{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.1664,
  "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": 127.02848815917969,
      "learning_rate": 5.000000000000001e-07,
      "loss": 8.9894,
      "step": 1
    },
    {
      "epoch": 0.0004,
      "grad_norm": 597.7093505859375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 11.8861,
      "step": 2
    },
    {
      "epoch": 0.0006,
      "grad_norm": 672.9093627929688,
      "learning_rate": 1.5e-06,
      "loss": 22.2202,
      "step": 3
    },
    {
      "epoch": 0.0008,
      "grad_norm": 203.70701599121094,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 9.8539,
      "step": 4
    },
    {
      "epoch": 0.001,
      "grad_norm": 621.4810180664062,
      "learning_rate": 2.5e-06,
      "loss": 14.1359,
      "step": 5
    },
    {
      "epoch": 0.0012,
      "grad_norm": 97.86430358886719,
      "learning_rate": 3e-06,
      "loss": 4.5683,
      "step": 6
    },
    {
      "epoch": 0.0014,
      "grad_norm": 134.84072875976562,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 9.4148,
      "step": 7
    },
    {
      "epoch": 0.0016,
      "grad_norm": 221.8789825439453,
      "learning_rate": 4.000000000000001e-06,
      "loss": 8.9489,
      "step": 8
    },
    {
      "epoch": 0.0018,
      "grad_norm": 574.8094482421875,
      "learning_rate": 4.5e-06,
      "loss": 10.2934,
      "step": 9
    },
    {
      "epoch": 0.002,
      "grad_norm": 211.35472106933594,
      "learning_rate": 5e-06,
      "loss": 10.3731,
      "step": 10
    },
    {
      "epoch": 0.0022,
      "grad_norm": 235.362060546875,
      "learning_rate": 5.500000000000001e-06,
      "loss": 9.9291,
      "step": 11
    },
    {
      "epoch": 0.0024,
      "grad_norm": 94.09746551513672,
      "learning_rate": 6e-06,
      "loss": 4.4554,
      "step": 12
    },
    {
      "epoch": 0.0026,
      "grad_norm": 83.25800323486328,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 7.3829,
      "step": 13
    },
    {
      "epoch": 0.0028,
      "grad_norm": 491.2541809082031,
      "learning_rate": 7.000000000000001e-06,
      "loss": 22.5135,
      "step": 14
    },
    {
      "epoch": 0.003,
      "grad_norm": 180.1479949951172,
      "learning_rate": 7.5e-06,
      "loss": 9.7904,
      "step": 15
    },
    {
      "epoch": 0.0032,
      "grad_norm": 1482.721435546875,
      "learning_rate": 8.000000000000001e-06,
      "loss": 15.6138,
      "step": 16
    },
    {
      "epoch": 0.0034,
      "grad_norm": 141.54798889160156,
      "learning_rate": 8.500000000000002e-06,
      "loss": 8.4214,
      "step": 17
    },
    {
      "epoch": 0.0036,
      "grad_norm": 265.2538757324219,
      "learning_rate": 9e-06,
      "loss": 8.1851,
      "step": 18
    },
    {
      "epoch": 0.0038,
      "grad_norm": 506.762451171875,
      "learning_rate": 9.5e-06,
      "loss": 18.3503,
      "step": 19
    },
    {
      "epoch": 0.004,
      "grad_norm": 533.5099487304688,
      "learning_rate": 1e-05,
      "loss": 9.0975,
      "step": 20
    },
    {
      "epoch": 0.0042,
      "grad_norm": 84.2190933227539,
      "learning_rate": 1.05e-05,
      "loss": 8.2995,
      "step": 21
    },
    {
      "epoch": 0.0044,
      "grad_norm": 1646.957275390625,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 16.4511,
      "step": 22
    },
    {
      "epoch": 0.0046,
      "grad_norm": 73.73599243164062,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 3.7298,
      "step": 23
    },
    {
      "epoch": 0.0048,
      "grad_norm": 487.625244140625,
      "learning_rate": 1.2e-05,
      "loss": 17.5225,
      "step": 24
    },
    {
      "epoch": 0.005,
      "grad_norm": 59.372535705566406,
      "learning_rate": 1.25e-05,
      "loss": 7.8804,
      "step": 25
    },
    {
      "epoch": 0.0052,
      "grad_norm": 97.49443054199219,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 7.7473,
      "step": 26
    },
    {
      "epoch": 0.0054,
      "grad_norm": 100.4667739868164,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 8.4293,
      "step": 27
    },
    {
      "epoch": 0.0056,
      "grad_norm": 44.721805572509766,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 7.0762,
      "step": 28
    },
    {
      "epoch": 0.0058,
      "grad_norm": 920.8779907226562,
      "learning_rate": 1.45e-05,
      "loss": 11.9611,
      "step": 29
    },
    {
      "epoch": 0.006,
      "grad_norm": 270.84088134765625,
      "learning_rate": 1.5e-05,
      "loss": 8.3698,
      "step": 30
    },
    {
      "epoch": 0.0062,
      "grad_norm": 55.78865432739258,
      "learning_rate": 1.55e-05,
      "loss": 7.5168,
      "step": 31
    },
    {
      "epoch": 0.0064,
      "grad_norm": 54.78589630126953,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 3.2809,
      "step": 32
    },
    {
      "epoch": 0.0066,
      "grad_norm": 99.09492492675781,
      "learning_rate": 1.65e-05,
      "loss": 7.8991,
      "step": 33
    },
    {
      "epoch": 0.0068,
      "grad_norm": 102.16423034667969,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.4389,
      "step": 34
    },
    {
      "epoch": 0.007,
      "grad_norm": 53.18633270263672,
      "learning_rate": 1.75e-05,
      "loss": 7.7285,
      "step": 35
    },
    {
      "epoch": 0.0072,
      "grad_norm": 2191.484130859375,
      "learning_rate": 1.8e-05,
      "loss": 11.3803,
      "step": 36
    },
    {
      "epoch": 0.0074,
      "grad_norm": 48.38761901855469,
      "learning_rate": 1.85e-05,
      "loss": 6.9352,
      "step": 37
    },
    {
      "epoch": 0.0076,
      "grad_norm": 37.02681350708008,
      "learning_rate": 1.9e-05,
      "loss": 7.463,
      "step": 38
    },
    {
      "epoch": 0.0078,
      "grad_norm": 804.8701782226562,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 10.1294,
      "step": 39
    },
    {
      "epoch": 0.008,
      "grad_norm": 33.863460540771484,
      "learning_rate": 2e-05,
      "loss": 6.993,
      "step": 40
    },
    {
      "epoch": 0.0082,
      "grad_norm": 122.3696060180664,
      "learning_rate": 2.05e-05,
      "loss": 7.8665,
      "step": 41
    },
    {
      "epoch": 0.0084,
      "grad_norm": 237.02120971679688,
      "learning_rate": 2.1e-05,
      "loss": 10.9977,
      "step": 42
    },
    {
      "epoch": 0.0086,
      "grad_norm": 114.43867492675781,
      "learning_rate": 2.15e-05,
      "loss": 6.6069,
      "step": 43
    },
    {
      "epoch": 0.0088,
      "grad_norm": 36.79835891723633,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 7.1009,
      "step": 44
    },
    {
      "epoch": 0.009,
      "grad_norm": 205.56387329101562,
      "learning_rate": 2.25e-05,
      "loss": 10.1884,
      "step": 45
    },
    {
      "epoch": 0.0092,
      "grad_norm": 43.2474479675293,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 6.7291,
      "step": 46
    },
    {
      "epoch": 0.0094,
      "grad_norm": 47.37004089355469,
      "learning_rate": 2.35e-05,
      "loss": 6.4736,
      "step": 47
    },
    {
      "epoch": 0.0096,
      "grad_norm": 542.9366455078125,
      "learning_rate": 2.4e-05,
      "loss": 10.2724,
      "step": 48
    },
    {
      "epoch": 0.0098,
      "grad_norm": 99.73185729980469,
      "learning_rate": 2.45e-05,
      "loss": 5.6246,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 38.3469123840332,
      "learning_rate": 2.5e-05,
      "loss": 6.3691,
      "step": 50
    },
    {
      "epoch": 0.0102,
      "grad_norm": 76.59410095214844,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 7.3026,
      "step": 51
    },
    {
      "epoch": 0.0104,
      "grad_norm": 234.27630615234375,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 8.6945,
      "step": 52
    },
    {
      "epoch": 0.0106,
      "grad_norm": 44.871097564697266,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 6.2657,
      "step": 53
    },
    {
      "epoch": 0.0108,
      "grad_norm": 60.8604850769043,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 6.6361,
      "step": 54
    },
    {
      "epoch": 0.011,
      "grad_norm": 62.86309051513672,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 4.5686,
      "step": 55
    },
    {
      "epoch": 0.0112,
      "grad_norm": 56.52326583862305,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 5.9374,
      "step": 56
    },
    {
      "epoch": 0.0114,
      "grad_norm": 29.06716537475586,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 6.9688,
      "step": 57
    },
    {
      "epoch": 0.0116,
      "grad_norm": 44.0078125,
      "learning_rate": 2.9e-05,
      "loss": 4.9339,
      "step": 58
    },
    {
      "epoch": 0.0118,
      "grad_norm": 42.00264358520508,
      "learning_rate": 2.95e-05,
      "loss": 5.7515,
      "step": 59
    },
    {
      "epoch": 0.012,
      "grad_norm": 140.99119567871094,
      "learning_rate": 3e-05,
      "loss": 6.9327,
      "step": 60
    },
    {
      "epoch": 0.0122,
      "grad_norm": 39.872074127197266,
      "learning_rate": 3.05e-05,
      "loss": 5.49,
      "step": 61
    },
    {
      "epoch": 0.0124,
      "grad_norm": 29.44024658203125,
      "learning_rate": 3.1e-05,
      "loss": 6.6595,
      "step": 62
    },
    {
      "epoch": 0.0126,
      "grad_norm": 29.17344856262207,
      "learning_rate": 3.15e-05,
      "loss": 4.0179,
      "step": 63
    },
    {
      "epoch": 0.0128,
      "grad_norm": 118.6063003540039,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 4.1218,
      "step": 64
    },
    {
      "epoch": 0.013,
      "grad_norm": 33.67353820800781,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 4.605,
      "step": 65
    },
    {
      "epoch": 0.0132,
      "grad_norm": 31.440397262573242,
      "learning_rate": 3.3e-05,
      "loss": 4.8518,
      "step": 66
    },
    {
      "epoch": 0.0134,
      "grad_norm": 17.99005699157715,
      "learning_rate": 3.35e-05,
      "loss": 3.2932,
      "step": 67
    },
    {
      "epoch": 0.0136,
      "grad_norm": 26.642446517944336,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 6.224,
      "step": 68
    },
    {
      "epoch": 0.0138,
      "grad_norm": 311.91741943359375,
      "learning_rate": 3.45e-05,
      "loss": 7.6243,
      "step": 69
    },
    {
      "epoch": 0.014,
      "grad_norm": 82.06395721435547,
      "learning_rate": 3.5e-05,
      "loss": 7.1435,
      "step": 70
    },
    {
      "epoch": 0.0142,
      "grad_norm": 262.7768249511719,
      "learning_rate": 3.55e-05,
      "loss": 4.886,
      "step": 71
    },
    {
      "epoch": 0.0144,
      "grad_norm": 37.81349563598633,
      "learning_rate": 3.6e-05,
      "loss": 6.8874,
      "step": 72
    },
    {
      "epoch": 0.0146,
      "grad_norm": 74.8050765991211,
      "learning_rate": 3.65e-05,
      "loss": 2.0786,
      "step": 73
    },
    {
      "epoch": 0.0148,
      "grad_norm": 90.38041687011719,
      "learning_rate": 3.7e-05,
      "loss": 3.6252,
      "step": 74
    },
    {
      "epoch": 0.015,
      "grad_norm": 60.22389602661133,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 4.9044,
      "step": 75
    },
    {
      "epoch": 0.0152,
      "grad_norm": 340.64166259765625,
      "learning_rate": 3.8e-05,
      "loss": 8.3815,
      "step": 76
    },
    {
      "epoch": 0.0154,
      "grad_norm": 27.283594131469727,
      "learning_rate": 3.85e-05,
      "loss": 5.8438,
      "step": 77
    },
    {
      "epoch": 0.0156,
      "grad_norm": 14.617252349853516,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 3.4718,
      "step": 78
    },
    {
      "epoch": 0.0158,
      "grad_norm": 20.60955810546875,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 4.0038,
      "step": 79
    },
    {
      "epoch": 0.016,
      "grad_norm": 24.20742416381836,
      "learning_rate": 4e-05,
      "loss": 1.6708,
      "step": 80
    },
    {
      "epoch": 0.0162,
      "grad_norm": 9.425049781799316,
      "learning_rate": 4.05e-05,
      "loss": 3.2246,
      "step": 81
    },
    {
      "epoch": 0.0164,
      "grad_norm": 133.04257202148438,
      "learning_rate": 4.1e-05,
      "loss": 6.8735,
      "step": 82
    },
    {
      "epoch": 0.0166,
      "grad_norm": 34.120025634765625,
      "learning_rate": 4.15e-05,
      "loss": 1.512,
      "step": 83
    },
    {
      "epoch": 0.0168,
      "grad_norm": 95.99542999267578,
      "learning_rate": 4.2e-05,
      "loss": 4.0492,
      "step": 84
    },
    {
      "epoch": 0.017,
      "grad_norm": 130.52586364746094,
      "learning_rate": 4.25e-05,
      "loss": 7.1804,
      "step": 85
    },
    {
      "epoch": 0.0172,
      "grad_norm": 50.44832229614258,
      "learning_rate": 4.3e-05,
      "loss": 5.8466,
      "step": 86
    },
    {
      "epoch": 0.0174,
      "grad_norm": 13.494194984436035,
      "learning_rate": 4.35e-05,
      "loss": 1.1162,
      "step": 87
    },
    {
      "epoch": 0.0176,
      "grad_norm": 32.111392974853516,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 5.315,
      "step": 88
    },
    {
      "epoch": 0.0178,
      "grad_norm": 54.81203842163086,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 3.1099,
      "step": 89
    },
    {
      "epoch": 0.018,
      "grad_norm": 19.034406661987305,
      "learning_rate": 4.5e-05,
      "loss": 2.7665,
      "step": 90
    },
    {
      "epoch": 0.0182,
      "grad_norm": 31.54140281677246,
      "learning_rate": 4.55e-05,
      "loss": 2.5888,
      "step": 91
    },
    {
      "epoch": 0.0184,
      "grad_norm": 11.20257568359375,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.1534,
      "step": 92
    },
    {
      "epoch": 0.0186,
      "grad_norm": 32.47492218017578,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 4.7617,
      "step": 93
    },
    {
      "epoch": 0.0188,
      "grad_norm": 34.09675216674805,
      "learning_rate": 4.7e-05,
      "loss": 2.0035,
      "step": 94
    },
    {
      "epoch": 0.019,
      "grad_norm": 37.695125579833984,
      "learning_rate": 4.75e-05,
      "loss": 5.3329,
      "step": 95
    },
    {
      "epoch": 0.0192,
      "grad_norm": 20.07520294189453,
      "learning_rate": 4.8e-05,
      "loss": 2.5175,
      "step": 96
    },
    {
      "epoch": 0.0194,
      "grad_norm": 415.35711669921875,
      "learning_rate": 4.85e-05,
      "loss": 10.5392,
      "step": 97
    },
    {
      "epoch": 0.0196,
      "grad_norm": 41.11747741699219,
      "learning_rate": 4.9e-05,
      "loss": 4.6094,
      "step": 98
    },
    {
      "epoch": 0.0198,
      "grad_norm": 42.81120300292969,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 6.263,
      "step": 99
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.508249282836914,
      "learning_rate": 5e-05,
      "loss": 0.8818,
      "step": 100
    },
    {
      "epoch": 0.0202,
      "grad_norm": 10.983814239501953,
      "learning_rate": 4.998979591836735e-05,
      "loss": 0.7669,
      "step": 101
    },
    {
      "epoch": 0.0204,
      "grad_norm": 9.199824333190918,
      "learning_rate": 4.9979591836734694e-05,
      "loss": 1.0418,
      "step": 102
    },
    {
      "epoch": 0.0206,
      "grad_norm": 66.83099365234375,
      "learning_rate": 4.996938775510204e-05,
      "loss": 1.6225,
      "step": 103
    },
    {
      "epoch": 0.0208,
      "grad_norm": 7.638471603393555,
      "learning_rate": 4.995918367346939e-05,
      "loss": 0.4936,
      "step": 104
    },
    {
      "epoch": 0.021,
      "grad_norm": 10.665667533874512,
      "learning_rate": 4.9948979591836735e-05,
      "loss": 0.8371,
      "step": 105
    },
    {
      "epoch": 0.0212,
      "grad_norm": 10.239150047302246,
      "learning_rate": 4.9938775510204084e-05,
      "loss": 0.8516,
      "step": 106
    },
    {
      "epoch": 0.0214,
      "grad_norm": 34.41289138793945,
      "learning_rate": 4.992857142857143e-05,
      "loss": 2.7268,
      "step": 107
    },
    {
      "epoch": 0.0216,
      "grad_norm": 107.72695922851562,
      "learning_rate": 4.9918367346938776e-05,
      "loss": 6.4053,
      "step": 108
    },
    {
      "epoch": 0.0218,
      "grad_norm": 64.1345443725586,
      "learning_rate": 4.9908163265306125e-05,
      "loss": 3.2117,
      "step": 109
    },
    {
      "epoch": 0.022,
      "grad_norm": 11.564123153686523,
      "learning_rate": 4.9897959183673474e-05,
      "loss": 0.8914,
      "step": 110
    },
    {
      "epoch": 0.0222,
      "grad_norm": 23.555273056030273,
      "learning_rate": 4.9887755102040816e-05,
      "loss": 2.3173,
      "step": 111
    },
    {
      "epoch": 0.0224,
      "grad_norm": 26.109325408935547,
      "learning_rate": 4.9877551020408165e-05,
      "loss": 2.4154,
      "step": 112
    },
    {
      "epoch": 0.0226,
      "grad_norm": 37.607215881347656,
      "learning_rate": 4.986734693877551e-05,
      "loss": 4.3035,
      "step": 113
    },
    {
      "epoch": 0.0228,
      "grad_norm": 8.989594459533691,
      "learning_rate": 4.985714285714286e-05,
      "loss": 2.5046,
      "step": 114
    },
    {
      "epoch": 0.023,
      "grad_norm": 27.475799560546875,
      "learning_rate": 4.9846938775510206e-05,
      "loss": 2.4623,
      "step": 115
    },
    {
      "epoch": 0.0232,
      "grad_norm": 5.941082954406738,
      "learning_rate": 4.983673469387755e-05,
      "loss": 0.7323,
      "step": 116
    },
    {
      "epoch": 0.0234,
      "grad_norm": 6.239040374755859,
      "learning_rate": 4.98265306122449e-05,
      "loss": 0.7388,
      "step": 117
    },
    {
      "epoch": 0.0236,
      "grad_norm": 5.513604640960693,
      "learning_rate": 4.981632653061225e-05,
      "loss": 0.6584,
      "step": 118
    },
    {
      "epoch": 0.0238,
      "grad_norm": 43.922508239746094,
      "learning_rate": 4.980612244897959e-05,
      "loss": 5.6676,
      "step": 119
    },
    {
      "epoch": 0.024,
      "grad_norm": 8.907512664794922,
      "learning_rate": 4.979591836734694e-05,
      "loss": 2.0489,
      "step": 120
    },
    {
      "epoch": 0.0242,
      "grad_norm": 6.571248531341553,
      "learning_rate": 4.978571428571429e-05,
      "loss": 2.4868,
      "step": 121
    },
    {
      "epoch": 0.0244,
      "grad_norm": 9.025981903076172,
      "learning_rate": 4.977551020408163e-05,
      "loss": 3.183,
      "step": 122
    },
    {
      "epoch": 0.0246,
      "grad_norm": 82.9317398071289,
      "learning_rate": 4.9765306122448986e-05,
      "loss": 6.5242,
      "step": 123
    },
    {
      "epoch": 0.0248,
      "grad_norm": 43.22480010986328,
      "learning_rate": 4.975510204081633e-05,
      "loss": 5.2471,
      "step": 124
    },
    {
      "epoch": 0.025,
      "grad_norm": 5.8927154541015625,
      "learning_rate": 4.974489795918368e-05,
      "loss": 6.9892,
      "step": 125
    },
    {
      "epoch": 0.0252,
      "grad_norm": 12.013327598571777,
      "learning_rate": 4.973469387755103e-05,
      "loss": 1.9226,
      "step": 126
    },
    {
      "epoch": 0.0254,
      "grad_norm": 15.207947731018066,
      "learning_rate": 4.972448979591837e-05,
      "loss": 6.7809,
      "step": 127
    },
    {
      "epoch": 0.0256,
      "grad_norm": 17.63144302368164,
      "learning_rate": 4.971428571428572e-05,
      "loss": 1.7614,
      "step": 128
    },
    {
      "epoch": 0.0258,
      "grad_norm": 28.31899642944336,
      "learning_rate": 4.970408163265307e-05,
      "loss": 2.4694,
      "step": 129
    },
    {
      "epoch": 0.026,
      "grad_norm": 5.781543254852295,
      "learning_rate": 4.969387755102041e-05,
      "loss": 2.5552,
      "step": 130
    },
    {
      "epoch": 0.0262,
      "grad_norm": 80.65481567382812,
      "learning_rate": 4.968367346938776e-05,
      "loss": 3.0708,
      "step": 131
    },
    {
      "epoch": 0.0264,
      "grad_norm": 4.071949005126953,
      "learning_rate": 4.967346938775511e-05,
      "loss": 0.2991,
      "step": 132
    },
    {
      "epoch": 0.0266,
      "grad_norm": 4.599740982055664,
      "learning_rate": 4.966326530612245e-05,
      "loss": 2.3899,
      "step": 133
    },
    {
      "epoch": 0.0268,
      "grad_norm": 20.408233642578125,
      "learning_rate": 4.96530612244898e-05,
      "loss": 6.7909,
      "step": 134
    },
    {
      "epoch": 0.027,
      "grad_norm": 4.217963218688965,
      "learning_rate": 4.964285714285715e-05,
      "loss": 2.3584,
      "step": 135
    },
    {
      "epoch": 0.0272,
      "grad_norm": 120.89624786376953,
      "learning_rate": 4.963265306122449e-05,
      "loss": 7.0599,
      "step": 136
    },
    {
      "epoch": 0.0274,
      "grad_norm": 4.268696308135986,
      "learning_rate": 4.962244897959184e-05,
      "loss": 0.3368,
      "step": 137
    },
    {
      "epoch": 0.0276,
      "grad_norm": 8.89269733428955,
      "learning_rate": 4.961224489795919e-05,
      "loss": 1.674,
      "step": 138
    },
    {
      "epoch": 0.0278,
      "grad_norm": 4.60618257522583,
      "learning_rate": 4.960204081632653e-05,
      "loss": 0.5549,
      "step": 139
    },
    {
      "epoch": 0.028,
      "grad_norm": 37.98936462402344,
      "learning_rate": 4.959183673469388e-05,
      "loss": 5.3535,
      "step": 140
    },
    {
      "epoch": 0.0282,
      "grad_norm": 8.033515930175781,
      "learning_rate": 4.958163265306123e-05,
      "loss": 1.5668,
      "step": 141
    },
    {
      "epoch": 0.0284,
      "grad_norm": 7.632925033569336,
      "learning_rate": 4.957142857142857e-05,
      "loss": 1.5609,
      "step": 142
    },
    {
      "epoch": 0.0286,
      "grad_norm": 3.267777681350708,
      "learning_rate": 4.956122448979592e-05,
      "loss": 0.2475,
      "step": 143
    },
    {
      "epoch": 0.0288,
      "grad_norm": 4.545289039611816,
      "learning_rate": 4.9551020408163265e-05,
      "loss": 2.3144,
      "step": 144
    },
    {
      "epoch": 0.029,
      "grad_norm": 19.45534324645996,
      "learning_rate": 4.9540816326530614e-05,
      "loss": 1.4105,
      "step": 145
    },
    {
      "epoch": 0.0292,
      "grad_norm": 12.32528018951416,
      "learning_rate": 4.953061224489796e-05,
      "loss": 0.6523,
      "step": 146
    },
    {
      "epoch": 0.0294,
      "grad_norm": 115.36129760742188,
      "learning_rate": 4.9520408163265305e-05,
      "loss": 6.924,
      "step": 147
    },
    {
      "epoch": 0.0296,
      "grad_norm": 16.953712463378906,
      "learning_rate": 4.9510204081632655e-05,
      "loss": 3.345,
      "step": 148
    },
    {
      "epoch": 0.0298,
      "grad_norm": 50.63832473754883,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 4.9704,
      "step": 149
    },
    {
      "epoch": 0.03,
      "grad_norm": 32.5130615234375,
      "learning_rate": 4.9489795918367346e-05,
      "loss": 6.6223,
      "step": 150
    },
    {
      "epoch": 0.0302,
      "grad_norm": 4.348920822143555,
      "learning_rate": 4.9479591836734695e-05,
      "loss": 0.5413,
      "step": 151
    },
    {
      "epoch": 0.0304,
      "grad_norm": 3.998990774154663,
      "learning_rate": 4.9469387755102045e-05,
      "loss": 0.4476,
      "step": 152
    },
    {
      "epoch": 0.0306,
      "grad_norm": 23.424474716186523,
      "learning_rate": 4.945918367346939e-05,
      "loss": 3.7908,
      "step": 153
    },
    {
      "epoch": 0.0308,
      "grad_norm": 22.823986053466797,
      "learning_rate": 4.9448979591836736e-05,
      "loss": 3.9092,
      "step": 154
    },
    {
      "epoch": 0.031,
      "grad_norm": 7.079718589782715,
      "learning_rate": 4.9438775510204085e-05,
      "loss": 2.9751,
      "step": 155
    },
    {
      "epoch": 0.0312,
      "grad_norm": 52.237789154052734,
      "learning_rate": 4.942857142857143e-05,
      "loss": 5.3007,
      "step": 156
    },
    {
      "epoch": 0.0314,
      "grad_norm": 6.196706295013428,
      "learning_rate": 4.941836734693878e-05,
      "loss": 2.8203,
      "step": 157
    },
    {
      "epoch": 0.0316,
      "grad_norm": 4.051186561584473,
      "learning_rate": 4.9408163265306126e-05,
      "loss": 0.5189,
      "step": 158
    },
    {
      "epoch": 0.0318,
      "grad_norm": 3.7578818798065186,
      "learning_rate": 4.939795918367347e-05,
      "loss": 0.4583,
      "step": 159
    },
    {
      "epoch": 0.032,
      "grad_norm": 22.257274627685547,
      "learning_rate": 4.938775510204082e-05,
      "loss": 3.8465,
      "step": 160
    },
    {
      "epoch": 0.0322,
      "grad_norm": 3.206167221069336,
      "learning_rate": 4.937755102040817e-05,
      "loss": 0.2594,
      "step": 161
    },
    {
      "epoch": 0.0324,
      "grad_norm": 9.934144973754883,
      "learning_rate": 4.936734693877551e-05,
      "loss": 1.4664,
      "step": 162
    },
    {
      "epoch": 0.0326,
      "grad_norm": 71.399169921875,
      "learning_rate": 4.935714285714286e-05,
      "loss": 6.7523,
      "step": 163
    },
    {
      "epoch": 0.0328,
      "grad_norm": 12.2349214553833,
      "learning_rate": 4.93469387755102e-05,
      "loss": 0.4401,
      "step": 164
    },
    {
      "epoch": 0.033,
      "grad_norm": 3.647381067276001,
      "learning_rate": 4.933673469387755e-05,
      "loss": 0.4522,
      "step": 165
    },
    {
      "epoch": 0.0332,
      "grad_norm": 2.9874470233917236,
      "learning_rate": 4.93265306122449e-05,
      "loss": 0.4223,
      "step": 166
    },
    {
      "epoch": 0.0334,
      "grad_norm": 15.757845878601074,
      "learning_rate": 4.931632653061224e-05,
      "loss": 3.0744,
      "step": 167
    },
    {
      "epoch": 0.0336,
      "grad_norm": 2.7883803844451904,
      "learning_rate": 4.930612244897959e-05,
      "loss": 0.1997,
      "step": 168
    },
    {
      "epoch": 0.0338,
      "grad_norm": 14.830641746520996,
      "learning_rate": 4.929591836734695e-05,
      "loss": 2.8235,
      "step": 169
    },
    {
      "epoch": 0.034,
      "grad_norm": 8.174346923828125,
      "learning_rate": 4.928571428571429e-05,
      "loss": 1.4324,
      "step": 170
    },
    {
      "epoch": 0.0342,
      "grad_norm": 9.390400886535645,
      "learning_rate": 4.927551020408164e-05,
      "loss": 1.1507,
      "step": 171
    },
    {
      "epoch": 0.0344,
      "grad_norm": 3.5765488147735596,
      "learning_rate": 4.926530612244899e-05,
      "loss": 0.2527,
      "step": 172
    },
    {
      "epoch": 0.0346,
      "grad_norm": 16.015581130981445,
      "learning_rate": 4.925510204081633e-05,
      "loss": 2.0474,
      "step": 173
    },
    {
      "epoch": 0.0348,
      "grad_norm": 13.759570121765137,
      "learning_rate": 4.924489795918368e-05,
      "loss": 7.0157,
      "step": 174
    },
    {
      "epoch": 0.035,
      "grad_norm": 45.530189514160156,
      "learning_rate": 4.923469387755102e-05,
      "loss": 4.6881,
      "step": 175
    },
    {
      "epoch": 0.0352,
      "grad_norm": 3.162175178527832,
      "learning_rate": 4.922448979591837e-05,
      "loss": 0.2335,
      "step": 176
    },
    {
      "epoch": 0.0354,
      "grad_norm": 2.960479497909546,
      "learning_rate": 4.921428571428572e-05,
      "loss": 0.4245,
      "step": 177
    },
    {
      "epoch": 0.0356,
      "grad_norm": 3.4160397052764893,
      "learning_rate": 4.920408163265306e-05,
      "loss": 0.4166,
      "step": 178
    },
    {
      "epoch": 0.0358,
      "grad_norm": 2.5386452674865723,
      "learning_rate": 4.919387755102041e-05,
      "loss": 0.3643,
      "step": 179
    },
    {
      "epoch": 0.036,
      "grad_norm": 2.4958510398864746,
      "learning_rate": 4.918367346938776e-05,
      "loss": 0.3616,
      "step": 180
    },
    {
      "epoch": 0.0362,
      "grad_norm": 7.223433017730713,
      "learning_rate": 4.91734693877551e-05,
      "loss": 1.4523,
      "step": 181
    },
    {
      "epoch": 0.0364,
      "grad_norm": 16.580181121826172,
      "learning_rate": 4.916326530612245e-05,
      "loss": 2.6113,
      "step": 182
    },
    {
      "epoch": 0.0366,
      "grad_norm": 3.077042818069458,
      "learning_rate": 4.91530612244898e-05,
      "loss": 0.4027,
      "step": 183
    },
    {
      "epoch": 0.0368,
      "grad_norm": 7.351260185241699,
      "learning_rate": 4.9142857142857144e-05,
      "loss": 2.8672,
      "step": 184
    },
    {
      "epoch": 0.037,
