-06,
      "loss": 2.2477,
      "step": 4532
    },
    {
      "epoch": 0.9066,
      "grad_norm": 0.04359735548496246,
      "learning_rate": 5.290169500525577e-06,
      "loss": 2.246,
      "step": 4533
    },
    {
      "epoch": 0.9068,
      "grad_norm": 0.04225108399987221,
      "learning_rate": 5.267786491738569e-06,
      "loss": 2.2473,
      "step": 4534
    },
    {
      "epoch": 0.907,
      "grad_norm": 0.04267553612589836,
      "learning_rate": 5.245449654281631e-06,
      "loss": 2.2468,
      "step": 4535
    },
    {
      "epoch": 0.9072,
      "grad_norm": 0.044603824615478516,
      "learning_rate": 5.223158999041444e-06,
      "loss": 2.2463,
      "step": 4536
    },
    {
      "epoch": 0.9074,
      "grad_norm": 0.04522203654050827,
      "learning_rate": 5.200914536882185e-06,
      "loss": 2.249,
      "step": 4537
    },
    {
      "epoch": 0.9076,
      "grad_norm": 0.04203332960605621,
      "learning_rate": 5.178716278645535e-06,
      "loss": 2.2463,
      "step": 4538
    },
    {
      "epoch": 0.9078,
      "grad_norm": 0.042629893869161606,
      "learning_rate": 5.1565642351506845e-06,
      "loss": 2.2452,
      "step": 4539
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.044222209602594376,
      "learning_rate": 5.134458417194254e-06,
      "loss": 2.2456,
      "step": 4540
    },
    {
      "epoch": 0.9082,
      "grad_norm": 0.042899370193481445,
      "learning_rate": 5.1123988355503475e-06,
      "loss": 2.2449,
      "step": 4541
    },
    {
      "epoch": 0.9084,
      "grad_norm": 0.042594246566295624,
      "learning_rate": 5.0903855009705514e-06,
      "loss": 2.2469,
      "step": 4542
    },
    {
      "epoch": 0.9086,
      "grad_norm": 0.04250936582684517,
      "learning_rate": 5.068418424183874e-06,
      "loss": 2.2464,
      "step": 4543
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.043413691222667694,
      "learning_rate": 5.046497615896806e-06,
      "loss": 2.2488,
      "step": 4544
    },
    {
      "epoch": 0.909,
      "grad_norm": 0.043816957622766495,
      "learning_rate": 5.024623086793323e-06,
      "loss": 2.247,
      "step": 4545
    },
    {
      "epoch": 0.9092,
      "grad_norm": 0.043012745678424835,
      "learning_rate": 5.002794847534764e-06,
      "loss": 2.2492,
      "step": 4546
    },
    {
      "epoch": 0.9094,
      "grad_norm": 0.04216570407152176,
      "learning_rate": 4.981012908759941e-06,
      "loss": 2.2456,
      "step": 4547
    },
    {
      "epoch": 0.9096,
      "grad_norm": 0.04172011837363243,
      "learning_rate": 4.959277281085129e-06,
      "loss": 2.2455,
      "step": 4548
    },
    {
      "epoch": 0.9098,
      "grad_norm": 0.04236491024494171,
      "learning_rate": 4.937587975103996e-06,
      "loss": 2.248,
      "step": 4549
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.04317000135779381,
      "learning_rate": 4.915945001387667e-06,
      "loss": 2.2451,
      "step": 4550
    },
    {
      "epoch": 0.9102,
      "grad_norm": 0.04235965013504028,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 2.2466,
      "step": 4551
    },
    {
      "epoch": 0.9104,
      "grad_norm": 0.0432877391576767,
      "learning_rate": 4.872798092920872e-06,
      "loss": 2.2489,
      "step": 4552
    },
    {
      "epoch": 0.9106,
      "grad_norm": 0.04253983497619629,
      "learning_rate": 4.8512941791996726e-06,
      "loss": 2.2465,
      "step": 4553
    },
    {
      "epoch": 0.9108,
      "grad_norm": 0.043307818472385406,
      "learning_rate": 4.829836639801843e-06,
      "loss": 2.2458,
      "step": 4554
    },
    {
      "epoch": 0.911,
      "grad_norm": 0.043061513453722,
      "learning_rate": 4.808425485185486e-06,
      "loss": 2.247,
      "step": 4555
    },
    {
      "epoch": 0.9112,
      "grad_norm": 0.042426466941833496,
      "learning_rate": 4.7870607257861415e-06,
      "loss": 2.248,
      "step": 4556
    },
    {
      "epoch": 0.9114,
      "grad_norm": 0.04382331296801567,
      "learning_rate": 4.765742372016735e-06,
      "loss": 2.2492,
      "step": 4557
    },
    {
      "epoch": 0.9116,
      "grad_norm": 0.042736560106277466,
      "learning_rate": 4.744470434267567e-06,
      "loss": 2.2473,
      "step": 4558
    },
    {
      "epoch": 0.9118,
      "grad_norm": 0.04181554168462753,
      "learning_rate": 4.723244922906356e-06,
      "loss": 2.2481,
      "step": 4559
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.042567212134599686,
      "learning_rate": 4.702065848278126e-06,
      "loss": 2.2488,
      "step": 4560
    },
    {
      "epoch": 0.9122,
      "grad_norm": 0.042348623275756836,
      "learning_rate": 4.680933220705308e-06,
      "loss": 2.2465,
      "step": 4561
    },
    {
      "epoch": 0.9124,
      "grad_norm": 0.04197364300489426,
      "learning_rate": 4.659847050487687e-06,
      "loss": 2.2474,
      "step": 4562
    },
    {
      "epoch": 0.9126,
      "grad_norm": 0.042128194123506546,
      "learning_rate": 4.638807347902408e-06,
      "loss": 2.2478,
      "step": 4563
    },
    {
      "epoch": 0.9128,
      "grad_norm": 0.04227472469210625,
      "learning_rate": 4.617814123203967e-06,
      "loss": 2.2475,
      "step": 4564
    },
    {
      "epoch": 0.913,
      "grad_norm": 0.04196309670805931,
      "learning_rate": 4.596867386624215e-06,
      "loss": 2.2454,
      "step": 4565
    },
    {
      "epoch": 0.9132,
      "grad_norm": 0.04249652102589607,
      "learning_rate": 4.575967148372317e-06,
      "loss": 2.2469,
      "step": 4566
    },
    {
      "epoch": 0.9134,
      "grad_norm": 0.04207258298993111,
      "learning_rate": 4.555113418634805e-06,
      "loss": 2.2455,
      "step": 4567
    },
    {
      "epoch": 0.9136,
      "grad_norm": 0.04224136471748352,
      "learning_rate": 4.534306207575545e-06,
      "loss": 2.2458,
      "step": 4568
    },
    {
      "epoch": 0.9138,
      "grad_norm": 0.04315311461687088,
      "learning_rate": 4.513545525335705e-06,
      "loss": 2.2483,
      "step": 4569
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.04249391332268715,
      "learning_rate": 4.492831382033791e-06,
      "loss": 2.247,
      "step": 4570
    },
    {
      "epoch": 0.9142,
      "grad_norm": 0.04203025996685028,
      "learning_rate": 4.4721637877656375e-06,
      "loss": 2.2472,
      "step": 4571
    },
    {
      "epoch": 0.9144,
      "grad_norm": 0.0427931472659111,
      "learning_rate": 4.451542752604365e-06,
      "loss": 2.2469,
      "step": 4572
    },
    {
      "epoch": 0.9146,
      "grad_norm": 0.041936200112104416,
      "learning_rate": 4.4309682866004125e-06,
      "loss": 2.2466,
      "step": 4573
    },
    {
      "epoch": 0.9148,
      "grad_norm": 0.041768066585063934,
      "learning_rate": 4.410440399781534e-06,
      "loss": 2.2461,
      "step": 4574
    },
    {
      "epoch": 0.915,
      "grad_norm": 0.04176061227917671,
      "learning_rate": 4.389959102152774e-06,
      "loss": 2.2459,
      "step": 4575
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.04205342009663582,
      "learning_rate": 4.369524403696457e-06,
      "loss": 2.2458,
      "step": 4576
    },
    {
      "epoch": 0.9154,
      "grad_norm": 0.042066242545843124,
      "learning_rate": 4.349136314372204e-06,
      "loss": 2.2468,
      "step": 4577
    },
    {
      "epoch": 0.9156,
      "grad_norm": 0.04268717020750046,
      "learning_rate": 4.328794844116946e-06,
      "loss": 2.2485,
      "step": 4578
    },
    {
      "epoch": 0.9158,
      "grad_norm": 0.042421575635671616,
      "learning_rate": 4.308500002844862e-06,
      "loss": 2.2472,
      "step": 4579
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.042485058307647705,
      "learning_rate": 4.288251800447385e-06,
      "loss": 2.2459,
      "step": 4580
    },
    {
      "epoch": 0.9162,
      "grad_norm": 0.04203041270375252,
      "learning_rate": 4.268050246793276e-06,
      "loss": 2.2463,
      "step": 4581
    },
    {
      "epoch": 0.9164,
      "grad_norm": 0.04211852326989174,
      "learning_rate": 4.247895351728504e-06,
      "loss": 2.2474,
      "step": 4582
    },
    {
      "epoch": 0.9166,
      "grad_norm": 0.04225442558526993,
      "learning_rate": 4.227787125076332e-06,
      "loss": 2.2468,
      "step": 4583
    },
    {
      "epoch": 0.9168,
      "grad_norm": 0.042000673711299896,
      "learning_rate": 4.207725576637256e-06,
      "loss": 2.2461,
      "step": 4584
    },
    {
      "epoch": 0.917,
      "grad_norm": 0.042117029428482056,
      "learning_rate": 4.187710716189042e-06,
      "loss": 2.2469,
      "step": 4585
    },
    {
      "epoch": 0.9172,
      "grad_norm": 0.042087722569704056,
      "learning_rate": 4.167742553486675e-06,
      "loss": 2.2459,
      "step": 4586
    },
    {
      "epoch": 0.9174,
      "grad_norm": 0.042040880769491196,
      "learning_rate": 4.147821098262405e-06,
      "loss": 2.246,
      "step": 4587
    },
    {
      "epoch": 0.9176,
      "grad_norm": 0.04225427284836769,
      "learning_rate": 4.127946360225721e-06,
      "loss": 2.247,
      "step": 4588
    },
    {
      "epoch": 0.9178,
      "grad_norm": 0.04163091629743576,
      "learning_rate": 4.108118349063306e-06,
      "loss": 2.2459,
      "step": 4589
    },
    {
      "epoch": 0.918,
      "grad_norm": 0.04199833422899246,
      "learning_rate": 4.088337074439097e-06,
      "loss": 2.2449,
      "step": 4590
    },
    {
      "epoch": 0.9182,
      "grad_norm": 0.042134348303079605,
      "learning_rate": 4.068602545994249e-06,
      "loss": 2.2467,
      "step": 4591
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.04216643422842026,
      "learning_rate": 4.048914773347134e-06,
      "loss": 2.2463,
      "step": 4592
    },
    {
      "epoch": 0.9186,
      "grad_norm": 0.04202716052532196,
      "learning_rate": 4.029273766093333e-06,
      "loss": 2.2451,
      "step": 4593
    },
    {
      "epoch": 0.9188,
      "grad_norm": 0.04209074005484581,
      "learning_rate": 4.009679533805633e-06,
      "loss": 2.2463,
      "step": 4594
    },
    {
      "epoch": 0.919,
      "grad_norm": 0.04290100559592247,
      "learning_rate": 3.990132086034026e-06,
      "loss": 2.2471,
      "step": 4595
    },
    {
      "epoch": 0.9192,
      "grad_norm": 0.04319092258810997,
      "learning_rate": 3.970631432305694e-06,
      "loss": 2.2463,
      "step": 4596
    },
    {
      "epoch": 0.9194,
      "grad_norm": 0.0419539175927639,
      "learning_rate": 3.951177582125021e-06,
      "loss": 2.2463,
      "step": 4597
    },
    {
      "epoch": 0.9196,
      "grad_norm": 0.04207827150821686,
      "learning_rate": 3.931770544973601e-06,
      "loss": 2.2468,
      "step": 4598
    },
    {
      "epoch": 0.9198,
      "grad_norm": 0.042830538004636765,
      "learning_rate": 3.912410330310156e-06,
      "loss": 2.2475,
      "step": 4599
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.043677590787410736,
      "learning_rate": 3.893096947570618e-06,
      "loss": 2.2471,
      "step": 4600
    },
    {
      "epoch": 0.9202,
      "grad_norm": 0.041998688131570816,
      "learning_rate": 3.873830406168111e-06,
      "loss": 2.2469,
      "step": 4601
    },
    {
      "epoch": 0.9204,
      "grad_norm": 0.0421612486243248,
      "learning_rate": 3.8546107154929235e-06,
      "loss": 2.2449,
      "step": 4602
    },
    {
      "epoch": 0.9206,
      "grad_norm": 0.04204260930418968,
      "learning_rate": 3.835437884912474e-06,
      "loss": 2.248,
      "step": 4603
    },
    {
      "epoch": 0.9208,
      "grad_norm": 0.042348798364400864,
      "learning_rate": 3.816311923771387e-06,
      "loss": 2.2474,
      "step": 4604
    },
    {
      "epoch": 0.921,
      "grad_norm": 0.04235577583312988,
      "learning_rate": 3.797232841391407e-06,
      "loss": 2.2468,
      "step": 4605
    },
    {
      "epoch": 0.9212,
      "grad_norm": 0.04235254228115082,
      "learning_rate": 3.7782006470714616e-06,
      "loss": 2.2454,
      "step": 4606
    },
    {
      "epoch": 0.9214,
      "grad_norm": 0.04237660765647888,
      "learning_rate": 3.759215350087619e-06,
      "loss": 2.2464,
      "step": 4607
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.041864294558763504,
      "learning_rate": 3.7402769596930563e-06,
      "loss": 2.2463,
      "step": 4608
    },
    {
      "epoch": 0.9218,
      "grad_norm": 0.042129356414079666,
      "learning_rate": 3.721385485118123e-06,
      "loss": 2.2477,
      "step": 4609
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.041834037750959396,
      "learning_rate": 3.7025409355702976e-06,
      "loss": 2.2458,
      "step": 4610
    },
    {
      "epoch": 0.9222,
      "grad_norm": 0.042171675711870193,
      "learning_rate": 3.68374332023419e-06,
      "loss": 2.2462,
      "step": 4611
    },
    {
      "epoch": 0.9224,
      "grad_norm": 0.042219169437885284,
      "learning_rate": 3.664992648271526e-06,
      "loss": 2.2454,
      "step": 4612
    },
    {
      "epoch": 0.9226,
      "grad_norm": 0.04255953058600426,
      "learning_rate": 3.6462889288211512e-06,
      "loss": 2.2478,
      "step": 4613
    },
    {
      "epoch": 0.9228,
      "grad_norm": 0.042440857738256454,
      "learning_rate": 3.627632170999029e-06,
      "loss": 2.2457,
      "step": 4614
    },
    {
      "epoch": 0.923,
      "grad_norm": 0.04235079512000084,
      "learning_rate": 3.6090223838982417e-06,
      "loss": 2.2478,
      "step": 4615
    },
    {
      "epoch": 0.9232,
      "grad_norm": 0.04212510213255882,
      "learning_rate": 3.5904595765890005e-06,
      "loss": 2.2478,
      "step": 4616
    },
    {
      "epoch": 0.9234,
      "grad_norm": 0.041679929941892624,
      "learning_rate": 3.5719437581185454e-06,
      "loss": 2.2449,
      "step": 4617
    },
    {
      "epoch": 0.9236,
      "grad_norm": 0.04196219518780708,
      "learning_rate": 3.553474937511281e-06,
      "loss": 2.2471,
      "step": 4618
    },
    {
      "epoch": 0.9238,
      "grad_norm": 0.04251885041594505,
      "learning_rate": 3.5350531237686724e-06,
      "loss": 2.2481,
      "step": 4619
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.04200327768921852,
      "learning_rate": 3.516678325869316e-06,
      "loss": 2.2454,
      "step": 4620
    },
    {
      "epoch": 0.9242,
      "grad_norm": 0.042492154985666275,
      "learning_rate": 3.4983505527688586e-06,
      "loss": 2.2462,
      "step": 4621
    },
    {
      "epoch": 0.9244,
      "grad_norm": 0.04213618487119675,
      "learning_rate": 3.480069813400022e-06,
      "loss": 2.2442,
      "step": 4622
    },
    {
      "epoch": 0.9246,
      "grad_norm": 0.04219068959355354,
      "learning_rate": 3.461836116672612e-06,
      "loss": 2.2468,
      "step": 4623
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.042059607803821564,
      "learning_rate": 3.443649471473531e-06,
      "loss": 2.2448,
      "step": 4624
    },
    {
      "epoch": 0.925,
      "grad_norm": 0.04188152775168419,
      "learning_rate": 3.425509886666711e-06,
      "loss": 2.2476,
      "step": 4625
    },
    {
      "epoch": 0.9252,
      "grad_norm": 0.04183971509337425,
      "learning_rate": 3.40741737109318e-06,
      "loss": 2.2468,
      "step": 4626
    },
    {
      "epoch": 0.9254,
      "grad_norm": 0.042108550667762756,
      "learning_rate": 3.389371933570995e-06,
      "loss": 2.2463,
      "step": 4627
    },
    {
      "epoch": 0.9256,
      "grad_norm": 0.042171090841293335,
      "learning_rate": 3.3713735828952987e-06,
      "loss": 2.2478,
      "step": 4628
    },
    {
      "epoch": 0.9258,
      "grad_norm": 0.04210466891527176,
      "learning_rate": 3.3534223278382405e-06,
      "loss": 2.2471,
      "step": 4629
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.04185418412089348,
      "learning_rate": 3.3355181771490772e-06,
      "loss": 2.2474,
      "step": 4630
    },
    {
      "epoch": 0.9262,
      "grad_norm": 0.042330868542194366,
      "learning_rate": 3.3176611395540626e-06,
      "loss": 2.2456,
      "step": 4631
    },
    {
      "epoch": 0.9264,
      "grad_norm": 0.04176975414156914,
      "learning_rate": 3.2998512237565005e-06,
      "loss": 2.2422,
      "step": 4632
    },
    {
      "epoch": 0.9266,
      "grad_norm": 0.04215880110859871,
      "learning_rate": 3.2820884384367146e-06,
      "loss": 2.2475,
      "step": 4633
    },
    {
      "epoch": 0.9268,
      "grad_norm": 0.04277411475777626,
      "learning_rate": 3.2643727922520906e-06,
      "loss": 2.2495,
      "step": 4634
    },
    {
      "epoch": 0.927,
      "grad_norm": 0.04231472685933113,
      "learning_rate": 3.246704293837011e-06,
      "loss": 2.2446,
      "step": 4635
    },
    {
      "epoch": 0.9272,
      "grad_norm": 0.042001064866781235,
      "learning_rate": 3.2290829518028862e-06,
      "loss": 2.2455,
      "step": 4636
    },
    {
      "epoch": 0.9274,
      "grad_norm": 0.0423799566924572,
      "learning_rate": 3.2115087747381366e-06,
      "loss": 2.2458,
      "step": 4637
    },
    {
      "epoch": 0.9276,
      "grad_norm": 0.04296949505805969,
      "learning_rate": 3.19398177120821e-06,
      "loss": 2.2465,
      "step": 4638
    },
    {
      "epoch": 0.9278,
      "grad_norm": 0.042394936084747314,
      "learning_rate": 3.1765019497555616e-06,
      "loss": 2.2448,
      "step": 4639
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.04363377019762993,
      "learning_rate": 3.1590693188996323e-06,
      "loss": 2.2472,
      "step": 4640
    },
    {
      "epoch": 0.9282,
      "grad_norm": 0.041970714926719666,
      "learning_rate": 3.1416838871368924e-06,
      "loss": 2.2458,
      "step": 4641
    },
    {
      "epoch": 0.9284,
      "grad_norm": 0.04176197573542595,
      "learning_rate": 3.124345662940764e-06,
      "loss": 2.2446,
      "step": 4642
    },
    {
      "epoch": 0.9286,
      "grad_norm": 0.04186378791928291,
      "learning_rate": 3.1070546547617095e-06,
      "loss": 2.2465,
      "step": 4643
    },
    {
      "epoch": 0.9288,
      "grad_norm": 0.04181055352091789,
      "learning_rate": 3.0898108710271434e-06,
      "loss": 2.2442,
      "step": 4644
    },
    {
      "epoch": 0.929,
      "grad_norm": 0.041941311210393906,
      "learning_rate": 3.072614320141487e-06,
      "loss": 2.2462,
      "step": 4645
    },
    {
      "epoch": 0.9292,
      "grad_norm": 0.04188625141978264,
      "learning_rate": 3.0554650104861136e-06,
      "loss": 2.2473,
      "step": 4646
    },
    {
      "epoch": 0.9294,
      "grad_norm": 0.04298453405499458,
      "learning_rate": 3.0383629504194046e-06,
      "loss": 2.247,
      "step": 4647
    },
    {
      "epoch": 0.9296,
      "grad_norm": 0.045267533510923386,
      "learning_rate": 3.0213081482766805e-06,
      "loss": 2.2475,
      "step": 4648
    },
    {
      "epoch": 0.9298,
      "grad_norm": 0.04182538390159607,
      "learning_rate": 3.0043006123702697e-06,
      "loss": 2.2469,
      "step": 4649