      "grad_norm": 5.3265886306762695,
      "learning_rate": 4.913265306122449e-05,
      "loss": 0.2311,
      "step": 185
    },
    {
      "epoch": 0.0372,
      "grad_norm": 25.74429702758789,
      "learning_rate": 4.912244897959184e-05,
      "loss": 1.7426,
      "step": 186
    },
    {
      "epoch": 0.0374,
      "grad_norm": 2.879579782485962,
      "learning_rate": 4.9112244897959185e-05,
      "loss": 0.3941,
      "step": 187
    },
    {
      "epoch": 0.0376,
      "grad_norm": 13.189726829528809,
      "learning_rate": 4.9102040816326534e-05,
      "loss": 0.5893,
      "step": 188
    },
    {
      "epoch": 0.0378,
      "grad_norm": 3.6334128379821777,
      "learning_rate": 4.909183673469388e-05,
      "loss": 0.2001,
      "step": 189
    },
    {
      "epoch": 0.038,
      "grad_norm": 14.311576843261719,
      "learning_rate": 4.9081632653061225e-05,
      "loss": 1.6499,
      "step": 190
    },
    {
      "epoch": 0.0382,
      "grad_norm": 24.885698318481445,
      "learning_rate": 4.9071428571428574e-05,
      "loss": 3.9402,
      "step": 191
    },
    {
      "epoch": 0.0384,
      "grad_norm": 2.6309475898742676,
      "learning_rate": 4.9061224489795924e-05,
      "loss": 0.3384,
      "step": 192
    },
    {
      "epoch": 0.0386,
      "grad_norm": 10.061433792114258,
      "learning_rate": 4.9051020408163266e-05,
      "loss": 1.5038,
      "step": 193
    },
    {
      "epoch": 0.0388,
      "grad_norm": 20.588462829589844,
      "learning_rate": 4.9040816326530615e-05,
      "loss": 2.2994,
      "step": 194
    },
    {
      "epoch": 0.039,
      "grad_norm": 8.214776992797852,
      "learning_rate": 4.903061224489796e-05,
      "loss": 1.6767,
      "step": 195
    },
    {
      "epoch": 0.0392,
      "grad_norm": 2.4332640171051025,
      "learning_rate": 4.902040816326531e-05,
      "loss": 0.3212,
      "step": 196
    },
    {
      "epoch": 0.0394,
      "grad_norm": 15.954568862915039,
      "learning_rate": 4.9010204081632656e-05,
      "loss": 2.0944,
      "step": 197
    },
    {
      "epoch": 0.0396,
      "grad_norm": 3.0381927490234375,
      "learning_rate": 4.9e-05,
      "loss": 0.4318,
      "step": 198
    },
    {
      "epoch": 0.0398,
      "grad_norm": 2.6343955993652344,
      "learning_rate": 4.898979591836735e-05,
      "loss": 0.1892,
      "step": 199
    },
    {
      "epoch": 0.04,
      "grad_norm": 6.825660705566406,
      "learning_rate": 4.89795918367347e-05,
      "loss": 1.2455,
      "step": 200
    },
    {
      "epoch": 0.0402,
      "grad_norm": 15.805678367614746,
      "learning_rate": 4.896938775510204e-05,
      "loss": 3.0766,
      "step": 201
    },
    {
      "epoch": 0.0404,
      "grad_norm": 22.658878326416016,
      "learning_rate": 4.895918367346939e-05,
      "loss": 2.6953,
      "step": 202
    },
    {
      "epoch": 0.0406,
      "grad_norm": 10.367082595825195,
      "learning_rate": 4.894897959183674e-05,
      "loss": 1.3568,
      "step": 203
    },
    {
      "epoch": 0.0408,
      "grad_norm": 14.985063552856445,
      "learning_rate": 4.893877551020408e-05,
      "loss": 3.11,
      "step": 204
    },
    {
      "epoch": 0.041,
      "grad_norm": 4.432097434997559,
      "learning_rate": 4.892857142857143e-05,
      "loss": 0.6922,
      "step": 205
    },
    {
      "epoch": 0.0412,
      "grad_norm": 74.39356994628906,
      "learning_rate": 4.891836734693878e-05,
      "loss": 4.0891,
      "step": 206
    },
    {
      "epoch": 0.0414,
      "grad_norm": 5.9803996086120605,
      "learning_rate": 4.890816326530612e-05,
      "loss": 1.1988,
      "step": 207
    },
    {
      "epoch": 0.0416,
      "grad_norm": 5.758410930633545,
      "learning_rate": 4.889795918367347e-05,
      "loss": 2.229,
      "step": 208
    },
    {
      "epoch": 0.0418,
      "grad_norm": 11.344877243041992,
      "learning_rate": 4.888775510204082e-05,
      "loss": 1.8269,
      "step": 209
    },
    {
      "epoch": 0.042,
      "grad_norm": 3.3845725059509277,
      "learning_rate": 4.887755102040816e-05,
      "loss": 0.4241,
      "step": 210
    },
    {
      "epoch": 0.0422,
      "grad_norm": 3.4041435718536377,
      "learning_rate": 4.886734693877551e-05,
      "loss": 0.4123,
      "step": 211
    },
    {
      "epoch": 0.0424,
      "grad_norm": 3.4436275959014893,
      "learning_rate": 4.885714285714286e-05,
      "loss": 0.4752,
      "step": 212
    },
    {
      "epoch": 0.0426,
      "grad_norm": 11.167293548583984,
      "learning_rate": 4.88469387755102e-05,
      "loss": 2.8301,
      "step": 213
    },
    {
      "epoch": 0.0428,
      "grad_norm": 25.37110137939453,
      "learning_rate": 4.883673469387756e-05,
      "loss": 3.3613,
      "step": 214
    },
    {
      "epoch": 0.043,
      "grad_norm": 2.6816182136535645,
      "learning_rate": 4.88265306122449e-05,
      "loss": 0.1673,
      "step": 215
    },
    {
      "epoch": 0.0432,
      "grad_norm": 5.956937789916992,
      "learning_rate": 4.881632653061225e-05,
      "loss": 0.2408,
      "step": 216
    },
    {
      "epoch": 0.0434,
      "grad_norm": 8.719182014465332,
      "learning_rate": 4.88061224489796e-05,
      "loss": 1.2296,
      "step": 217
    },
    {
      "epoch": 0.0436,
      "grad_norm": 7.700140953063965,
      "learning_rate": 4.879591836734694e-05,
      "loss": 2.7421,
      "step": 218
    },
    {
      "epoch": 0.0438,
      "grad_norm": 1.9446169137954712,
      "learning_rate": 4.878571428571429e-05,
      "loss": 0.1474,
      "step": 219
    },
    {
      "epoch": 0.044,
      "grad_norm": 87.47311401367188,
      "learning_rate": 4.877551020408164e-05,
      "loss": 4.962,
      "step": 220
    },
    {
      "epoch": 0.0442,
      "grad_norm": 2.5825252532958984,
      "learning_rate": 4.876530612244898e-05,
      "loss": 0.178,
      "step": 221
    },
    {
      "epoch": 0.0444,
      "grad_norm": 54.101078033447266,
      "learning_rate": 4.875510204081633e-05,
      "loss": 6.6264,
      "step": 222
    },
    {
      "epoch": 0.0446,
      "grad_norm": 6.605436325073242,
      "learning_rate": 4.874489795918368e-05,
      "loss": 1.5507,
      "step": 223
    },
    {
      "epoch": 0.0448,
      "grad_norm": 14.785659790039062,
      "learning_rate": 4.873469387755102e-05,
      "loss": 3.2945,
      "step": 224
    },
    {
      "epoch": 0.045,
      "grad_norm": 9.988936424255371,
      "learning_rate": 4.872448979591837e-05,
      "loss": 2.7765,
      "step": 225
    },
    {
      "epoch": 0.0452,
      "grad_norm": 40.442596435546875,
      "learning_rate": 4.8714285714285714e-05,
      "loss": 4.506,
      "step": 226
    },
    {
      "epoch": 0.0454,
      "grad_norm": 2.5142595767974854,
      "learning_rate": 4.8704081632653064e-05,
      "loss": 0.1866,
      "step": 227
    },
    {
      "epoch": 0.0456,
      "grad_norm": 4.363778591156006,
      "learning_rate": 4.869387755102041e-05,
      "loss": 2.0775,
      "step": 228
    },
    {
      "epoch": 0.0458,
      "grad_norm": 2.496397018432617,
      "learning_rate": 4.8683673469387755e-05,
      "loss": 0.1748,
      "step": 229
    },
    {
      "epoch": 0.046,
      "grad_norm": 3.2190587520599365,
      "learning_rate": 4.8673469387755104e-05,
      "loss": 0.4083,
      "step": 230
    },
    {
      "epoch": 0.0462,
      "grad_norm": 13.004916191101074,
      "learning_rate": 4.8663265306122454e-05,
      "loss": 0.826,
      "step": 231
    },
    {
      "epoch": 0.0464,
      "grad_norm": 3.257899045944214,
      "learning_rate": 4.8653061224489796e-05,
      "loss": 0.4073,
      "step": 232
    },
    {
      "epoch": 0.0466,
      "grad_norm": 22.85171890258789,
      "learning_rate": 4.8642857142857145e-05,
      "loss": 2.5843,
      "step": 233
    },
    {
      "epoch": 0.0468,
      "grad_norm": 23.22861671447754,
      "learning_rate": 4.8632653061224494e-05,
      "loss": 3.0827,
      "step": 234
    },
    {
      "epoch": 0.047,
      "grad_norm": 2.492215633392334,
      "learning_rate": 4.862244897959184e-05,
      "loss": 0.315,
      "step": 235
    },
    {
      "epoch": 0.0472,
      "grad_norm": 2.859203338623047,
      "learning_rate": 4.8612244897959186e-05,
      "loss": 0.4144,
      "step": 236
    },
    {
      "epoch": 0.0474,
      "grad_norm": 10.443507194519043,
      "learning_rate": 4.8602040816326535e-05,
      "loss": 1.4981,
      "step": 237
    },
    {
      "epoch": 0.0476,
      "grad_norm": 5.941455364227295,
      "learning_rate": 4.859183673469388e-05,
      "loss": 1.4664,
      "step": 238
    },
    {
      "epoch": 0.0478,
      "grad_norm": 2.6655406951904297,
      "learning_rate": 4.858163265306123e-05,
      "loss": 0.327,
      "step": 239
    },
    {
      "epoch": 0.048,
      "grad_norm": 5.939995288848877,
      "learning_rate": 4.8571428571428576e-05,
      "loss": 1.3496,
      "step": 240
    },
    {
      "epoch": 0.0482,
      "grad_norm": 36.7309684753418,
      "learning_rate": 4.856122448979592e-05,
      "loss": 4.0816,
      "step": 241
    },
    {
      "epoch": 0.0484,
      "grad_norm": 5.255819797515869,
      "learning_rate": 4.855102040816327e-05,
      "loss": 0.2191,
      "step": 242
    },
    {
      "epoch": 0.0486,
      "grad_norm": 24.00787925720215,
      "learning_rate": 4.854081632653062e-05,
      "loss": 3.8993,
      "step": 243
    },
    {
      "epoch": 0.0488,
      "grad_norm": 6.310253143310547,
      "learning_rate": 4.853061224489796e-05,
      "loss": 2.7461,
      "step": 244
    },
    {
      "epoch": 0.049,
      "grad_norm": 2.4759926795959473,
      "learning_rate": 4.852040816326531e-05,
      "loss": 0.312,
      "step": 245
    },
    {
      "epoch": 0.0492,
      "grad_norm": 10.137598037719727,
      "learning_rate": 4.851020408163266e-05,
      "loss": 0.6682,
      "step": 246
    },
    {
      "epoch": 0.0494,
      "grad_norm": 7.681159019470215,
      "learning_rate": 4.85e-05,
      "loss": 1.4522,
      "step": 247
    },
    {
      "epoch": 0.0496,
      "grad_norm": 7.122126579284668,
      "learning_rate": 4.848979591836735e-05,
      "loss": 1.34,
      "step": 248
    },
    {
      "epoch": 0.0498,
      "grad_norm": 3.2464118003845215,
      "learning_rate": 4.847959183673469e-05,
      "loss": 0.4424,
      "step": 249
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.2065978050231934,
      "learning_rate": 4.846938775510204e-05,
      "loss": 0.3443,
      "step": 250
    },
    {
      "epoch": 0.0502,
      "grad_norm": 12.5082426071167,
      "learning_rate": 4.845918367346939e-05,
      "loss": 1.12,
      "step": 251
    },
    {
      "epoch": 0.0504,
      "grad_norm": 18.6636962890625,
      "learning_rate": 4.844897959183673e-05,
      "loss": 4.2972,
      "step": 252
    },
    {
      "epoch": 0.0506,
      "grad_norm": 4.454346656799316,
      "learning_rate": 4.843877551020408e-05,
      "loss": 2.1657,
      "step": 253
    },
    {
      "epoch": 0.0508,
      "grad_norm": 9.119332313537598,
      "learning_rate": 4.842857142857143e-05,
      "loss": 1.2732,
      "step": 254
    },
    {
      "epoch": 0.051,
      "grad_norm": 2.056427240371704,
      "learning_rate": 4.841836734693877e-05,
      "loss": 0.3077,
      "step": 255
    },
    {
      "epoch": 0.0512,
      "grad_norm": 11.680416107177734,
      "learning_rate": 4.840816326530612e-05,
      "loss": 2.7134,
      "step": 256
    },
    {
      "epoch": 0.0514,
      "grad_norm": 3.6624484062194824,
      "learning_rate": 4.839795918367347e-05,
      "loss": 0.1659,
      "step": 257
    },
    {
      "epoch": 0.0516,
      "grad_norm": 31.827489852905273,
      "learning_rate": 4.8387755102040814e-05,
      "loss": 2.0273,
      "step": 258
    },
    {
      "epoch": 0.0518,
      "grad_norm": 115.13412475585938,
      "learning_rate": 4.837755102040817e-05,
      "loss": 5.7663,
      "step": 259
    },
    {
      "epoch": 0.052,
      "grad_norm": 6.4298810958862305,
      "learning_rate": 4.836734693877551e-05,
      "loss": 1.1484,
      "step": 260
    },
    {
      "epoch": 0.0522,
      "grad_norm": 28.639739990234375,
      "learning_rate": 4.835714285714286e-05,
      "loss": 2.9524,
      "step": 261
    },
    {
      "epoch": 0.0524,
      "grad_norm": 22.621858596801758,
      "learning_rate": 4.834693877551021e-05,
      "loss": 6.4986,
      "step": 262
    },
    {
      "epoch": 0.0526,
      "grad_norm": 12.89719009399414,
      "learning_rate": 4.833673469387755e-05,
      "loss": 6.5437,
      "step": 263
    },
    {
      "epoch": 0.0528,
      "grad_norm": 14.225550651550293,
      "learning_rate": 4.83265306122449e-05,
      "loss": 1.9275,
      "step": 264
    },
    {
      "epoch": 0.053,
      "grad_norm": 25.970439910888672,
      "learning_rate": 4.831632653061225e-05,
      "loss": 6.3776,
      "step": 265
    },
    {
      "epoch": 0.0532,
      "grad_norm": 4.544527053833008,
      "learning_rate": 4.8306122448979594e-05,
      "loss": 2.1623,
      "step": 266
    },
    {
      "epoch": 0.0534,
      "grad_norm": 5.257410049438477,
      "learning_rate": 4.829591836734694e-05,
      "loss": 0.2056,
      "step": 267
    },
    {
      "epoch": 0.0536,
      "grad_norm": 6.435934066772461,
      "learning_rate": 4.828571428571429e-05,
      "loss": 1.2436,
      "step": 268
    },
    {
      "epoch": 0.0538,
      "grad_norm": 2.09723162651062,
      "learning_rate": 4.8275510204081634e-05,
      "loss": 0.281,
      "step": 269
    },
    {
      "epoch": 0.054,
      "grad_norm": 2.2552835941314697,
      "learning_rate": 4.8265306122448984e-05,
      "loss": 0.3421,
      "step": 270
    },
    {
      "epoch": 0.0542,
      "grad_norm": 23.66434669494629,
      "learning_rate": 4.825510204081633e-05,
      "loss": 7.0027,
      "step": 271
    },
    {
      "epoch": 0.0544,
      "grad_norm": 9.996148109436035,
      "learning_rate": 4.8244897959183675e-05,
      "loss": 1.566,
      "step": 272
    },
    {
      "epoch": 0.0546,
      "grad_norm": 15.409833908081055,
      "learning_rate": 4.8234693877551024e-05,
      "loss": 1.5482,
      "step": 273
    },
    {
      "epoch": 0.0548,
      "grad_norm": 11.011585235595703,
      "learning_rate": 4.8224489795918373e-05,
      "loss": 1.4062,
      "step": 274
    },
    {
      "epoch": 0.055,
      "grad_norm": 4.346341609954834,
      "learning_rate": 4.8214285714285716e-05,
      "loss": 2.0908,
      "step": 275
    },
    {
      "epoch": 0.0552,
      "grad_norm": 30.071130752563477,
      "learning_rate": 4.8204081632653065e-05,
      "loss": 6.2823,
      "step": 276
    },
    {
      "epoch": 0.0554,
      "grad_norm": 4.043562889099121,
      "learning_rate": 4.8193877551020414e-05,
      "loss": 0.1602,
      "step": 277
    },
    {
      "epoch": 0.0556,
      "grad_norm": 14.959501266479492,
      "learning_rate": 4.818367346938776e-05,
      "loss": 1.2077,
      "step": 278
    },
    {
      "epoch": 0.0558,
      "grad_norm": 2.124439239501953,
      "learning_rate": 4.8173469387755106e-05,
      "loss": 0.2668,
      "step": 279
    },
    {
      "epoch": 0.056,
      "grad_norm": 33.473724365234375,
      "learning_rate": 4.816326530612245e-05,
      "loss": 6.1955,
      "step": 280
    },
    {
      "epoch": 0.0562,
      "grad_norm": 2.401884078979492,
      "learning_rate": 4.81530612244898e-05,
      "loss": 0.1679,
      "step": 281
    },
    {
      "epoch": 0.0564,
      "grad_norm": 14.710826873779297,
      "learning_rate": 4.8142857142857147e-05,
      "loss": 1.228,
      "step": 282
    },
    {
      "epoch": 0.0566,
      "grad_norm": 2.006013870239258,
      "learning_rate": 4.813265306122449e-05,
      "loss": 0.2674,
      "step": 283
    },
    {
      "epoch": 0.0568,
      "grad_norm": 2.3197853565216064,
      "learning_rate": 4.812244897959184e-05,
      "loss": 0.3204,
      "step": 284
    },
    {
      "epoch": 0.057,
      "grad_norm": 2.53511118888855,
      "learning_rate": 4.811224489795919e-05,
      "loss": 0.1583,
      "step": 285
    },
    {
      "epoch": 0.0572,
      "grad_norm": 2.7353696823120117,
      "learning_rate": 4.810204081632653e-05,
      "loss": 0.4002,
      "step": 286
    },
    {
      "epoch": 0.0574,
      "grad_norm": 2.9320266246795654,
      "learning_rate": 4.809183673469388e-05,
      "loss": 0.3796,
      "step": 287
    },
    {
      "epoch": 0.0576,
      "grad_norm": 6.147301197052002,
      "learning_rate": 4.808163265306123e-05,
      "loss": 2.6707,
      "step": 288
    },
    {
      "epoch": 0.0578,
      "grad_norm": 2.694920063018799,
      "learning_rate": 4.807142857142857e-05,
      "loss": 0.384,
      "step": 289
    },
    {
      "epoch": 0.058,
      "grad_norm": 12.652451515197754,
      "learning_rate": 4.806122448979592e-05,
      "loss": 2.7884,
      "step": 290
    },
    {
      "epoch": 0.0582,
      "grad_norm": 5.94713020324707,
      "learning_rate": 4.805102040816327e-05,
      "loss": 1.1616,
      "step": 291
    },
    {
      "epoch": 0.0584,
      "grad_norm": 2.474952459335327,
      "learning_rate": 4.804081632653061e-05,
      "loss": 0.3208,
      "step": 292
    },
    {
      "epoch": 0.0586,
      "grad_norm": 6.267833232879639,
      "learning_rate": 4.803061224489796e-05,
      "loss": 0.5163,
      "step": 293
    },
    {
      "epoch": 0.0588,
      "grad_norm": 4.102102279663086,
      "learning_rate": 4.802040816326531e-05,
      "loss": 0.1298,
      "step": 294
    },
    {
      "epoch": 0.059,
      "grad_norm": 4.8158135414123535,
      "learning_rate": 4.801020408163265e-05,
      "loss": 2.5453,
      "step": 295
    },
    {
      "epoch": 0.0592,
      "grad_norm": 4.911422252655029,
      "learning_rate": 4.8e-05,
      "loss": 0.4622,
      "step": 296
    },
    {
      "epoch": 0.0594,
      "grad_norm": 10.098528861999512,
      "learning_rate": 4.798979591836735e-05,
      "loss": 2.6648,
      "step": 297
    },
    {
      "epoch": 0.0596,
      "grad_norm": 6.3386640548706055,
      "learning_rate": 4.797959183673469e-05,
      "loss": 0.4538,
      "step": 298
    },
    {
      "epoch": 0.0598,
      "grad_norm": 6.769515514373779,
      "learning_rate": 4.796938775510204e-05,
      "loss": 1.3707,
      "step": 299
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.17781400680542,
      "learning_rate": 4.795918367346939e-05,
      "loss": 0.1102,
      "step": 300
    },
    {
      "epoch": 0.0602,
      "grad_norm": 3.4535491466522217,
      "learning_rate": 4.7948979591836734e-05,
      "loss": 0.1158,
      "step": 301
    },
    {
      "epoch": 0.0604,
      "grad_norm": 25.51836585998535,
      "learning_rate": 4.793877551020408e-05,
      "loss": 6.5621,
      "step": 302
    },
    {
      "epoch": 0.0606,
      "grad_norm": 3.729757308959961,
      "learning_rate": 4.7928571428571425e-05,
      "loss": 0.1819,
      "step": 303
    },
    {
      "epoch": 0.0608,
      "grad_norm": 2.9387378692626953,
      "learning_rate": 4.7918367346938774e-05,
      "loss": 0.2026,
      "step": 304
    },
    {
      "epoch": 0.061,
      "grad_norm": 2.278538703918457,
      "learning_rate": 4.790816326530613e-05,
      "loss": 0.1594,
      "step": 305
    },
    {
      "epoch": 0.0612,
      "grad_norm": 6.259002685546875,
      "learning_rate": 4.789795918367347e-05,
      "loss": 1.0925,
      "step": 306
    },
    {
      "epoch": 0.0614,
      "grad_norm": 13.990778923034668,
      "learning_rate": 4.788775510204082e-05,
      "loss": 6.5765,
      "step": 307
    },
    {
      "epoch": 0.0616,
      "grad_norm": 2.2181642055511475,
      "learning_rate": 4.787755102040817e-05,
      "loss": 0.3021,
      "step": 308
    },
    {
      "epoch": 0.0618,
      "grad_norm": 5.557339191436768,
      "learning_rate": 4.7867346938775513e-05,
      "loss": 2.5857,
      "step": 309
    },
    {
      "epoch": 0.062,
      "grad_norm": 3.1741716861724854,
      "learning_rate": 4.785714285714286e-05,
      "loss": 0.3761,
      "step": 310
    },
    {
      "epoch": 0.0622,
      "grad_norm": 7.62800407409668,
      "learning_rate": 4.7846938775510205e-05,
      "loss": 1.4748,
      "step": 311
    },
    {
      "epoch": 0.0624,
      "grad_norm": 52.55562210083008,
      "learning_rate": 4.7836734693877554e-05,
      "loss": 5.4467,
      "step": 312
    },
    {
      "epoch": 0.0626,
      "grad_norm": 2.609224557876587,
      "learning_rate": 4.7826530612244903e-05,
      "loss": 0.2835,
      "step": 313
    },
    {
      "epoch": 0.0628,
      "grad_norm": 13.866734504699707,
      "learning_rate": 4.7816326530612246e-05,
      "loss": 1.8371,
      "step": 314
    },
    {
      "epoch": 0.063,
      "grad_norm": 1.6872336864471436,
      "learning_rate": 4.7806122448979595e-05,
      "loss": 0.2887,
      "step": 315
    },
    {
      "epoch": 0.0632,
      "grad_norm": 9.54610824584961,
      "learning_rate": 4.7795918367346944e-05,
      "loss": 2.5558,
      "step": 316
    },
    {
      "epoch": 0.0634,
      "grad_norm": 5.848460674285889,
      "learning_rate": 4.7785714285714287e-05,
      "loss": 1.2751,
      "step": 317
    },
    {
      "epoch": 0.0636,
      "grad_norm": 11.767894744873047,
      "learning_rate": 4.7775510204081636e-05,
      "loss": 1.2123,
      "step": 318
    },
    {
      "epoch": 0.0638,
      "grad_norm": 2.920628786087036,
      "learning_rate": 4.7765306122448985e-05,
      "loss": 0.3535,
      "step": 319
    },
    {
      "epoch": 0.064,
      "grad_norm": 26.431793212890625,
      "learning_rate": 4.775510204081633e-05,
      "loss": 4.3793,
      "step": 320
    },
    {
      "epoch": 0.0642,
      "grad_norm": 3.030142307281494,
      "learning_rate": 4.7744897959183677e-05,
      "loss": 0.1938,
      "step": 321
    },
    {
      "epoch": 0.0644,
      "grad_norm": 7.9036126136779785,
      "learning_rate": 4.7734693877551026e-05,
      "loss": 2.4549,
      "step": 322
    },
    {
      "epoch": 0.0646,
      "grad_norm": 2.295531988143921,
      "learning_rate": 4.772448979591837e-05,
      "loss": 0.1455,
      "step": 323
    },
    {
      "epoch": 0.0648,
      "grad_norm": 2.797222137451172,
      "learning_rate": 4.771428571428572e-05,
      "loss": 0.3548,
      "step": 324
    },
    {
      "epoch": 0.065,
      "grad_norm": 4.948246002197266,
      "learning_rate": 4.7704081632653066e-05,
      "loss": 2.1088,
      "step": 325
    },
    {
      "epoch": 0.0652,
      "grad_norm": 3.0309126377105713,
      "learning_rate": 4.769387755102041e-05,
      "loss": 0.3759,
      "step": 326
    },
    {
      "epoch": 0.0654,
      "grad_norm": 2.7186086177825928,
      "learning_rate": 4.768367346938776e-05,
      "loss": 0.3967,
      "step": 327
    },
    {
      "epoch": 0.0656,
      "grad_norm": 2.1337838172912598,
      "learning_rate": 4.767346938775511e-05,
      "loss": 0.1388,
      "step": 328
    },
    {
      "epoch": 0.0658,
      "grad_norm": 7.373128890991211,
      "learning_rate": 4.766326530612245e-05,
      "loss": 1.0983,
      "step": 329
    },
    {
      "epoch": 0.066,
      "grad_norm": 4.437475204467773,
      "learning_rate": 4.76530612244898e-05,
      "loss": 2.0878,
      "step": 330
    },
    {
      "epoch": 0.0662,
      "grad_norm": 21.72403907775879,
      "learning_rate": 4.764285714285715e-05,
      "loss": 1.1818,
      "step": 331
    },
    {
      "epoch": 0.0664,
      "grad_norm": 5.763108730316162,
      "learning_rate": 4.763265306122449e-05,
      "loss": 1.274,
      "step": 332
    },
    {
      "epoch": 0.0666,
      "grad_norm": 13.025044441223145,
      "learning_rate": 4.762244897959184e-05,
      "loss": 1.723,
      "step": 333
    },
    {
      "epoch": 0.0668,
      "grad_norm": 3.968433380126953,
      "learning_rate": 4.761224489795918e-05,
      "loss": 2.1153,
      "step": 334
    },
    {
      "epoch": 0.067,
      "grad_norm": 7.378103733062744,
      "learning_rate": 4.760204081632653e-05,
      "loss": 0.4202,
      "step": 335
    },
    {
      "epoch": 0.0672,
      "grad_norm": 2.830109119415283,
      "learning_rate": 4.759183673469388e-05,
      "loss": 0.3293,
      "step": 336
    },
    {
      "epoch": 0.0674,
      "grad_norm": 3.178539514541626,
      "learning_rate": 4.758163265306122e-05,
      "loss": 0.3486,
      "step": 337
    },
    {
      "epoch": 0.0676,
      "grad_norm": 7.1868181228637695,
      "learning_rate": 4.757142857142857e-05,
      "loss": 1.1632,
      "step": 338
    },
    {
      "epoch": 0.0678,
      "grad_norm": 2.4396426677703857,
      "learning_rate": 4.756122448979592e-05,
      "loss": 0.1441,
      "step": 339
    },
    {
      "epoch": 0.068,
      "grad_norm": 15.88539981842041,
      "learning_rate": 4.7551020408163263e-05,
      "loss": 2.5294,
      "step": 340
    },
    {
      "epoch": 0.0682,
      "grad_norm": 16.781665802001953,
      "learning_rate": 4.754081632653061e-05,
      "loss": 2.8364,
      "step": 341
    },
    {
      "epoch": 0.0684,
      "grad_norm": 10.077573776245117,
      "learning_rate": 4.753061224489796e-05,
      "loss": 0.8778,
      "step": 342
    },
    {
      "epoch": 0.0686,
      "grad_norm": 5.763261318206787,
      "learning_rate": 4.7520408163265304e-05,
      "loss": 2.6801,
      "step": 343
    },
    {
      "epoch": 0.0688,
      "grad_norm": 2.591111660003662,
      "learning_rate": 4.7510204081632653e-05,
      "loss": 0.341,
      "step": 344
    },
    {
      "epoch": 0.069,
      "grad_norm": 5.386274337768555,
      "learning_rate": 4.75e-05,
      "loss": 2.5503,
      "step": 345
    },
    {
      "epoch": 0.0692,
      "grad_norm": 24.671098709106445,
      "learning_rate": 4.7489795918367345e-05,
      "loss": 6.5451,
      "step": 346
    },
    {
      "epoch": 0.0694,
      "grad_norm": 1.994725227355957,
      "learning_rate": 4.7479591836734694e-05,
      "loss": 0.2798,
      "step": 347
    },
    {
      "epoch": 0.0696,
      "grad_norm": 2.557914972305298,
      "learning_rate": 4.746938775510204e-05,
      "loss": 0.3585,
      "step": 348
    },
    {
      "epoch": 0.0698,
      "grad_norm": 34.20411682128906,
      "learning_rate": 4.7459183673469386e-05,
      "loss": 5.0002,
      "step": 349
    },
    {
      "epoch": 0.07,
      "grad_norm": 17.687604904174805,
      "learning_rate": 4.744897959183674e-05,
      "loss": 6.5777,
      "step": 350
    },
    {
      "epoch": 0.0702,
      "grad_norm": 7.226384162902832,
      "learning_rate": 4.7438775510204084e-05,
      "loss": 1.1322,
      "step": 351
    },
    {
      "epoch": 0.0704,
      "grad_norm": 5.94661283493042,
      "learning_rate": 4.742857142857143e-05,
      "loss": 1.0266,
      "step": 352
    },
    {
      "epoch": 0.0706,
      "grad_norm": 2.209641218185425,
      "learning_rate": 4.741836734693878e-05,
      "loss": 0.1366,
      "step": 353
    },
    {
      "epoch": 0.0708,
      "grad_norm": 3.1439688205718994,
      "learning_rate": 4.7408163265306125e-05,
      "loss": 0.1135,
      "step": 354
    },
    {
      "epoch": 0.071,
      "grad_norm": 2.3960354328155518,
      "learning_rate": 4.7397959183673474e-05,
      "loss": 0.2808,
      "step": 355
    },
    {
      "epoch": 0.0712,
      "grad_norm": 2.367637872695923,
      "learning_rate": 4.738775510204082e-05,
      "loss": 0.1627,
      "step": 356
    },
    {
      "epoch": 0.0714,
      "grad_norm": 2.480921745300293,
      "learning_rate": 4.7377551020408166e-05,
      "loss": 0.3506,
      "step": 357
    },
    {
      "epoch": 0.0716,
      "grad_norm": 8.608354568481445,
      "learning_rate": 4.7367346938775515e-05,
      "loss": 0.4904,
      "step": 358
    },
    {
      "epoch": 0.0718,
      "grad_norm": 2.8946304321289062,
      "learning_rate": 4.7357142857142864e-05,
      "loss": 0.1005,
      "step": 359
    },
    {
      "epoch": 0.072,
      "grad_norm": 2.059964656829834,
      "learning_rate": 4.7346938775510206e-05,
      "loss": 0.1328,
      "step": 360
    },
    {
      "epoch": 0.0722,
      "grad_norm": 2.6717636585235596,
      "learning_rate": 4.7336734693877556e-05,
      "loss": 0.3791,
      "step": 361
    },
    {
      "epoch": 0.0724,