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.04211350530385971,
      "learning_rate": 2.9873403509894203e-06,
      "loss": 2.2466,
      "step": 4650
    },
    {
      "epoch": 0.9302,
      "grad_norm": 0.04425599053502083,
      "learning_rate": 2.970427372400353e-06,
      "loss": 2.2494,
      "step": 4651
    },
    {
      "epoch": 0.9304,
      "grad_norm": 0.04222286865115166,
      "learning_rate": 2.953561684846262e-06,
      "loss": 2.2437,
      "step": 4652
    },
    {
      "epoch": 0.9306,
      "grad_norm": 0.04247282072901726,
      "learning_rate": 2.936743296547273e-06,
      "loss": 2.2442,
      "step": 4653
    },
    {
      "epoch": 0.9308,
      "grad_norm": 0.04234054312109947,
      "learning_rate": 2.9199722157004616e-06,
      "loss": 2.2457,
      "step": 4654
    },
    {
      "epoch": 0.931,
      "grad_norm": 0.041991520673036575,
      "learning_rate": 2.9032484504798452e-06,
      "loss": 2.2445,
      "step": 4655
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.04267240688204765,
      "learning_rate": 2.8865720090364034e-06,
      "loss": 2.2453,
      "step": 4656
    },
    {
      "epoch": 0.9314,
      "grad_norm": 0.04213079810142517,
      "learning_rate": 2.8699428994980017e-06,
      "loss": 2.2456,
      "step": 4657
    },
    {
      "epoch": 0.9316,
      "grad_norm": 0.04203270003199577,
      "learning_rate": 2.8533611299694783e-06,
      "loss": 2.2457,
      "step": 4658
    },
    {
      "epoch": 0.9318,
      "grad_norm": 0.041567519307136536,
      "learning_rate": 2.836826708532603e-06,
      "loss": 2.2453,
      "step": 4659
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.04196970537304878,
      "learning_rate": 2.8203396432460506e-06,
      "loss": 2.2464,
      "step": 4660
    },
    {
      "epoch": 0.9322,
      "grad_norm": 0.042941004037857056,
      "learning_rate": 2.8038999421453826e-06,
      "loss": 2.2474,
      "step": 4661
    },
    {
      "epoch": 0.9324,
      "grad_norm": 0.04242083430290222,
      "learning_rate": 2.7875076132431344e-06,
      "loss": 2.2468,
      "step": 4662
    },
    {
      "epoch": 0.9326,
      "grad_norm": 0.04224457964301109,
      "learning_rate": 2.771162664528726e-06,
      "loss": 2.2451,
      "step": 4663
    },
    {
      "epoch": 0.9328,
      "grad_norm": 0.04223548620939255,
      "learning_rate": 2.7548651039684846e-06,
      "loss": 2.2473,
      "step": 4664
    },
    {
      "epoch": 0.933,
      "grad_norm": 0.04189779981970787,
      "learning_rate": 2.738614939505646e-06,
      "loss": 2.2459,
      "step": 4665
    },
    {
      "epoch": 0.9332,
      "grad_norm": 0.04208115115761757,
      "learning_rate": 2.7224121790603517e-06,
      "loss": 2.2465,
      "step": 4666
    },
    {
      "epoch": 0.9334,
      "grad_norm": 0.041899874806404114,
      "learning_rate": 2.7062568305295965e-06,
      "loss": 2.2458,
      "step": 4667
    },
    {
      "epoch": 0.9336,
      "grad_norm": 0.04176611825823784,
      "learning_rate": 2.690148901787337e-06,
      "loss": 2.2451,
      "step": 4668
    },
    {
      "epoch": 0.9338,
      "grad_norm": 0.04225778579711914,
      "learning_rate": 2.6740884006843825e-06,
      "loss": 2.2461,
      "step": 4669
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.041933443397283554,
      "learning_rate": 2.6580753350484046e-06,
      "loss": 2.2469,
      "step": 4670
    },
    {
      "epoch": 0.9342,
      "grad_norm": 0.042197734117507935,
      "learning_rate": 2.6421097126839712e-06,
      "loss": 2.2479,
      "step": 4671
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.04269314929842949,
      "learning_rate": 2.6261915413725578e-06,
      "loss": 2.2464,
      "step": 4672
    },
    {
      "epoch": 0.9346,
      "grad_norm": 0.04215075075626373,
      "learning_rate": 2.610320828872481e-06,
      "loss": 2.2458,
      "step": 4673
    },
    {
      "epoch": 0.9348,
      "grad_norm": 0.041819095611572266,
      "learning_rate": 2.59449758291892e-06,
      "loss": 2.2467,
      "step": 4674
    },
    {
      "epoch": 0.935,
      "grad_norm": 0.04178544506430626,
      "learning_rate": 2.5787218112239496e-06,
      "loss": 2.2472,
      "step": 4675
    },
    {
      "epoch": 0.9352,
      "grad_norm": 0.04200144112110138,
      "learning_rate": 2.5629935214764865e-06,
      "loss": 2.2474,
      "step": 4676
    },
    {
      "epoch": 0.9354,
      "grad_norm": 0.042338889092206955,
      "learning_rate": 2.5473127213422763e-06,
      "loss": 2.2465,
      "step": 4677
    },
    {
      "epoch": 0.9356,
      "grad_norm": 0.0425640307366848,
      "learning_rate": 2.531679418464006e-06,
      "loss": 2.2471,
      "step": 4678
    },
    {
      "epoch": 0.9358,
      "grad_norm": 0.042013294994831085,
      "learning_rate": 2.516093620461124e-06,
      "loss": 2.2464,
      "step": 4679
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.04219992831349373,
      "learning_rate": 2.5005553349299547e-06,
      "loss": 2.248,
      "step": 4680
    },
    {
      "epoch": 0.9362,
      "grad_norm": 0.04252691566944122,
      "learning_rate": 2.4850645694436736e-06,
      "loss": 2.2472,
      "step": 4681
    },
    {
      "epoch": 0.9364,
      "grad_norm": 0.04198553413152695,
      "learning_rate": 2.4696213315523074e-06,
      "loss": 2.2464,
      "step": 4682
    },
    {
      "epoch": 0.9366,
      "grad_norm": 0.042256344109773636,
      "learning_rate": 2.4542256287826914e-06,
      "loss": 2.248,
      "step": 4683
    },
    {
      "epoch": 0.9368,
      "grad_norm": 0.04276254400610924,
      "learning_rate": 2.4388774686385007e-06,
      "loss": 2.2458,
      "step": 4684
    },
    {
      "epoch": 0.937,
      "grad_norm": 0.04218049347400665,
      "learning_rate": 2.4235768586002517e-06,
      "loss": 2.2454,
      "step": 4685
    },
    {
      "epoch": 0.9372,
      "grad_norm": 0.0423363633453846,
      "learning_rate": 2.4083238061252567e-06,
      "loss": 2.2469,
      "step": 4686
    },
    {
      "epoch": 0.9374,
      "grad_norm": 0.04195069521665573,
      "learning_rate": 2.3931183186477026e-06,
      "loss": 2.2478,
      "step": 4687
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.04219464585185051,
      "learning_rate": 2.3779604035785273e-06,
      "loss": 2.2477,
      "step": 4688
    },
    {
      "epoch": 0.9378,
      "grad_norm": 0.04184878244996071,
      "learning_rate": 2.3628500683055222e-06,
      "loss": 2.2454,
      "step": 4689
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.04187074676156044,
      "learning_rate": 2.3477873201932734e-06,
      "loss": 2.2432,
      "step": 4690
    },
    {
      "epoch": 0.9382,
      "grad_norm": 0.042079389095306396,
      "learning_rate": 2.332772166583208e-06,
      "loss": 2.2455,
      "step": 4691
    },
    {
      "epoch": 0.9384,
      "grad_norm": 0.04153715446591377,
      "learning_rate": 2.3178046147935175e-06,
      "loss": 2.2451,
      "step": 4692
    },
    {
      "epoch": 0.9386,
      "grad_norm": 0.042866580188274384,
      "learning_rate": 2.3028846721191876e-06,
      "loss": 2.2472,
      "step": 4693
    },
    {
      "epoch": 0.9388,
      "grad_norm": 0.04183875769376755,
      "learning_rate": 2.288012345832047e-06,
      "loss": 2.245,
      "step": 4694
    },
    {
      "epoch": 0.939,
      "grad_norm": 0.04219052568078041,
      "learning_rate": 2.273187643180652e-06,
      "loss": 2.248,
      "step": 4695
    },
    {
      "epoch": 0.9392,
      "grad_norm": 0.04302806034684181,
      "learning_rate": 2.2584105713904125e-06,
      "loss": 2.2487,
      "step": 4696
    },
    {
      "epoch": 0.9394,
      "grad_norm": 0.04217345640063286,
      "learning_rate": 2.2436811376634893e-06,
      "loss": 2.2465,
      "step": 4697
    },
    {
      "epoch": 0.9396,
      "grad_norm": 0.041779909282922745,
      "learning_rate": 2.2289993491788064e-06,
      "loss": 2.2458,
      "step": 4698
    },
    {
      "epoch": 0.9398,
      "grad_norm": 0.042276255786418915,
      "learning_rate": 2.2143652130921176e-06,
      "loss": 2.2467,
      "step": 4699
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.04200126603245735,
      "learning_rate": 2.1997787365358958e-06,
      "loss": 2.2471,
      "step": 4700
    },
    {
      "epoch": 0.9402,
      "grad_norm": 0.04195806756615639,
      "learning_rate": 2.1852399266194314e-06,
      "loss": 2.2461,
      "step": 4701
    },
    {
      "epoch": 0.9404,
      "grad_norm": 0.042331669479608536,
      "learning_rate": 2.1707487904287672e-06,
      "loss": 2.2473,
      "step": 4702
    },
    {
      "epoch": 0.9406,
      "grad_norm": 0.04185601696372032,
      "learning_rate": 2.156305335026698e-06,
      "loss": 2.2461,
      "step": 4703
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.04195520654320717,
      "learning_rate": 2.141909567452793e-06,
      "loss": 2.2465,
      "step": 4704
    },
    {
      "epoch": 0.941,
      "grad_norm": 0.0420578233897686,
      "learning_rate": 2.1275614947233624e-06,
      "loss": 2.2468,
      "step": 4705
    },
    {
      "epoch": 0.9412,
      "grad_norm": 0.04214233160018921,
      "learning_rate": 2.1132611238315003e-06,
      "loss": 2.2478,
      "step": 4706
    },
    {
      "epoch": 0.9414,
      "grad_norm": 0.042973898351192474,
      "learning_rate": 2.0990084617470206e-06,
      "loss": 2.2472,
      "step": 4707
    },
    {
      "epoch": 0.9416,
      "grad_norm": 0.042841289192438126,
      "learning_rate": 2.084803515416511e-06,
      "loss": 2.2488,
      "step": 4708
    },
    {
      "epoch": 0.9418,
      "grad_norm": 0.04184092581272125,
      "learning_rate": 2.0706462917632673e-06,
      "loss": 2.2472,
      "step": 4709
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.04203999787569046,
      "learning_rate": 2.0565367976873584e-06,
      "loss": 2.2472,
      "step": 4710
    },
    {
      "epoch": 0.9422,
      "grad_norm": 0.041852716356515884,
      "learning_rate": 2.0424750400655947e-06,
      "loss": 2.2468,
      "step": 4711
    },
    {
      "epoch": 0.9424,
      "grad_norm": 0.04229775816202164,
      "learning_rate": 2.0284610257514937e-06,
      "loss": 2.2478,
      "step": 4712
    },
    {
      "epoch": 0.9426,
      "grad_norm": 0.041917651891708374,
      "learning_rate": 2.014494761575314e-06,
      "loss": 2.2458,
      "step": 4713
    },
    {
      "epoch": 0.9428,
      "grad_norm": 0.04194018989801407,
      "learning_rate": 2.0005762543440445e-06,
      "loss": 2.2467,
      "step": 4714
    },
    {
      "epoch": 0.943,
      "grad_norm": 0.04186386242508888,
      "learning_rate": 1.986705510841402e-06,
      "loss": 2.2457,
      "step": 4715
    },
    {
      "epoch": 0.9432,
      "grad_norm": 0.04219166189432144,
      "learning_rate": 1.9728825378278246e-06,
      "loss": 2.2438,
      "step": 4716
    },
    {
      "epoch": 0.9434,
      "grad_norm": 0.0419410839676857,
      "learning_rate": 1.9591073420404337e-06,
      "loss": 2.2474,
      "step": 4717
    },
    {
      "epoch": 0.9436,
      "grad_norm": 0.04265453293919563,
      "learning_rate": 1.945379930193125e-06,
      "loss": 2.2472,
      "step": 4718
    },
    {
      "epoch": 0.9438,
      "grad_norm": 0.041999947279691696,
      "learning_rate": 1.931700308976436e-06,
      "loss": 2.2463,
      "step": 4719
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.04199992120265961,
      "learning_rate": 1.918068485057689e-06,
      "loss": 2.244,
      "step": 4720
    },
    {
      "epoch": 0.9442,
      "grad_norm": 0.0416669026017189,
      "learning_rate": 1.904484465080847e-06,
      "loss": 2.2431,
      "step": 4721
    },
    {
      "epoch": 0.9444,
      "grad_norm": 0.04216381162405014,
      "learning_rate": 1.8909482556666024e-06,
      "loss": 2.2476,
      "step": 4722
    },
    {
      "epoch": 0.9446,
      "grad_norm": 0.04207403212785721,
      "learning_rate": 1.8774598634123232e-06,
      "loss": 2.2469,
      "step": 4723
    },
    {
      "epoch": 0.9448,
      "grad_norm": 0.04221693426370621,
      "learning_rate": 1.8640192948921054e-06,
      "loss": 2.2456,
      "step": 4724
    },
    {
      "epoch": 0.945,
      "grad_norm": 0.04189835488796234,
      "learning_rate": 1.8506265566567094e-06,
      "loss": 2.2459,
      "step": 4725
    },
    {
      "epoch": 0.9452,
      "grad_norm": 0.04270515590906143,
      "learning_rate": 1.8372816552336026e-06,
      "loss": 2.2472,
      "step": 4726
    },
    {
      "epoch": 0.9454,
      "grad_norm": 0.042527515441179276,
      "learning_rate": 1.8239845971269266e-06,
      "loss": 2.2483,
      "step": 4727
    },
    {
      "epoch": 0.9456,
      "grad_norm": 0.04274747520685196,
      "learning_rate": 1.8107353888175083e-06,
      "loss": 2.2474,
      "step": 4728
    },
    {
      "epoch": 0.9458,
      "grad_norm": 0.04273593798279762,
      "learning_rate": 1.7975340367628268e-06,
      "loss": 2.2476,
      "step": 4729
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.042976297438144684,
      "learning_rate": 1.7843805473970798e-06,
      "loss": 2.2483,
      "step": 4730
    },
    {
      "epoch": 0.9462,
      "grad_norm": 0.04295085370540619,
      "learning_rate": 1.771274927131139e-06,
      "loss": 2.2485,
      "step": 4731
    },
    {
      "epoch": 0.9464,
      "grad_norm": 0.042283158749341965,
      "learning_rate": 1.7582171823524951e-06,
      "loss": 2.247,
      "step": 4732
    },
    {
      "epoch": 0.9466,
      "grad_norm": 0.04179101437330246,
      "learning_rate": 1.7452073194253238e-06,
      "loss": 2.2468,
      "step": 4733
    },
    {
      "epoch": 0.9468,
      "grad_norm": 0.04205218330025673,
      "learning_rate": 1.7322453446905085e-06,
      "loss": 2.2483,
      "step": 4734
    },
    {
      "epoch": 0.947,
      "grad_norm": 0.04232469201087952,
      "learning_rate": 1.719331264465529e-06,
      "loss": 2.2475,
      "step": 4735
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.04187312722206116,
      "learning_rate": 1.706465085044584e-06,
      "loss": 2.2454,
      "step": 4736
    },
    {
      "epoch": 0.9474,
      "grad_norm": 0.04196391627192497,
      "learning_rate": 1.6936468126984572e-06,
      "loss": 2.246,
      "step": 4737
    },
    {
      "epoch": 0.9476,
      "grad_norm": 0.0426512248814106,
      "learning_rate": 1.680876453674629e-06,
      "loss": 2.2474,
      "step": 4738
    },
    {
      "epoch": 0.9478,
      "grad_norm": 0.043836966156959534,
      "learning_rate": 1.6681540141972429e-06,
      "loss": 2.2482,
      "step": 4739
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.04253261163830757,
      "learning_rate": 1.6554795004670388e-06,
      "loss": 2.2478,
      "step": 4740
    },
    {
      "epoch": 0.9482,
      "grad_norm": 0.042087193578481674,
      "learning_rate": 1.6428529186614195e-06,
      "loss": 2.2459,
      "step": 4741
    },
    {
      "epoch": 0.9484,
      "grad_norm": 0.04205450788140297,
      "learning_rate": 1.6302742749344291e-06,
      "loss": 2.2464,
      "step": 4742
    },
    {
      "epoch": 0.9486,
      "grad_norm": 0.042523209005594254,
      "learning_rate": 1.6177435754167415e-06,
      "loss": 2.2485,
      "step": 4743
    },
    {
      "epoch": 0.9488,
      "grad_norm": 0.04214242473244667,
      "learning_rate": 1.605260826215682e-06,
      "loss": 2.2489,
      "step": 4744
    },
    {
      "epoch": 0.949,
      "grad_norm": 0.04294848442077637,
      "learning_rate": 1.5928260334151845e-06,
      "loss": 2.2492,
      "step": 4745
    },
    {
      "epoch": 0.9492,
      "grad_norm": 0.042182423174381256,
      "learning_rate": 1.580439203075812e-06,
      "loss": 2.246,
      "step": 4746
    },
    {
      "epoch": 0.9494,
      "grad_norm": 0.0423334464430809,
      "learning_rate": 1.5681003412347572e-06,
      "loss": 2.2459,
      "step": 4747
    },
    {
      "epoch": 0.9496,
      "grad_norm": 0.041907209903001785,
      "learning_rate": 1.555809453905821e-06,
      "loss": 2.2469,
      "step": 4748
    },
    {
      "epoch": 0.9498,
      "grad_norm": 0.04273667186498642,
      "learning_rate": 1.543566547079467e-06,
      "loss": 2.2495,
      "step": 4749
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.042989760637283325,
      "learning_rate": 1.5313716267226996e-06,
      "loss": 2.2463,
      "step": 4750
    },
    {
      "epoch": 0.9502,
      "grad_norm": 0.042957838624715805,
      "learning_rate": 1.5192246987791981e-06,
      "loss": 2.2482,
      "step": 4751
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.04173266515135765,
      "learning_rate": 1.5071257691692153e-06,
      "loss": 2.2464,
      "step": 4752
    },
    {
      "epoch": 0.9506,
      "grad_norm": 0.041637081652879715,
      "learning_rate": 1.4950748437896234e-06,
      "loss": 2.2466,
      "step": 4753
    },
    {
      "epoch": 0.9508,
      "grad_norm": 0.04191829264163971,
      "learning_rate": 1.4830719285139127e-06,
      "loss": 2.2453,
      "step": 4754
    },
    {
      "epoch": 0.951,
      "grad_norm": 0.04174181446433067,
      "learning_rate": 1.4711170291921484e-06,
      "loss": 2.2464,
      "step": 4755
    },
    {
      "epoch": 0.9512,
      "grad_norm": 0.041958168148994446,
      "learning_rate": 1.4592101516509914e-06,
      "loss": 2.2449,
      "step": 4756
    },
    {
      "epoch": 0.9514,
      "grad_norm": 0.04320523515343666,
      "learning_rate": 1.4473513016937223e-06,
      "loss": 2.2481,
      "step": 4757
    },
    {
      "epoch": 0.9516,
      "grad_norm": 0.04201063886284828,
      "learning_rate": 1.4355404851001952e-06,
      "loss": 2.2456,
      "step": 4758
    },
    {
      "epoch": 0.9518,
      "grad_norm": 0.04208441823720932,
      "learning_rate": 1.423777707626872e-06,
      "loss": 2.2466,
      "step": 4759
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.04209692403674126,
      "learning_rate": 1.412062975006767e-06,
      "loss": 2.2465,
      "step": 4760
    },
    {
      "epoch": 0.9522,
      "grad_norm": 0.04256878048181534,
      "learning_rate": 1.400396292949513e-06,
      "loss": 2.2463,
      "step": 4761
    },
    {
      "epoch": 0.9524,
      "grad_norm": 0.042691297829151154,
      "learning_rate": 1.3887776671412943e-06,
      "loss": 2.2467,
      "step": 4762
    },
    {
      "epoch": 0.9526,
      "grad_norm": 0.04227031022310257,
      "learning_rate": 1.377207103244904e-06,
      "loss": 2.2463,
      "step": 4763
    },
    {
      "epoch": 0.9528,
      "grad_norm": 0.04241124168038368,
      "learning_rate": 1.3656846068996976e-06,
      "loss": 2.2466,
      "step": 4764
    },
    {
      "epoch": 0.953,
      "grad_norm": 0.04263845458626747,
      "learning_rate": 1.3542101837215826e-06,
      "loss": 2.2494,
      "step": 4765
    },
    {
      "epoch": 0.9532,
      "grad_norm": 0.042334817349910736,
      "learning_rate": 1.3427838393030633e-06,
      "loss": 2.2459,
      "step": 4766
    },