      "grad_norm": 2.0760722160339355,
      "learning_rate": 4.7326530612244905e-05,
      "loss": 0.1078,
      "step": 362
    },
    {
      "epoch": 0.0726,
      "grad_norm": 53.7850341796875,
      "learning_rate": 4.731632653061225e-05,
      "loss": 2.9198,
      "step": 363
    },
    {
      "epoch": 0.0728,
      "grad_norm": 2.0827722549438477,
      "learning_rate": 4.7306122448979596e-05,
      "loss": 0.2764,
      "step": 364
    },
    {
      "epoch": 0.073,
      "grad_norm": 9.983550071716309,
      "learning_rate": 4.729591836734694e-05,
      "loss": 2.7176,
      "step": 365
    },
    {
      "epoch": 0.0732,
      "grad_norm": 13.311184883117676,
      "learning_rate": 4.728571428571429e-05,
      "loss": 2.3279,
      "step": 366
    },
    {
      "epoch": 0.0734,
      "grad_norm": 26.359949111938477,
      "learning_rate": 4.727551020408164e-05,
      "loss": 2.3328,
      "step": 367
    },
    {
      "epoch": 0.0736,
      "grad_norm": 5.651286602020264,
      "learning_rate": 4.726530612244898e-05,
      "loss": 0.3716,
      "step": 368
    },
    {
      "epoch": 0.0738,
      "grad_norm": 1.93024480342865,
      "learning_rate": 4.725510204081633e-05,
      "loss": 0.2244,
      "step": 369
    },
    {
      "epoch": 0.074,
      "grad_norm": 6.500854969024658,
      "learning_rate": 4.724489795918368e-05,
      "loss": 0.3689,
      "step": 370
    },
    {
      "epoch": 0.0742,
      "grad_norm": 6.808442115783691,
      "learning_rate": 4.723469387755102e-05,
      "loss": 0.3618,
      "step": 371
    },
    {
      "epoch": 0.0744,
      "grad_norm": 2.819575071334839,
      "learning_rate": 4.722448979591837e-05,
      "loss": 0.1492,
      "step": 372
    },
    {
      "epoch": 0.0746,
      "grad_norm": 2.6610302925109863,
      "learning_rate": 4.721428571428572e-05,
      "loss": 0.2873,
      "step": 373
    },
    {
      "epoch": 0.0748,
      "grad_norm": 27.070655822753906,
      "learning_rate": 4.720408163265306e-05,
      "loss": 3.0837,
      "step": 374
    },
    {
      "epoch": 0.075,
      "grad_norm": 2.9257850646972656,
      "learning_rate": 4.719387755102041e-05,
      "loss": 0.2015,
      "step": 375
    },
    {
      "epoch": 0.0752,
      "grad_norm": 1.9699163436889648,
      "learning_rate": 4.718367346938776e-05,
      "loss": 0.2495,
      "step": 376
    },
    {
      "epoch": 0.0754,
      "grad_norm": 31.752384185791016,
      "learning_rate": 4.71734693877551e-05,
      "loss": 4.9408,
      "step": 377
    },
    {
      "epoch": 0.0756,
      "grad_norm": 1.8258370161056519,
      "learning_rate": 4.716326530612245e-05,
      "loss": 0.254,
      "step": 378
    },
    {
      "epoch": 0.0758,
      "grad_norm": 18.534879684448242,
      "learning_rate": 4.71530612244898e-05,
      "loss": 3.016,
      "step": 379
    },
    {
      "epoch": 0.076,
      "grad_norm": 1.9348527193069458,
      "learning_rate": 4.714285714285714e-05,
      "loss": 0.1144,
      "step": 380
    },
    {
      "epoch": 0.0762,
      "grad_norm": 6.82661247253418,
      "learning_rate": 4.713265306122449e-05,
      "loss": 2.5575,
      "step": 381
    },
    {
      "epoch": 0.0764,
      "grad_norm": 6.1823883056640625,
      "learning_rate": 4.712244897959184e-05,
      "loss": 2.626,
      "step": 382
    },
    {
      "epoch": 0.0766,
      "grad_norm": 11.660492897033691,
      "learning_rate": 4.711224489795918e-05,
      "loss": 0.9745,
      "step": 383
    },
    {
      "epoch": 0.0768,
      "grad_norm": 2.0270681381225586,
      "learning_rate": 4.710204081632653e-05,
      "loss": 0.2937,
      "step": 384
    },
    {
      "epoch": 0.077,
      "grad_norm": 56.424224853515625,
      "learning_rate": 4.7091836734693875e-05,
      "loss": 6.583,
      "step": 385
    },
    {
      "epoch": 0.0772,
      "grad_norm": 8.324662208557129,
      "learning_rate": 4.7081632653061224e-05,
      "loss": 1.3856,
      "step": 386
    },
    {
      "epoch": 0.0774,
      "grad_norm": 8.9666748046875,
      "learning_rate": 4.707142857142857e-05,
      "loss": 2.572,
      "step": 387
    },
    {
      "epoch": 0.0776,
      "grad_norm": 2.187340259552002,
      "learning_rate": 4.7061224489795916e-05,
      "loss": 0.2333,
      "step": 388
    },
    {
      "epoch": 0.0778,
      "grad_norm": 13.370061874389648,
      "learning_rate": 4.7051020408163265e-05,
      "loss": 1.9124,
      "step": 389
    },
    {
      "epoch": 0.078,
      "grad_norm": 2.4363479614257812,
      "learning_rate": 4.7040816326530614e-05,
      "loss": 0.1063,
      "step": 390
    },
    {
      "epoch": 0.0782,
      "grad_norm": 9.3132963180542,
      "learning_rate": 4.7030612244897956e-05,
      "loss": 2.6552,
      "step": 391
    },
    {
      "epoch": 0.0784,
      "grad_norm": 55.959781646728516,
      "learning_rate": 4.7020408163265306e-05,
      "loss": 6.6248,
      "step": 392
    },
    {
      "epoch": 0.0786,
      "grad_norm": 18.18315315246582,
      "learning_rate": 4.7010204081632655e-05,
      "loss": 2.1259,
      "step": 393
    },
    {
      "epoch": 0.0788,
      "grad_norm": 25.41035270690918,
      "learning_rate": 4.7e-05,
      "loss": 6.2054,
      "step": 394
    },
    {
      "epoch": 0.079,
      "grad_norm": 2.9915242195129395,
      "learning_rate": 4.698979591836735e-05,
      "loss": 0.3402,
      "step": 395
    },
    {
      "epoch": 0.0792,
      "grad_norm": 5.823123455047607,
      "learning_rate": 4.6979591836734696e-05,
      "loss": 1.2686,
      "step": 396
    },
    {
      "epoch": 0.0794,
      "grad_norm": 2.028001308441162,
      "learning_rate": 4.6969387755102045e-05,
      "loss": 0.2905,
      "step": 397
    },
    {
      "epoch": 0.0796,
      "grad_norm": 42.19456100463867,
      "learning_rate": 4.6959183673469394e-05,
      "loss": 6.2213,
      "step": 398
    },
    {
      "epoch": 0.0798,
      "grad_norm": 2.543256998062134,
      "learning_rate": 4.6948979591836736e-05,
      "loss": 0.3686,
      "step": 399
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.2356905937194824,
      "learning_rate": 4.6938775510204086e-05,
      "loss": 0.1324,
      "step": 400
    },
    {
      "epoch": 0.0802,
      "grad_norm": 2.372619867324829,
      "learning_rate": 4.6928571428571435e-05,
      "loss": 0.3644,
      "step": 401
    },
    {
      "epoch": 0.0804,
      "grad_norm": 1.9613386392593384,
      "learning_rate": 4.691836734693878e-05,
      "loss": 0.241,
      "step": 402
    },
    {
      "epoch": 0.0806,
      "grad_norm": 8.247029304504395,
      "learning_rate": 4.6908163265306126e-05,
      "loss": 6.9809,
      "step": 403
    },
    {
      "epoch": 0.0808,
      "grad_norm": 8.159348487854004,
      "learning_rate": 4.6897959183673475e-05,
      "loss": 1.5399,
      "step": 404
    },
    {
      "epoch": 0.081,
      "grad_norm": 10.0965576171875,
      "learning_rate": 4.688775510204082e-05,
      "loss": 2.3527,
      "step": 405
    },
    {
      "epoch": 0.0812,
      "grad_norm": 5.929891109466553,
      "learning_rate": 4.687755102040817e-05,
      "loss": 0.9952,
      "step": 406
    },
    {
      "epoch": 0.0814,
      "grad_norm": 2.4128918647766113,
      "learning_rate": 4.6867346938775516e-05,
      "loss": 0.3466,
      "step": 407
    },
    {
      "epoch": 0.0816,
      "grad_norm": 1.925405502319336,
      "learning_rate": 4.685714285714286e-05,
      "loss": 0.2287,
      "step": 408
    },
    {
      "epoch": 0.0818,
      "grad_norm": 7.814586162567139,
      "learning_rate": 4.684693877551021e-05,
      "loss": 1.0657,
      "step": 409
    },
    {
      "epoch": 0.082,
      "grad_norm": 2.403630495071411,
      "learning_rate": 4.683673469387756e-05,
      "loss": 0.3243,
      "step": 410
    },
    {
      "epoch": 0.0822,
      "grad_norm": 9.4711275100708,
      "learning_rate": 4.68265306122449e-05,
      "loss": 1.6438,
      "step": 411
    },
    {
      "epoch": 0.0824,
      "grad_norm": 48.67058181762695,
      "learning_rate": 4.681632653061225e-05,
      "loss": 5.4859,
      "step": 412
    },
    {
      "epoch": 0.0826,
      "grad_norm": 2.594776153564453,
      "learning_rate": 4.68061224489796e-05,
      "loss": 0.1547,
      "step": 413
    },
    {
      "epoch": 0.0828,
      "grad_norm": 49.50006103515625,
      "learning_rate": 4.679591836734694e-05,
      "loss": 6.3499,
      "step": 414
    },
    {
      "epoch": 0.083,
      "grad_norm": 43.77144241333008,
      "learning_rate": 4.678571428571429e-05,
      "loss": 7.4436,
      "step": 415
    },
    {
      "epoch": 0.0832,
      "grad_norm": 1.756616234779358,
      "learning_rate": 4.677551020408163e-05,
      "loss": 0.2482,
      "step": 416
    },
    {
      "epoch": 0.0834,
      "grad_norm": 10.484140396118164,
      "learning_rate": 4.676530612244898e-05,
      "loss": 1.5785,
      "step": 417
    },
    {
      "epoch": 0.0836,
      "grad_norm": 9.244158744812012,
      "learning_rate": 4.675510204081633e-05,
      "loss": 0.8072,
      "step": 418
    },
    {
      "epoch": 0.0838,
      "grad_norm": 8.533461570739746,
      "learning_rate": 4.674489795918367e-05,
      "loss": 1.5188,
      "step": 419
    },
    {
      "epoch": 0.084,
      "grad_norm": 2.3646106719970703,
      "learning_rate": 4.673469387755102e-05,
      "loss": 0.1298,
      "step": 420
    },
    {
      "epoch": 0.0842,
      "grad_norm": 5.251696586608887,
      "learning_rate": 4.672448979591837e-05,
      "loss": 0.3743,
      "step": 421
    },
    {
      "epoch": 0.0844,
      "grad_norm": 2.401188611984253,
      "learning_rate": 4.671428571428571e-05,
      "loss": 0.3042,
      "step": 422
    },
    {
      "epoch": 0.0846,
      "grad_norm": 4.875900745391846,
      "learning_rate": 4.670408163265306e-05,
      "loss": 2.55,
      "step": 423
    },
    {
      "epoch": 0.0848,
      "grad_norm": 6.620153903961182,
      "learning_rate": 4.669387755102041e-05,
      "loss": 1.0621,
      "step": 424
    },
    {
      "epoch": 0.085,
      "grad_norm": 2.050922155380249,
      "learning_rate": 4.6683673469387754e-05,
      "loss": 0.1177,
      "step": 425
    },
    {
      "epoch": 0.0852,
      "grad_norm": 4.2134222984313965,
      "learning_rate": 4.66734693877551e-05,
      "loss": 2.5296,
      "step": 426
    },
    {
      "epoch": 0.0854,
      "grad_norm": 2.0711166858673096,
      "learning_rate": 4.666326530612245e-05,
      "loss": 0.1249,
      "step": 427
    },
    {
      "epoch": 0.0856,
      "grad_norm": 6.235434055328369,
      "learning_rate": 4.6653061224489795e-05,
      "loss": 1.0692,
      "step": 428
    },
    {
      "epoch": 0.0858,
      "grad_norm": 6.365238189697266,
      "learning_rate": 4.6642857142857144e-05,
      "loss": 1.1687,
      "step": 429
    },
    {
      "epoch": 0.086,
      "grad_norm": 9.341511726379395,
      "learning_rate": 4.663265306122449e-05,
      "loss": 0.3599,
      "step": 430
    },
    {
      "epoch": 0.0862,
      "grad_norm": 11.925809860229492,
      "learning_rate": 4.6622448979591836e-05,
      "loss": 2.4904,
      "step": 431
    },
    {
      "epoch": 0.0864,
      "grad_norm": 10.085716247558594,
      "learning_rate": 4.6612244897959185e-05,
      "loss": 2.4821,
      "step": 432
    },
    {
      "epoch": 0.0866,
      "grad_norm": 39.38417434692383,
      "learning_rate": 4.6602040816326534e-05,
      "loss": 5.7205,
      "step": 433
    },
    {
      "epoch": 0.0868,
      "grad_norm": 5.331692695617676,
      "learning_rate": 4.6591836734693876e-05,
      "loss": 2.591,
      "step": 434
    },
    {
      "epoch": 0.087,
      "grad_norm": 7.388650417327881,
      "learning_rate": 4.6581632653061226e-05,
      "loss": 1.1156,
      "step": 435
    },
    {
      "epoch": 0.0872,
      "grad_norm": 21.16617774963379,
      "learning_rate": 4.6571428571428575e-05,
      "loss": 6.2198,
      "step": 436
    },
    {
      "epoch": 0.0874,
      "grad_norm": 4.8475213050842285,
      "learning_rate": 4.656122448979592e-05,
      "loss": 0.2768,
      "step": 437
    },
    {
      "epoch": 0.0876,
      "grad_norm": 4.950862407684326,
      "learning_rate": 4.6551020408163266e-05,
      "loss": 1.0098,
      "step": 438
    },
    {
      "epoch": 0.0878,
      "grad_norm": 4.628479957580566,
      "learning_rate": 4.654081632653061e-05,
      "loss": 2.423,
      "step": 439
    },
    {
      "epoch": 0.088,
      "grad_norm": 4.182015419006348,
      "learning_rate": 4.653061224489796e-05,
      "loss": 2.3441,
      "step": 440
    },
    {
      "epoch": 0.0882,
      "grad_norm": 2.807716131210327,
      "learning_rate": 4.6520408163265314e-05,
      "loss": 0.5297,
      "step": 441
    },
    {
      "epoch": 0.0884,
      "grad_norm": 5.62799072265625,
      "learning_rate": 4.6510204081632656e-05,
      "loss": 1.1379,
      "step": 442
    },
    {
      "epoch": 0.0886,
      "grad_norm": 2.8704583644866943,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 0.3582,
      "step": 443
    },
    {
      "epoch": 0.0888,
      "grad_norm": 1.8881853818893433,
      "learning_rate": 4.6489795918367355e-05,
      "loss": 0.2726,
      "step": 444
    },
    {
      "epoch": 0.089,
      "grad_norm": 1.7147490978240967,
      "learning_rate": 4.64795918367347e-05,
      "loss": 0.0869,
      "step": 445
    },
    {
      "epoch": 0.0892,
      "grad_norm": 5.187413692474365,
      "learning_rate": 4.6469387755102046e-05,
      "loss": 0.9787,
      "step": 446
    },
    {
      "epoch": 0.0894,
      "grad_norm": 2.709937334060669,
      "learning_rate": 4.645918367346939e-05,
      "loss": 0.341,
      "step": 447
    },
    {
      "epoch": 0.0896,
      "grad_norm": 1.8514126539230347,
      "learning_rate": 4.644897959183674e-05,
      "loss": 0.2127,
      "step": 448
    },
    {
      "epoch": 0.0898,
      "grad_norm": 5.530944347381592,
      "learning_rate": 4.643877551020409e-05,
      "loss": 1.0078,
      "step": 449
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.764533996582031,
      "learning_rate": 4.642857142857143e-05,
      "loss": 2.3106,
      "step": 450
    },
    {
      "epoch": 0.0902,
      "grad_norm": 16.673175811767578,
      "learning_rate": 4.641836734693878e-05,
      "loss": 3.5456,
      "step": 451
    },
    {
      "epoch": 0.0904,
      "grad_norm": 2.0734784603118896,
      "learning_rate": 4.640816326530613e-05,
      "loss": 0.2862,
      "step": 452
    },
    {
      "epoch": 0.0906,
      "grad_norm": 21.004438400268555,
      "learning_rate": 4.639795918367347e-05,
      "loss": 5.2117,
      "step": 453
    },
    {
      "epoch": 0.0908,
      "grad_norm": 8.92627239227295,
      "learning_rate": 4.638775510204082e-05,
      "loss": 1.619,
      "step": 454
    },
    {
      "epoch": 0.091,
      "grad_norm": 2.5494918823242188,
      "learning_rate": 4.637755102040817e-05,
      "loss": 0.3257,
      "step": 455
    },
    {
      "epoch": 0.0912,
      "grad_norm": 19.945465087890625,
      "learning_rate": 4.636734693877551e-05,
      "loss": 2.0669,
      "step": 456
    },
    {
      "epoch": 0.0914,
      "grad_norm": 1.8492319583892822,
      "learning_rate": 4.635714285714286e-05,
      "loss": 0.3081,
      "step": 457
    },
    {
      "epoch": 0.0916,
      "grad_norm": 1.979044795036316,
      "learning_rate": 4.634693877551021e-05,
      "loss": 0.2047,
      "step": 458
    },
    {
      "epoch": 0.0918,
      "grad_norm": 2.6501336097717285,
      "learning_rate": 4.633673469387755e-05,
      "loss": 0.1064,
      "step": 459
    },
    {
      "epoch": 0.092,
      "grad_norm": 6.860795021057129,
      "learning_rate": 4.63265306122449e-05,
      "loss": 0.9122,
      "step": 460
    },
    {
      "epoch": 0.0922,
      "grad_norm": 12.75924301147461,
      "learning_rate": 4.631632653061225e-05,
      "loss": 2.1535,
      "step": 461
    },
    {
      "epoch": 0.0924,
      "grad_norm": 2.1344172954559326,
      "learning_rate": 4.630612244897959e-05,
      "loss": 0.1228,
      "step": 462
    },
    {
      "epoch": 0.0926,
      "grad_norm": 2.5231246948242188,
      "learning_rate": 4.629591836734694e-05,
      "loss": 0.1514,
      "step": 463
    },
    {
      "epoch": 0.0928,
      "grad_norm": 2.375951051712036,
      "learning_rate": 4.628571428571429e-05,
      "loss": 0.3321,
      "step": 464
    },
    {
      "epoch": 0.093,
      "grad_norm": 3.2040934562683105,
      "learning_rate": 4.627551020408163e-05,
      "loss": 0.1209,
      "step": 465
    },
    {
      "epoch": 0.0932,
      "grad_norm": 6.255486011505127,
      "learning_rate": 4.626530612244898e-05,
      "loss": 2.4753,
      "step": 466
    },
    {
      "epoch": 0.0934,
      "grad_norm": 12.446625709533691,
      "learning_rate": 4.625510204081633e-05,
      "loss": 1.9803,
      "step": 467
    },
    {
      "epoch": 0.0936,
      "grad_norm": 10.153313636779785,
      "learning_rate": 4.6244897959183674e-05,
      "loss": 1.6769,
      "step": 468
    },
    {
      "epoch": 0.0938,
      "grad_norm": 17.7565975189209,
      "learning_rate": 4.623469387755102e-05,
      "loss": 1.7344,
      "step": 469
    },
    {
      "epoch": 0.094,
      "grad_norm": 13.362611770629883,
      "learning_rate": 4.6224489795918366e-05,
      "loss": 1.4584,
      "step": 470
    },
    {
      "epoch": 0.0942,
      "grad_norm": 1.8224835395812988,
      "learning_rate": 4.6214285714285715e-05,
      "loss": 0.1938,
      "step": 471
    },
    {
      "epoch": 0.0944,
      "grad_norm": 2.838259696960449,
      "learning_rate": 4.6204081632653064e-05,
      "loss": 0.1336,
      "step": 472
    },
    {
      "epoch": 0.0946,
      "grad_norm": 2.060246706008911,
      "learning_rate": 4.6193877551020406e-05,
      "loss": 0.2559,
      "step": 473
    },
    {
      "epoch": 0.0948,
      "grad_norm": 2.429246664047241,
      "learning_rate": 4.6183673469387755e-05,
      "loss": 0.14,
      "step": 474
    },
    {
      "epoch": 0.095,
      "grad_norm": 10.238293647766113,
      "learning_rate": 4.6173469387755105e-05,
      "loss": 2.5295,
      "step": 475
    },
    {
      "epoch": 0.0952,
      "grad_norm": 5.1625285148620605,
      "learning_rate": 4.616326530612245e-05,
      "loss": 0.3549,
      "step": 476
    },
    {
      "epoch": 0.0954,
      "grad_norm": 2.8266282081604004,
      "learning_rate": 4.6153061224489796e-05,
      "loss": 0.0985,
      "step": 477
    },
    {
      "epoch": 0.0956,
      "grad_norm": 14.649356842041016,
      "learning_rate": 4.6142857142857145e-05,
      "loss": 4.0913,
      "step": 478
    },
    {
      "epoch": 0.0958,
      "grad_norm": 2.2226009368896484,
      "learning_rate": 4.613265306122449e-05,
      "loss": 0.2541,
      "step": 479
    },
    {
      "epoch": 0.096,
      "grad_norm": 4.22063684463501,
      "learning_rate": 4.612244897959184e-05,
      "loss": 0.4131,
      "step": 480
    },
    {
      "epoch": 0.0962,
      "grad_norm": 5.823008060455322,
      "learning_rate": 4.6112244897959186e-05,
      "loss": 1.0252,
      "step": 481
    },
    {
      "epoch": 0.0964,
      "grad_norm": 10.060904502868652,
      "learning_rate": 4.610204081632653e-05,
      "loss": 1.5716,
      "step": 482
    },
    {
      "epoch": 0.0966,
      "grad_norm": 2.1902894973754883,
      "learning_rate": 4.609183673469388e-05,
      "loss": 0.1208,
      "step": 483
    },
    {
      "epoch": 0.0968,
      "grad_norm": 15.719635963439941,
      "learning_rate": 4.608163265306123e-05,
      "loss": 7.0908,
      "step": 484
    },
    {
      "epoch": 0.097,
      "grad_norm": 8.193558692932129,
      "learning_rate": 4.607142857142857e-05,
      "loss": 1.26,
      "step": 485
    },
    {
      "epoch": 0.0972,
      "grad_norm": 6.4959306716918945,
      "learning_rate": 4.6061224489795925e-05,
      "loss": 0.9872,
      "step": 486
    },
    {
      "epoch": 0.0974,
      "grad_norm": 6.141908168792725,
      "learning_rate": 4.605102040816327e-05,
      "loss": 1.0136,
      "step": 487
    },
    {
      "epoch": 0.0976,
      "grad_norm": 32.68397903442383,
      "learning_rate": 4.604081632653062e-05,
      "loss": 4.1119,
      "step": 488
    },
    {
      "epoch": 0.0978,
      "grad_norm": 5.456779956817627,
      "learning_rate": 4.6030612244897966e-05,
      "loss": 1.3421,
      "step": 489
    },
    {
      "epoch": 0.098,
      "grad_norm": 1.829673409461975,
      "learning_rate": 4.602040816326531e-05,
      "loss": 0.277,
      "step": 490
    },
    {
      "epoch": 0.0982,
      "grad_norm": 4.504250526428223,
      "learning_rate": 4.601020408163266e-05,
      "loss": 2.4248,
      "step": 491
    },
    {
      "epoch": 0.0984,
      "grad_norm": 4.908005714416504,
      "learning_rate": 4.600000000000001e-05,
      "loss": 2.3024,
      "step": 492
    },
    {
      "epoch": 0.0986,
      "grad_norm": 4.181675434112549,
      "learning_rate": 4.598979591836735e-05,
      "loss": 2.4278,
      "step": 493
    },
    {
      "epoch": 0.0988,
      "grad_norm": 2.944359064102173,
      "learning_rate": 4.59795918367347e-05,
      "loss": 0.4215,
      "step": 494
    },
    {
      "epoch": 0.099,
      "grad_norm": 34.008079528808594,
      "learning_rate": 4.596938775510205e-05,
      "loss": 4.4627,
      "step": 495
    },
    {
      "epoch": 0.0992,
      "grad_norm": 2.167393922805786,
      "learning_rate": 4.595918367346939e-05,
      "loss": 0.0745,
      "step": 496
    },
    {
      "epoch": 0.0994,
      "grad_norm": 2.0184450149536133,
      "learning_rate": 4.594897959183674e-05,
      "loss": 0.3072,
      "step": 497
    },
    {
      "epoch": 0.0996,
      "grad_norm": 7.414123058319092,
      "learning_rate": 4.593877551020409e-05,
      "loss": 0.985,
      "step": 498
    },
    {
      "epoch": 0.0998,
      "grad_norm": 2.57053804397583,
      "learning_rate": 4.592857142857143e-05,
      "loss": 0.3009,
      "step": 499
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.8467835187911987,
      "learning_rate": 4.591836734693878e-05,
      "loss": 0.2618,
      "step": 500
    },
    {
      "epoch": 0.1002,
      "grad_norm": 6.0907111167907715,
      "learning_rate": 4.590816326530612e-05,
      "loss": 0.9765,
      "step": 501
    },
    {
      "epoch": 0.1004,
      "grad_norm": 2.033963441848755,
      "learning_rate": 4.589795918367347e-05,
      "loss": 0.1242,
      "step": 502
    },
    {
      "epoch": 0.1006,
      "grad_norm": 1.8364888429641724,
      "learning_rate": 4.588775510204082e-05,
      "loss": 0.2615,
      "step": 503
    },
    {
      "epoch": 0.1008,
      "grad_norm": 5.7680864334106445,
      "learning_rate": 4.587755102040816e-05,
      "loss": 0.9865,
      "step": 504
    },
    {
      "epoch": 0.101,
      "grad_norm": 7.0686116218566895,
      "learning_rate": 4.586734693877551e-05,
      "loss": 0.7085,
      "step": 505
    },
    {
      "epoch": 0.1012,
      "grad_norm": 2.136955499649048,
      "learning_rate": 4.585714285714286e-05,
      "loss": 0.0715,
      "step": 506
    },
    {
      "epoch": 0.1014,
      "grad_norm": 1.9805129766464233,
      "learning_rate": 4.5846938775510204e-05,
      "loss": 0.1,
      "step": 507
    },
    {
      "epoch": 0.1016,
      "grad_norm": 8.698697090148926,
      "learning_rate": 4.583673469387755e-05,
      "loss": 1.4424,
      "step": 508
    },
    {
      "epoch": 0.1018,
      "grad_norm": 7.782326698303223,
      "learning_rate": 4.58265306122449e-05,
      "loss": 1.4647,
      "step": 509
    },
    {
      "epoch": 0.102,
      "grad_norm": 41.67585754394531,
      "learning_rate": 4.5816326530612245e-05,
      "loss": 6.3981,
      "step": 510
    },
    {
      "epoch": 0.1022,
      "grad_norm": 1.8698790073394775,
      "learning_rate": 4.5806122448979594e-05,
      "loss": 0.1129,
      "step": 511
    },
    {
      "epoch": 0.1024,
      "grad_norm": 4.760601043701172,
      "learning_rate": 4.579591836734694e-05,
      "loss": 0.3221,
      "step": 512
    },
    {
      "epoch": 0.1026,
      "grad_norm": 33.31980895996094,
      "learning_rate": 4.5785714285714285e-05,
      "loss": 1.7785,
      "step": 513
    },
    {
      "epoch": 0.1028,
      "grad_norm": 17.92771339416504,
      "learning_rate": 4.5775510204081635e-05,
      "loss": 5.4287,
      "step": 514
    },
    {
      "epoch": 0.103,
      "grad_norm": 2.300755023956299,
      "learning_rate": 4.5765306122448984e-05,
      "loss": 0.123,
      "step": 515
    },
    {
      "epoch": 0.1032,
      "grad_norm": 4.174624443054199,
      "learning_rate": 4.5755102040816326e-05,
      "loss": 0.3054,
      "step": 516
    },
    {
      "epoch": 0.1034,
      "grad_norm": 2.479468822479248,
      "learning_rate": 4.5744897959183675e-05,
      "loss": 0.3015,
      "step": 517
    },
    {
      "epoch": 0.1036,
      "grad_norm": 2.323305606842041,
      "learning_rate": 4.5734693877551025e-05,
      "loss": 0.2464,
      "step": 518
    },
    {
      "epoch": 0.1038,
      "grad_norm": 2.432375192642212,
      "learning_rate": 4.572448979591837e-05,
      "loss": 0.2387,
      "step": 519
    },
    {
      "epoch": 0.104,
      "grad_norm": 2.466305732727051,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 0.1291,
      "step": 520
    },
    {
      "epoch": 0.1042,
      "grad_norm": 2.67108416557312,
      "learning_rate": 4.570408163265306e-05,
      "loss": 0.3753,
      "step": 521
    },
    {
      "epoch": 0.1044,
      "grad_norm": 2.484483242034912,
      "learning_rate": 4.569387755102041e-05,
      "loss": 0.4598,
      "step": 522
    },
    {
      "epoch": 0.1046,
      "grad_norm": 16.740785598754883,
      "learning_rate": 4.568367346938776e-05,
      "loss": 2.6504,
      "step": 523
    },
    {
      "epoch": 0.1048,
      "grad_norm": 2.645599842071533,
      "learning_rate": 4.56734693877551e-05,
      "loss": 0.3468,
      "step": 524
    },
    {
      "epoch": 0.105,
      "grad_norm": 9.68822956085205,
      "learning_rate": 4.566326530612245e-05,
      "loss": 2.6367,
      "step": 525
    },
    {
      "epoch": 0.1052,
      "grad_norm": 2.621138095855713,
      "learning_rate": 4.56530612244898e-05,
      "loss": 0.1783,
      "step": 526
    },
    {
      "epoch": 0.1054,
      "grad_norm": 2.0223898887634277,
      "learning_rate": 4.564285714285714e-05,
      "loss": 0.1192,
      "step": 527
    },
    {
      "epoch": 0.1056,
      "grad_norm": 22.834697723388672,
      "learning_rate": 4.563265306122449e-05,
      "loss": 5.2872,
      "step": 528
    },
    {
      "epoch": 0.1058,
      "grad_norm": 2.804661989212036,
      "learning_rate": 4.562244897959184e-05,
      "loss": 0.3319,
      "step": 529
    },
    {
      "epoch": 0.106,
      "grad_norm": 2.925588846206665,
      "learning_rate": 4.561224489795918e-05,
      "loss": 0.3356,
      "step": 530
    },
    {
      "epoch": 0.1062,
      "grad_norm": 10.556026458740234,
      "learning_rate": 4.560204081632653e-05,
      "loss": 2.0116,
      "step": 531
    },
    {
      "epoch": 0.1064,
      "grad_norm": 70.50692749023438,
      "learning_rate": 4.559183673469388e-05,
      "loss": 7.9615,
      "step": 532
    },
    {
      "epoch": 0.1066,
      "grad_norm": 2.5424160957336426,
      "learning_rate": 4.558163265306123e-05,
      "loss": 0.3561,
      "step": 533
    },
    {
      "epoch": 0.1068,
      "grad_norm": 8.875993728637695,
      "learning_rate": 4.557142857142858e-05,
      "loss": 1.0426,
      "step": 534
    },
    {
      "epoch": 0.107,
      "grad_norm": 2.2023794651031494,
      "learning_rate": 4.556122448979592e-05,
      "loss": 0.1573,
      "step": 535
    },
    {
      "epoch": 0.1072,
      "grad_norm": 4.783979415893555,
      "learning_rate": 4.555102040816327e-05,
      "loss": 0.2794,
      "step": 536
    },
    {
      "epoch": 0.1074,
      "grad_norm": 2.1055617332458496,
      "learning_rate": 4.554081632653062e-05,
      "loss": 0.1174,
      "step": 537
    },
    {
      "epoch": 0.1076,
      "grad_norm": 2.1910979747772217,
      "learning_rate": 4.553061224489796e-05,
      "loss": 0.249,
      "step": 538
    },
    {
      "epoch": 0.1078,
      "grad_norm": 13.680757522583008,