    {
      "epoch": 0.9534,
      "grad_norm": 0.04222381114959717,
      "learning_rate": 1.3314055792131964e-06,
      "loss": 2.2461,
      "step": 4767
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.04194162040948868,
      "learning_rate": 1.320075408997612e-06,
      "loss": 2.2471,
      "step": 4768
    },
    {
      "epoch": 0.9538,
      "grad_norm": 0.042759090662002563,
      "learning_rate": 1.308793334178493e-06,
      "loss": 2.2484,
      "step": 4769
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.04395074024796486,
      "learning_rate": 1.2975593602545965e-06,
      "loss": 2.2491,
      "step": 4770
    },
    {
      "epoch": 0.9542,
      "grad_norm": 0.04197677969932556,
      "learning_rate": 1.2863734927012095e-06,
      "loss": 2.2468,
      "step": 4771
    },
    {
      "epoch": 0.9544,
      "grad_norm": 0.04197382554411888,
      "learning_rate": 1.275235736970193e-06,
      "loss": 2.2484,
      "step": 4772
    },
    {
      "epoch": 0.9546,
      "grad_norm": 0.042031411081552505,
      "learning_rate": 1.26414609848996e-06,
      "loss": 2.247,
      "step": 4773
    },
    {
      "epoch": 0.9548,
      "grad_norm": 0.04249570518732071,
      "learning_rate": 1.2531045826654653e-06,
      "loss": 2.2474,
      "step": 4774
    },
    {
      "epoch": 0.955,
      "grad_norm": 0.04244305193424225,
      "learning_rate": 1.2421111948782149e-06,
      "loss": 2.2492,
      "step": 4775
    },
    {
      "epoch": 0.9552,
      "grad_norm": 0.04225420206785202,
      "learning_rate": 1.231165940486234e-06,
      "loss": 2.2474,
      "step": 4776
    },
    {
      "epoch": 0.9554,
      "grad_norm": 0.04264970123767853,
      "learning_rate": 1.2202688248241112e-06,
      "loss": 2.2463,
      "step": 4777
    },
    {
      "epoch": 0.9556,
      "grad_norm": 0.042469896376132965,
      "learning_rate": 1.2094198532029755e-06,
      "loss": 2.2479,
      "step": 4778
    },
    {
      "epoch": 0.9558,
      "grad_norm": 0.04206443578004837,
      "learning_rate": 1.1986190309104861e-06,
      "loss": 2.248,
      "step": 4779
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.0420866534113884,
      "learning_rate": 1.1878663632108322e-06,
      "loss": 2.2475,
      "step": 4780
    },
    {
      "epoch": 0.9562,
      "grad_norm": 0.04182295873761177,
      "learning_rate": 1.1771618553447216e-06,
      "loss": 2.2462,
      "step": 4781
    },
    {
      "epoch": 0.9564,
      "grad_norm": 0.042519908398389816,
      "learning_rate": 1.1665055125294033e-06,
      "loss": 2.2459,
      "step": 4782
    },
    {
      "epoch": 0.9566,
      "grad_norm": 0.042504988610744476,
      "learning_rate": 1.155897339958667e-06,
      "loss": 2.2463,
      "step": 4783
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.04249581694602966,
      "learning_rate": 1.1453373428027992e-06,
      "loss": 2.2474,
      "step": 4784
    },
    {
      "epoch": 0.957,
      "grad_norm": 0.04200626537203789,
      "learning_rate": 1.134825526208605e-06,
      "loss": 2.2443,
      "step": 4785
    },
    {
      "epoch": 0.9572,
      "grad_norm": 0.04209687560796738,
      "learning_rate": 1.1243618952994195e-06,
      "loss": 2.2449,
      "step": 4786
    },
    {
      "epoch": 0.9574,
      "grad_norm": 0.0417964942753315,
      "learning_rate": 1.1139464551750856e-06,
      "loss": 2.2446,
      "step": 4787
    },
    {
      "epoch": 0.9576,
      "grad_norm": 0.04190913215279579,
      "learning_rate": 1.103579210911976e-06,
      "loss": 2.2463,
      "step": 4788
    },
    {
      "epoch": 0.9578,
      "grad_norm": 0.042010046541690826,
      "learning_rate": 1.0932601675629595e-06,
      "loss": 2.2467,
      "step": 4789
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.04205063357949257,
      "learning_rate": 1.0829893301573913e-06,
      "loss": 2.2461,
      "step": 4790
    },
    {
      "epoch": 0.9582,
      "grad_norm": 0.04210396483540535,
      "learning_rate": 1.0727667037011668e-06,
      "loss": 2.2465,
      "step": 4791
    },
    {
      "epoch": 0.9584,
      "grad_norm": 0.04253307357430458,
      "learning_rate": 1.0625922931766785e-06,
      "loss": 2.2479,
      "step": 4792
    },
    {
      "epoch": 0.9586,
      "grad_norm": 0.042413078248500824,
      "learning_rate": 1.052466103542793e-06,
      "loss": 2.246,
      "step": 4793
    },
    {
      "epoch": 0.9588,
      "grad_norm": 0.041886311024427414,
      "learning_rate": 1.0423881397349068e-06,
      "loss": 2.2448,
      "step": 4794
    },
    {
      "epoch": 0.959,
      "grad_norm": 0.04200904816389084,
      "learning_rate": 1.0323584066648795e-06,
      "loss": 2.2459,
      "step": 4795
    },
    {
      "epoch": 0.9592,
      "grad_norm": 0.042207058519124985,
      "learning_rate": 1.0223769092211012e-06,
      "loss": 2.2468,
      "step": 4796
    },
    {
      "epoch": 0.9594,
      "grad_norm": 0.04229293763637543,
      "learning_rate": 1.0124436522684243e-06,
      "loss": 2.2466,
      "step": 4797
    },
    {
      "epoch": 0.9596,
      "grad_norm": 0.0419418029487133,
      "learning_rate": 1.002558640648199e-06,
      "loss": 2.2456,
      "step": 4798
    },
    {
      "epoch": 0.9598,
      "grad_norm": 0.042017728090286255,
      "learning_rate": 9.927218791782599e-07,
      "loss": 2.2464,
      "step": 4799
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.04238438978791237,
      "learning_rate": 9.829333726529056e-07,
      "loss": 2.2447,
      "step": 4800
    },
    {
      "epoch": 0.9602,
      "grad_norm": 0.04255027696490288,
      "learning_rate": 9.731931258429638e-07,
      "loss": 2.2461,
      "step": 4801
    },
    {
      "epoch": 0.9604,
      "grad_norm": 0.041962720453739166,
      "learning_rate": 9.635011434957152e-07,
      "loss": 2.2475,
      "step": 4802
    },
    {
      "epoch": 0.9606,
      "grad_norm": 0.04187700152397156,
      "learning_rate": 9.538574303348813e-07,
      "loss": 2.2453,
      "step": 4803
    },
    {
      "epoch": 0.9608,
      "grad_norm": 0.042064692825078964,
      "learning_rate": 9.442619910607131e-07,
      "loss": 2.2482,
      "step": 4804
    },
    {
      "epoch": 0.961,
      "grad_norm": 0.04204823076725006,
      "learning_rate": 9.347148303499142e-07,
      "loss": 2.2468,
      "step": 4805
    },
    {
      "epoch": 0.9612,
      "grad_norm": 0.0421004593372345,
      "learning_rate": 9.252159528556403e-07,
      "loss": 2.2466,
      "step": 4806
    },
    {
      "epoch": 0.9614,
      "grad_norm": 0.042116519063711166,
      "learning_rate": 9.157653632075436e-07,
      "loss": 2.248,
      "step": 4807
    },
    {
      "epoch": 0.9616,
      "grad_norm": 0.042330533266067505,
      "learning_rate": 9.063630660117173e-07,
      "loss": 2.2464,
      "step": 4808
    },
    {
      "epoch": 0.9618,
      "grad_norm": 0.04187636449933052,
      "learning_rate": 8.970090658507291e-07,
      "loss": 2.2439,
      "step": 4809
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.04784221202135086,
      "learning_rate": 8.877033672835988e-07,
      "loss": 2.2471,
      "step": 4810
    },
    {
      "epoch": 0.9622,
      "grad_norm": 0.04589470103383064,
      "learning_rate": 8.784459748458318e-07,
      "loss": 2.2456,
      "step": 4811
    },
    {
      "epoch": 0.9624,
      "grad_norm": 0.047020263969898224,
      "learning_rate": 8.692368930493521e-07,
      "loss": 2.247,
      "step": 4812
    },
    {
      "epoch": 0.9626,
      "grad_norm": 0.04291694983839989,
      "learning_rate": 8.600761263825474e-07,
      "loss": 2.2472,
      "step": 4813
    },
    {
      "epoch": 0.9628,
      "grad_norm": 0.04418307542800903,
      "learning_rate": 8.509636793102682e-07,
      "loss": 2.2473,
      "step": 4814
    },
    {
      "epoch": 0.963,
      "grad_norm": 0.04323112964630127,
      "learning_rate": 8.418995562738285e-07,
      "loss": 2.2466,
      "step": 4815
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.04612420126795769,
      "learning_rate": 8.328837616909613e-07,
      "loss": 2.2486,
      "step": 4816
    },
    {
      "epoch": 0.9634,
      "grad_norm": 0.04358498752117157,
      "learning_rate": 8.239162999558403e-07,
      "loss": 2.2488,
      "step": 4817
    },
    {
      "epoch": 0.9636,
      "grad_norm": 0.04747925326228142,
      "learning_rate": 8.14997175439125e-07,
      "loss": 2.2475,
      "step": 4818
    },
    {
      "epoch": 0.9638,
      "grad_norm": 0.050820592790842056,
      "learning_rate": 8.061263924878604e-07,
      "loss": 2.2482,
      "step": 4819
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.05080116167664528,
      "learning_rate": 7.973039554255768e-07,
      "loss": 2.249,
      "step": 4820
    },
    {
      "epoch": 0.9642,
      "grad_norm": 0.04889367148280144,
      "learning_rate": 7.885298685522235e-07,
      "loss": 2.2488,
      "step": 4821
    },
    {
      "epoch": 0.9644,
      "grad_norm": 0.04437240585684776,
      "learning_rate": 7.798041361441688e-07,
      "loss": 2.2461,
      "step": 4822
    },
    {
      "epoch": 0.9646,
      "grad_norm": 0.047719504684209824,
      "learning_rate": 7.711267624542329e-07,
      "loss": 2.2466,
      "step": 4823
    },
    {
      "epoch": 0.9648,
      "grad_norm": 0.0487961508333683,
      "learning_rate": 7.624977517116772e-07,
      "loss": 2.2508,
      "step": 4824
    },
    {
      "epoch": 0.965,
      "grad_norm": 0.049368131905794144,
      "learning_rate": 7.539171081221596e-07,
      "loss": 2.2462,
      "step": 4825
    },
    {
      "epoch": 0.9652,
      "grad_norm": 0.04779953137040138,
      "learning_rate": 7.453848358678017e-07,
      "loss": 2.2455,
      "step": 4826
    },
    {
      "epoch": 0.9654,
      "grad_norm": 0.04666581749916077,
      "learning_rate": 7.369009391070992e-07,
      "loss": 2.2431,
      "step": 4827
    },
    {
      "epoch": 0.9656,
      "grad_norm": 0.04547097533941269,
      "learning_rate": 7.284654219750331e-07,
      "loss": 2.248,
      "step": 4828
    },
    {
      "epoch": 0.9658,
      "grad_norm": 0.042896818369627,
      "learning_rate": 7.200782885829482e-07,
      "loss": 2.2429,
      "step": 4829
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.04215206578373909,
      "learning_rate": 7.117395430186414e-07,
      "loss": 2.2452,
      "step": 4830
    },
    {
      "epoch": 0.9662,
      "grad_norm": 0.042052220553159714,
      "learning_rate": 7.034491893463058e-07,
      "loss": 2.2459,
      "step": 4831
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.04223154857754707,
      "learning_rate": 6.952072316065761e-07,
      "loss": 2.2465,
      "step": 4832
    },
    {
      "epoch": 0.9666,
      "grad_norm": 0.04330229014158249,
      "learning_rate": 6.870136738164612e-07,
      "loss": 2.2465,
      "step": 4833
    },
    {
      "epoch": 0.9668,
      "grad_norm": 0.04382220655679703,
      "learning_rate": 6.788685199694222e-07,
      "loss": 2.2469,
      "step": 4834
    },
    {
      "epoch": 0.967,
      "grad_norm": 0.04382426664233208,
      "learning_rate": 6.707717740353059e-07,
      "loss": 2.2455,
      "step": 4835
    },
    {
      "epoch": 0.9672,
      "grad_norm": 0.045259300619363785,
      "learning_rate": 6.627234399603555e-07,
      "loss": 2.246,
      "step": 4836
    },
    {
      "epoch": 0.9674,
      "grad_norm": 0.04443197324872017,
      "learning_rate": 6.547235216672443e-07,
      "loss": 2.2455,
      "step": 4837
    },
    {
      "epoch": 0.9676,
      "grad_norm": 0.0438527911901474,
      "learning_rate": 6.4677202305502e-07,
      "loss": 2.245,
      "step": 4838
    },
    {
      "epoch": 0.9678,
      "grad_norm": 0.04428362846374512,
      "learning_rate": 6.388689479991605e-07,
      "loss": 2.2479,
      "step": 4839
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.043319109827280045,
      "learning_rate": 6.310143003515179e-07,
      "loss": 2.2473,
      "step": 4840
    },
    {
      "epoch": 0.9682,
      "grad_norm": 0.04208577796816826,
      "learning_rate": 6.232080839403631e-07,
      "loss": 2.2461,
      "step": 4841
    },
    {
      "epoch": 0.9684,
      "grad_norm": 0.04179921746253967,
      "learning_rate": 6.154503025703417e-07,
      "loss": 2.2459,
      "step": 4842
    },
    {
      "epoch": 0.9686,
      "grad_norm": 0.041661910712718964,
      "learning_rate": 6.07740960022507e-07,
      "loss": 2.244,
      "step": 4843
    },
    {
      "epoch": 0.9688,
      "grad_norm": 0.04207811877131462,
      "learning_rate": 6.000800600542977e-07,
      "loss": 2.2468,
      "step": 4844
    },
    {
      "epoch": 0.969,
      "grad_norm": 0.04209699481725693,
      "learning_rate": 5.924676063995382e-07,
      "loss": 2.2469,
      "step": 4845
    },
    {
      "epoch": 0.9692,
      "grad_norm": 0.04203713685274124,
      "learning_rate": 5.849036027684606e-07,
      "loss": 2.247,
      "step": 4846
    },
    {
      "epoch": 0.9694,
      "grad_norm": 0.0427197702229023,
      "learning_rate": 5.773880528476494e-07,
      "loss": 2.2461,
      "step": 4847
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.0425562784075737,
      "learning_rate": 5.699209603001076e-07,
      "loss": 2.2446,
      "step": 4848
    },
    {
      "epoch": 0.9698,
      "grad_norm": 0.042173415422439575,
      "learning_rate": 5.62502328765202e-07,
      "loss": 2.246,
      "step": 4849
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.04235592484474182,
      "learning_rate": 5.551321618586736e-07,
      "loss": 2.2461,
      "step": 4850
    },
    {
      "epoch": 0.9702,
      "grad_norm": 0.04204482585191727,
      "learning_rate": 5.478104631726711e-07,
      "loss": 2.246,
      "step": 4851
    },
    {
      "epoch": 0.9704,
      "grad_norm": 0.04162456467747688,
      "learning_rate": 5.405372362756734e-07,
      "loss": 2.2463,
      "step": 4852
    },
    {
      "epoch": 0.9706,
      "grad_norm": 0.04180959239602089,
      "learning_rate": 5.333124847125892e-07,
      "loss": 2.2466,
      "step": 4853
    },
    {
      "epoch": 0.9708,
      "grad_norm": 0.04191577434539795,
      "learning_rate": 5.261362120046686e-07,
      "loss": 2.2443,
      "step": 4854
    },
    {
      "epoch": 0.971,
      "grad_norm": 0.04195530340075493,
      "learning_rate": 5.190084216495361e-07,
      "loss": 2.2471,
      "step": 4855
    },
    {
      "epoch": 0.9712,
      "grad_norm": 0.04230710491538048,
      "learning_rate": 5.119291171211793e-07,
      "loss": 2.2468,
      "step": 4856
    },
    {
      "epoch": 0.9714,
      "grad_norm": 0.044103801250457764,
      "learning_rate": 5.048983018699827e-07,
      "loss": 2.2491,
      "step": 4857
    },
    {
      "epoch": 0.9716,
      "grad_norm": 0.04408625140786171,
      "learning_rate": 4.979159793226718e-07,
      "loss": 2.2487,
      "step": 4858
    },
    {
      "epoch": 0.9718,
      "grad_norm": 0.042919356375932693,
      "learning_rate": 4.909821528823577e-07,
      "loss": 2.2472,
      "step": 4859
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.041963789612054825,
      "learning_rate": 4.840968259284817e-07,
      "loss": 2.2456,
      "step": 4860
    },
    {
      "epoch": 0.9722,
      "grad_norm": 0.04208457097411156,
      "learning_rate": 4.772600018168816e-07,
      "loss": 2.2463,
      "step": 4861
    },
    {
      "epoch": 0.9724,
      "grad_norm": 0.042056597769260406,
      "learning_rate": 4.704716838797363e-07,
      "loss": 2.2461,
      "step": 4862
    },
    {
      "epoch": 0.9726,
      "grad_norm": 0.04207644611597061,
      "learning_rate": 4.6373187542561035e-07,
      "loss": 2.2467,
      "step": 4863
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.04184737801551819,
      "learning_rate": 4.570405797393762e-07,
      "loss": 2.2452,
      "step": 4864
    },
    {
      "epoch": 0.973,
      "grad_norm": 0.04197438061237335,
      "learning_rate": 4.503978000823028e-07,
      "loss": 2.2456,
      "step": 4865
    },
    {
      "epoch": 0.9732,
      "grad_norm": 0.042449094355106354,
      "learning_rate": 4.438035396920004e-07,
      "loss": 2.2481,
      "step": 4866
    },
    {
      "epoch": 0.9734,
      "grad_norm": 0.04210434854030609,
      "learning_rate": 4.372578017824314e-07,
      "loss": 2.246,
      "step": 4867
    },
    {
      "epoch": 0.9736,
      "grad_norm": 0.04214822128415108,
      "learning_rate": 4.307605895439104e-07,
      "loss": 2.2456,
      "step": 4868
    },
    {
      "epoch": 0.9738,
      "grad_norm": 0.04222250357270241,
      "learning_rate": 4.2431190614309335e-07,
      "loss": 2.2452,
      "step": 4869
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.042234282940626144,
      "learning_rate": 4.179117547229883e-07,
      "loss": 2.2463,
      "step": 4870
    },
    {
      "epoch": 0.9742,
      "grad_norm": 0.041887130588293076,
      "learning_rate": 4.115601384029666e-07,
      "loss": 2.2475,
      "step": 4871
    },
    {
      "epoch": 0.9744,
      "grad_norm": 0.042101189494132996,
      "learning_rate": 4.0525706027870756e-07,
      "loss": 2.2467,
      "step": 4872
    },
    {
      "epoch": 0.9746,
      "grad_norm": 0.04210363328456879,
      "learning_rate": 3.9900252342228717e-07,
      "loss": 2.2472,
      "step": 4873
    },
    {
      "epoch": 0.9748,
      "grad_norm": 0.0419105589389801,
      "learning_rate": 3.9279653088205584e-07,
      "loss": 2.2453,
      "step": 4874
    },
    {
      "epoch": 0.975,
      "grad_norm": 0.042503345757722855,
      "learning_rate": 3.866390856827495e-07,
      "loss": 2.2482,
      "step": 4875
    },
    {
      "epoch": 0.9752,
      "grad_norm": 0.04240279644727707,
      "learning_rate": 3.805301908254455e-07,
      "loss": 2.2479,
      "step": 4876
    },
    {
      "epoch": 0.9754,
      "grad_norm": 0.04226323589682579,
      "learning_rate": 3.744698492875398e-07,
      "loss": 2.2465,
      "step": 4877
    },
    {
      "epoch": 0.9756,
      "grad_norm": 0.04206984117627144,
      "learning_rate": 3.6845806402275863e-07,
      "loss": 2.2467,
      "step": 4878
    },
    {
      "epoch": 0.9758,
      "grad_norm": 0.042128633707761765,
      "learning_rate": 3.6249483796116924e-07,
      "loss": 2.247,
      "step": 4879
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.0421300046145916,
      "learning_rate": 3.565801740092023e-07,
      "loss": 2.2475,
      "step": 4880
    },
    {
      "epoch": 0.9762,
      "grad_norm": 0.0416780523955822,
      "learning_rate": 3.50714075049563e-07,
      "loss": 2.2452,
      "step": 4881
    },
    {
      "epoch": 0.9764,
      "grad_norm": 0.0418848842382431,
      "learning_rate": 3.4489654394134205e-07,
      "loss": 2.2445,
      "step": 4882
    },
    {
      "epoch": 0.9766,
      "grad_norm": 0.041983406990766525,
      "learning_rate": 3.3912758351991593e-07,
      "loss": 2.2466,
      "step": 4883
    },
    {
      "epoch": 0.9768,
      "grad_norm": 0.04248931631445885,