      "learning_rate": 4.552040816326531e-05,
      "loss": 3.8584,
      "step": 539
    },
    {
      "epoch": 0.108,
      "grad_norm": 2.199683666229248,
      "learning_rate": 4.551020408163266e-05,
      "loss": 0.2536,
      "step": 540
    },
    {
      "epoch": 0.1082,
      "grad_norm": 1.8234713077545166,
      "learning_rate": 4.55e-05,
      "loss": 0.1149,
      "step": 541
    },
    {
      "epoch": 0.1084,
      "grad_norm": 2.0882201194763184,
      "learning_rate": 4.548979591836735e-05,
      "loss": 0.2728,
      "step": 542
    },
    {
      "epoch": 0.1086,
      "grad_norm": 72.3479995727539,
      "learning_rate": 4.54795918367347e-05,
      "loss": 6.6544,
      "step": 543
    },
    {
      "epoch": 0.1088,
      "grad_norm": 2.8216774463653564,
      "learning_rate": 4.546938775510204e-05,
      "loss": 0.1134,
      "step": 544
    },
    {
      "epoch": 0.109,
      "grad_norm": 26.27626609802246,
      "learning_rate": 4.545918367346939e-05,
      "loss": 2.9438,
      "step": 545
    },
    {
      "epoch": 0.1092,
      "grad_norm": 2.255739688873291,
      "learning_rate": 4.544897959183674e-05,
      "loss": 0.2993,
      "step": 546
    },
    {
      "epoch": 0.1094,
      "grad_norm": 2.0638649463653564,
      "learning_rate": 4.543877551020408e-05,
      "loss": 0.2525,
      "step": 547
    },
    {
      "epoch": 0.1096,
      "grad_norm": 1.93026602268219,
      "learning_rate": 4.542857142857143e-05,
      "loss": 0.2609,
      "step": 548
    },
    {
      "epoch": 0.1098,
      "grad_norm": 2.0406620502471924,
      "learning_rate": 4.541836734693878e-05,
      "loss": 0.2668,
      "step": 549
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.5131540298461914,
      "learning_rate": 4.5408163265306124e-05,
      "loss": 0.3013,
      "step": 550
    },
    {
      "epoch": 0.1102,
      "grad_norm": 5.5197858810424805,
      "learning_rate": 4.539795918367347e-05,
      "loss": 1.2905,
      "step": 551
    },
    {
      "epoch": 0.1104,
      "grad_norm": 1.8763465881347656,
      "learning_rate": 4.538775510204082e-05,
      "loss": 0.2596,
      "step": 552
    },
    {
      "epoch": 0.1106,
      "grad_norm": 11.363729476928711,
      "learning_rate": 4.5377551020408164e-05,
      "loss": 1.8338,
      "step": 553
    },
    {
      "epoch": 0.1108,
      "grad_norm": 20.515701293945312,
      "learning_rate": 4.5367346938775514e-05,
      "loss": 6.1151,
      "step": 554
    },
    {
      "epoch": 0.111,
      "grad_norm": 2.955251455307007,
      "learning_rate": 4.5357142857142856e-05,
      "loss": 0.0994,
      "step": 555
    },
    {
      "epoch": 0.1112,
      "grad_norm": 2.0138936042785645,
      "learning_rate": 4.5346938775510205e-05,
      "loss": 0.2722,
      "step": 556
    },
    {
      "epoch": 0.1114,
      "grad_norm": 5.070435523986816,
      "learning_rate": 4.5336734693877554e-05,
      "loss": 2.5984,
      "step": 557
    },
    {
      "epoch": 0.1116,
      "grad_norm": 5.005916118621826,
      "learning_rate": 4.53265306122449e-05,
      "loss": 1.2611,
      "step": 558
    },
    {
      "epoch": 0.1118,
      "grad_norm": 1.90200674533844,
      "learning_rate": 4.5316326530612246e-05,
      "loss": 0.1851,
      "step": 559
    },
    {
      "epoch": 0.112,
      "grad_norm": 2.2471718788146973,
      "learning_rate": 4.5306122448979595e-05,
      "loss": 0.3086,
      "step": 560
    },
    {
      "epoch": 0.1122,
      "grad_norm": 2.431797504425049,
      "learning_rate": 4.529591836734694e-05,
      "loss": 0.0901,
      "step": 561
    },
    {
      "epoch": 0.1124,
      "grad_norm": 17.897838592529297,
      "learning_rate": 4.528571428571429e-05,
      "loss": 4.6469,
      "step": 562
    },
    {
      "epoch": 0.1126,
      "grad_norm": 2.6925182342529297,
      "learning_rate": 4.5275510204081636e-05,
      "loss": 0.2705,
      "step": 563
    },
    {
      "epoch": 0.1128,
      "grad_norm": 7.428674697875977,
      "learning_rate": 4.526530612244898e-05,
      "loss": 1.2483,
      "step": 564
    },
    {
      "epoch": 0.113,
      "grad_norm": 33.47559356689453,
      "learning_rate": 4.525510204081633e-05,
      "loss": 4.067,
      "step": 565
    },
    {
      "epoch": 0.1132,
      "grad_norm": 2.309847354888916,
      "learning_rate": 4.524489795918368e-05,
      "loss": 0.3087,
      "step": 566
    },
    {
      "epoch": 0.1134,
      "grad_norm": 2.301074504852295,
      "learning_rate": 4.523469387755102e-05,
      "loss": 0.262,
      "step": 567
    },
    {
      "epoch": 0.1136,
      "grad_norm": 2.009781837463379,
      "learning_rate": 4.522448979591837e-05,
      "loss": 0.0959,
      "step": 568
    },
    {
      "epoch": 0.1138,
      "grad_norm": 9.266844749450684,
      "learning_rate": 4.521428571428572e-05,
      "loss": 2.4829,
      "step": 569
    },
    {
      "epoch": 0.114,
      "grad_norm": 2.2796411514282227,
      "learning_rate": 4.520408163265306e-05,
      "loss": 0.1412,
      "step": 570
    },
    {
      "epoch": 0.1142,
      "grad_norm": 2.4037399291992188,
      "learning_rate": 4.519387755102041e-05,
      "loss": 0.1177,
      "step": 571
    },
    {
      "epoch": 0.1144,
      "grad_norm": 1.8711189031600952,
      "learning_rate": 4.518367346938776e-05,
      "loss": 0.1035,
      "step": 572
    },
    {
      "epoch": 0.1146,
      "grad_norm": 6.985625743865967,
      "learning_rate": 4.51734693877551e-05,
      "loss": 2.3783,
      "step": 573
    },
    {
      "epoch": 0.1148,
      "grad_norm": 15.73539924621582,
      "learning_rate": 4.516326530612245e-05,
      "loss": 4.3117,
      "step": 574
    },
    {
      "epoch": 0.115,
      "grad_norm": 8.829004287719727,
      "learning_rate": 4.515306122448979e-05,
      "loss": 1.3898,
      "step": 575
    },
    {
      "epoch": 0.1152,
      "grad_norm": 8.513986587524414,
      "learning_rate": 4.514285714285714e-05,
      "loss": 1.6092,
      "step": 576
    },
    {
      "epoch": 0.1154,
      "grad_norm": 4.182492256164551,
      "learning_rate": 4.51326530612245e-05,
      "loss": 0.2672,
      "step": 577
    },
    {
      "epoch": 0.1156,
      "grad_norm": 1.9439010620117188,
      "learning_rate": 4.512244897959184e-05,
      "loss": 0.1175,
      "step": 578
    },
    {
      "epoch": 0.1158,
      "grad_norm": 2.382523775100708,
      "learning_rate": 4.511224489795919e-05,
      "loss": 0.309,
      "step": 579
    },
    {
      "epoch": 0.116,
      "grad_norm": 15.121125221252441,
      "learning_rate": 4.510204081632654e-05,
      "loss": 4.528,
      "step": 580
    },
    {
      "epoch": 0.1162,
      "grad_norm": 1.8058350086212158,
      "learning_rate": 4.509183673469388e-05,
      "loss": 0.2308,
      "step": 581
    },
    {
      "epoch": 0.1164,
      "grad_norm": 5.423115253448486,
      "learning_rate": 4.508163265306123e-05,
      "loss": 1.2338,
      "step": 582
    },
    {
      "epoch": 0.1166,
      "grad_norm": 2.032620429992676,
      "learning_rate": 4.507142857142858e-05,
      "loss": 0.1202,
      "step": 583
    },
    {
      "epoch": 0.1168,
      "grad_norm": 2.276237964630127,
      "learning_rate": 4.506122448979592e-05,
      "loss": 0.3036,
      "step": 584
    },
    {
      "epoch": 0.117,
      "grad_norm": 5.8232526779174805,
      "learning_rate": 4.505102040816327e-05,
      "loss": 2.4301,
      "step": 585
    },
    {
      "epoch": 0.1172,
      "grad_norm": 4.476560592651367,
      "learning_rate": 4.504081632653061e-05,
      "loss": 1.0937,
      "step": 586
    },
    {
      "epoch": 0.1174,
      "grad_norm": 5.9133992195129395,
      "learning_rate": 4.503061224489796e-05,
      "loss": 0.9071,
      "step": 587
    },
    {
      "epoch": 0.1176,
      "grad_norm": 1.6806764602661133,
      "learning_rate": 4.502040816326531e-05,
      "loss": 0.1801,
      "step": 588
    },
    {
      "epoch": 0.1178,
      "grad_norm": 32.35234069824219,
      "learning_rate": 4.5010204081632654e-05,
      "loss": 4.8664,
      "step": 589
    },
    {
      "epoch": 0.118,
      "grad_norm": 2.2517266273498535,
      "learning_rate": 4.5e-05,
      "loss": 0.2928,
      "step": 590
    },
    {
      "epoch": 0.1182,
      "grad_norm": 2.939851760864258,
      "learning_rate": 4.498979591836735e-05,
      "loss": 0.4259,
      "step": 591
    },
    {
      "epoch": 0.1184,
      "grad_norm": 5.593673229217529,
      "learning_rate": 4.4979591836734694e-05,
      "loss": 0.3221,
      "step": 592
    },
    {
      "epoch": 0.1186,
      "grad_norm": 2.246220350265503,
      "learning_rate": 4.4969387755102044e-05,
      "loss": 0.0967,
      "step": 593
    },
    {
      "epoch": 0.1188,
      "grad_norm": 1.910014033317566,
      "learning_rate": 4.495918367346939e-05,
      "loss": 0.3013,
      "step": 594
    },
    {
      "epoch": 0.119,
      "grad_norm": 1.9520102739334106,
      "learning_rate": 4.4948979591836735e-05,
      "loss": 0.2674,
      "step": 595
    },
    {
      "epoch": 0.1192,
      "grad_norm": 2.66032338142395,
      "learning_rate": 4.4938775510204084e-05,
      "loss": 0.3248,
      "step": 596
    },
    {
      "epoch": 0.1194,
      "grad_norm": 9.406121253967285,
      "learning_rate": 4.4928571428571434e-05,
      "loss": 1.4535,
      "step": 597
    },
    {
      "epoch": 0.1196,
      "grad_norm": 8.665762901306152,
      "learning_rate": 4.4918367346938776e-05,
      "loss": 0.7553,
      "step": 598
    },
    {
      "epoch": 0.1198,
      "grad_norm": 4.330859184265137,
      "learning_rate": 4.4908163265306125e-05,
      "loss": 2.3006,
      "step": 599
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.626834750175476,
      "learning_rate": 4.4897959183673474e-05,
      "loss": 0.098,
      "step": 600
    },
    {
      "epoch": 0.1202,
      "grad_norm": 12.482990264892578,
      "learning_rate": 4.488775510204082e-05,
      "loss": 6.0159,
      "step": 601
    },
    {
      "epoch": 0.1204,
      "grad_norm": 1.843930959701538,
      "learning_rate": 4.4877551020408166e-05,
      "loss": 0.1119,
      "step": 602
    },
    {
      "epoch": 0.1206,
      "grad_norm": 6.070778846740723,
      "learning_rate": 4.4867346938775515e-05,
      "loss": 0.6427,
      "step": 603
    },
    {
      "epoch": 0.1208,
      "grad_norm": 6.043375015258789,
      "learning_rate": 4.485714285714286e-05,
      "loss": 1.1888,
      "step": 604
    },
    {
      "epoch": 0.121,
      "grad_norm": 38.47969436645508,
      "learning_rate": 4.484693877551021e-05,
      "loss": 4.228,
      "step": 605
    },
    {
      "epoch": 0.1212,
      "grad_norm": 5.892314434051514,
      "learning_rate": 4.483673469387755e-05,
      "loss": 0.33,
      "step": 606
    },
    {
      "epoch": 0.1214,
      "grad_norm": 15.078666687011719,
      "learning_rate": 4.48265306122449e-05,
      "loss": 2.3711,
      "step": 607
    },
    {
      "epoch": 0.1216,
      "grad_norm": 3.097940444946289,
      "learning_rate": 4.481632653061225e-05,
      "loss": 0.1112,
      "step": 608
    },
    {
      "epoch": 0.1218,
      "grad_norm": 12.760353088378906,
      "learning_rate": 4.480612244897959e-05,
      "loss": 4.0837,
      "step": 609
    },
    {
      "epoch": 0.122,
      "grad_norm": 11.902802467346191,
      "learning_rate": 4.479591836734694e-05,
      "loss": 1.5209,
      "step": 610
    },
    {
      "epoch": 0.1222,
      "grad_norm": 2.2661423683166504,
      "learning_rate": 4.478571428571429e-05,
      "loss": 0.076,
      "step": 611
    },
    {
      "epoch": 0.1224,
      "grad_norm": 8.932852745056152,
      "learning_rate": 4.477551020408163e-05,
      "loss": 1.6964,
      "step": 612
    },
    {
      "epoch": 0.1226,
      "grad_norm": 13.332756996154785,
      "learning_rate": 4.476530612244898e-05,
      "loss": 2.6943,
      "step": 613
    },
    {
      "epoch": 0.1228,
      "grad_norm": 1.782543659210205,
      "learning_rate": 4.475510204081633e-05,
      "loss": 0.2552,
      "step": 614
    },
    {
      "epoch": 0.123,
      "grad_norm": 9.798274040222168,
      "learning_rate": 4.474489795918367e-05,
      "loss": 1.7567,
      "step": 615
    },
    {
      "epoch": 0.1232,
      "grad_norm": 2.503276824951172,
      "learning_rate": 4.473469387755102e-05,
      "loss": 0.3502,
      "step": 616
    },
    {
      "epoch": 0.1234,
      "grad_norm": 1.6842453479766846,
      "learning_rate": 4.472448979591837e-05,
      "loss": 0.2415,
      "step": 617
    },
    {
      "epoch": 0.1236,
      "grad_norm": 4.441090106964111,
      "learning_rate": 4.471428571428571e-05,
      "loss": 0.4642,
      "step": 618
    },
    {
      "epoch": 0.1238,
      "grad_norm": 4.461360454559326,
      "learning_rate": 4.470408163265306e-05,
      "loss": 1.1069,
      "step": 619
    },
    {
      "epoch": 0.124,
      "grad_norm": 4.818694591522217,
      "learning_rate": 4.469387755102041e-05,
      "loss": 0.2668,
      "step": 620
    },
    {
      "epoch": 0.1242,
      "grad_norm": 14.786673545837402,
      "learning_rate": 4.468367346938775e-05,
      "loss": 1.2223,
      "step": 621
    },
    {
      "epoch": 0.1244,
      "grad_norm": 2.0519139766693115,
      "learning_rate": 4.467346938775511e-05,
      "loss": 0.1196,
      "step": 622
    },
    {
      "epoch": 0.1246,
      "grad_norm": 5.332647323608398,
      "learning_rate": 4.466326530612245e-05,
      "loss": 0.2951,
      "step": 623
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.252056360244751,
      "learning_rate": 4.46530612244898e-05,
      "loss": 0.325,
      "step": 624
    },
    {
      "epoch": 0.125,
      "grad_norm": 5.384517669677734,
      "learning_rate": 4.464285714285715e-05,
      "loss": 0.8607,
      "step": 625
    },
    {
      "epoch": 0.1252,
      "grad_norm": 10.462895393371582,
      "learning_rate": 4.463265306122449e-05,
      "loss": 2.4684,
      "step": 626
    },
    {
      "epoch": 0.1254,
      "grad_norm": 1.8529956340789795,
      "learning_rate": 4.462244897959184e-05,
      "loss": 0.1103,
      "step": 627
    },
    {
      "epoch": 0.1256,
      "grad_norm": 9.0777006149292,
      "learning_rate": 4.461224489795919e-05,
      "loss": 1.6302,
      "step": 628
    },
    {
      "epoch": 0.1258,
      "grad_norm": 7.788472652435303,
      "learning_rate": 4.460204081632653e-05,
      "loss": 1.4777,
      "step": 629
    },
    {
      "epoch": 0.126,
      "grad_norm": 11.898658752441406,
      "learning_rate": 4.459183673469388e-05,
      "loss": 6.0391,
      "step": 630
    },
    {
      "epoch": 0.1262,
      "grad_norm": 13.46484661102295,
      "learning_rate": 4.458163265306123e-05,
      "loss": 2.6615,
      "step": 631
    },
    {
      "epoch": 0.1264,
      "grad_norm": 4.574545860290527,
      "learning_rate": 4.4571428571428574e-05,
      "loss": 1.0586,
      "step": 632
    },
    {
      "epoch": 0.1266,
      "grad_norm": 1.9689745903015137,
      "learning_rate": 4.456122448979592e-05,
      "loss": 0.2847,
      "step": 633
    },
    {
      "epoch": 0.1268,
      "grad_norm": 13.859519958496094,
      "learning_rate": 4.455102040816327e-05,
      "loss": 1.6259,
      "step": 634
    },
    {
      "epoch": 0.127,
      "grad_norm": 1.6551806926727295,
      "learning_rate": 4.4540816326530614e-05,
      "loss": 0.1856,
      "step": 635
    },
    {
      "epoch": 0.1272,
      "grad_norm": 5.734321594238281,
      "learning_rate": 4.4530612244897963e-05,
      "loss": 0.9242,
      "step": 636
    },
    {
      "epoch": 0.1274,
      "grad_norm": 2.1179840564727783,
      "learning_rate": 4.4520408163265306e-05,
      "loss": 0.2987,
      "step": 637
    },
    {
      "epoch": 0.1276,
      "grad_norm": 1.6108742952346802,
      "learning_rate": 4.4510204081632655e-05,
      "loss": 0.165,
      "step": 638
    },
    {
      "epoch": 0.1278,
      "grad_norm": 5.832553386688232,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 1.0757,
      "step": 639
    },
    {
      "epoch": 0.128,
      "grad_norm": 2.1176013946533203,
      "learning_rate": 4.448979591836735e-05,
      "loss": 0.1239,
      "step": 640
    },
    {
      "epoch": 0.1282,
      "grad_norm": 5.174328804016113,
      "learning_rate": 4.4479591836734696e-05,
      "loss": 0.3201,
      "step": 641
    },
    {
      "epoch": 0.1284,
      "grad_norm": 2.0192956924438477,
      "learning_rate": 4.4469387755102045e-05,
      "loss": 0.134,
      "step": 642
    },
    {
      "epoch": 0.1286,
      "grad_norm": 2.0334508419036865,
      "learning_rate": 4.445918367346939e-05,
      "loss": 0.2273,
      "step": 643
    },
    {
      "epoch": 0.1288,
      "grad_norm": 2.3532967567443848,
      "learning_rate": 4.4448979591836737e-05,
      "loss": 0.3139,
      "step": 644
    },
    {
      "epoch": 0.129,
      "grad_norm": 5.952183723449707,
      "learning_rate": 4.4438775510204086e-05,
      "loss": 1.0241,
      "step": 645
    },
    {
      "epoch": 0.1292,
      "grad_norm": 56.45768356323242,
      "learning_rate": 4.442857142857143e-05,
      "loss": 4.0514,
      "step": 646
    },
    {
      "epoch": 0.1294,
      "grad_norm": 15.219496726989746,
      "learning_rate": 4.441836734693878e-05,
      "loss": 6.0333,
      "step": 647
    },
    {
      "epoch": 0.1296,
      "grad_norm": 11.211031913757324,
      "learning_rate": 4.4408163265306127e-05,
      "loss": 1.6167,
      "step": 648
    },
    {
      "epoch": 0.1298,
      "grad_norm": 31.7654972076416,
      "learning_rate": 4.439795918367347e-05,
      "loss": 1.9553,
      "step": 649
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.0561766624450684,
      "learning_rate": 4.438775510204082e-05,
      "loss": 0.2417,
      "step": 650
    },
    {
      "epoch": 0.1302,
      "grad_norm": 2.283371686935425,
      "learning_rate": 4.437755102040817e-05,
      "loss": 0.2687,
      "step": 651
    },
    {
      "epoch": 0.1304,
      "grad_norm": 1.6434128284454346,
      "learning_rate": 4.436734693877551e-05,
      "loss": 0.1885,
      "step": 652
    },
    {
      "epoch": 0.1306,
      "grad_norm": 7.3232340812683105,
      "learning_rate": 4.435714285714286e-05,
      "loss": 2.4361,
      "step": 653
    },
    {
      "epoch": 0.1308,
      "grad_norm": 7.41826868057251,
      "learning_rate": 4.434693877551021e-05,
      "loss": 2.3235,
      "step": 654
    },
    {
      "epoch": 0.131,
      "grad_norm": 6.322724342346191,
      "learning_rate": 4.433673469387755e-05,
      "loss": 1.433,
      "step": 655
    },
    {
      "epoch": 0.1312,
      "grad_norm": 2.0319886207580566,
      "learning_rate": 4.43265306122449e-05,
      "loss": 0.1222,
      "step": 656
    },
    {
      "epoch": 0.1314,
      "grad_norm": 5.4758524894714355,
      "learning_rate": 4.431632653061225e-05,
      "loss": 1.5181,
      "step": 657
    },
    {
      "epoch": 0.1316,
      "grad_norm": 5.770097732543945,
      "learning_rate": 4.430612244897959e-05,
      "loss": 1.5341,
      "step": 658
    },
    {
      "epoch": 0.1318,
      "grad_norm": 1.8182827234268188,
      "learning_rate": 4.429591836734694e-05,
      "loss": 0.1104,
      "step": 659
    },
    {
      "epoch": 0.132,
      "grad_norm": 8.128026962280273,
      "learning_rate": 4.428571428571428e-05,
      "loss": 2.4333,
      "step": 660
    },
    {
      "epoch": 0.1322,
      "grad_norm": 1.7597488164901733,
      "learning_rate": 4.427551020408163e-05,
      "loss": 0.1998,
      "step": 661
    },
    {
      "epoch": 0.1324,
      "grad_norm": 1.8226596117019653,
      "learning_rate": 4.426530612244898e-05,
      "loss": 0.2336,
      "step": 662
    },
    {
      "epoch": 0.1326,
      "grad_norm": 18.413177490234375,
      "learning_rate": 4.4255102040816324e-05,
      "loss": 4.6057,
      "step": 663
    },
    {
      "epoch": 0.1328,
      "grad_norm": 5.388815879821777,
      "learning_rate": 4.424489795918367e-05,
      "loss": 2.4815,
      "step": 664
    },
    {
      "epoch": 0.133,
      "grad_norm": 1.821333408355713,
      "learning_rate": 4.423469387755102e-05,
      "loss": 0.2417,
      "step": 665
    },
    {
      "epoch": 0.1332,
      "grad_norm": 8.400188446044922,
      "learning_rate": 4.4224489795918364e-05,
      "loss": 1.4933,
      "step": 666
    },
    {
      "epoch": 0.1334,
      "grad_norm": 4.5805983543396,
      "learning_rate": 4.4214285714285714e-05,
      "loss": 1.1053,
      "step": 667
    },
    {
      "epoch": 0.1336,
      "grad_norm": 1.690073013305664,
      "learning_rate": 4.420408163265306e-05,
      "loss": 0.1809,
      "step": 668
    },
    {
      "epoch": 0.1338,
      "grad_norm": 5.437297344207764,
      "learning_rate": 4.419387755102041e-05,
      "loss": 0.9489,
      "step": 669
    },
    {
      "epoch": 0.134,
      "grad_norm": 2.3655149936676025,
      "learning_rate": 4.418367346938776e-05,
      "loss": 0.1009,
      "step": 670
    },
    {
      "epoch": 0.1342,
      "grad_norm": 2.305649518966675,
      "learning_rate": 4.4173469387755103e-05,
      "loss": 0.2685,
      "step": 671
    },
    {
      "epoch": 0.1344,
      "grad_norm": 2.1379716396331787,
      "learning_rate": 4.416326530612245e-05,
      "loss": 0.2611,
      "step": 672
    },
    {
      "epoch": 0.1346,
      "grad_norm": 1.7722504138946533,
      "learning_rate": 4.41530612244898e-05,
      "loss": 0.1145,
      "step": 673
    },
    {
      "epoch": 0.1348,
      "grad_norm": 9.311881065368652,
      "learning_rate": 4.4142857142857144e-05,
      "loss": 1.4772,
      "step": 674
    },
    {
      "epoch": 0.135,
      "grad_norm": 2.3688912391662598,
      "learning_rate": 4.4132653061224493e-05,
      "loss": 0.3648,
      "step": 675
    },
    {
      "epoch": 0.1352,
      "grad_norm": 7.750619411468506,
      "learning_rate": 4.412244897959184e-05,
      "loss": 0.6366,
      "step": 676
    },
    {
      "epoch": 0.1354,
      "grad_norm": 1.9949312210083008,
      "learning_rate": 4.4112244897959185e-05,
      "loss": 0.2289,
      "step": 677
    },
    {
      "epoch": 0.1356,
      "grad_norm": 2.4722976684570312,
      "learning_rate": 4.4102040816326534e-05,
      "loss": 0.0601,
      "step": 678
    },
    {
      "epoch": 0.1358,
      "grad_norm": 5.0097761154174805,
      "learning_rate": 4.409183673469388e-05,
      "loss": 0.8654,
      "step": 679
    },
    {
      "epoch": 0.136,
      "grad_norm": 1.8627954721450806,
      "learning_rate": 4.4081632653061226e-05,
      "loss": 0.0965,
      "step": 680
    },
    {
      "epoch": 0.1362,
      "grad_norm": 7.933848857879639,
      "learning_rate": 4.4071428571428575e-05,
      "loss": 1.6723,
      "step": 681
    },
    {
      "epoch": 0.1364,
      "grad_norm": 10.013572692871094,
      "learning_rate": 4.4061224489795924e-05,
      "loss": 1.65,
      "step": 682
    },
    {
      "epoch": 0.1366,
      "grad_norm": 1.8995143175125122,
      "learning_rate": 4.4051020408163267e-05,
      "loss": 0.2396,
      "step": 683
    },
    {
      "epoch": 0.1368,
      "grad_norm": 1.689717411994934,
      "learning_rate": 4.4040816326530616e-05,
      "loss": 0.0887,
      "step": 684
    },
    {
      "epoch": 0.137,
      "grad_norm": 4.838736057281494,
      "learning_rate": 4.4030612244897965e-05,
      "loss": 1.0061,
      "step": 685
    },
    {
      "epoch": 0.1372,
      "grad_norm": 5.647104263305664,
      "learning_rate": 4.402040816326531e-05,
      "loss": 0.3167,
      "step": 686
    },
    {
      "epoch": 0.1374,
      "grad_norm": 2.3787808418273926,
      "learning_rate": 4.4010204081632656e-05,
      "loss": 0.2948,
      "step": 687
    },
    {
      "epoch": 0.1376,
      "grad_norm": 1.9067801237106323,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2273,
      "step": 688
    },
    {
      "epoch": 0.1378,
      "grad_norm": 2.5516650676727295,
      "learning_rate": 4.398979591836735e-05,
      "loss": 0.0622,
      "step": 689
    },
    {
      "epoch": 0.138,
      "grad_norm": 2.4554383754730225,
      "learning_rate": 4.39795918367347e-05,
      "loss": 0.3739,
      "step": 690
    },
    {
      "epoch": 0.1382,
      "grad_norm": 2.3099822998046875,
      "learning_rate": 4.396938775510204e-05,
      "loss": 0.2943,
      "step": 691
    },
    {
      "epoch": 0.1384,
      "grad_norm": 2.306360960006714,
      "learning_rate": 4.395918367346939e-05,
      "loss": 0.34,
      "step": 692
    },
    {
      "epoch": 0.1386,
      "grad_norm": 8.243119239807129,
      "learning_rate": 4.394897959183674e-05,
      "loss": 0.6953,
      "step": 693
    },
    {
      "epoch": 0.1388,
      "grad_norm": 1.6192740201950073,
      "learning_rate": 4.393877551020408e-05,
      "loss": 0.0911,
      "step": 694
    },
    {
      "epoch": 0.139,
      "grad_norm": 1.8218029737472534,
      "learning_rate": 4.392857142857143e-05,
      "loss": 0.2832,
      "step": 695
    },
    {
      "epoch": 0.1392,
      "grad_norm": 2.514047622680664,
      "learning_rate": 4.391836734693878e-05,
      "loss": 0.4452,
      "step": 696
    },
    {
      "epoch": 0.1394,
      "grad_norm": 12.909881591796875,
      "learning_rate": 4.390816326530612e-05,
      "loss": 0.7189,
      "step": 697
    },
    {
      "epoch": 0.1396,
      "grad_norm": 6.974311351776123,
      "learning_rate": 4.389795918367347e-05,
      "loss": 1.5038,
      "step": 698
    },
    {
      "epoch": 0.1398,
      "grad_norm": 2.2936313152313232,
      "learning_rate": 4.388775510204082e-05,
      "loss": 0.2825,
      "step": 699
    },
    {
      "epoch": 0.14,
      "grad_norm": 11.503886222839355,
      "learning_rate": 4.387755102040816e-05,
      "loss": 5.795,
      "step": 700
    },
    {
      "epoch": 0.1402,
      "grad_norm": 1.8265719413757324,
      "learning_rate": 4.386734693877551e-05,
      "loss": 0.1091,
      "step": 701
    },
    {
      "epoch": 0.1404,
      "grad_norm": 2.218329906463623,
      "learning_rate": 4.385714285714286e-05,
      "loss": 0.0533,
      "step": 702
    },
    {
      "epoch": 0.1406,
      "grad_norm": 2.1319103240966797,
      "learning_rate": 4.38469387755102e-05,
      "loss": 0.2347,
      "step": 703
    },
    {
      "epoch": 0.1408,
      "grad_norm": 2.2720017433166504,
      "learning_rate": 4.383673469387755e-05,
      "loss": 0.2567,
      "step": 704
    },
    {
      "epoch": 0.141,
      "grad_norm": 2.6647064685821533,
      "learning_rate": 4.38265306122449e-05,
      "loss": 0.0915,
      "step": 705
    },
    {
      "epoch": 0.1412,
      "grad_norm": 1.7370988130569458,
      "learning_rate": 4.3816326530612243e-05,
      "loss": 0.1033,
      "step": 706
    },
    {
      "epoch": 0.1414,
      "grad_norm": 7.157041072845459,
      "learning_rate": 4.380612244897959e-05,
      "loss": 0.2841,
      "step": 707
    },
    {
      "epoch": 0.1416,
      "grad_norm": 6.928675651550293,
      "learning_rate": 4.379591836734694e-05,
      "loss": 1.4424,
      "step": 708
    },
    {
      "epoch": 0.1418,
      "grad_norm": 2.09725022315979,
      "learning_rate": 4.3785714285714284e-05,
      "loss": 0.0728,
      "step": 709
    },
    {
      "epoch": 0.142,
      "grad_norm": 1.9853891134262085,
      "learning_rate": 4.377551020408163e-05,
      "loss": 0.0603,
      "step": 710
    },
    {
      "epoch": 0.1422,
      "grad_norm": 9.301671981811523,
      "learning_rate": 4.3765306122448976e-05,
      "loss": 2.4837,
      "step": 711
    },
    {
      "epoch": 0.1424,
      "grad_norm": 5.64785099029541,
      "learning_rate": 4.3755102040816325e-05,
      "loss": 0.8947,
      "step": 712
    },
    {
      "epoch": 0.1426,
      "grad_norm": 1.9208048582077026,
      "learning_rate": 4.374489795918368e-05,
      "loss": 0.1086,
      "step": 713
    },
    {
      "epoch": 0.1428,
      "grad_norm": 4.717841625213623,
      "learning_rate": 4.373469387755102e-05,
      "loss": 1.1693,
      "step": 714
    },
    {
      "epoch": 0.143,
      "grad_norm": 9.996125221252441,
      "learning_rate": 4.372448979591837e-05,
      "loss": 0.9532,
      "step": 715
    },
    {
      "epoch": 0.1432,
      "grad_norm": 2.3046953678131104,