      "learning_rate": 3.3340719659701313e-07,
      "loss": 2.2476,
      "step": 4884
    },
    {
      "epoch": 0.977,
      "grad_norm": 0.0424676351249218,
      "learning_rate": 3.277353859606813e-07,
      "loss": 2.2491,
      "step": 4885
    },
    {
      "epoch": 0.9772,
      "grad_norm": 0.04249142110347748,
      "learning_rate": 3.2211215437528694e-07,
      "loss": 2.2491,
      "step": 4886
    },
    {
      "epoch": 0.9774,
      "grad_norm": 0.04207305982708931,
      "learning_rate": 3.1653750458152666e-07,
      "loss": 2.2487,
      "step": 4887
    },
    {
      "epoch": 0.9776,
      "grad_norm": 0.04196580499410629,
      "learning_rate": 3.1101143929641585e-07,
      "loss": 2.245,
      "step": 4888
    },
    {
      "epoch": 0.9778,
      "grad_norm": 0.04178526997566223,
      "learning_rate": 3.0553396121330013e-07,
      "loss": 2.2451,
      "step": 4889
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.041999634355306625,
      "learning_rate": 3.001050730018218e-07,
      "loss": 2.2482,
      "step": 4890
    },
    {
      "epoch": 0.9782,
      "grad_norm": 0.04255033656954765,
      "learning_rate": 2.947247773079753e-07,
      "loss": 2.2487,
      "step": 4891
    },
    {
      "epoch": 0.9784,
      "grad_norm": 0.041793983429670334,
      "learning_rate": 2.893930767540298e-07,
      "loss": 2.2448,
      "step": 4892
    },
    {
      "epoch": 0.9786,
      "grad_norm": 0.04245176538825035,
      "learning_rate": 2.841099739386066e-07,
      "loss": 2.2484,
      "step": 4893
    },
    {
      "epoch": 0.9788,
      "grad_norm": 0.04197253659367561,
      "learning_rate": 2.7887547143662373e-07,
      "loss": 2.2477,
      "step": 4894
    },
    {
      "epoch": 0.979,
      "grad_norm": 0.042157307267189026,
      "learning_rate": 2.73689571799296e-07,
      "loss": 2.2455,
      "step": 4895
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.04202910140156746,
      "learning_rate": 2.685522775541904e-07,
      "loss": 2.245,
      "step": 4896
    },
    {
      "epoch": 0.9794,
      "grad_norm": 0.042129673063755035,
      "learning_rate": 2.6346359120514863e-07,
      "loss": 2.2454,
      "step": 4897
    },
    {
      "epoch": 0.9796,
      "grad_norm": 0.04243868216872215,
      "learning_rate": 2.584235152323422e-07,
      "loss": 2.2488,
      "step": 4898
    },
    {
      "epoch": 0.9798,
      "grad_norm": 0.04231812804937363,
      "learning_rate": 2.534320520922506e-07,
      "loss": 2.2461,
      "step": 4899
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.04229908809065819,
      "learning_rate": 2.484892042176279e-07,
      "loss": 2.2473,
      "step": 4900
    },
    {
      "epoch": 0.9802,
      "grad_norm": 0.042307525873184204,
      "learning_rate": 2.4359497401758024e-07,
      "loss": 2.2483,
      "step": 4901
    },
    {
      "epoch": 0.9804,
      "grad_norm": 0.04215418919920921,
      "learning_rate": 2.387493638774774e-07,
      "loss": 2.2477,
      "step": 4902
    },
    {
      "epoch": 0.9806,
      "grad_norm": 0.04213734716176987,
      "learning_rate": 2.339523761590301e-07,
      "loss": 2.2487,
      "step": 4903
    },
    {
      "epoch": 0.9808,
      "grad_norm": 0.04251346364617348,
      "learning_rate": 2.2920401320022378e-07,
      "loss": 2.248,
      "step": 4904
    },
    {
      "epoch": 0.981,
      "grad_norm": 0.04204172268509865,
      "learning_rate": 2.2450427731534053e-07,
      "loss": 2.2467,
      "step": 4905
    },
    {
      "epoch": 0.9812,
      "grad_norm": 0.042006634175777435,
      "learning_rate": 2.1985317079500356e-07,
      "loss": 2.2482,
      "step": 4906
    },
    {
      "epoch": 0.9814,
      "grad_norm": 0.04286862164735794,
      "learning_rate": 2.1525069590607737e-07,
      "loss": 2.2488,
      "step": 4907
    },
    {
      "epoch": 0.9816,
      "grad_norm": 0.042507242411375046,
      "learning_rate": 2.106968548917676e-07,
      "loss": 2.2487,
      "step": 4908
    },
    {
      "epoch": 0.9818,
      "grad_norm": 0.04215283319354057,
      "learning_rate": 2.0619164997155438e-07,
      "loss": 2.247,
      "step": 4909
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.04254566505551338,
      "learning_rate": 2.017350833412146e-07,
      "loss": 2.2481,
      "step": 4910
    },
    {
      "epoch": 0.9822,
      "grad_norm": 0.04242175444960594,
      "learning_rate": 1.973271571728441e-07,
      "loss": 2.2484,
      "step": 4911
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.04197784140706062,
      "learning_rate": 1.9296787361480216e-07,
      "loss": 2.2478,
      "step": 4912
    },
    {
      "epoch": 0.9826,
      "grad_norm": 0.04223606362938881,
      "learning_rate": 1.8865723479173368e-07,
      "loss": 2.2465,
      "step": 4913
    },
    {
      "epoch": 0.9828,
      "grad_norm": 0.042069245129823685,
      "learning_rate": 1.8439524280462472e-07,
      "loss": 2.2493,
      "step": 4914
    },
    {
      "epoch": 0.983,
      "grad_norm": 0.0420907698571682,
      "learning_rate": 1.8018189973069143e-07,
      "loss": 2.2489,
      "step": 4915
    },
    {
      "epoch": 0.9832,
      "grad_norm": 0.04202284663915634,
      "learning_rate": 1.7601720762346897e-07,
      "loss": 2.2481,
      "step": 4916
    },
    {
      "epoch": 0.9834,
      "grad_norm": 0.04199298843741417,
      "learning_rate": 1.7190116851280026e-07,
      "loss": 2.2478,
      "step": 4917
    },
    {
      "epoch": 0.9836,
      "grad_norm": 0.042496249079704285,
      "learning_rate": 1.678337844047695e-07,
      "loss": 2.2477,
      "step": 4918
    },
    {
      "epoch": 0.9838,
      "grad_norm": 0.04262417182326317,
      "learning_rate": 1.6381505728176872e-07,
      "loss": 2.2481,
      "step": 4919
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.04194940999150276,
      "learning_rate": 1.598449891024978e-07,
      "loss": 2.2453,
      "step": 4920
    },
    {
      "epoch": 0.9842,
      "grad_norm": 0.041907161474227905,
      "learning_rate": 1.5592358180189782e-07,
      "loss": 2.2461,
      "step": 4921
    },
    {
      "epoch": 0.9844,
      "grad_norm": 0.04231199249625206,
      "learning_rate": 1.520508372912288e-07,
      "loss": 2.2469,
      "step": 4922
    },
    {
      "epoch": 0.9846,
      "grad_norm": 0.041992004960775375,
      "learning_rate": 1.4822675745801429e-07,
      "loss": 2.2479,
      "step": 4923
    },
    {
      "epoch": 0.9848,
      "grad_norm": 0.041790351271629333,
      "learning_rate": 1.4445134416607442e-07,
      "loss": 2.2447,
      "step": 4924
    },
    {
      "epoch": 0.985,
      "grad_norm": 0.04204342141747475,
      "learning_rate": 1.4072459925548177e-07,
      "loss": 2.2459,
      "step": 4925
    },
    {
      "epoch": 0.9852,
      "grad_norm": 0.04201235994696617,
      "learning_rate": 1.3704652454261668e-07,
      "loss": 2.2481,
      "step": 4926
    },
    {
      "epoch": 0.9854,
      "grad_norm": 0.04205546900629997,
      "learning_rate": 1.33417121820123e-07,
      "loss": 2.2456,
      "step": 4927
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.041941724717617035,
      "learning_rate": 1.2983639285693017e-07,
      "loss": 2.2461,
      "step": 4928
    },
    {
      "epoch": 0.9858,
      "grad_norm": 0.042218420654535294,
      "learning_rate": 1.2630433939825327e-07,
      "loss": 2.2474,
      "step": 4929
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.042197588831186295,
      "learning_rate": 1.2282096316554858e-07,
      "loss": 2.2459,
      "step": 4930
    },
    {
      "epoch": 0.9862,
      "grad_norm": 0.042172159999608994,
      "learning_rate": 1.193862658566025e-07,
      "loss": 2.2486,
      "step": 4931
    },
    {
      "epoch": 0.9864,
      "grad_norm": 0.04178261756896973,
      "learning_rate": 1.160002491454093e-07,
      "loss": 2.2465,
      "step": 4932
    },
    {
      "epoch": 0.9866,
      "grad_norm": 0.04198949784040451,
      "learning_rate": 1.1266291468229328e-07,
      "loss": 2.2471,
      "step": 4933
    },
    {
      "epoch": 0.9868,
      "grad_norm": 0.04186876490712166,
      "learning_rate": 1.0937426409384221e-07,
      "loss": 2.246,
      "step": 4934
    },
    {
      "epoch": 0.987,
      "grad_norm": 0.04203135147690773,
      "learning_rate": 1.0613429898287398e-07,
      "loss": 2.2454,
      "step": 4935
    },
    {
      "epoch": 0.9872,
      "grad_norm": 0.0421082004904747,
      "learning_rate": 1.0294302092853647e-07,
      "loss": 2.2466,
      "step": 4936
    },
    {
      "epoch": 0.9874,
      "grad_norm": 0.0422077551484108,
      "learning_rate": 9.980043148619667e-08,
      "loss": 2.248,
      "step": 4937
    },
    {
      "epoch": 0.9876,
      "grad_norm": 0.042047977447509766,
      "learning_rate": 9.670653218752934e-08,
      "loss": 2.2486,
      "step": 4938
    },
    {
      "epoch": 0.9878,
      "grad_norm": 0.042026594281196594,
      "learning_rate": 9.366132454046162e-08,
      "loss": 2.2478,
      "step": 4939
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.042867422103881836,
      "learning_rate": 9.066481002918403e-08,
      "loss": 2.2495,
      "step": 4940
    },
    {
      "epoch": 0.9882,
      "grad_norm": 0.04187886044383049,
      "learning_rate": 8.771699011416168e-08,
      "loss": 2.2451,
      "step": 4941
    },
    {
      "epoch": 0.9884,
      "grad_norm": 0.041927359998226166,
      "learning_rate": 8.481786623214528e-08,
      "loss": 2.2467,
      "step": 4942
    },
    {