      "learning_rate": 4.371428571428572e-05,
      "loss": 0.331,
      "step": 716
    },
    {
      "epoch": 0.1434,
      "grad_norm": 1.901801586151123,
      "learning_rate": 4.3704081632653064e-05,
      "loss": 0.272,
      "step": 717
    },
    {
      "epoch": 0.1436,
      "grad_norm": 24.127500534057617,
      "learning_rate": 4.369387755102041e-05,
      "loss": 4.3717,
      "step": 718
    },
    {
      "epoch": 0.1438,
      "grad_norm": 6.888923645019531,
      "learning_rate": 4.368367346938776e-05,
      "loss": 1.3725,
      "step": 719
    },
    {
      "epoch": 0.144,
      "grad_norm": 6.642779350280762,
      "learning_rate": 4.3673469387755105e-05,
      "loss": 1.2854,
      "step": 720
    },
    {
      "epoch": 0.1442,
      "grad_norm": 4.80898904800415,
      "learning_rate": 4.3663265306122454e-05,
      "loss": 0.2967,
      "step": 721
    },
    {
      "epoch": 0.1444,
      "grad_norm": 4.025364398956299,
      "learning_rate": 4.3653061224489796e-05,
      "loss": 0.9348,
      "step": 722
    },
    {
      "epoch": 0.1446,
      "grad_norm": 1.866579532623291,
      "learning_rate": 4.3642857142857146e-05,
      "loss": 0.1253,
      "step": 723
    },
    {
      "epoch": 0.1448,
      "grad_norm": 5.358083724975586,
      "learning_rate": 4.3632653061224495e-05,
      "loss": 2.4192,
      "step": 724
    },
    {
      "epoch": 0.145,
      "grad_norm": 12.656335830688477,
      "learning_rate": 4.362244897959184e-05,
      "loss": 2.5382,
      "step": 725
    },
    {
      "epoch": 0.1452,
      "grad_norm": 2.922243356704712,
      "learning_rate": 4.3612244897959186e-05,
      "loss": 0.8893,
      "step": 726
    },
    {
      "epoch": 0.1454,
      "grad_norm": 2.257392406463623,
      "learning_rate": 4.3602040816326536e-05,
      "loss": 0.2155,
      "step": 727
    },
    {
      "epoch": 0.1456,
      "grad_norm": 1.99644935131073,
      "learning_rate": 4.359183673469388e-05,
      "loss": 0.2558,
      "step": 728
    },
    {
      "epoch": 0.1458,
      "grad_norm": 1.754607915878296,
      "learning_rate": 4.358163265306123e-05,
      "loss": 0.2429,
      "step": 729
    },
    {
      "epoch": 0.146,
      "grad_norm": 1.7153891324996948,
      "learning_rate": 4.3571428571428576e-05,
      "loss": 0.1047,
      "step": 730
    },
    {
      "epoch": 0.1462,
      "grad_norm": 1.9489779472351074,
      "learning_rate": 4.356122448979592e-05,
      "loss": 0.1325,
      "step": 731
    },
    {
      "epoch": 0.1464,
      "grad_norm": 1.736901879310608,
      "learning_rate": 4.355102040816327e-05,
      "loss": 0.1266,
      "step": 732
    },
    {
      "epoch": 0.1466,
      "grad_norm": 25.8922176361084,
      "learning_rate": 4.354081632653062e-05,
      "loss": 3.2462,
      "step": 733
    },
    {
      "epoch": 0.1468,
      "grad_norm": 8.070469856262207,
      "learning_rate": 4.353061224489796e-05,
      "loss": 2.528,
      "step": 734
    },
    {
      "epoch": 0.147,
      "grad_norm": 6.357401371002197,
      "learning_rate": 4.352040816326531e-05,
      "loss": 0.6996,
      "step": 735
    },
    {
      "epoch": 0.1472,
      "grad_norm": 1.9338181018829346,
      "learning_rate": 4.351020408163266e-05,
      "loss": 0.2494,
      "step": 736
    },
    {
      "epoch": 0.1474,
      "grad_norm": 7.206000804901123,
      "learning_rate": 4.35e-05,
      "loss": 1.5507,
      "step": 737
    },
    {
      "epoch": 0.1476,
      "grad_norm": 5.9976019859313965,
      "learning_rate": 4.348979591836735e-05,
      "loss": 0.9014,
      "step": 738
    },
    {
      "epoch": 0.1478,
      "grad_norm": 4.735384464263916,
      "learning_rate": 4.34795918367347e-05,
      "loss": 0.816,
      "step": 739
    },
    {
      "epoch": 0.148,
      "grad_norm": 1.800541639328003,
      "learning_rate": 4.346938775510204e-05,
      "loss": 0.1901,
      "step": 740
    },
    {
      "epoch": 0.1482,
      "grad_norm": 2.0132083892822266,
      "learning_rate": 4.345918367346939e-05,
      "loss": 0.2509,
      "step": 741
    },
    {
      "epoch": 0.1484,
      "grad_norm": 1.8128314018249512,
      "learning_rate": 4.344897959183673e-05,
      "loss": 0.051,
      "step": 742
    },
    {
      "epoch": 0.1486,
      "grad_norm": 2.1387152671813965,
      "learning_rate": 4.343877551020408e-05,
      "loss": 0.2722,
      "step": 743
    },
    {
      "epoch": 0.1488,
      "grad_norm": 2.3190178871154785,
      "learning_rate": 4.342857142857143e-05,
      "loss": 0.2977,
      "step": 744
    },
    {
      "epoch": 0.149,
      "grad_norm": 7.440555572509766,
      "learning_rate": 4.341836734693877e-05,
      "loss": 1.4205,
      "step": 745
    },
    {
      "epoch": 0.1492,
      "grad_norm": 1.9111486673355103,
      "learning_rate": 4.340816326530612e-05,
      "loss": 0.1122,
      "step": 746
    },
    {
      "epoch": 0.1494,
      "grad_norm": 13.870625495910645,
      "learning_rate": 4.339795918367347e-05,
      "loss": 2.2386,
      "step": 747
    },
    {
      "epoch": 0.1496,
      "grad_norm": 1.77839994430542,
      "learning_rate": 4.3387755102040814e-05,
      "loss": 0.1634,
      "step": 748
    },
    {
      "epoch": 0.1498,
      "grad_norm": 7.970048427581787,
      "learning_rate": 4.337755102040816e-05,
      "loss": 0.7243,
      "step": 749
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.571451187133789,
      "learning_rate": 4.336734693877551e-05,
      "loss": 0.3174,
      "step": 750
    },
    {
      "epoch": 0.1502,
      "grad_norm": 1.6975948810577393,
      "learning_rate": 4.3357142857142855e-05,
      "loss": 0.1841,
      "step": 751
    },
    {
      "epoch": 0.1504,
      "grad_norm": 7.2954816818237305,
      "learning_rate": 4.3346938775510204e-05,
      "loss": 1.3127,
      "step": 752
    },
    {
      "epoch": 0.1506,
      "grad_norm": 4.603607177734375,
      "learning_rate": 4.333673469387755e-05,
      "loss": 1.1817,
      "step": 753
    },
    {
      "epoch": 0.1508,
      "grad_norm": 1.546514630317688,
      "learning_rate": 4.3326530612244896e-05,
      "loss": 0.2038,
      "step": 754
    },
    {
      "epoch": 0.151,
      "grad_norm": 1.9785537719726562,
      "learning_rate": 4.3316326530612245e-05,
      "loss": 0.2896,
      "step": 755
    },
    {
      "epoch": 0.1512,
      "grad_norm": 4.313318252563477,
      "learning_rate": 4.3306122448979594e-05,
      "loss": 1.1541,
      "step": 756
    },
    {
      "epoch": 0.1514,
      "grad_norm": 4.420673370361328,
      "learning_rate": 4.3295918367346936e-05,
      "loss": 1.0872,
      "step": 757
    },
    {
      "epoch": 0.1516,
      "grad_norm": 4.9215288162231445,
      "learning_rate": 4.328571428571429e-05,
      "loss": 0.8623,
      "step": 758
    },
    {
      "epoch": 0.1518,
      "grad_norm": 1.6973063945770264,
      "learning_rate": 4.3275510204081635e-05,
      "loss": 0.0893,
      "step": 759
    },
    {
      "epoch": 0.152,
      "grad_norm": 1.7262623310089111,
      "learning_rate": 4.3265306122448984e-05,
      "loss": 0.2229,
      "step": 760
    },
    {
      "epoch": 0.1522,
      "grad_norm": 2.2670018672943115,
      "learning_rate": 4.325510204081633e-05,
      "loss": 0.2578,
      "step": 761
    },
    {
      "epoch": 0.1524,
      "grad_norm": 8.593221664428711,
      "learning_rate": 4.3244897959183676e-05,
      "loss": 2.3867,
      "step": 762
    },
    {
      "epoch": 0.1526,
      "grad_norm": 9.047572135925293,
      "learning_rate": 4.3234693877551025e-05,
      "loss": 2.3965,
      "step": 763
    },
    {
      "epoch": 0.1528,
      "grad_norm": 1.6796197891235352,
      "learning_rate": 4.3224489795918374e-05,
      "loss": 0.1077,
      "step": 764
    },
    {
      "epoch": 0.153,
      "grad_norm": 1.9067870378494263,
      "learning_rate": 4.3214285714285716e-05,
      "loss": 0.1235,
      "step": 765
    },
    {
      "epoch": 0.1532,
      "grad_norm": 6.224389553070068,
      "learning_rate": 4.3204081632653065e-05,
      "loss": 1.4195,
      "step": 766
    },
    {
      "epoch": 0.1534,
      "grad_norm": 6.695178508758545,
      "learning_rate": 4.3193877551020415e-05,
      "loss": 1.2996,
      "step": 767
    },
    {
      "epoch": 0.1536,
      "grad_norm": 3.7735090255737305,
      "learning_rate": 4.318367346938776e-05,
      "loss": 0.8999,
      "step": 768
    },
    {
      "epoch": 0.1538,
      "grad_norm": 18.430770874023438,
      "learning_rate": 4.3173469387755106e-05,
      "loss": 5.8124,
      "step": 769
    },
    {
      "epoch": 0.154,
      "grad_norm": 13.440892219543457,
      "learning_rate": 4.3163265306122455e-05,
      "loss": 2.2063,
      "step": 770
    },
    {
      "epoch": 0.1542,
      "grad_norm": 1.6782939434051514,
      "learning_rate": 4.31530612244898e-05,
      "loss": 0.1103,
      "step": 771
    },
    {
      "epoch": 0.1544,
      "grad_norm": 2.325686454772949,
      "learning_rate": 4.314285714285715e-05,
      "loss": 0.0827,
      "step": 772
    },
    {
      "epoch": 0.1546,
      "grad_norm": 7.990361213684082,
      "learning_rate": 4.313265306122449e-05,
      "loss": 1.2616,
      "step": 773
    },
    {
      "epoch": 0.1548,
      "grad_norm": 1.9885655641555786,
      "learning_rate": 4.312244897959184e-05,
      "loss": 0.1165,
      "step": 774
    },
    {
      "epoch": 0.155,
      "grad_norm": 2.0096423625946045,
      "learning_rate": 4.311224489795919e-05,
      "loss": 0.3091,
      "step": 775
    },
    {
      "epoch": 0.1552,
      "grad_norm": 2.642979621887207,
      "learning_rate": 4.310204081632653e-05,
      "loss": 0.3795,
      "step": 776
    },
    {
      "epoch": 0.1554,
      "grad_norm": 5.770168781280518,
      "learning_rate": 4.309183673469388e-05,
      "loss": 0.8668,
      "step": 777
    },
    {
      "epoch": 0.1556,
      "grad_norm": 1.6523147821426392,
      "learning_rate": 4.308163265306123e-05,
      "loss": 0.1001,
      "step": 778
    },
    {
      "epoch": 0.1558,
      "grad_norm": 2.008592128753662,
      "learning_rate": 4.307142857142857e-05,
      "loss": 0.1983,
      "step": 779
    },
    {
      "epoch": 0.156,
      "grad_norm": 2.2496085166931152,
      "learning_rate": 4.306122448979592e-05,
      "loss": 0.2928,
      "step": 780
    },
    {
      "epoch": 0.1562,
      "grad_norm": 2.215184211730957,
      "learning_rate": 4.305102040816327e-05,
      "loss": 0.2607,
      "step": 781
    },
    {
      "epoch": 0.1564,
      "grad_norm": 4.5969953536987305,
      "learning_rate": 4.304081632653061e-05,
      "loss": 1.0656,
      "step": 782
    },
    {
      "epoch": 0.1566,
      "grad_norm": 2.250063896179199,
      "learning_rate": 4.303061224489796e-05,
      "loss": 0.1636,
      "step": 783
    },
    {
      "epoch": 0.1568,
      "grad_norm": 5.640593528747559,
      "learning_rate": 4.302040816326531e-05,
      "loss": 2.3175,
      "step": 784
    },
    {
      "epoch": 0.157,
      "grad_norm": 15.217977523803711,
      "learning_rate": 4.301020408163265e-05,
      "loss": 1.4736,
      "step": 785
    },
    {
      "epoch": 0.1572,
      "grad_norm": 1.8735414743423462,
      "learning_rate": 4.3e-05,
      "loss": 0.1066,
      "step": 786
    },
    {
      "epoch": 0.1574,
      "grad_norm": 24.6962890625,
      "learning_rate": 4.298979591836735e-05,
      "loss": 4.4351,
      "step": 787
    },
    {
      "epoch": 0.1576,
      "grad_norm": 19.73663330078125,
      "learning_rate": 4.297959183673469e-05,
      "loss": 4.5139,
      "step": 788
    },
    {
      "epoch": 0.1578,
      "grad_norm": 12.441691398620605,
      "learning_rate": 4.296938775510204e-05,
      "loss": 2.562,
      "step": 789
    },
    {
      "epoch": 0.158,
      "grad_norm": 4.711108684539795,
      "learning_rate": 4.295918367346939e-05,
      "loss": 0.8065,
      "step": 790
    },
    {
      "epoch": 0.1582,
      "grad_norm": 2.1329903602600098,
      "learning_rate": 4.2948979591836734e-05,
      "loss": 0.0514,
      "step": 791
    },
    {
      "epoch": 0.1584,
      "grad_norm": 1.7468369007110596,
      "learning_rate": 4.293877551020408e-05,
      "loss": 0.1048,
      "step": 792
    },
    {
      "epoch": 0.1586,
      "grad_norm": 2.452786445617676,
      "learning_rate": 4.292857142857143e-05,
      "loss": 0.2923,
      "step": 793
    },
    {
      "epoch": 0.1588,
      "grad_norm": 3.929147243499756,
      "learning_rate": 4.2918367346938775e-05,
      "loss": 2.3275,
      "step": 794
    },
    {
      "epoch": 0.159,
      "grad_norm": 1.9393938779830933,
      "learning_rate": 4.2908163265306124e-05,
      "loss": 0.1292,
      "step": 795
    },
    {
      "epoch": 0.1592,
      "grad_norm": 7.826672554016113,
      "learning_rate": 4.2897959183673466e-05,
      "loss": 1.2891,
      "step": 796
    },
    {
      "epoch": 0.1594,
      "grad_norm": 1.6239943504333496,
      "learning_rate": 4.2887755102040816e-05,
      "loss": 0.1988,
      "step": 797
    },
    {
      "epoch": 0.1596,
      "grad_norm": 2.3447299003601074,
      "learning_rate": 4.2877551020408165e-05,
      "loss": 0.3015,
      "step": 798
    },
    {
      "epoch": 0.1598,
      "grad_norm": 2.075815439224243,
      "learning_rate": 4.286734693877551e-05,
      "loss": 0.2568,
      "step": 799
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.7597588300704956,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.104,
      "step": 800
    },
    {
      "epoch": 0.1602,
      "grad_norm": 4.3136725425720215,
      "learning_rate": 4.2846938775510205e-05,
      "loss": 2.4686,
      "step": 801
    },
    {
      "epoch": 0.1604,
      "grad_norm": 1.7578059434890747,
      "learning_rate": 4.283673469387755e-05,
      "loss": 0.0796,
      "step": 802
    },
    {
      "epoch": 0.1606,
      "grad_norm": 2.104520797729492,
      "learning_rate": 4.28265306122449e-05,
      "loss": 0.0624,
      "step": 803
    },
    {
      "epoch": 0.1608,
      "grad_norm": 1.956167459487915,
      "learning_rate": 4.281632653061225e-05,
      "loss": 0.2198,
      "step": 804
    },
    {
      "epoch": 0.161,
      "grad_norm": 2.076270341873169,
      "learning_rate": 4.2806122448979595e-05,
      "loss": 0.0752,
      "step": 805
    },
    {
      "epoch": 0.1612,
      "grad_norm": 1.9125944375991821,
      "learning_rate": 4.2795918367346945e-05,
      "loss": 0.2213,
      "step": 806
    },
    {
      "epoch": 0.1614,
      "grad_norm": 1.775360345840454,
      "learning_rate": 4.278571428571429e-05,
      "loss": 0.109,
      "step": 807
    },
    {
      "epoch": 0.1616,
      "grad_norm": 4.13633394241333,
      "learning_rate": 4.2775510204081636e-05,
      "loss": 2.4488,
      "step": 808
    },
    {
      "epoch": 0.1618,
      "grad_norm": 8.054276466369629,
      "learning_rate": 4.2765306122448985e-05,
      "loss": 2.2969,
      "step": 809
    },
    {
      "epoch": 0.162,
      "grad_norm": 6.543287754058838,
      "learning_rate": 4.275510204081633e-05,
      "loss": 0.7256,
      "step": 810
    },
    {
      "epoch": 0.1622,
      "grad_norm": 3.911162853240967,
      "learning_rate": 4.274489795918368e-05,
      "loss": 2.3648,
      "step": 811
    },
    {
      "epoch": 0.1624,
      "grad_norm": 3.8917007446289062,
      "learning_rate": 4.2734693877551026e-05,
      "loss": 2.3915,
      "step": 812
    },
    {
      "epoch": 0.1626,
      "grad_norm": 3.657830238342285,
      "learning_rate": 4.272448979591837e-05,
      "loss": 2.2125,
      "step": 813
    },
    {
      "epoch": 0.1628,
      "grad_norm": 46.86573791503906,
      "learning_rate": 4.271428571428572e-05,
      "loss": 6.0395,
      "step": 814
    },
    {
      "epoch": 0.163,
      "grad_norm": 7.47802734375,
      "learning_rate": 4.270408163265307e-05,
      "loss": 1.3507,
      "step": 815
    },
    {
      "epoch": 0.1632,
      "grad_norm": 1.9284485578536987,
      "learning_rate": 4.269387755102041e-05,
      "loss": 0.158,
      "step": 816
    },
    {
      "epoch": 0.1634,
      "grad_norm": 2.250680446624756,
      "learning_rate": 4.268367346938776e-05,
      "loss": 0.3714,
      "step": 817
    },
    {
      "epoch": 0.1636,
      "grad_norm": 6.096149921417236,
      "learning_rate": 4.267346938775511e-05,
      "loss": 1.1214,
      "step": 818
    },
    {
      "epoch": 0.1638,
      "grad_norm": 4.140794277191162,
      "learning_rate": 4.266326530612245e-05,
      "loss": 2.4074,
      "step": 819
    },
    {
      "epoch": 0.164,
      "grad_norm": 7.302412509918213,
      "learning_rate": 4.26530612244898e-05,
      "loss": 2.3707,
      "step": 820
    },
    {
      "epoch": 0.1642,
      "grad_norm": 22.319246292114258,
      "learning_rate": 4.264285714285715e-05,
      "loss": 4.4901,
      "step": 821
    },
    {
      "epoch": 0.1644,
      "grad_norm": 6.420026779174805,
      "learning_rate": 4.263265306122449e-05,
      "loss": 2.4446,
      "step": 822
    },
    {
      "epoch": 0.1646,
      "grad_norm": 15.629339218139648,
      "learning_rate": 4.262244897959184e-05,
      "loss": 4.5816,
      "step": 823
    },
    {
      "epoch": 0.1648,
      "grad_norm": 2.6500132083892822,
      "learning_rate": 4.261224489795919e-05,
      "loss": 0.1216,
      "step": 824
    },
    {
      "epoch": 0.165,
      "grad_norm": 6.980380535125732,
      "learning_rate": 4.260204081632653e-05,
      "loss": 1.2799,
      "step": 825
    },
    {
      "epoch": 0.1652,
      "grad_norm": 1.9646185636520386,
      "learning_rate": 4.259183673469388e-05,
      "loss": 0.1346,
      "step": 826
    },
    {
      "epoch": 0.1654,
      "grad_norm": 3.1223630905151367,
      "learning_rate": 4.258163265306122e-05,
      "loss": 0.8811,
      "step": 827
    },
    {
      "epoch": 0.1656,
      "grad_norm": 2.060901165008545,
      "learning_rate": 4.257142857142857e-05,
      "loss": 0.1289,
      "step": 828
    },
    {
      "epoch": 0.1658,
      "grad_norm": 3.2158594131469727,
      "learning_rate": 4.256122448979592e-05,
      "loss": 0.0913,
      "step": 829
    },
    {
      "epoch": 0.166,
      "grad_norm": 50.38131332397461,
      "learning_rate": 4.2551020408163264e-05,
      "loss": 5.8206,
      "step": 830
    },
    {
      "epoch": 0.1662,
      "grad_norm": 3.1545493602752686,
      "learning_rate": 4.254081632653061e-05,
      "loss": 0.0734,
      "step": 831
    },
    {
      "epoch": 0.1664,
      "grad_norm": 1.6991525888442993,
      "learning_rate": 4.253061224489796e-05,
      "loss": 0.2189,
      "step": 832
    },
    {
      "epoch": 0.1666,
      "grad_norm": 2.1904635429382324,
      "learning_rate": 4.2520408163265305e-05,
      "loss": 0.1192,
      "step": 833
    },
    {
      "epoch": 0.1668,
      "grad_norm": 1.8767436742782593,
      "learning_rate": 4.2510204081632654e-05,
      "loss": 0.188,
      "step": 834
    },
    {
      "epoch": 0.167,
      "grad_norm": 5.210190773010254,
      "learning_rate": 4.25e-05,
      "loss": 1.0994,
      "step": 835
    },
    {
      "epoch": 0.1672,
      "grad_norm": 2.0396203994750977,
      "learning_rate": 4.2489795918367345e-05,
      "loss": 0.1005,
      "step": 836
    },
    {
      "epoch": 0.1674,
      "grad_norm": 1.7658389806747437,
      "learning_rate": 4.2479591836734695e-05,
      "loss": 0.0898,
      "step": 837
    },
    {
      "epoch": 0.1676,
      "grad_norm": 1.6944715976715088,
      "learning_rate": 4.2469387755102044e-05,
      "loss": 0.2445,
      "step": 838
    },
    {
      "epoch": 0.1678,
      "grad_norm": 1.6534080505371094,
      "learning_rate": 4.2459183673469386e-05,
      "loss": 0.0917,
      "step": 839
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.8207001686096191,
      "learning_rate": 4.2448979591836735e-05,
      "loss": 0.1746,
      "step": 840
    },
    {
      "epoch": 0.1682,
      "grad_norm": 1.6604434251785278,
      "learning_rate": 4.2438775510204085e-05,
      "loss": 0.1816,
      "step": 841
    },
    {
      "epoch": 0.1684,
      "grad_norm": 1.7828645706176758,
      "learning_rate": 4.242857142857143e-05,
      "loss": 0.2107,
      "step": 842
    },
    {
      "epoch": 0.1686,
      "grad_norm": 4.550353527069092,
      "learning_rate": 4.2418367346938776e-05,
      "loss": 1.1876,
      "step": 843
    },
    {
      "epoch": 0.1688,
      "grad_norm": 4.328961372375488,
      "learning_rate": 4.2408163265306125e-05,
      "loss": 2.4182,
      "step": 844
    },
    {
      "epoch": 0.169,
      "grad_norm": 2.939178943634033,
      "learning_rate": 4.239795918367347e-05,
      "loss": 0.8846,
      "step": 845
    },
    {
      "epoch": 0.1692,
      "grad_norm": 4.374268054962158,
      "learning_rate": 4.238775510204082e-05,
      "loss": 0.962,
      "step": 846
    },
    {
      "epoch": 0.1694,
      "grad_norm": 1.9265333414077759,
      "learning_rate": 4.237755102040816e-05,
      "loss": 0.162,
      "step": 847
    },
    {
      "epoch": 0.1696,
      "grad_norm": 1.5972001552581787,
      "learning_rate": 4.236734693877551e-05,
      "loss": 0.1947,
      "step": 848
    },
    {
      "epoch": 0.1698,
      "grad_norm": 5.813956260681152,
      "learning_rate": 4.2357142857142864e-05,
      "loss": 1.3463,
      "step": 849
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.302892684936523,
      "learning_rate": 4.234693877551021e-05,
      "loss": 1.6315,
      "step": 850
    },
    {
      "epoch": 0.1702,
      "grad_norm": 1.6612507104873657,
      "learning_rate": 4.2336734693877556e-05,
      "loss": 0.1486,
      "step": 851
    },
    {
      "epoch": 0.1704,
      "grad_norm": 1.7442834377288818,
      "learning_rate": 4.2326530612244905e-05,
      "loss": 0.2076,
      "step": 852
    },
    {
      "epoch": 0.1706,
      "grad_norm": 8.239985466003418,
      "learning_rate": 4.231632653061225e-05,
      "loss": 2.4089,
      "step": 853
    },
    {
      "epoch": 0.1708,
      "grad_norm": 1.635135531425476,
      "learning_rate": 4.23061224489796e-05,
      "loss": 0.1433,
      "step": 854
    },
    {
      "epoch": 0.171,
      "grad_norm": 11.539763450622559,
      "learning_rate": 4.2295918367346946e-05,
      "loss": 2.2066,
      "step": 855
    },
    {
      "epoch": 0.1712,
      "grad_norm": 2.6980319023132324,
      "learning_rate": 4.228571428571429e-05,
      "loss": 0.082,
      "step": 856
    },
    {
      "epoch": 0.1714,
      "grad_norm": 2.544710397720337,
      "learning_rate": 4.227551020408164e-05,
      "loss": 0.1421,
      "step": 857
    },
    {
      "epoch": 0.1716,
      "grad_norm": 6.985103130340576,
      "learning_rate": 4.226530612244898e-05,
      "loss": 1.4134,
      "step": 858
    },
    {
      "epoch": 0.1718,
      "grad_norm": 3.9861574172973633,
      "learning_rate": 4.225510204081633e-05,
      "loss": 0.7055,
      "step": 859
    },
    {
      "epoch": 0.172,
      "grad_norm": 1.906098484992981,
      "learning_rate": 4.224489795918368e-05,
      "loss": 0.111,
      "step": 860
    },
    {
      "epoch": 0.1722,
      "grad_norm": 2.332689046859741,
      "learning_rate": 4.223469387755102e-05,
      "loss": 0.1167,
      "step": 861
    },
    {
      "epoch": 0.1724,
      "grad_norm": 15.099422454833984,
      "learning_rate": 4.222448979591837e-05,
      "loss": 3.3177,
      "step": 862
    },
    {
      "epoch": 0.1726,
      "grad_norm": 6.7551984786987305,
      "learning_rate": 4.221428571428572e-05,
      "loss": 2.307,
      "step": 863
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.085442066192627,
      "learning_rate": 4.220408163265306e-05,
      "loss": 0.1908,
      "step": 864
    },
    {
      "epoch": 0.173,
      "grad_norm": 38.784244537353516,
      "learning_rate": 4.219387755102041e-05,
      "loss": 5.5188,
      "step": 865
    },
    {
      "epoch": 0.1732,
      "grad_norm": 2.1101486682891846,
      "learning_rate": 4.218367346938776e-05,
      "loss": 0.2555,
      "step": 866
    },
    {
      "epoch": 0.1734,
      "grad_norm": 4.37405252456665,
      "learning_rate": 4.21734693877551e-05,
      "loss": 2.2608,
      "step": 867
    },
    {
      "epoch": 0.1736,
      "grad_norm": 12.383613586425781,
      "learning_rate": 4.216326530612245e-05,
      "loss": 2.5355,
      "step": 868
    },
    {
      "epoch": 0.1738,
      "grad_norm": 1.730635166168213,
      "learning_rate": 4.21530612244898e-05,
      "loss": 0.2009,
      "step": 869
    },
    {
      "epoch": 0.174,
      "grad_norm": 8.86620807647705,
      "learning_rate": 4.214285714285714e-05,
      "loss": 1.3526,
      "step": 870
    },
    {
      "epoch": 0.1742,
      "grad_norm": 4.039647102355957,
      "learning_rate": 4.213265306122449e-05,
      "loss": 2.3077,
      "step": 871
    },
    {
      "epoch": 0.1744,
      "grad_norm": 2.9431746006011963,
      "learning_rate": 4.212244897959184e-05,
      "loss": 0.1349,
      "step": 872
    },
    {
      "epoch": 0.1746,
      "grad_norm": 1.9896126985549927,
      "learning_rate": 4.2112244897959184e-05,
      "loss": 0.0804,
      "step": 873
    },
    {
      "epoch": 0.1748,
      "grad_norm": 6.88995361328125,
      "learning_rate": 4.210204081632653e-05,
      "loss": 0.6664,
      "step": 874
    },
    {
      "epoch": 0.175,
      "grad_norm": 2.000009298324585,
      "learning_rate": 4.209183673469388e-05,
      "loss": 0.2,
      "step": 875
    },
    {
      "epoch": 0.1752,
      "grad_norm": 1.6612294912338257,
      "learning_rate": 4.2081632653061225e-05,
      "loss": 0.1047,
      "step": 876
    },
    {
      "epoch": 0.1754,
      "grad_norm": 4.192002296447754,
      "learning_rate": 4.2071428571428574e-05,
      "loss": 2.411,
      "step": 877
    },
    {
      "epoch": 0.1756,
      "grad_norm": 1.8390638828277588,
      "learning_rate": 4.206122448979592e-05,
      "loss": 0.214,
      "step": 878
    },
    {
      "epoch": 0.1758,
      "grad_norm": 1.7344233989715576,
      "learning_rate": 4.2051020408163265e-05,
      "loss": 0.2782,
      "step": 879
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.545967698097229,
      "learning_rate": 4.2040816326530615e-05,
      "loss": 0.16,
      "step": 880
    },
    {
      "epoch": 0.1762,
      "grad_norm": 1.775578498840332,
      "learning_rate": 4.203061224489796e-05,
      "loss": 0.2152,
      "step": 881
    },
    {
      "epoch": 0.1764,
      "grad_norm": 1.6764472723007202,
      "learning_rate": 4.2020408163265306e-05,
      "loss": 0.1382,
      "step": 882
    },
    {
      "epoch": 0.1766,
      "grad_norm": 1.90451979637146,
      "learning_rate": 4.2010204081632655e-05,
      "loss": 0.2481,
      "step": 883
    },
    {
      "epoch": 0.1768,
      "grad_norm": 7.125704765319824,
      "learning_rate": 4.2e-05,
      "loss": 2.3166,
      "step": 884
    },
    {
      "epoch": 0.177,
      "grad_norm": 1.615654706954956,
      "learning_rate": 4.198979591836735e-05,
      "loss": 0.1979,
      "step": 885
    },
    {
      "epoch": 0.1772,
      "grad_norm": 6.812422752380371,
      "learning_rate": 4.1979591836734696e-05,
      "loss": 0.543,
      "step": 886
    },
    {
      "epoch": 0.1774,
      "grad_norm": 4.927462577819824,
      "learning_rate": 4.196938775510204e-05,
      "loss": 0.8311,
      "step": 887
    },
    {
      "epoch": 0.1776,
      "grad_norm": 32.130638122558594,
      "learning_rate": 4.195918367346939e-05,
      "loss": 3.5849,
      "step": 888
    },
    {
      "epoch": 0.1778,
      "grad_norm": 9.701715469360352,
      "learning_rate": 4.194897959183674e-05,
      "loss": 1.5411,
      "step": 889
    },
    {
      "epoch": 0.178,
      "grad_norm": 1.8125908374786377,
      "learning_rate": 4.193877551020408e-05,
      "loss": 0.1623,
      "step": 890
    },
    {
      "epoch": 0.1782,
      "grad_norm": 27.293272018432617,
      "learning_rate": 4.192857142857143e-05,
      "loss": 5.7133,
      "step": 891
    },
    {
      "epoch": 0.1784,
      "grad_norm": 4.58138370513916,
      "learning_rate": 4.191836734693878e-05,
      "loss": 2.2323,
      "step": 892
    },
    {
      "epoch": 0.1786,
      "grad_norm": 1.850812554359436,
      "learning_rate": 4.190816326530612e-05,
      "loss": 0.209,