      "epoch": 0.9886,
      "grad_norm": 0.042105838656425476,
      "learning_rate": 8.196743979610455e-08,
      "loss": 2.2481,
      "step": 4943
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.041942398995161057,
      "learning_rate": 7.91657121953171e-08,
      "loss": 2.2472,
      "step": 4944
    },
    {
      "epoch": 0.989,
      "grad_norm": 0.04212789610028267,
      "learning_rate": 7.641268479531283e-08,
      "loss": 2.2449,
      "step": 4945
    },
    {
      "epoch": 0.9892,
      "grad_norm": 0.04188912734389305,
      "learning_rate": 7.370835893788508e-08,
      "loss": 2.2462,
      "step": 4946
    },
    {
      "epoch": 0.9894,
      "grad_norm": 0.0420597568154335,
      "learning_rate": 7.105273594107953e-08,
      "loss": 2.246,
      "step": 4947
    },
    {
      "epoch": 0.9896,
      "grad_norm": 0.041969235986471176,
      "learning_rate": 6.844581709921638e-08,
      "loss": 2.2456,
      "step": 4948
    },
    {
      "epoch": 0.9898,
      "grad_norm": 0.04207475483417511,
      "learning_rate": 6.588760368287928e-08,
      "loss": 2.2478,
      "step": 4949
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.04185843840241432,
      "learning_rate": 6.337809693891527e-08,
      "loss": 2.2453,
      "step": 4950
    },
    {
      "epoch": 0.9902,
      "grad_norm": 0.042119644582271576,
      "learning_rate": 6.09172980904238e-08,
      "loss": 2.2469,
      "step": 4951
    },
    {
      "epoch": 0.9904,
      "grad_norm": 0.04207319766283035,
      "learning_rate": 5.850520833676765e-08,
      "loss": 2.2468,
      "step": 4952
    },
    {
      "epoch": 0.9906,
      "grad_norm": 0.04200761020183563,
      "learning_rate": 5.6141828853573106e-08,
      "loss": 2.2464,
      "step": 4953
    },
    {
      "epoch": 0.9908,
      "grad_norm": 0.04207742214202881,
      "learning_rate": 5.382716079271877e-08,
      "loss": 2.245,
      "step": 4954
    },
    {
      "epoch": 0.991,
      "grad_norm": 0.04238390922546387,
      "learning_rate": 5.1561205282335547e-08,
      "loss": 2.2462,
      "step": 4955
    },
    {
      "epoch": 0.9912,
      "grad_norm": 0.045365430414676666,
      "learning_rate": 4.934396342684e-08,
      "loss": 2.2478,
      "step": 4956
    },
    {
      "epoch": 0.9914,
      "grad_norm": 0.04507150128483772,
      "learning_rate": 4.717543630688992e-08,
      "loss": 2.2478,
      "step": 4957
    },
    {
      "epoch": 0.9916,
      "grad_norm": 0.04232514277100563,
      "learning_rate": 4.5055624979384316e-08,
      "loss": 2.2458,
      "step": 4958
    },
    {
      "epoch": 0.9918,
      "grad_norm": 0.04182828962802887,
      "learning_rate": 4.298453047749673e-08,
      "loss": 2.2461,
      "step": 4959
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.04178604483604431,
      "learning_rate": 4.096215381066415e-08,
      "loss": 2.2466,
      "step": 4960
    },
    {
      "epoch": 0.9922,
      "grad_norm": 0.041768014430999756,
      "learning_rate": 3.898849596456478e-08,
      "loss": 2.2461,
      "step": 4961
    },
    {
      "epoch": 0.9924,
      "grad_norm": 0.04197957366704941,
      "learning_rate": 3.7063557901129144e-08,
      "loss": 2.246,
      "step": 4962
    },
    {
      "epoch": 0.9926,
      "grad_norm": 0.04182039946317673,
      "learning_rate": 3.518734055855122e-08,
      "loss": 2.2443,
      "step": 4963
    },
    {
      "epoch": 0.9928,
      "grad_norm": 0.04194793105125427,
      "learning_rate": 3.33598448512773e-08,
      "loss": 2.2458,
      "step": 4964
    },
    {
      "epoch": 0.993,
      "grad_norm": 0.04182804003357887,
      "learning_rate": 3.1581071670006015e-08,
      "loss": 2.2478,
      "step": 4965
    },
    {
      "epoch": 0.9932,
      "grad_norm": 0.041852936148643494,
      "learning_rate": 2.985102188168831e-08,
      "loss": 2.2454,
      "step": 4966
    },
    {
      "epoch": 0.9934,
      "grad_norm": 0.04188346490263939,
      "learning_rate": 2.8169696329527483e-08,
      "loss": 2.2451,
      "step": 4967
    },
    {
      "epoch": 0.9936,
      "grad_norm": 0.04234989732503891,
      "learning_rate": 2.6537095832990245e-08,
      "loss": 2.2475,
      "step": 4968
    },
    {
      "epoch": 0.9938,
      "grad_norm": 0.04199014604091644,
      "learning_rate": 2.4953221187784537e-08,
      "loss": 2.2459,
      "step": 4969
    },
    {
      "epoch": 0.994,
      "grad_norm": 0.042086031287908554,
      "learning_rate": 2.3418073165870636e-08,
      "loss": 2.2461,
      "step": 4970
    },
    {
      "epoch": 0.9942,
      "grad_norm": 0.042029235512018204,
      "learning_rate": 2.193165251545004e-08,
      "loss": 2.2462,
      "step": 4971
    },
    {
      "epoch": 0.9944,
      "grad_norm": 0.04222570359706879,
      "learning_rate": 2.0493959960998787e-08,
      "loss": 2.2472,
      "step": 4972
    },
    {
      "epoch": 0.9946,
      "grad_norm": 0.04194052889943123,
      "learning_rate": 1.9104996203223037e-08,
      "loss": 2.2465,
      "step": 4973
    },
    {
      "epoch": 0.9948,
      "grad_norm": 0.04254184290766716,
      "learning_rate": 1.7764761919103477e-08,
      "loss": 2.2455,
      "step": 4974
    },
    {
      "epoch": 0.995,
      "grad_norm": 0.04185968264937401,
      "learning_rate": 1.6473257761828732e-08,
      "loss": 2.2439,
      "step": 4975
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.04181914031505585,
      "learning_rate": 1.5230484360873044e-08,
      "loss": 2.2448,
      "step": 4976
    },
    {
      "epoch": 0.9954,
      "grad_norm": 0.04268046095967293,
      "learning_rate": 1.4036442321962995e-08,
      "loss": 2.2476,
      "step": 4977
    },
    {
      "epoch": 0.9956,
      "grad_norm": 0.042745109647512436,
      "learning_rate": 1.2891132227033087e-08,
      "loss": 2.2479,
      "step": 4978
    },
    {
      "epoch": 0.9958,
      "grad_norm": 0.04354769363999367,
      "learning_rate": 1.1794554634314558e-08,
      "loss": 2.2493,
      "step": 4979
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.0429508239030838,
      "learning_rate": 1.074671007825767e-08,
      "loss": 2.2479,
      "step": 4980
    },
    {
      "epoch": 0.9962,
      "grad_norm": 0.043742530047893524,
      "learning_rate": 9.747599069576119e-09,
      "loss": 2.2496,
      "step": 4981
    },
    {
      "epoch": 0.9964,
      "grad_norm": 0.04309924691915512,
      "learning_rate": 8.797222095224823e-09,
      "loss": 2.247,
      "step": 4982
    },
    {
      "epoch": 0.9966,
      "grad_norm": 0.042034752666950226,
      "learning_rate": 7.895579618388827e-09,
      "loss": 2.2463,
      "step": 4983
    },
    {
      "epoch": 0.9968,
      "grad_norm": 0.04313487932085991,
      "learning_rate": 7.042672078527713e-09,
      "loss": 2.2471,
      "step": 4984
    },
    {
      "epoch": 0.997,
      "grad_norm": 0.042644500732421875,
      "learning_rate": 6.238499891353389e-09,
      "loss": 2.2476,
      "step": 4985
    },
    {
      "epoch": 0.9972,
      "grad_norm": 0.043875738978385925,
      "learning_rate": 5.483063448785686e-09,
      "loss": 2.2483,
      "step": 4986
    },
    {
      "epoch": 0.9974,
      "grad_norm": 0.043871596455574036,
      "learning_rate": 4.776363119030069e-09,
      "loss": 2.2472,
      "step": 4987
    },
    {
      "epoch": 0.9976,
      "grad_norm": 0.04385942220687866,
      "learning_rate": 4.1183992465221315e-09,
      "loss": 2.2477,
      "step": 4988
    },
    {
      "epoch": 0.9978,
      "grad_norm": 0.04326163977384567,
      "learning_rate": 3.509172151938689e-09,
      "loss": 2.2479,
      "step": 4989
    },
    {
      "epoch": 0.998,
      "grad_norm": 0.043807774782180786,
      "learning_rate": 2.948682132208891e-09,
      "loss": 2.2469,
      "step": 4990
    },
    {
      "epoch": 0.9982,
      "grad_norm": 0.04346576705574989,
      "learning_rate": 2.4369294605253166e-09,
      "loss": 2.2479,
      "step": 4991
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.04307727888226509,
      "learning_rate": 1.973914386288467e-09,
      "loss": 2.2474,
      "step": 4992
    },
    {
      "epoch": 0.9986,
      "grad_norm": 0.043591246008872986,
      "learning_rate": 1.5596371351733752e-09,
      "loss": 2.2485,
      "step": 4993
    },
    {
      "epoch": 0.9988,
      "grad_norm": 0.043508537113666534,
      "learning_rate": 1.1940979091074056e-09,
      "loss": 2.2478,
      "step": 4994
    },
    {
      "epoch": 0.999,
      "grad_norm": 0.0436728298664093,
      "learning_rate": 8.772968862369446e-10,
      "loss": 2.2485,
      "step": 4995
    },
    {
      "epoch": 0.9992,
      "grad_norm": 0.043184902518987656,
      "learning_rate": 6.092342209607083e-10,
      "loss": 2.2479,
      "step": 4996
    },
    {
      "epoch": 0.9994,
      "grad_norm": 0.042752817273139954,
      "learning_rate": 3.899100439408443e-10,
      "loss": 2.2474,
      "step": 4997
    },
    {
      "epoch": 0.9996,
      "grad_norm": 0.0422513447701931,
      "learning_rate": 2.1932446206962553e-10,
      "loss": 2.2487,
      "step": 4998
    },
    {
      "epoch": 0.9998,
      "grad_norm": 0.041733987629413605,
      "learning_rate": 9.74775584916543e-11,
      "loss": 2.2451,
      "step": 4999
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.04232018440961838,
      "learning_rate": 2.4369392592760164e-11,
      "loss": 2.2455,
      "step": 5000
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.022422523904e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}