      "step": 893
    },
    {
      "epoch": 0.1788,
      "grad_norm": 13.71081256866455,
      "learning_rate": 4.1897959183673476e-05,
      "loss": 1.3616,
      "step": 894
    },
    {
      "epoch": 0.179,
      "grad_norm": 4.414346218109131,
      "learning_rate": 4.188775510204082e-05,
      "loss": 0.765,
      "step": 895
    },
    {
      "epoch": 0.1792,
      "grad_norm": 1.5059181451797485,
      "learning_rate": 4.187755102040817e-05,
      "loss": 0.2032,
      "step": 896
    },
    {
      "epoch": 0.1794,
      "grad_norm": 1.4246588945388794,
      "learning_rate": 4.186734693877552e-05,
      "loss": 0.1675,
      "step": 897
    },
    {
      "epoch": 0.1796,
      "grad_norm": 1.410630464553833,
      "learning_rate": 4.185714285714286e-05,
      "loss": 0.1632,
      "step": 898
    },
    {
      "epoch": 0.1798,
      "grad_norm": 9.267348289489746,
      "learning_rate": 4.184693877551021e-05,
      "loss": 0.6752,
      "step": 899
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.1020305156707764,
      "learning_rate": 4.183673469387756e-05,
      "loss": 0.0524,
      "step": 900
    },
    {
      "epoch": 0.1802,
      "grad_norm": 1.746822714805603,
      "learning_rate": 4.18265306122449e-05,
      "loss": 0.1791,
      "step": 901
    },
    {
      "epoch": 0.1804,
      "grad_norm": 3.0003411769866943,
      "learning_rate": 4.181632653061225e-05,
      "loss": 0.8223,
      "step": 902
    },
    {
      "epoch": 0.1806,
      "grad_norm": 1.722442626953125,
      "learning_rate": 4.18061224489796e-05,
      "loss": 0.1501,
      "step": 903
    },
    {
      "epoch": 0.1808,
      "grad_norm": 1.9100525379180908,
      "learning_rate": 4.179591836734694e-05,
      "loss": 0.1688,
      "step": 904
    },
    {
      "epoch": 0.181,
      "grad_norm": 2.5086095333099365,
      "learning_rate": 4.178571428571429e-05,
      "loss": 0.0947,
      "step": 905
    },
    {
      "epoch": 0.1812,
      "grad_norm": 4.063119411468506,
      "learning_rate": 4.177551020408164e-05,
      "loss": 0.9714,
      "step": 906
    },
    {
      "epoch": 0.1814,
      "grad_norm": 6.684601783752441,
      "learning_rate": 4.176530612244898e-05,
      "loss": 1.3965,
      "step": 907
    },
    {
      "epoch": 0.1816,
      "grad_norm": 1.689247727394104,
      "learning_rate": 4.175510204081633e-05,
      "loss": 0.1731,
      "step": 908
    },
    {
      "epoch": 0.1818,
      "grad_norm": 14.05957317352295,
      "learning_rate": 4.174489795918368e-05,
      "loss": 2.3015,
      "step": 909
    },
    {
      "epoch": 0.182,
      "grad_norm": 18.073965072631836,
      "learning_rate": 4.173469387755102e-05,
      "loss": 3.9085,
      "step": 910
    },
    {
      "epoch": 0.1822,
      "grad_norm": 5.947948932647705,
      "learning_rate": 4.172448979591837e-05,
      "loss": 1.154,
      "step": 911
    },
    {
      "epoch": 0.1824,
      "grad_norm": 6.424638748168945,
      "learning_rate": 4.1714285714285714e-05,
      "loss": 1.2806,
      "step": 912
    },
    {
      "epoch": 0.1826,
      "grad_norm": 1.6589370965957642,
      "learning_rate": 4.170408163265306e-05,
      "loss": 0.12,
      "step": 913
    },
    {
      "epoch": 0.1828,
      "grad_norm": 2.2325034141540527,
      "learning_rate": 4.169387755102041e-05,
      "loss": 0.1171,
      "step": 914
    },
    {
      "epoch": 0.183,
      "grad_norm": 6.6812849044799805,
      "learning_rate": 4.1683673469387754e-05,
      "loss": 2.3485,
      "step": 915
    },
    {
      "epoch": 0.1832,
      "grad_norm": 4.563431739807129,
      "learning_rate": 4.1673469387755104e-05,
      "loss": 2.3878,
      "step": 916
    },
    {
      "epoch": 0.1834,
      "grad_norm": 1.316593885421753,
      "learning_rate": 4.166326530612245e-05,
      "loss": 0.1204,
      "step": 917
    },
    {
      "epoch": 0.1836,
      "grad_norm": 2.0545766353607178,
      "learning_rate": 4.1653061224489795e-05,
      "loss": 0.1217,
      "step": 918
    },
    {
      "epoch": 0.1838,
      "grad_norm": 1.4564555883407593,
      "learning_rate": 4.1642857142857144e-05,
      "loss": 0.1337,
      "step": 919
    },
    {
      "epoch": 0.184,
      "grad_norm": 5.850950717926025,
      "learning_rate": 4.1632653061224494e-05,
      "loss": 1.317,
      "step": 920
    },
    {
      "epoch": 0.1842,
      "grad_norm": 5.424844741821289,
      "learning_rate": 4.1622448979591836e-05,
      "loss": 1.0097,
      "step": 921
    },
    {
      "epoch": 0.1844,
      "grad_norm": 4.220058441162109,
      "learning_rate": 4.1612244897959185e-05,
      "loss": 2.2679,
      "step": 922
    },
    {
      "epoch": 0.1846,
      "grad_norm": 1.9250975847244263,
      "learning_rate": 4.1602040816326534e-05,
      "loss": 0.1915,
      "step": 923
    },
    {
      "epoch": 0.1848,
      "grad_norm": 1.4541889429092407,
      "learning_rate": 4.159183673469388e-05,
      "loss": 0.1724,
      "step": 924
    },
    {
      "epoch": 0.185,
      "grad_norm": 81.74874877929688,
      "learning_rate": 4.1581632653061226e-05,
      "loss": 6.8632,
      "step": 925
    },
    {
      "epoch": 0.1852,
      "grad_norm": 18.632627487182617,
      "learning_rate": 4.1571428571428575e-05,
      "loss": 0.3878,
      "step": 926
    },
    {
      "epoch": 0.1854,
      "grad_norm": 1.8790223598480225,
      "learning_rate": 4.156122448979592e-05,
      "loss": 0.2882,
      "step": 927
    },
    {
      "epoch": 0.1856,
      "grad_norm": 1.3396214246749878,
      "learning_rate": 4.155102040816327e-05,
      "loss": 0.1173,
      "step": 928
    },
    {
      "epoch": 0.1858,
      "grad_norm": 12.873229026794434,
      "learning_rate": 4.1540816326530616e-05,
      "loss": 4.3862,
      "step": 929
    },
    {
      "epoch": 0.186,
      "grad_norm": 17.393413543701172,
      "learning_rate": 4.153061224489796e-05,
      "loss": 4.3354,
      "step": 930
    },
    {
      "epoch": 0.1862,
      "grad_norm": 12.224554061889648,
      "learning_rate": 4.152040816326531e-05,
      "loss": 1.6314,
      "step": 931
    },
    {
      "epoch": 0.1864,
      "grad_norm": 4.5745439529418945,
      "learning_rate": 4.151020408163265e-05,
      "loss": 1.005,
      "step": 932
    },
    {
      "epoch": 0.1866,
      "grad_norm": 1.3737479448318481,
      "learning_rate": 4.15e-05,
      "loss": 0.1172,
      "step": 933
    },
    {
      "epoch": 0.1868,
      "grad_norm": 4.484574794769287,
      "learning_rate": 4.148979591836735e-05,
      "loss": 2.1438,
      "step": 934
    },
    {
      "epoch": 0.187,
      "grad_norm": 13.013484001159668,
      "learning_rate": 4.147959183673469e-05,
      "loss": 1.3852,
      "step": 935
    },
    {
      "epoch": 0.1872,
      "grad_norm": 1.4868980646133423,
      "learning_rate": 4.146938775510204e-05,
      "loss": 0.129,
      "step": 936
    },
    {
      "epoch": 0.1874,
      "grad_norm": 2.100748300552368,
      "learning_rate": 4.145918367346939e-05,
      "loss": 0.1455,
      "step": 937
    },
    {
      "epoch": 0.1876,
      "grad_norm": 17.68741798400879,
      "learning_rate": 4.144897959183673e-05,
      "loss": 4.359,
      "step": 938
    },
    {
      "epoch": 0.1878,
      "grad_norm": 4.203042030334473,
      "learning_rate": 4.143877551020408e-05,
      "loss": 1.1126,
      "step": 939
    },
    {
      "epoch": 0.188,
      "grad_norm": 7.120700836181641,
      "learning_rate": 4.1428571428571437e-05,
      "loss": 1.2638,
      "step": 940
    },
    {
      "epoch": 0.1882,
      "grad_norm": 3.8899881839752197,
      "learning_rate": 4.141836734693878e-05,
      "loss": 0.9725,
      "step": 941
    },
    {
      "epoch": 0.1884,
      "grad_norm": 9.656298637390137,
      "learning_rate": 4.140816326530613e-05,
      "loss": 0.6918,
      "step": 942
    },
    {
      "epoch": 0.1886,
      "grad_norm": 6.113008499145508,
      "learning_rate": 4.139795918367347e-05,
      "loss": 1.315,
      "step": 943
    },
    {
      "epoch": 0.1888,
      "grad_norm": 8.569879531860352,
      "learning_rate": 4.138775510204082e-05,
      "loss": 1.3466,
      "step": 944
    },
    {
      "epoch": 0.189,
      "grad_norm": 7.1254353523254395,
      "learning_rate": 4.137755102040817e-05,
      "loss": 2.4282,
      "step": 945
    },
    {
      "epoch": 0.1892,
      "grad_norm": 1.4450511932373047,
      "learning_rate": 4.136734693877551e-05,
      "loss": 0.1443,
      "step": 946
    },
    {
      "epoch": 0.1894,
      "grad_norm": 3.5518486499786377,
      "learning_rate": 4.135714285714286e-05,
      "loss": 0.3862,
      "step": 947
    },
    {
      "epoch": 0.1896,
      "grad_norm": 6.623416423797607,
      "learning_rate": 4.134693877551021e-05,
      "loss": 1.2714,
      "step": 948
    },
    {
      "epoch": 0.1898,
      "grad_norm": 1.5826369524002075,
      "learning_rate": 4.133673469387755e-05,
      "loss": 0.1646,
      "step": 949
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6537444591522217,
      "learning_rate": 4.13265306122449e-05,
      "loss": 0.0755,
      "step": 950
    },
    {
      "epoch": 0.1902,
      "grad_norm": 33.40952682495117,
      "learning_rate": 4.131632653061225e-05,
      "loss": 0.794,
      "step": 951
    },
    {
      "epoch": 0.1904,
      "grad_norm": 18.153114318847656,
      "learning_rate": 4.130612244897959e-05,
      "loss": 2.4001,
      "step": 952
    },
    {
      "epoch": 0.1906,
      "grad_norm": 11.340116500854492,
      "learning_rate": 4.129591836734694e-05,
      "loss": 2.5849,
      "step": 953
    },
    {
      "epoch": 0.1908,
      "grad_norm": 8.346242904663086,
      "learning_rate": 4.128571428571429e-05,
      "loss": 2.2724,
      "step": 954
    },
    {
      "epoch": 0.191,
      "grad_norm": 1.4004164934158325,
      "learning_rate": 4.1275510204081634e-05,
      "loss": 0.1659,
      "step": 955
    },
    {
      "epoch": 0.1912,
      "grad_norm": 2.0775654315948486,
      "learning_rate": 4.126530612244898e-05,
      "loss": 0.2222,
      "step": 956
    },
    {
      "epoch": 0.1914,
      "grad_norm": 1.3837953805923462,
      "learning_rate": 4.125510204081633e-05,
      "loss": 0.1013,
      "step": 957
    },
    {
      "epoch": 0.1916,
      "grad_norm": 1.9300739765167236,
      "learning_rate": 4.1244897959183674e-05,
      "loss": 0.0995,
      "step": 958
    },
    {
      "epoch": 0.1918,
      "grad_norm": 1.8797160387039185,
      "learning_rate": 4.1234693877551024e-05,
      "loss": 0.097,
      "step": 959
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.974260926246643,
      "learning_rate": 4.122448979591837e-05,
      "loss": 0.1699,
      "step": 960
    },
    {
      "epoch": 0.1922,
      "grad_norm": 9.13991928100586,
      "learning_rate": 4.1214285714285715e-05,
      "loss": 1.3085,
      "step": 961
    },
    {
      "epoch": 0.1924,
      "grad_norm": 1.9515283107757568,
      "learning_rate": 4.1204081632653064e-05,
      "loss": 0.1082,
      "step": 962
    },
    {
      "epoch": 0.1926,
      "grad_norm": 1.9016188383102417,
      "learning_rate": 4.119387755102041e-05,
      "loss": 0.1154,
      "step": 963
    },
    {
      "epoch": 0.1928,
      "grad_norm": 7.120417594909668,
      "learning_rate": 4.1183673469387756e-05,
      "loss": 2.2656,
      "step": 964
    },
    {
      "epoch": 0.193,
      "grad_norm": 8.615710258483887,
      "learning_rate": 4.1173469387755105e-05,
      "loss": 1.2942,
      "step": 965
    },
    {
      "epoch": 0.1932,
      "grad_norm": 10.473421096801758,
      "learning_rate": 4.116326530612245e-05,
      "loss": 0.8338,
      "step": 966
    },
    {
      "epoch": 0.1934,
      "grad_norm": 1.2581517696380615,
      "learning_rate": 4.11530612244898e-05,
      "loss": 0.1018,
      "step": 967
    },
    {
      "epoch": 0.1936,
      "grad_norm": 1.8771737813949585,
      "learning_rate": 4.1142857142857146e-05,
      "loss": 0.1048,
      "step": 968
    },
    {
      "epoch": 0.1938,
      "grad_norm": 1.536227822303772,
      "learning_rate": 4.113265306122449e-05,
      "loss": 0.0699,
      "step": 969
    },
    {
      "epoch": 0.194,
      "grad_norm": 5.4825520515441895,
      "learning_rate": 4.112244897959184e-05,
      "loss": 2.3103,
      "step": 970
    },
    {
      "epoch": 0.1942,
      "grad_norm": 6.8128790855407715,
      "learning_rate": 4.1112244897959187e-05,
      "loss": 0.7666,
      "step": 971
    },
    {
      "epoch": 0.1944,
      "grad_norm": 5.419755935668945,
      "learning_rate": 4.110204081632653e-05,
      "loss": 1.3184,
      "step": 972
    },
    {
      "epoch": 0.1946,
      "grad_norm": 1.8374887704849243,
      "learning_rate": 4.109183673469388e-05,
      "loss": 0.1427,
      "step": 973
    },
    {
      "epoch": 0.1948,
      "grad_norm": 4.784745693206787,
      "learning_rate": 4.108163265306123e-05,
      "loss": 1.1199,
      "step": 974
    },
    {
      "epoch": 0.195,
      "grad_norm": 1.6024447679519653,
      "learning_rate": 4.107142857142857e-05,
      "loss": 0.0935,
      "step": 975
    },
    {
      "epoch": 0.1952,
      "grad_norm": 2.622997760772705,
      "learning_rate": 4.106122448979592e-05,
      "loss": 0.8585,
      "step": 976
    },
    {
      "epoch": 0.1954,
      "grad_norm": 1.4445765018463135,
      "learning_rate": 4.105102040816327e-05,
      "loss": 0.1321,
      "step": 977
    },
    {
      "epoch": 0.1956,
      "grad_norm": 1.9010988473892212,
      "learning_rate": 4.104081632653061e-05,
      "loss": 0.1113,
      "step": 978
    },
    {
      "epoch": 0.1958,
      "grad_norm": 1.8826258182525635,
      "learning_rate": 4.103061224489796e-05,
      "loss": 0.1027,
      "step": 979
    },
    {
      "epoch": 0.196,
      "grad_norm": 1.7285670042037964,
      "learning_rate": 4.102040816326531e-05,
      "loss": 0.096,
      "step": 980
    },
    {
      "epoch": 0.1962,
      "grad_norm": 24.584623336791992,
      "learning_rate": 4.101020408163265e-05,
      "loss": 1.6457,
      "step": 981
    },
    {
      "epoch": 0.1964,
      "grad_norm": 33.21518325805664,
      "learning_rate": 4.1e-05,
      "loss": 5.8246,
      "step": 982
    },
    {
      "epoch": 0.1966,
      "grad_norm": 2.0982282161712646,
      "learning_rate": 4.098979591836735e-05,
      "loss": 0.1859,
      "step": 983
    },
    {
      "epoch": 0.1968,
      "grad_norm": 1.2820332050323486,
      "learning_rate": 4.097959183673469e-05,
      "loss": 0.0985,
      "step": 984
    },
    {
      "epoch": 0.197,
      "grad_norm": 8.416690826416016,
      "learning_rate": 4.096938775510205e-05,
      "loss": 1.5457,
      "step": 985
    },
    {
      "epoch": 0.1972,
      "grad_norm": 1.9515358209609985,
      "learning_rate": 4.095918367346939e-05,
      "loss": 0.1537,
      "step": 986
    },
    {
      "epoch": 0.1974,
      "grad_norm": 1.6056196689605713,
      "learning_rate": 4.094897959183674e-05,
      "loss": 0.1138,
      "step": 987
    },
    {
      "epoch": 0.1976,
      "grad_norm": 10.302643775939941,
      "learning_rate": 4.093877551020409e-05,
      "loss": 3.2709,
      "step": 988
    },
    {
      "epoch": 0.1978,
      "grad_norm": 5.601901531219482,
      "learning_rate": 4.092857142857143e-05,
      "loss": 0.774,
      "step": 989
    },
    {
      "epoch": 0.198,
      "grad_norm": 1.9346834421157837,
      "learning_rate": 4.091836734693878e-05,
      "loss": 0.1553,
      "step": 990
    },
    {
      "epoch": 0.1982,
      "grad_norm": 2.078249454498291,
      "learning_rate": 4.090816326530613e-05,
      "loss": 0.1278,
      "step": 991
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.7769577503204346,
      "learning_rate": 4.089795918367347e-05,
      "loss": 0.0883,
      "step": 992
    },
    {
      "epoch": 0.1986,
      "grad_norm": 1.2814565896987915,
      "learning_rate": 4.088775510204082e-05,
      "loss": 0.1313,
      "step": 993
    },
    {
      "epoch": 0.1988,
      "grad_norm": 6.84991455078125,
      "learning_rate": 4.0877551020408164e-05,
      "loss": 1.2505,
      "step": 994
    },
    {
      "epoch": 0.199,
      "grad_norm": 1.601853609085083,
      "learning_rate": 4.086734693877551e-05,
      "loss": 0.157,
      "step": 995
    },
    {
      "epoch": 0.1992,
      "grad_norm": 8.058499336242676,
      "learning_rate": 4.085714285714286e-05,
      "loss": 0.6271,
      "step": 996
    },
    {
      "epoch": 0.1994,
      "grad_norm": 1.5115468502044678,
      "learning_rate": 4.0846938775510204e-05,
      "loss": 0.2509,
      "step": 997
    },
    {
      "epoch": 0.1996,
      "grad_norm": 1.6834666728973389,
      "learning_rate": 4.0836734693877553e-05,
      "loss": 0.0936,
      "step": 998
    },
    {
      "epoch": 0.1998,
      "grad_norm": 9.186635971069336,
      "learning_rate": 4.08265306122449e-05,
      "loss": 2.1009,
      "step": 999
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.8956700563430786,
      "learning_rate": 4.0816326530612245e-05,
      "loss": 0.1993,
      "step": 1000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 1.4122716188430786,
      "learning_rate": 4.0806122448979594e-05,
      "loss": 0.1491,
      "step": 1001
    },
    {
      "epoch": 0.2004,
      "grad_norm": 11.156471252441406,
      "learning_rate": 4.0795918367346943e-05,
      "loss": 3.8965,
      "step": 1002
    },
    {
      "epoch": 0.2006,
      "grad_norm": 1.5105379819869995,
      "learning_rate": 4.0785714285714286e-05,
      "loss": 0.1107,
      "step": 1003
    },
    {
      "epoch": 0.2008,
      "grad_norm": 70.52470397949219,
      "learning_rate": 4.0775510204081635e-05,
      "loss": 6.2614,
      "step": 1004
    },
    {
      "epoch": 0.201,
      "grad_norm": 4.5841498374938965,
      "learning_rate": 4.0765306122448984e-05,
      "loss": 0.8138,
      "step": 1005
    },
    {
      "epoch": 0.2012,
      "grad_norm": 3.8832404613494873,
      "learning_rate": 4.0755102040816327e-05,
      "loss": 0.9416,
      "step": 1006
    },
    {
      "epoch": 0.2014,
      "grad_norm": 7.785485744476318,
      "learning_rate": 4.0744897959183676e-05,
      "loss": 2.2909,
      "step": 1007
    },
    {
      "epoch": 0.2016,
      "grad_norm": 6.2355780601501465,
      "learning_rate": 4.0734693877551025e-05,
      "loss": 1.1862,
      "step": 1008
    },
    {
      "epoch": 0.2018,
      "grad_norm": 7.021670818328857,
      "learning_rate": 4.072448979591837e-05,
      "loss": 0.5645,
      "step": 1009
    },
    {
      "epoch": 0.202,
      "grad_norm": 1.8375966548919678,
      "learning_rate": 4.0714285714285717e-05,
      "loss": 0.0982,
      "step": 1010
    },
    {
      "epoch": 0.2022,
      "grad_norm": 1.345582365989685,
      "learning_rate": 4.0704081632653066e-05,
      "loss": 0.0887,
      "step": 1011
    },
    {
      "epoch": 0.2024,
      "grad_norm": 1.925706148147583,
      "learning_rate": 4.069387755102041e-05,
      "loss": 0.1029,
      "step": 1012
    },
    {
      "epoch": 0.2026,
      "grad_norm": 21.210050582885742,
      "learning_rate": 4.068367346938776e-05,
      "loss": 3.965,
      "step": 1013
    },
    {
      "epoch": 0.2028,
      "grad_norm": 1.2436354160308838,
      "learning_rate": 4.0673469387755106e-05,
      "loss": 0.0701,
      "step": 1014
    },
    {
      "epoch": 0.203,
      "grad_norm": 11.792455673217773,
      "learning_rate": 4.066326530612245e-05,
      "loss": 5.489,
      "step": 1015
    },
    {
      "epoch": 0.2032,
      "grad_norm": 1.7196590900421143,
      "learning_rate": 4.06530612244898e-05,
      "loss": 0.1133,
      "step": 1016
    },
    {
      "epoch": 0.2034,
      "grad_norm": 1.7486034631729126,
      "learning_rate": 4.064285714285714e-05,
      "loss": 0.0889,
      "step": 1017
    },
    {
      "epoch": 0.2036,
      "grad_norm": 1.5281075239181519,
      "learning_rate": 4.063265306122449e-05,
      "loss": 0.1184,
      "step": 1018
    },
    {
      "epoch": 0.2038,
      "grad_norm": 1.5642445087432861,
      "learning_rate": 4.062244897959184e-05,
      "loss": 0.0774,
      "step": 1019
    },
    {
      "epoch": 0.204,
      "grad_norm": 1.566006064414978,
      "learning_rate": 4.061224489795918e-05,
      "loss": 0.1203,
      "step": 1020
    },
    {
      "epoch": 0.2042,
      "grad_norm": 9.783367156982422,
      "learning_rate": 4.060204081632653e-05,
      "loss": 2.1491,
      "step": 1021
    },
    {
      "epoch": 0.2044,
      "grad_norm": 16.899147033691406,
      "learning_rate": 4.059183673469388e-05,
      "loss": 4.2063,
      "step": 1022
    },
    {
      "epoch": 0.2046,
      "grad_norm": 1.4680163860321045,
      "learning_rate": 4.058163265306122e-05,
      "loss": 0.0917,
      "step": 1023
    },
    {
      "epoch": 0.2048,
      "grad_norm": 4.537769317626953,
      "learning_rate": 4.057142857142857e-05,
      "loss": 0.7457,
      "step": 1024
    },
    {
      "epoch": 0.205,
      "grad_norm": 1.6647769212722778,
      "learning_rate": 4.056122448979592e-05,
      "loss": 0.0705,
      "step": 1025
    },
    {
      "epoch": 0.2052,
      "grad_norm": 8.374734878540039,
      "learning_rate": 4.055102040816326e-05,
      "loss": 4.0424,
      "step": 1026
    },
    {
      "epoch": 0.2054,
      "grad_norm": 1.743696689605713,
      "learning_rate": 4.054081632653061e-05,
      "loss": 0.1013,
      "step": 1027
    },
    {
      "epoch": 0.2056,
      "grad_norm": 6.450601577758789,
      "learning_rate": 4.053061224489796e-05,
      "loss": 1.207,
      "step": 1028
    },
    {
      "epoch": 0.2058,
      "grad_norm": 14.592948913574219,
      "learning_rate": 4.0520408163265304e-05,
      "loss": 4.1025,
      "step": 1029
    },
    {
      "epoch": 0.206,
      "grad_norm": 1.6705548763275146,
      "learning_rate": 4.051020408163265e-05,
      "loss": 0.084,
      "step": 1030
    },
    {
      "epoch": 0.2062,
      "grad_norm": 9.311860084533691,
      "learning_rate": 4.05e-05,
      "loss": 0.6941,
      "step": 1031
    },
    {
      "epoch": 0.2064,
      "grad_norm": 1.5666495561599731,
      "learning_rate": 4.048979591836735e-05,
      "loss": 0.1536,
      "step": 1032
    },
    {
      "epoch": 0.2066,
      "grad_norm": 1.5592368841171265,
      "learning_rate": 4.04795918367347e-05,
      "loss": 0.1401,
      "step": 1033
    },
    {
      "epoch": 0.2068,
      "grad_norm": 2.0136380195617676,
      "learning_rate": 4.046938775510204e-05,
      "loss": 0.114,
      "step": 1034
    },
    {
      "epoch": 0.207,
      "grad_norm": 1.42483651638031,
      "learning_rate": 4.045918367346939e-05,
      "loss": 0.1033,
      "step": 1035
    },
    {
      "epoch": 0.2072,
      "grad_norm": 1.5242152214050293,
      "learning_rate": 4.044897959183674e-05,
      "loss": 0.074,
      "step": 1036
    },
    {
      "epoch": 0.2074,
      "grad_norm": 7.547857761383057,
      "learning_rate": 4.0438775510204083e-05,
      "loss": 2.3123,
      "step": 1037
    },
    {
      "epoch": 0.2076,
      "grad_norm": 1.7124192714691162,
      "learning_rate": 4.042857142857143e-05,
      "loss": 0.1537,
      "step": 1038
    },
    {
      "epoch": 0.2078,
      "grad_norm": 1.7321916818618774,
      "learning_rate": 4.041836734693878e-05,
      "loss": 0.1755,
      "step": 1039
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.7035274505615234,
      "learning_rate": 4.0408163265306124e-05,
      "loss": 0.1048,
      "step": 1040
    },
    {
      "epoch": 0.2082,
      "grad_norm": 7.646542549133301,
      "learning_rate": 4.039795918367347e-05,
      "loss": 0.5286,
      "step": 1041
    },
    {
      "epoch": 0.2084,
      "grad_norm": 30.4531307220459,
      "learning_rate": 4.038775510204082e-05,
      "loss": 3.739,
      "step": 1042
    },
    {
      "epoch": 0.2086,
      "grad_norm": 1.4997987747192383,
      "learning_rate": 4.0377551020408165e-05,
      "loss": 0.1062,
      "step": 1043
    },
    {
      "epoch": 0.2088,
      "grad_norm": 1.8914159536361694,
      "learning_rate": 4.0367346938775514e-05,
      "loss": 0.0973,
      "step": 1044
    },
    {
      "epoch": 0.209,
      "grad_norm": 60.22584533691406,
      "learning_rate": 4.035714285714286e-05,
      "loss": 5.7281,
      "step": 1045
    },
    {
      "epoch": 0.2092,
      "grad_norm": 4.276406764984131,
      "learning_rate": 4.0346938775510206e-05,
      "loss": 0.7713,
      "step": 1046
    },
    {
      "epoch": 0.2094,
      "grad_norm": 10.648032188415527,
      "learning_rate": 4.0336734693877555e-05,
      "loss": 1.5118,
      "step": 1047
    },
    {
      "epoch": 0.2096,
      "grad_norm": 1.9461679458618164,
      "learning_rate": 4.03265306122449e-05,
      "loss": 0.1174,
      "step": 1048
    },
    {
      "epoch": 0.2098,
      "grad_norm": 4.108247756958008,
      "learning_rate": 4.0316326530612246e-05,
      "loss": 1.0173,
      "step": 1049
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.454129934310913,
      "learning_rate": 4.0306122448979596e-05,
      "loss": 0.1232,
      "step": 1050
    },
    {
      "epoch": 0.2102,
      "grad_norm": 1.7871676683425903,
      "learning_rate": 4.029591836734694e-05,
      "loss": 0.0944,
      "step": 1051
    },
    {
      "epoch": 0.2104,
      "grad_norm": 5.2860307693481445,
      "learning_rate": 4.028571428571429e-05,
      "loss": 1.1604,
      "step": 1052
    },
    {
      "epoch": 0.2106,
      "grad_norm": 4.047928810119629,
      "learning_rate": 4.0275510204081636e-05,
      "loss": 0.7197,
      "step": 1053
    },
    {
      "epoch": 0.2108,
      "grad_norm": 28.927228927612305,
      "learning_rate": 4.026530612244898e-05,
      "loss": 4.3561,
      "step": 1054
    },
    {
      "epoch": 0.211,
      "grad_norm": 29.201982498168945,
      "learning_rate": 4.025510204081633e-05,
      "loss": 1.1396,
      "step": 1055
    },
    {
      "epoch": 0.2112,
      "grad_norm": 1.7634601593017578,
      "learning_rate": 4.024489795918368e-05,
      "loss": 0.1091,
      "step": 1056
    },
    {
      "epoch": 0.2114,
      "grad_norm": 6.9372124671936035,
      "learning_rate": 4.023469387755102e-05,
      "loss": 1.2503,
      "step": 1057
    },
    {
      "epoch": 0.2116,
      "grad_norm": 2.028412103652954,
      "learning_rate": 4.022448979591837e-05,
      "loss": 0.1896,
      "step": 1058
    },
    {
      "epoch": 0.2118,
      "grad_norm": 1.4823191165924072,
      "learning_rate": 4.021428571428572e-05,
      "loss": 0.2048,
      "step": 1059
    },
    {
      "epoch": 0.212,
      "grad_norm": 1.8830736875534058,
      "learning_rate": 4.020408163265306e-05,
      "loss": 0.1155,
      "step": 1060
    },
    {
      "epoch": 0.2122,
      "grad_norm": 8.2816162109375,
      "learning_rate": 4.019387755102041e-05,
      "loss": 0.5329,
      "step": 1061
    },
    {
      "epoch": 0.2124,
      "grad_norm": 3.7619011402130127,
      "learning_rate": 4.018367346938776e-05,
      "loss": 0.8836,
      "step": 1062
    },
    {
      "epoch": 0.2126,
      "grad_norm": 1.6038355827331543,
      "learning_rate": 4.01734693877551e-05,
      "loss": 0.0724,
      "step": 1063
    },
    {
      "epoch": 0.2128,
      "grad_norm": 2.4276578426361084,
      "learning_rate": 4.016326530612245e-05,
      "loss": 0.2589,
      "step": 1064
    },
    {
      "epoch": 0.213,
      "grad_norm": 6.890028953552246,
      "learning_rate": 4.01530612244898e-05,
      "loss": 1.2599,
      "step": 1065
    },
    {
      "epoch": 0.2132,
      "grad_norm": 3.703716993331909,
      "learning_rate": 4.014285714285714e-05,
      "loss": 0.8813,
      "step": 1066
    },
    {
      "epoch": 0.2134,
      "grad_norm": 1.9085053205490112,
      "learning_rate": 4.013265306122449e-05,
      "loss": 0.1695,
      "step": 1067
    },
    {
      "epoch": 0.2136,
      "grad_norm": 4.963956356048584,
      "learning_rate": 4.0122448979591833e-05,
      "loss": 1.195,
      "step": 1068
    },
    {
      "epoch": 0.2138,
      "grad_norm": 1.4116706848144531,
      "learning_rate": 4.011224489795918e-05,
      "loss": 0.0697,
      "step": 1069
    },
    {
      "epoch": 0.214,
      "grad_norm": 6.491616725921631,
      "learning_rate": 4.010204081632653e-05,
      "loss": 1.1058,
      "step": 1070
    },
    {
      "epoch": 0.2142,
      "grad_norm": 17.118547439575195,
      "learning_rate": 4.0091836734693874e-05,
      "loss": 2.6657,
      "step": 1071
    },
    {
      "epoch": 0.2144,
      "grad_norm": 1.7242052555084229,
      "learning_rate": 4.008163265306122e-05,
      "loss": 0.0941,
      "step": 1072
    },
    {
      "epoch": 0.2146,
      "grad_norm": 22.678550720214844,
      "learning_rate": 4.007142857142857e-05,
      "loss": 4.1993,
      "step": 1073
    },
    {
      "epoch": 0.2148,
      "grad_norm": 3.679158926010132,
      "learning_rate": 4.0061224489795915e-05,
      "loss": 0.9252,
      "step": 1074
    },
    {
      "epoch": 0.215,
      "grad_norm": 1.7487565279006958,
      "learning_rate": 4.0051020408163264e-05,
      "loss": 0.0862,
      "step": 1075
    },
    {
      "epoch": 0.2152,
      "grad_norm": 1.4549202919006348,
      "learning_rate": 4.004081632653062e-05,
      "loss": 0.1285,
      "step": 1076
    },
    {
      "epoch": 0.2154,
      "grad_norm": 21.872774124145508,
      "learning_rate": 4.003061224489796e-05,
      "loss": 4.259,
      "step": 1077
    },
    {
      "epoch": 0.2156,
      "grad_norm": 1.3242027759552002,
      "learning_rate": 4.002040816326531e-05,
      "loss": 0.1155,
      "step": 1078
    },
    {
      "epoch": 0.2158,
      "grad_norm": 1.9196478128433228,
      "learning_rate": 4.0010204081632654e-05,
      "loss": 0.1505,
      "step": 1079
    },
    {
      "epoch": 0.216,
      "grad_norm": 1.3824145793914795,
      "learning_rate": 4e-05,
      "loss": 0.1283,
      "step": 1080
    },
    {
      "epoch": 0.2162,
      "grad_norm": 1.5698976516723633,
      "learning_rate": 3.998979591836735e-05,
      "loss": 0.0639,
      "step": 1081
    },
    {
      "epoch": 0.2164,
      "grad_norm": 1.4868955612182617,
      "learning_rate": 3.9979591836734695e-05,
      "loss": 0.125,
      "step": 1082
    },
    {
      "epoch": 0.2166,
      "grad_norm": 1.6327617168426514,
      "learning_rate": 3.9969387755102044e-05,
      "loss": 0.1406,
      "step": 1083
    },
    {
      "epoch": 0.2168,
      "grad_norm": 3.6507341861724854,
      "learning_rate": 3.995918367346939e-05,
      "loss": 0.9076,
      "step": 1084
    },
    {
      "epoch": 0.217,
      "grad_norm": 1.826192021369934,
      "learning_rate": 3.9948979591836736e-05,
      "loss": 0.1481,
      "step": 1085
    },
    {
      "epoch": 0.2172,
      "grad_norm": 7.602888584136963,
      "learning_rate": 3.9938775510204085e-05,
      "loss": 2.2922,
      "step": 1086
    },
    {
      "epoch": 0.2174,
      "grad_norm": 8.769946098327637,
      "learning_rate": 3.9928571428571434e-05,
      "loss": 1.3738,
      "step": 1087
    },
    {
      "epoch": 0.2176,
      "grad_norm": 1.91317880153656,
      "learning_rate": 3.9918367346938776e-05,
      "loss": 0.1733,
      "step": 1088
    },
    {
      "epoch": 0.2178,
      "grad_norm": 1.6258636713027954,
      "learning_rate": 3.9908163265306126e-05,
      "loss": 0.145,
      "step": 1089
    },
    {
      "epoch": 0.218,
      "grad_norm": 22.450666427612305,
      "learning_rate": 3.9897959183673475e-05,
      "loss": 4.5677,
      "step": 1090
    },
    {
      "epoch": 0.2182,
      "grad_norm": 20.471054077148438,
      "learning_rate": 3.988775510204082e-05,
      "loss": 2.4693,
      "step": 1091
    },
    {
      "epoch": 0.2184,
      "grad_norm": 1.944197177886963,
      "learning_rate": 3.9877551020408166e-05,
      "loss": 0.0889,
      "step": 1092
    },
    {
      "epoch": 0.2186,
      "grad_norm": 5.251967906951904,
      "learning_rate": 3.9867346938775516e-05,
      "loss": 0.6365,
      "step": 1093
    },
    {
      "epoch": 0.2188,
      "grad_norm": 6.404879093170166,
      "learning_rate": 3.985714285714286e-05,
      "loss": 1.2409,
      "step": 1094
    },
    {
      "epoch": 0.219,
      "grad_norm": 1.7063329219818115,
      "learning_rate": 3.984693877551021e-05,
      "loss": 0.1512,
      "step": 1095
    },
    {
      "epoch": 0.2192,
      "grad_norm": 1.841525912284851,
      "learning_rate": 3.9836734693877556e-05,
      "loss": 0.1593,
      "step": 1096
    },
    {
      "epoch": 0.2194,
      "grad_norm": 7.1279826164245605,
      "learning_rate": 3.98265306122449e-05,
      "loss": 1.1394,
      "step": 1097
    },
    {
      "epoch": 0.2196,
      "grad_norm": 4.164473533630371,
      "learning_rate": 3.981632653061225e-05,
      "loss": 0.752,
      "step": 1098
    },
    {
      "epoch": 0.2198,
      "grad_norm": 9.854219436645508,
      "learning_rate": 3.980612244897959e-05,
      "loss": 3.5586,
      "step": 1099
    },
    {
      "epoch": 0.22,
      "grad_norm": 6.977601528167725,
      "learning_rate": 3.979591836734694e-05,
      "loss": 2.3759,
      "step": 1100
    },
    {
      "epoch": 0.2202,
      "grad_norm": 31.07599449157715,
      "learning_rate": 3.978571428571429e-05,
      "loss": 3.3308,
      "step": 1101
    },
    {
      "epoch": 0.2204,
      "grad_norm": 1.951879620552063,
      "learning_rate": 3.977551020408163e-05,
      "loss": 0.1847,
      "step": 1102
    },
    {
      "epoch": 0.2206,
      "grad_norm": 2.1651017665863037,
      "learning_rate": 3.976530612244898e-05,
      "loss": 0.197,
      "step": 1103
    },
    {
      "epoch": 0.2208,
      "grad_norm": 1.6172817945480347,
      "learning_rate": 3.975510204081633e-05,
      "loss": 0.152,
      "step": 1104
    },
    {
      "epoch": 0.221,
      "grad_norm": 1.4493610858917236,
      "learning_rate": 3.974489795918367e-05,
      "loss": 0.1325,
      "step": 1105
    },
    {
      "epoch": 0.2212,
      "grad_norm": 4.297125339508057,
      "learning_rate": 3.973469387755102e-05,
      "loss": 1.1716,
      "step": 1106
    },
    {
      "epoch": 0.2214,
      "grad_norm": 5.791296482086182,
      "learning_rate": 3.972448979591837e-05,
      "loss": 1.1987,
      "step": 1107
    },
    {
      "epoch": 0.2216,
      "grad_norm": 4.080999374389648,
      "learning_rate": 3.971428571428571e-05,
      "loss": 1.0897,
      "step": 1108
    },
    {
      "epoch": 0.2218,
      "grad_norm": 1.6441861391067505,
      "learning_rate": 3.970408163265306e-05,
      "loss": 0.1664,
      "step": 1109
    },
    {
      "epoch": 0.222,
      "grad_norm": 9.482796669006348,
      "learning_rate": 3.969387755102041e-05,
      "loss": 3.391,
      "step": 1110
    },
    {
      "epoch": 0.2222,
      "grad_norm": 6.508853435516357,
      "learning_rate": 3.968367346938775e-05,
      "loss": 2.2502,
      "step": 1111
    },
    {
      "epoch": 0.2224,
      "grad_norm": 4.284785747528076,
      "learning_rate": 3.96734693877551e-05,
      "loss": 1.0736,
      "step": 1112
    },
    {
      "epoch": 0.2226,
      "grad_norm": 1.7649943828582764,
      "learning_rate": 3.966326530612245e-05,
      "loss": 0.1619,
      "step": 1113
    },
    {
      "epoch": 0.2228,
      "grad_norm": 2.027397394180298,
      "learning_rate": 3.9653061224489794e-05,
      "loss": 0.1351,
      "step": 1114
    },
    {
      "epoch": 0.223,
      "grad_norm": 1.4270291328430176,
      "learning_rate": 3.964285714285714e-05,
      "loss": 0.1335,
      "step": 1115
    },
    {
      "epoch": 0.2232,
      "grad_norm": 1.7953479290008545,
      "learning_rate": 3.963265306122449e-05,
      "loss": 0.1167,
      "step": 1116
    },
    {
      "epoch": 0.2234,
      "grad_norm": 1.2949645519256592,
      "learning_rate": 3.9622448979591835e-05,
      "loss": 0.1138,
      "step": 1117
    },
    {
      "epoch": 0.2236,
      "grad_norm": 1.8343255519866943,
      "learning_rate": 3.9612244897959184e-05,
      "loss": 0.1137,
      "step": 1118
    },
    {
      "epoch": 0.2238,
      "grad_norm": 2.346017360687256,
      "learning_rate": 3.960204081632653e-05,
      "loss": 0.2721,
      "step": 1119
    },
    {
      "epoch": 0.224,
      "grad_norm": 34.164283752441406,
      "learning_rate": 3.9591836734693876e-05,
      "loss": 3.905,
      "step": 1120
    },
    {
      "epoch": 0.2242,
      "grad_norm": 5.671075820922852,
      "learning_rate": 3.958163265306123e-05,
      "loss": 1.4206,
      "step": 1121
    },
    {
      "epoch": 0.2244,
      "grad_norm": 1.5829075574874878,
      "learning_rate": 3.9571428571428574e-05,
      "loss": 0.2156,
      "step": 1122
    },
    {
      "epoch": 0.2246,
      "grad_norm": 16.01406478881836,
      "learning_rate": 3.956122448979592e-05,
      "loss": 2.3045,
      "step": 1123
    },
    {
      "epoch": 0.2248,
      "grad_norm": 4.0922746658325195,
      "learning_rate": 3.955102040816327e-05,
      "loss": 0.9245,
      "step": 1124
    },
    {
      "epoch": 0.225,
      "grad_norm": 1.7736269235610962,
      "learning_rate": 3.9540816326530615e-05,
      "loss": 0.1719,
      "step": 1125
    },
    {
      "epoch": 0.2252,
      "grad_norm": 1.5497605800628662,
      "learning_rate": 3.9530612244897964e-05,
      "loss": 0.139,
      "step": 1126
    },
    {
      "epoch": 0.2254,
      "grad_norm": 1.831518292427063,
      "learning_rate": 3.952040816326531e-05,
      "loss": 0.1242,
      "step": 1127
    },
    {
      "epoch": 0.2256,
      "grad_norm": 1.4932825565338135,
      "learning_rate": 3.9510204081632655e-05,
      "loss": 0.1355,
      "step": 1128
    },
    {
      "epoch": 0.2258,
      "grad_norm": 1.522917628288269,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.163,
      "step": 1129
    },
    {
      "epoch": 0.226,
      "grad_norm": 3.278972625732422,
      "learning_rate": 3.9489795918367354e-05,
      "loss": 0.3247,
      "step": 1130
    },
    {
      "epoch": 0.2262,
      "grad_norm": 1.7440088987350464,
      "learning_rate": 3.9479591836734696e-05,
      "loss": 0.0979,
      "step": 1131
    },
    {
      "epoch": 0.2264,
      "grad_norm": 4.800971031188965,
      "learning_rate": 3.9469387755102045e-05,
      "loss": 1.2838,
      "step": 1132
    },
    {
      "epoch": 0.2266,
      "grad_norm": 8.162755012512207,
      "learning_rate": 3.945918367346939e-05,
      "loss": 1.3834,
      "step": 1133
    },
    {
      "epoch": 0.2268,
      "grad_norm": 36.66396713256836,
      "learning_rate": 3.944897959183674e-05,
      "loss": 4.7365,
      "step": 1134
    },
    {
      "epoch": 0.227,
      "grad_norm": 2.0371525287628174,
      "learning_rate": 3.9438775510204086e-05,
      "loss": 0.1467,
      "step": 1135
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.2436259984970093,
      "learning_rate": 3.942857142857143e-05,
      "loss": 0.1262,
      "step": 1136
    },
    {
      "epoch": 0.2274,
      "grad_norm": 6.373444080352783,
      "learning_rate": 3.941836734693878e-05,
      "loss": 1.164,
      "step": 1137
    },
    {
      "epoch": 0.2276,
      "grad_norm": 6.302332878112793,
      "learning_rate": 3.940816326530613e-05,
      "loss": 1.3267,
      "step": 1138
    },
    {
      "epoch": 0.2278,
      "grad_norm": 1.712341070175171,
      "learning_rate": 3.939795918367347e-05,
      "loss": 0.2145,
      "step": 1139
    },
    {
      "epoch": 0.228,
      "grad_norm": 3.593083620071411,
      "learning_rate": 3.938775510204082e-05,
      "loss": 0.3416,
      "step": 1140
    },
    {
      "epoch": 0.2282,
      "grad_norm": 1.5694092512130737,
      "learning_rate": 3.937755102040817e-05,
      "loss": 0.0872,
      "step": 1141
    },
    {
      "epoch": 0.2284,
      "grad_norm": 10.168475151062012,
      "learning_rate": 3.936734693877551e-05,
      "loss": 2.4081,
      "step": 1142
    },
    {
      "epoch": 0.2286,
      "grad_norm": 1.5657432079315186,
      "learning_rate": 3.935714285714286e-05,
      "loss": 0.1414,
      "step": 1143
    },
    {
      "epoch": 0.2288,
      "grad_norm": 4.660126686096191,
      "learning_rate": 3.934693877551021e-05,
      "loss": 1.1826,
      "step": 1144
    },
    {
      "epoch": 0.229,
      "grad_norm": 11.854291915893555,
      "learning_rate": 3.933673469387755e-05,
      "loss": 3.3537,
      "step": 1145
    },
    {
      "epoch": 0.2292,
      "grad_norm": 1.510237216949463,
      "learning_rate": 3.93265306122449e-05,
      "loss": 0.0829,
      "step": 1146
    },
    {
      "epoch": 0.2294,
      "grad_norm": 4.058131217956543,
      "learning_rate": 3.931632653061225e-05,
      "loss": 0.3575,
      "step": 1147
    },
    {
      "epoch": 0.2296,
      "grad_norm": 2.597877264022827,
      "learning_rate": 3.930612244897959e-05,
      "loss": 0.8784,
      "step": 1148
    },
    {
      "epoch": 0.2298,
      "grad_norm": 1.6106457710266113,
      "learning_rate": 3.929591836734694e-05,
      "loss": 0.1452,
      "step": 1149
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.7510958909988403,
      "learning_rate": 3.928571428571429e-05,
      "loss": 0.1684,
      "step": 1150
    },
    {
      "epoch": 0.2302,
      "grad_norm": 1.4825830459594727,
      "learning_rate": 3.927551020408163e-05,
      "loss": 0.1473,
      "step": 1151
    },
    {
      "epoch": 0.2304,
      "grad_norm": 4.590761184692383,
      "learning_rate": 3.926530612244898e-05,
      "loss": 1.0117,
      "step": 1152
    },
    {
      "epoch": 0.2306,
      "grad_norm": 9.946000099182129,
      "learning_rate": 3.9255102040816324e-05,
      "loss": 1.5746,
      "step": 1153
    },
    {
      "epoch": 0.2308,
      "grad_norm": 1.388819932937622,
      "learning_rate": 3.924489795918367e-05,
      "loss": 0.0816,
      "step": 1154
    },
    {
      "epoch": 0.231,
      "grad_norm": 18.854167938232422,
      "learning_rate": 3.923469387755102e-05,
      "loss": 1.3711,
      "step": 1155
    },
    {
      "epoch": 0.2312,
      "grad_norm": 1.3573522567749023,
      "learning_rate": 3.9224489795918365e-05,
      "loss": 0.1349,
      "step": 1156
    },
    {
      "epoch": 0.2314,
      "grad_norm": 8.461785316467285,
      "learning_rate": 3.9214285714285714e-05,
      "loss": 0.6804,
      "step": 1157
    },
    {
      "epoch": 0.2316,
      "grad_norm": 1.6372792720794678,
      "learning_rate": 3.920408163265306e-05,
      "loss": 0.1442,
      "step": 1158
    },
    {
      "epoch": 0.2318,
      "grad_norm": 6.600191116333008,
      "learning_rate": 3.9193877551020406e-05,
      "loss": 1.1584,
      "step": 1159
    },
    {
      "epoch": 0.232,
      "grad_norm": 7.225102424621582,
      "learning_rate": 3.9183673469387755e-05,
      "loss": 1.2314,
      "step": 1160
    },
    {
      "epoch": 0.2322,
      "grad_norm": 1.274695634841919,
      "learning_rate": 3.9173469387755104e-05,
      "loss": 0.1083,
      "step": 1161
    },
    {
      "epoch": 0.2324,
      "grad_norm": 1.7186917066574097,
      "learning_rate": 3.9163265306122446e-05,
      "loss": 0.0973,
      "step": 1162
    },
    {
      "epoch": 0.2326,
      "grad_norm": 1.218078851699829,
      "learning_rate": 3.9153061224489795e-05,
      "loss": 0.0565,
      "step": 1163
    },
    {
      "epoch": 0.2328,
      "grad_norm": 3.951833724975586,
      "learning_rate": 3.9142857142857145e-05,
      "loss": 0.9821,
      "step": 1164
    },
    {
      "epoch": 0.233,
      "grad_norm": 1.9798835515975952,
      "learning_rate": 3.913265306122449e-05,
      "loss": 0.2112,
      "step": 1165
    },
    {
      "epoch": 0.2332,
      "grad_norm": 1.6355785131454468,
      "learning_rate": 3.9122448979591836e-05,
      "loss": 0.1587,
      "step": 1166
    },
    {
      "epoch": 0.2334,
      "grad_norm": 5.939909934997559,
      "learning_rate": 3.9112244897959185e-05,
      "loss": 0.6297,
      "step": 1167
    },
    {
      "epoch": 0.2336,
      "grad_norm": 7.891049385070801,
      "learning_rate": 3.9102040816326535e-05,
      "loss": 1.3208,
      "step": 1168
    },
    {
      "epoch": 0.2338,
      "grad_norm": 5.886658668518066,
      "learning_rate": 3.9091836734693884e-05,
      "loss": 1.0827,
      "step": 1169
    },
    {
      "epoch": 0.234,
      "grad_norm": 1.9486569166183472,
      "learning_rate": 3.9081632653061226e-05,
      "loss": 0.1653,
      "step": 1170
    },
    {
      "epoch": 0.2342,
      "grad_norm": 1.8028507232666016,
      "learning_rate": 3.9071428571428575e-05,
      "loss": 0.1124,
      "step": 1171
    },
    {
      "epoch": 0.2344,
      "grad_norm": 1.5691555738449097,
      "learning_rate": 3.9061224489795925e-05,
      "loss": 0.1285,
      "step": 1172
    },
    {
      "epoch": 0.2346,
      "grad_norm": 3.233590841293335,
      "learning_rate": 3.905102040816327e-05,
      "loss": 0.3467,
      "step": 1173
    },
    {
      "epoch": 0.2348,
      "grad_norm": 13.754486083984375,
      "learning_rate": 3.9040816326530616e-05,
      "loss": 3.2339,
      "step": 1174
    },
    {
      "epoch": 0.235,
      "grad_norm": 2.6587777137756348,
      "learning_rate": 3.9030612244897965e-05,
      "loss": 0.82,
      "step": 1175
    },
    {
      "epoch": 0.2352,
      "grad_norm": 6.628676891326904,
      "learning_rate": 3.902040816326531e-05,
      "loss": 2.2773,
      "step": 1176
    },
    {
      "epoch": 0.2354,
      "grad_norm": 1.5130740404129028,
      "learning_rate": 3.901020408163266e-05,
      "loss": 0.062,
      "step": 1177
    },
    {
      "epoch": 0.2356,
      "grad_norm": 3.9275505542755127,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.9092,
      "step": 1178
    },
    {
      "epoch": 0.2358,
      "grad_norm": 1.7313053607940674,
      "learning_rate": 3.898979591836735e-05,
      "loss": 0.1316,
      "step": 1179
    },
    {
      "epoch": 0.236,
      "grad_norm": 24.386125564575195,
      "learning_rate": 3.89795918367347e-05,
      "loss": 5.702,
      "step": 1180
    },
    {
      "epoch": 0.2362,
      "grad_norm": 4.170772075653076,
      "learning_rate": 3.896938775510205e-05,
      "loss": 0.775,
      "step": 1181
    },
    {
      "epoch": 0.2364,
      "grad_norm": 1.594424843788147,
      "learning_rate": 3.895918367346939e-05,
      "loss": 0.1186,
      "step": 1182
    },
    {
      "epoch": 0.2366,
      "grad_norm": 6.332061767578125,
      "learning_rate": 3.894897959183674e-05,
      "loss": 1.2124,
      "step": 1183
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.5418387651443481,
      "learning_rate": 3.893877551020408e-05,
      "loss": 0.0678,
      "step": 1184
    },
    {
      "epoch": 0.237,
      "grad_norm": 1.4276838302612305,
      "learning_rate": 3.892857142857143e-05,
      "loss": 0.138,
      "step": 1185
    },
    {
      "epoch": 0.2372,
      "grad_norm": 1.4515182971954346,
      "learning_rate": 3.891836734693878e-05,
      "loss": 0.1521,
      "step": 1186
    },
    {
      "epoch": 0.2374,
      "grad_norm": 3.8713340759277344,
      "learning_rate": 3.890816326530612e-05,
      "loss": 0.9481,
      "step": 1187
    },
    {
      "epoch": 0.2376,
      "grad_norm": 1.266196846961975,
      "learning_rate": 3.889795918367347e-05,
      "loss": 0.1171,
      "step": 1188
    },
    {
      "epoch": 0.2378,
      "grad_norm": 1.7428760528564453,
      "learning_rate": 3.888775510204082e-05,
      "loss": 0.1543,
      "step": 1189
    },
    {
      "epoch": 0.238,
      "grad_norm": 6.290396690368652,
      "learning_rate": 3.887755102040816e-05,
      "loss": 2.354,
      "step": 1190
    },
    {
      "epoch": 0.2382,
      "grad_norm": 1.7652369737625122,
      "learning_rate": 3.886734693877551e-05,
      "loss": 0.1633,
      "step": 1191
    },
    {
      "epoch": 0.2384,
      "grad_norm": 1.5250771045684814,
      "learning_rate": 3.885714285714286e-05,
      "loss": 0.0776,
      "step": 1192
    },
    {
      "epoch": 0.2386,
      "grad_norm": 1.7794101238250732,
      "learning_rate": 3.88469387755102e-05,
      "loss": 0.1795,
      "step": 1193
    },
    {
      "epoch": 0.2388,
      "grad_norm": 1.3927475214004517,
      "learning_rate": 3.883673469387755e-05,
      "loss": 0.1239,
      "step": 1194
    },
    {
      "epoch": 0.239,
      "grad_norm": 1.7098913192749023,
      "learning_rate": 3.88265306122449e-05,
      "loss": 0.1516,
      "step": 1195
    },
    {
      "epoch": 0.2392,
      "grad_norm": 1.2571477890014648,
      "learning_rate": 3.8816326530612244e-05,
      "loss": 0.1057,
      "step": 1196
    },
    {
      "epoch": 0.2394,
      "grad_norm": 1.652297019958496,
      "learning_rate": 3.880612244897959e-05,
      "loss": 0.1436,
      "step": 1197
    },
    {
      "epoch": 0.2396,
      "grad_norm": 13.419703483581543,
      "learning_rate": 3.879591836734694e-05,
      "loss": 3.5707,
      "step": 1198
    },
    {
      "epoch": 0.2398,
      "grad_norm": 9.41876220703125,
      "learning_rate": 3.8785714285714285e-05,
      "loss": 2.1655,
      "step": 1199
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.557036876678467,
      "learning_rate": 3.8775510204081634e-05,
      "loss": 1.1251,
      "step": 1200
    },
    {
      "epoch": 0.2402,
      "grad_norm": 1.5032589435577393,
      "learning_rate": 3.876530612244898e-05,
      "loss": 0.1279,
      "step": 1201
    },
    {
      "epoch": 0.2404,
      "grad_norm": 1.44595468044281,
      "learning_rate": 3.8755102040816325e-05,
      "loss": 0.069,
      "step": 1202
    },
    {
      "epoch": 0.2406,
      "grad_norm": 4.187497138977051,
      "learning_rate": 3.8744897959183675e-05,
      "loss": 1.0626,
      "step": 1203
    },
    {
      "epoch": 0.2408,
      "grad_norm": 1.6708344221115112,
      "learning_rate": 3.8734693877551024e-05,
      "loss": 0.1072,
      "step": 1204
    },
    {
      "epoch": 0.241,
      "grad_norm": 6.50513219833374,
      "learning_rate": 3.8724489795918366e-05,
      "loss": 1.1696,
      "step": 1205
    },
    {
      "epoch": 0.2412,
      "grad_norm": 1.7861454486846924,
      "learning_rate": 3.8714285714285715e-05,
      "loss": 0.1797,
      "step": 1206
    },
    {
      "epoch": 0.2414,
      "grad_norm": 1.772229552268982,
      "learning_rate": 3.870408163265306e-05,
      "loss": 0.0919,
      "step": 1207
    },
    {
      "epoch": 0.2416,
      "grad_norm": 1.9311400651931763,
      "learning_rate": 3.869387755102041e-05,
      "loss": 0.1622,
      "step": 1208
    },
    {
      "epoch": 0.2418,
      "grad_norm": 1.8080766201019287,
      "learning_rate": 3.8683673469387756e-05,
      "loss": 0.1087,
      "step": 1209
    },
    {
      "epoch": 0.242,
      "grad_norm": 1.8979612588882446,
      "learning_rate": 3.86734693877551e-05,
      "loss": 0.0955,
      "step": 1210
    },
    {
      "epoch": 0.2422,
      "grad_norm": 1.825605869293213,
      "learning_rate": 3.866326530612245e-05,
      "loss": 0.1084,
      "step": 1211
    },
    {
      "epoch": 0.2424,
      "grad_norm": 1.3832318782806396,
      "learning_rate": 3.8653061224489804e-05,
      "loss": 0.0901,
      "step": 1212
    },
    {
      "epoch": 0.2426,
      "grad_norm": 16.540746688842773,
      "learning_rate": 3.8642857142857146e-05,
      "loss": 3.5837,
      "step": 1213
    },
    {
      "epoch": 0.2428,
      "grad_norm": 1.9575875997543335,
      "learning_rate": 3.8632653061224495e-05,
      "loss": 0.1763,
      "step": 1214
    },
    {
      "epoch": 0.243,
      "grad_norm": 1.6776007413864136,
      "learning_rate": 3.862244897959184e-05,
      "loss": 0.1061,
      "step": 1215
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.3324201107025146,
      "learning_rate": 3.861224489795919e-05,
      "loss": 0.1453,
      "step": 1216
    },
    {
      "epoch": 0.2434,
      "grad_norm": 1.635075330734253,
      "learning_rate": 3.8602040816326536e-05,
      "loss": 0.1365,
      "step": 1217
    },
    {
      "epoch": 0.2436,
      "grad_norm": 1.4818600416183472,
      "learning_rate": 3.859183673469388e-05,
      "loss": 0.1351,
      "step": 1218
    },
    {
      "epoch": 0.2438,
      "grad_norm": 48.662357330322266,
      "learning_rate": 3.858163265306123e-05,
      "loss": 5.8292,
      "step": 1219
    },
    {
      "epoch": 0.244,
      "grad_norm": 3.794050455093384,
      "learning_rate": 3.857142857142858e-05,
      "loss": 0.9183,
      "step": 1220
    },
    {
      "epoch": 0.2442,
      "grad_norm": 1.3571511507034302,
      "learning_rate": 3.856122448979592e-05,
      "loss": 0.1196,
      "step": 1221
    },
    {
      "epoch": 0.2444,
      "grad_norm": 1.4742392301559448,
      "learning_rate": 3.855102040816327e-05,
      "loss": 0.1399,
      "step": 1222
    },
    {
      "epoch": 0.2446,
      "grad_norm": 18.71346664428711,
      "learning_rate": 3.854081632653062e-05,
      "loss": 5.4186,
      "step": 1223
    },
    {
      "epoch": 0.2448,
      "grad_norm": 1.8707962036132812,
      "learning_rate": 3.853061224489796e-05,
      "loss": 0.1217,
      "step": 1224
    },
    {
      "epoch": 0.245,
      "grad_norm": 1.4898655414581299,
      "learning_rate": 3.852040816326531e-05,
      "loss": 0.141,
      "step": 1225
    },
    {
      "epoch": 0.2452,
      "grad_norm": 3.1395351886749268,
      "learning_rate": 3.851020408163266e-05,
      "loss": 0.8529,
      "step": 1226
    },
    {
      "epoch": 0.2454,
      "grad_norm": 1.0336296558380127,
      "learning_rate": 3.85e-05,
      "loss": 0.0862,
      "step": 1227
    },
    {
      "epoch": 0.2456,
      "grad_norm": 1.612135648727417,
      "learning_rate": 3.848979591836735e-05,
      "loss": 0.1439,
      "step": 1228
    },
    {
      "epoch": 0.2458,
      "grad_norm": 1.3654370307922363,
      "learning_rate": 3.84795918367347e-05,
      "loss": 0.1077,
      "step": 1229
    },
    {
      "epoch": 0.246,
      "grad_norm": 3.510406255722046,
      "learning_rate": 3.846938775510204e-05,
      "loss": 0.8932,
      "step": 1230
    },
    {
      "epoch": 0.2462,
      "grad_norm": 1.8323819637298584,
      "learning_rate": 3.845918367346939e-05,
      "loss": 0.1688,
      "step": 1231
    },
    {
      "epoch": 0.2464,
      "grad_norm": 3.9554245471954346,
      "learning_rate": 3.844897959183674e-05,
      "loss": 1.0654,
      "step": 1232
    },
    {
      "epoch": 0.2466,
      "grad_norm": 3.3375043869018555,
      "learning_rate": 3.843877551020408e-05,
      "loss": 0.865,
      "step": 1233
    },
    {
      "epoch": 0.2468,
      "grad_norm": 63.8090705871582,
      "learning_rate": 3.842857142857143e-05,
      "loss": 6.3251,
      "step": 1234
    },
    {
      "epoch": 0.247,
      "grad_norm": 1.4541479349136353,
      "learning_rate": 3.841836734693878e-05,
      "loss": 0.0772,
      "step": 1235
    },
    {
      "epoch": 0.2472,
      "grad_norm": 1.5204209089279175,
      "learning_rate": 3.840816326530612e-05,
      "loss": 0.1428,
      "step": 1236
    },
    {
      "epoch": 0.2474,
      "grad_norm": 6.136209011077881,
      "learning_rate": 3.839795918367347e-05,
      "loss": 1.1625,
      "step": 1237
    },
    {
      "epoch": 0.2476,
      "grad_norm": 1.3140615224838257,
      "learning_rate": 3.8387755102040815e-05,
      "loss": 0.1221,
      "step": 1238
    },
    {
      "epoch": 0.2478,
      "grad_norm": 1.2308504581451416,
      "learning_rate": 3.8377551020408164e-05,
      "loss": 0.1003,
      "step": 1239
    },
    {
      "epoch": 0.248,
      "grad_norm": 6.357258319854736,
      "learning_rate": 3.836734693877551e-05,
      "loss": 2.259,
      "step": 1240
    },
    {
      "epoch": 0.2482,
      "grad_norm": 1.3247007131576538,
      "learning_rate": 3.8357142857142855e-05,
      "loss": 0.0658,
      "step": 1241
    },
    {
      "epoch": 0.2484,
      "grad_norm": 1.5625752210617065,
      "learning_rate": 3.8346938775510205e-05,
      "loss": 0.1311,
      "step": 1242
    },
    {
      "epoch": 0.2486,
      "grad_norm": 1.7948263883590698,
      "learning_rate": 3.8336734693877554e-05,
      "loss": 0.1043,
      "step": 1243
    },
    {
      "epoch": 0.2488,
      "grad_norm": 4.615420818328857,
      "learning_rate": 3.8326530612244896e-05,
      "loss": 1.2485,
      "step": 1244
    },
    {
      "epoch": 0.249,
      "grad_norm": 3.948204517364502,
      "learning_rate": 3.8316326530612245e-05,
      "loss": 0.971,
      "step": 1245
    },
    {
      "epoch": 0.2492,
      "grad_norm": 10.756133079528809,
      "learning_rate": 3.8306122448979594e-05,
      "loss": 2.3515,
      "step": 1246
    },
    {
      "epoch": 0.2494,
      "grad_norm": 19.966793060302734,
      "learning_rate": 3.829591836734694e-05,
      "loss": 3.2037,
      "step": 1247
    },
    {
      "epoch": 0.2496,
      "grad_norm": 1.7232310771942139,
      "learning_rate": 3.8285714285714286e-05,
      "loss": 0.086,
      "step": 1248
    },
    {
      "epoch": 0.2498,
      "grad_norm": 15.161467552185059,
      "learning_rate": 3.8275510204081635e-05,
      "loss": 3.2497,
      "step": 1249
    },
    {
      "epoch": 0.25,
      "grad_norm": 5.99517297744751,
      "learning_rate": 3.826530612244898e-05,
      "loss": 2.1919,
      "step": 1250
    },
    {
      "epoch": 0.2502,
      "grad_norm": 1.4167498350143433,
      "learning_rate": 3.825510204081633e-05,
      "loss": 0.1137,
      "step": 1251
    },
    {
      "epoch": 0.2504,
      "grad_norm": 2.161652088165283,
      "learning_rate": 3.8244897959183676e-05,
      "loss": 0.215,
      "step": 1252
    },
    {
      "epoch": 0.2506,
      "grad_norm": 1.7977770566940308,
      "learning_rate": 3.823469387755102e-05,
      "loss": 0.1669,
      "step": 1253
    },
    {
      "epoch": 0.2508,
      "grad_norm": 2.0701310634613037,
      "learning_rate": 3.822448979591837e-05,
      "loss": 0.2376,
      "step": 1254
    },
    {
      "epoch": 0.251,
      "grad_norm": 4.391334533691406,
      "learning_rate": 3.821428571428572e-05,
      "loss": 1.3211,
      "step": 1255
    },
    {
      "epoch": 0.2512,
      "grad_norm": 1.8493633270263672,
      "learning_rate": 3.820408163265306e-05,
      "loss": 0.1635,
      "step": 1256
    },
    {
      "epoch": 0.2514,
      "grad_norm": 3.1207923889160156,
      "learning_rate": 3.8193877551020415e-05,
      "loss": 0.3237,
      "step": 1257
    },
    {
      "epoch": 0.2516,
      "grad_norm": 4.7217326164245605,
      "learning_rate": 3.818367346938776e-05,
      "loss": 0.7855,
      "step": 1258
    },
    {
      "epoch": 0.2518,
      "grad_norm": 4.3752827644348145,
      "learning_rate": 3.817346938775511e-05,
      "loss": 1.0889,
      "step": 1259
    },
    {
      "epoch": 0.252,
      "grad_norm": 1.588794231414795,
      "learning_rate": 3.8163265306122456e-05,
      "loss": 0.1937,
      "step": 1260
    },
    {
      "epoch": 0.2522,
      "grad_norm": 1.5312442779541016,
      "learning_rate": 3.81530612244898e-05,
      "loss": 0.1339,
      "step": 1261
    },
    {
      "epoch": 0.2524,
      "grad_norm": 54.07605743408203,
      "learning_rate": 3.814285714285715e-05,
      "loss": 4.1722,
      "step": 1262
    },
    {
      "epoch": 0.2526,
      "grad_norm": 1.6507775783538818,
      "learning_rate": 3.81326530612245e-05,
      "loss": 0.178,
      "step": 1263
    },
    {
      "epoch": 0.2528,
      "grad_norm": 3.3244357109069824,
      "learning_rate": 3.812244897959184e-05,
      "loss": 0.8281,
      "step": 1264
    },
    {
      "epoch": 0.253,
      "grad_norm": 2.177236795425415,
      "learning_rate": 3.811224489795919e-05,
      "loss": 0.124,
      "step": 1265
    },
    {
      "epoch": 0.2532,
      "grad_norm": 12.845377922058105,
      "learning_rate": 3.810204081632654e-05,
      "loss": 2.2881,
      "step": 1266
    },
    {
      "epoch": 0.2534,
      "grad_norm": 38.4703369140625,
      "learning_rate": 3.809183673469388e-05,
      "loss": 3.1562,
      "step": 1267
    },
    {
      "epoch": 0.2536,
      "grad_norm": 3.0963709354400635,
      "learning_rate": 3.808163265306123e-05,
      "loss": 0.8849,
      "step": 1268
    },
    {
      "epoch": 0.2538,
      "grad_norm": 1.7664254903793335,
      "learning_rate": 3.807142857142857e-05,
      "loss": 0.152,
      "step": 1269
    },
    {
      "epoch": 0.254,
      "grad_norm": 4.133462429046631,
      "learning_rate": 3.806122448979592e-05,
      "loss": 1.0243,
      "step": 1270
    },
    {
      "epoch": 0.2542,
      "grad_norm": 1.334006428718567,
      "learning_rate": 3.805102040816327e-05,
      "loss": 0.1479,
      "step": 1271
    },
    {
      "epoch": 0.2544,
      "grad_norm": 1.698577642440796,
      "learning_rate": 3.804081632653061e-05,
      "loss": 0.1708,
      "step": 1272
    },
    {
      "epoch": 0.2546,
      "grad_norm": 2.968167781829834,
      "learning_rate": 3.803061224489796e-05,
      "loss": 0.1021,
      "step": 1273
    },
    {
      "epoch": 0.2548,
      "grad_norm": 1.6637413501739502,
      "learning_rate": 3.802040816326531e-05,
      "loss": 0.1823,
      "step": 1274
    },
    {
      "epoch": 0.255,
      "grad_norm": 1.5605759620666504,
      "learning_rate": 3.801020408163265e-05,
      "loss": 0.0836,
      "step": 1275
    },
    {
      "epoch": 0.2552,
      "grad_norm": 1.4904472827911377,
      "learning_rate": 3.8e-05,
      "loss": 0.0866,
      "step": 1276
    },
    {
      "epoch": 0.2554,
      "grad_norm": 6.826018810272217,
      "learning_rate": 3.798979591836735e-05,
      "loss": 1.1779,
      "step": 1277
    },
    {
      "epoch": 0.2556,
      "grad_norm": 3.469442844390869,
      "learning_rate": 3.7979591836734694e-05,
      "loss": 0.866,
      "step": 1278
    },
    {
      "epoch": 0.2558,
      "grad_norm": 15.142460823059082,
      "learning_rate": 3.796938775510204e-05,
      "loss": 1.5539,
      "step": 1279
    },
    {
      "epoch": 0.256,
      "grad_norm": 1.6873462200164795,
      "learning_rate": 3.795918367346939e-05,
      "loss": 0.1636,
      "step": 1280
    },
    {
      "epoch": 0.2562,
      "grad_norm": 1.8198899030685425,
      "learning_rate": 3.7948979591836734e-05,
      "loss": 0.1531,
      "step": 1281
    },
    {
      "epoch": 0.2564,
      "grad_norm": 35.50392532348633,
      "learning_rate": 3.7938775510204084e-05,
      "loss": 5.2858,
      "step": 1282
    },
    {
      "epoch": 0.2566,
      "grad_norm": 4.1826491355896,
      "learning_rate": 3.792857142857143e-05,
      "loss": 0.7081,
      "step": 1283
    },
    {
      "epoch": 0.2568,
      "grad_norm": 1.536327838897705,
      "learning_rate": 3.7918367346938775e-05,
      "loss": 0.0921,
      "step": 1284
    },
    {
      "epoch": 0.257,
      "grad_norm": 7.594724655151367,
      "learning_rate": 3.7908163265306124e-05,
      "loss": 1.3357,
      "step": 1285
    },
    {
      "epoch": 0.2572,
      "grad_norm": 1.5905827283859253,
      "learning_rate": 3.7897959183673474e-05,
      "loss": 0.1269,
      "step": 1286
    },
    {
      "epoch": 0.2574,
      "grad_norm": 3.3080477714538574,
      "learning_rate": 3.7887755102040816e-05,
      "loss": 0.8568,
      "step": 1287
    },
    {
      "epoch": 0.2576,
      "grad_norm": 1.4824048280715942,
      "learning_rate": 3.7877551020408165e-05,
      "loss": 0.0757,
      "step": 1288
    },
    {
      "epoch": 0.2578,
      "grad_norm": 2.0729236602783203,
      "learning_rate": 3.786734693877551e-05,
      "loss": 0.1869,
      "step": 1289
    },
    {
      "epoch": 0.258,
      "grad_norm": 1.4475592374801636,
      "learning_rate": 3.785714285714286e-05,
      "loss": 0.1157,
      "step": 1290
    },
    {
      "epoch": 0.2582,
      "grad_norm": 2.0196914672851562,
      "learning_rate": 3.7846938775510206e-05,
      "loss": 0.0811,
      "step": 1291
    },
    {
      "epoch": 0.2584,
      "grad_norm": 3.5122108459472656,
      "learning_rate": 3.783673469387755e-05,
      "loss": 0.8649,
      "step": 1292
    },
    {
      "epoch": 0.2586,
      "grad_norm": 31.961008071899414,
      "learning_rate": 3.78265306122449e-05,
      "loss": 3.9754,
      "step": 1293
    },
    {
      "epoch": 0.2588,
      "grad_norm": 3.7600467205047607,
      "learning_rate": 3.781632653061225e-05,
      "loss": 0.287,
      "step": 1294
    },
    {
      "epoch": 0.259,
      "grad_norm": 1.7909352779388428,
      "learning_rate": 3.780612244897959e-05,
      "loss": 0.1029,
      "step": 1295
    },
    {
      "epoch": 0.2592,
      "grad_norm": 1.7947596311569214,
      "learning_rate": 3.779591836734694e-05,
      "loss": 0.1077,
      "step": 1296
    },
    {
      "epoch": 0.2594,
      "grad_norm": 1.495173692703247,
      "learning_rate": 3.778571428571429e-05,
      "loss": 0.1247,
      "step": 1297
    },
    {
      "epoch": 0.2596,
      "grad_norm": 1.5115487575531006,
      "learning_rate": 3.777551020408163e-05,
      "loss": 0.0777,
      "step": 1298
    },
    {
      "epoch": 0.2598,
      "grad_norm": 8.24157428741455,
      "learning_rate": 3.776530612244898e-05,
      "loss": 1.4188,
      "step": 1299
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.461794376373291,
      "learning_rate": 3.775510204081633e-05,
      "loss": 0.1331,
      "step": 1300
    },
    {
      "epoch": 0.2602,
      "grad_norm": 1.6198186874389648,
      "learning_rate": 3.774489795918367e-05,
      "loss": 0.0813,
      "step": 1301
    },
    {
      "epoch": 0.2604,
      "grad_norm": 1.730018138885498,
      "learning_rate": 3.773469387755102e-05,
      "loss": 0.0997,
      "step": 1302
    },
    {
      "epoch": 0.2606,
      "grad_norm": 1.579390287399292,
      "learning_rate": 3.772448979591837e-05,
      "loss": 0.085,
      "step": 1303
    },
    {
      "epoch": 0.2608,
      "grad_norm": 1.8548482656478882,
      "learning_rate": 3.771428571428572e-05,
      "loss": 0.1627,
      "step": 1304
    },
    {
      "epoch": 0.261,
      "grad_norm": 3.0801405906677246,
      "learning_rate": 3.770408163265307e-05,
      "loss": 0.3232,
      "step": 1305
    },
    {
      "epoch": 0.2612,
      "grad_norm": 2.0121631622314453,
      "learning_rate": 3.769387755102041e-05,
      "loss": 0.1992,
      "step": 1306
    },
    {
      "epoch": 0.2614,
      "grad_norm": 7.033249855041504,
      "learning_rate": 3.768367346938776e-05,
      "loss": 2.3145,
      "step": 1307
    },
    {
      "epoch": 0.2616,
      "grad_norm": 6.946961402893066,
      "learning_rate": 3.767346938775511e-05,
      "loss": 1.1799,
      "step": 1308
    },
    {
      "epoch": 0.2618,
      "grad_norm": 1.6728273630142212,
      "learning_rate": 3.766326530612245e-05,
      "loss": 0.1331,
      "step": 1309
    },
    {
      "epoch": 0.262,
      "grad_norm": 3.6165990829467773,
      "learning_rate": 3.76530612244898e-05,
      "loss": 0.8752,
      "step": 1310
    },
    {
      "epoch": 0.2622,
      "grad_norm": 2.263660430908203,
      "learning_rate": 3.764285714285715e-05,
      "loss": 0.2218,
      "step": 1311
    },
    {
      "epoch": 0.2624,
      "grad_norm": 23.657312393188477,
      "learning_rate": 3.763265306122449e-05,
      "loss": 5.4844,
      "step": 1312
    },
    {
      "epoch": 0.2626,
      "grad_norm": 6.865463733673096,
      "learning_rate": 3.762244897959184e-05,
      "loss": 1.2535,
      "step": 1313
    },
    {
      "epoch": 0.2628,
      "grad_norm": 18.283000946044922,
      "learning_rate": 3.761224489795919e-05,
      "loss": 4.1451,
      "step": 1314
    },
    {
      "epoch": 0.263,
      "grad_norm": 18.092365264892578,
      "learning_rate": 3.760204081632653e-05,
      "loss": 4.2177,
      "step": 1315
    },
    {
      "epoch": 0.2632,
      "grad_norm": 14.155285835266113,
      "learning_rate": 3.759183673469388e-05,
      "loss": 5.4509,
      "step": 1316
    },
    {
      "epoch": 0.2634,
      "grad_norm": 1.3538949489593506,
      "learning_rate": 3.758163265306123e-05,
      "loss": 0.1153,
      "step": 1317
    },
    {
      "epoch": 0.2636,
      "grad_norm": 4.133082389831543,
      "learning_rate": 3.757142857142857e-05,
      "loss": 0.738,
      "step": 1318
    },
    {
      "epoch": 0.2638,
      "grad_norm": 1.8121343851089478,
      "learning_rate": 3.756122448979592e-05,
      "loss": 0.1799,
      "step": 1319
    },
    {
      "epoch": 0.264,
      "grad_norm": 1.3098820447921753,
      "learning_rate": 3.7551020408163264e-05,
      "loss": 0.0949,
      "step": 1320
    },
    {
      "epoch": 0.2642,
      "grad_norm": 7.844435691833496,
      "learning_rate": 3.7540816326530614e-05,
      "loss": 1.4964,
      "step": 1321
    },
    {
      "epoch": 0.2644,
      "grad_norm": 5.921436786651611,
      "learning_rate": 3.753061224489796e-05,
      "loss": 1.2696,
      "step": 1322
    },
    {
      "epoch": 0.2646,
      "grad_norm": 1.8828604221343994,
      "learning_rate": 3.7520408163265305e-05,
      "loss": 0.1063,
      "step": 1323
    },
    {
      "epoch": 0.2648,
      "grad_norm": 4.386382102966309,
      "learning_rate": 3.7510204081632654e-05,
      "loss": 0.7707,
      "step": 1324
    },
    {
      "epoch": 0.265,
      "grad_norm": 6.917270660400391,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.2423,
      "step": 1325
    },
    {
      "epoch": 0.2652,
      "grad_norm": 1.765566110610962,
      "learning_rate": 3.7489795918367346e-05,
      "loss": 0.1591,
      "step": 1326
    },
    {
      "epoch": 0.2654,
      "grad_norm": 1.6731854677200317,
      "learning_rate": 3.7479591836734695e-05,
      "loss": 0.0748,
      "step": 1327
    },
    {
      "epoch": 0.2656,
      "grad_norm": 4.075490951538086,
      "learning_rate": 3.7469387755102044e-05,
      "loss": 0.9095,
      "step": 1328
    },
    {
      "epoch": 0.2658,
      "grad_norm": 11.133137702941895,
      "learning_rate": 3.745918367346939e-05,
      "loss": 1.4753,
      "step": 1329
    },
    {
      "epoch": 0.266,
      "grad_norm": 1.6695935726165771,
      "learning_rate": 3.7448979591836736e-05,
      "loss": 0.1598,
      "step": 1330
    },
    {
      "epoch": 0.2662,
      "grad_norm": 1.5798397064208984,
      "learning_rate": 3.7438775510204085e-05,
      "loss": 0.1487,
      "step": 1331
    },
    {
      "epoch": 0.2664,
      "grad_norm": 1.7983241081237793,
      "learning_rate": 3.742857142857143e-05,
      "loss": 0.1521,
      "step": 1332
    },
    {
      "epoch": 0.2666,
      "grad_norm": 10.795758247375488,
      "learning_rate": 3.7418367346938777e-05,
      "loss": 1.1869,
      "step": 1333
    },
    {
      "epoch": 0.2668,
      "grad_norm": 3.6086666584014893,
      "learning_rate": 3.7408163265306126e-05,
      "loss": 0.8384,
      "step": 1334
    },
    {
      "epoch": 0.267,
      "grad_norm": 5.579383373260498,
      "learning_rate": 3.739795918367347e-05,
      "loss": 1.0828,
      "step": 1335
    },
    {
      "epoch": 0.2672,
      "grad_norm": 1.3698099851608276,
      "learning_rate": 3.738775510204082e-05,
      "loss": 0.1315,
      "step": 1336
    },
    {
      "epoch": 0.2674,
      "grad_norm": 1.4228986501693726,
      "learning_rate": 3.7377551020408167e-05,
      "loss": 0.0627,
      "step": 1337
    },
    {
      "epoch": 0.2676,
      "grad_norm": 5.64273738861084,
      "learning_rate": 3.736734693877551e-05,
      "loss": 1.0982,
      "step": 1338
    },
    {
      "epoch": 0.2678,
      "grad_norm": 1.7403396368026733,
      "learning_rate": 3.735714285714286e-05,
      "loss": 0.0856,
      "step": 1339
    },
    {
      "epoch": 0.268,
      "grad_norm": 6.413857460021973,
      "learning_rate": 3.734693877551021e-05,
      "loss": 1.4826,
      "step": 1340
    },
    {
      "epoch": 0.2682,
      "grad_norm": 1.5933765172958374,
      "learning_rate": 3.733673469387755e-05,
      "loss": 0.0893,
      "step": 1341
    },
    {
      "epoch": 0.2684,
      "grad_norm": 1.4069416522979736,
      "learning_rate": 3.73265306122449e-05,
      "loss": 0.0718,
      "step": 1342
    },
    {
      "epoch": 0.2686,
      "grad_norm": 5.735523223876953,
      "learning_rate": 3.731632653061224e-05,
      "loss": 1.4975,
      "step": 1343
    },
    {
      "epoch": 0.2688,
      "grad_norm": 1.6116992235183716,
      "learning_rate": 3.730612244897959e-05,
      "loss": 0.1336,
      "step": 1344
    },
    {
      "epoch": 0.269,
      "grad_norm": 1.9418095350265503,
      "learning_rate": 3.729591836734694e-05,
      "loss": 0.0917,
      "step": 1345
    },
    {
      "epoch": 0.2692,
      "grad_norm": 1.7028329372406006,
      "learning_rate": 3.728571428571428e-05,
      "loss": 0.097,
      "step": 1346
    },
    {
      "epoch": 0.2694,
      "grad_norm": 1.7631217241287231,
      "learning_rate": 3.727551020408163e-05,
      "loss": 0.0936,
      "step": 1347
    },
    {
      "epoch": 0.2696,
      "grad_norm": 22.473411560058594,
      "learning_rate": 3.726530612244899e-05,
      "loss": 3.687,
      "step": 1348
    },
    {
      "epoch": 0.2698,
      "grad_norm": 1.4212217330932617,
      "learning_rate": 3.725510204081633e-05,
      "loss": 0.1405,
      "step": 1349
    },
    {
      "epoch": 0.27,
      "grad_norm": 3.952842950820923,
      "learning_rate": 3.724489795918368e-05,
      "loss": 0.6278,
      "step": 1350
    },
    {
      "epoch": 0.2702,
      "grad_norm": 1.5269299745559692,
      "learning_rate": 3.723469387755102e-05,
      "loss": 0.1122,
      "step": 1351
    },
    {
      "epoch": 0.2704,
      "grad_norm": 1.2545517683029175,
      "learning_rate": 3.722448979591837e-05,
      "loss": 0.101,
      "step": 1352
    },
    {
      "epoch": 0.2706,
      "grad_norm": 1.3637840747833252,
      "learning_rate": 3.721428571428572e-05,
      "loss": 0.1407,
      "step": 1353
    },
    {
      "epoch": 0.2708,
      "grad_norm": 1.5960081815719604,
      "learning_rate": 3.720408163265306e-05,
      "loss": 0.0814,
      "step": 1354
    },
    {
      "epoch": 0.271,
      "grad_norm": 1.6952459812164307,
      "learning_rate": 3.719387755102041e-05,
      "loss": 0.1448,
      "step": 1355
    },
    {
      "epoch": 0.2712,
      "grad_norm": 6.112464904785156,
      "learning_rate": 3.718367346938776e-05,
      "loss": 1.0341,
      "step": 1356
    },
    {
      "epoch": 0.2714,
      "grad_norm": 3.779942512512207,
      "learning_rate": 3.71734693877551e-05,
      "loss": 0.8778,
      "step": 1357
    },
    {
      "epoch": 0.2716,
      "grad_norm": 1.5804752111434937,
      "learning_rate": 3.716326530612245e-05,
      "loss": 0.0789,
      "step": 1358
    },
    {
      "epoch": 0.2718,
      "grad_norm": 1.850809931755066,
      "learning_rate": 3.71530612244898e-05,
      "loss": 0.2246,
      "step": 1359
    },
    {
      "epoch": 0.272,
      "grad_norm": 3.5643880367279053,
      "learning_rate": 3.7142857142857143e-05,
      "loss": 0.8722,
      "step": 1360
    },
    {
      "epoch": 0.2722,
      "grad_norm": 1.2772294282913208,
      "learning_rate": 3.713265306122449e-05,
      "loss": 0.1378,
      "step": 1361
    },
    {
      "epoch": 0.2724,
      "grad_norm": 1.5688132047653198,
      "learning_rate": 3.712244897959184e-05,
      "loss": 0.1225,
      "step": 1362
    },
    {
      "epoch": 0.2726,
      "grad_norm": 4.985512733459473,
      "learning_rate": 3.7112244897959184e-05,
      "loss": 1.0646,
      "step": 1363
    },
    {
      "epoch": 0.2728,
      "grad_norm": 1.7303472757339478,
      "learning_rate": 3.7102040816326533e-05,
      "loss": 0.0765,
      "step": 1364
    },
    {
      "epoch": 0.273,
      "grad_norm": 1.4362291097640991,
      "learning_rate": 3.709183673469388e-05,
      "loss": 0.0705,
      "step": 1365
    },
    {
      "epoch": 0.2732,
      "grad_norm": 4.591298580169678,
      "learning_rate": 3.7081632653061225e-05,
      "loss": 0.7663,
      "step": 1366
    },
    {
      "epoch": 0.2734,
      "grad_norm": 1.4391281604766846,
      "learning_rate": 3.7071428571428574e-05,
      "loss": 0.1558,
      "step": 1367
    },
    {
      "epoch": 0.2736,
      "grad_norm": 22.59450912475586,
      "learning_rate": 3.706122448979592e-05,
      "loss": 4.1362,
      "step": 1368
    },
    {
      "epoch": 0.2738,
      "grad_norm": 1.5219882726669312,
      "learning_rate": 3.7051020408163266e-05,
      "loss": 0.1387,
      "step": 1369
    },
    {
      "epoch": 0.274,
      "grad_norm": 1.5087683200836182,
      "learning_rate": 3.7040816326530615e-05,
      "loss": 0.1129,
      "step": 1370
    },
    {
      "epoch": 0.2742,
      "grad_norm": 1.8149199485778809,
      "learning_rate": 3.7030612244897964e-05,
      "loss": 0.1189,
      "step": 1371
    },
    {
      "epoch": 0.2744,
      "grad_norm": 9.546114921569824,
      "learning_rate": 3.7020408163265307e-05,
      "loss": 2.0722,
      "step": 1372
    },
    {
      "epoch": 0.2746,
      "grad_norm": 1.3203532695770264,
      "learning_rate": 3.7010204081632656e-05,
      "loss": 0.1022,
      "step": 1373
    },
    {
      "epoch": 0.2748,
      "grad_norm": 8.20767593383789,
      "learning_rate": 3.7e-05,
      "loss": 2.306,
      "step": 1374
    },
    {
      "epoch": 0.275,
      "grad_norm": 4.272182464599609,
      "learning_rate": 3.698979591836735e-05,
      "loss": 0.6594,
      "step": 1375
    },
    {
      "epoch": 0.2752,
      "grad_norm": 12.107483863830566,
      "learning_rate": 3.6979591836734696e-05,
      "loss": 4.1481,
      "step": 1376
    },
    {
      "epoch": 0.2754,
      "grad_norm": 6.796584606170654,
      "learning_rate": 3.696938775510204e-05,
      "loss": 2.4405,
      "step": 1377
    },
    {
      "epoch": 0.2756,
      "grad_norm": 10.60947036743164,
      "learning_rate": 3.695918367346939e-05,
      "loss": 4.4079,
      "step": 1378
    },
    {
      "epoch": 0.2758,
      "grad_norm": 74.99689483642578,
      "learning_rate": 3.694897959183674e-05,
      "loss": 5.6219,
      "step": 1379
    },
    {
      "epoch": 0.276,
      "grad_norm": 14.767340660095215,
      "learning_rate": 3.693877551020408e-05,
      "loss": 2.0869,
      "step": 1380
    },
    {
      "epoch": 0.2762,
      "grad_norm": 19.619022369384766,
      "learning_rate": 3.692857142857143e-05,
      "loss": 1.5368,
      "step": 1381
    },
    {
      "epoch": 0.2764,
      "grad_norm": 4.810040473937988,
      "learning_rate": 3.691836734693878e-05,
      "loss": 1.0624,
      "step": 1382
    },
    {
      "epoch": 0.2766,
      "grad_norm": 1.496702790260315,
      "learning_rate": 3.690816326530612e-05,
      "loss": 0.0921,
      "step": 1383
    },
    {
      "epoch": 0.2768,
      "grad_norm": 61.29378890991211,
      "learning_rate": 3.689795918367347e-05,
      "loss": 5.6475,
      "step": 1384
    },
    {
      "epoch": 0.277,
      "grad_norm": 1.4522367715835571,
      "learning_rate": 3.688775510204082e-05,
      "loss": 0.1096,
      "step": 1385
    },
    {
      "epoch": 0.2772,
      "grad_norm": 1.9087193012237549,
      "learning_rate": 3.687755102040816e-05,
      "loss": 0.0828,
      "step": 1386
    },
    {
      "epoch": 0.2774,
      "grad_norm": 35.92570495605469,
      "learning_rate": 3.686734693877551e-05,
      "loss": 5.3491,
      "step": 1387
    },
    {
      "epoch": 0.2776,
      "grad_norm": 2.087874412536621,
      "learning_rate": 3.685714285714286e-05,
      "loss": 0.2269,
      "step": 1388
    },
    {
      "epoch": 0.2778,
      "grad_norm": 21.29840850830078,
      "learning_rate": 3.68469387755102e-05,
      "loss": 4.7003,
      "step": 1389
    },
    {
      "epoch": 0.278,
      "grad_norm": 1.5787270069122314,
      "learning_rate": 3.683673469387755e-05,
      "loss": 0.1242,
      "step": 1390
    },
    {
      "epoch": 0.2782,
      "grad_norm": 2.447349786758423,
      "learning_rate": 3.68265306122449e-05,
      "loss": 0.0794,
      "step": 1391
    },
    {
      "epoch": 0.2784,
      "grad_norm": 1.8658487796783447,
      "learning_rate": 3.681632653061224e-05,
      "loss": 0.0941,
      "step": 1392
    },
    {
      "epoch": 0.2786,
      "grad_norm": 4.186209201812744,
      "learning_rate": 3.680612244897959e-05,
      "loss": 0.7502,
      "step": 1393
    },
    {
      "epoch": 0.2788,
      "grad_norm": 3.0826706886291504,
      "learning_rate": 3.679591836734694e-05,
      "loss": 0.0862,
      "step": 1394
    },
    {
      "epoch": 0.279,
      "grad_norm": 17.367568969726562,
      "learning_rate": 3.678571428571429e-05,
      "loss": 1.3791,
      "step": 1395
    },
    {
      "epoch": 0.2792,
      "grad_norm": 1.1378870010375977,
      "learning_rate": 3.677551020408164e-05,
      "loss": 0.0919,
      "step": 1396
    },
    {
      "epoch": 0.2794,
      "grad_norm": 1.4609671831130981,
      "learning_rate": 3.676530612244898e-05,
      "loss": 0.0793,
      "step": 1397
    },
    {
      "epoch": 0.2796,
      "grad_norm": 6.895241737365723,
      "learning_rate": 3.675510204081633e-05,
      "loss": 2.2601,
      "step": 1398
    },
    {
      "epoch": 0.2798,
      "grad_norm": 1.5115087032318115,
      "learning_rate": 3.674489795918368e-05,
      "loss": 0.0784,
      "step": 1399
    },
    {
      "epoch": 0.28,
      "grad_norm": 3.2029130458831787,
      "learning_rate": 3.673469387755102e-05,
      "loss": 0.3369,
      "step": 1400
    },
    {
      "epoch": 0.2802,
      "grad_norm": 1.7710728645324707,
      "learning_rate": 3.672448979591837e-05,
      "loss": 0.1016,
      "step": 1401
    },
    {
      "epoch": 0.2804,
      "grad_norm": 1.6437673568725586,
      "learning_rate": 3.671428571428572e-05,
      "loss": 0.0811,
      "step": 1402
    },
    {
      "epoch": 0.2806,
      "grad_norm": 1.339395523071289,
      "learning_rate": 3.670408163265306e-05,
      "loss": 0.1016,
      "step": 1403
    },
    {
      "epoch": 0.2808,
      "grad_norm": 12.196369171142578,
      "learning_rate": 3.669387755102041e-05,
      "loss": 4.7879,
      "step": 1404
    },
    {
      "epoch": 0.281,
      "grad_norm": 1.3435919284820557,
      "learning_rate": 3.6683673469387755e-05,
      "loss": 0.1578,
      "step": 1405
    },
    {
      "epoch": 0.2812,
      "grad_norm": 1.9169238805770874,
      "learning_rate": 3.6673469387755104e-05,
      "loss": 0.099,
      "step": 1406
    },
    {
      "epoch": 0.2814,
      "grad_norm": 6.315728187561035,
      "learning_rate": 3.666326530612245e-05,
      "loss": 1.0508,
      "step": 1407
    },
    {
      "epoch": 0.2816,
      "grad_norm": 23.689496994018555,
      "learning_rate": 3.6653061224489796e-05,
      "loss": 5.2598,
      "step": 1408
    },
    {
      "epoch": 0.2818,
      "grad_norm": 17.41273307800293,
      "learning_rate": 3.6642857142857145e-05,
      "loss": 1.1995,
      "step": 1409
    },
    {
      "epoch": 0.282,
      "grad_norm": 5.75798225402832,
      "learning_rate": 3.6632653061224494e-05,
      "loss": 1.074,
      "step": 1410
    },
    {
      "epoch": 0.2822,
      "grad_norm": 1.9065535068511963,
      "learning_rate": 3.6622448979591836e-05,
      "loss": 0.2362,
      "step": 1411
    },
    {
      "epoch": 0.2824,
      "grad_norm": 1.2875878810882568,
      "learning_rate": 3.6612244897959186e-05,
      "loss": 0.1163,
      "step": 1412
    },
    {
      "epoch": 0.2826,
      "grad_norm": 3.6460695266723633,
      "learning_rate": 3.6602040816326535e-05,
      "loss": 0.9125,
      "step": 1413
    },
    {
      "epoch": 0.2828,
      "grad_norm": 1.1291654109954834,
      "learning_rate": 3.659183673469388e-05,
      "loss": 0.0723,
      "step": 1414
    },
    {
      "epoch": 0.283,
      "grad_norm": 8.621419906616211,
      "learning_rate": 3.6581632653061226e-05,
      "loss": 1.9511,
      "step": 1415
    },
    {
      "epoch": 0.2832,
      "grad_norm": 4.533262252807617,
      "learning_rate": 3.6571428571428576e-05,
      "loss": 1.0461,
      "step": 1416
    },
    {
      "epoch": 0.2834,
      "grad_norm": 1.8045939207077026,
      "learning_rate": 3.656122448979592e-05,
      "loss": 0.1642,
      "step": 1417
    },
    {
      "epoch": 0.2836,
      "grad_norm": 5.960536956787109,
      "learning_rate": 3.655102040816327e-05,
      "loss": 2.2314,
      "step": 1418
    },
    {
      "epoch": 0.2838,
      "grad_norm": 1.9197490215301514,
      "learning_rate": 3.6540816326530616e-05,
      "loss": 0.1698,
      "step": 1419
    },
    {
      "epoch": 0.284,
      "grad_norm": 1.6662993431091309,
      "learning_rate": 3.653061224489796e-05,
      "loss": 0.075,
      "step": 1420
    },
    {
      "epoch": 0.2842,
      "grad_norm": 1.8599644899368286,
      "learning_rate": 3.652040816326531e-05,
      "loss": 0.2059,
      "step": 1421
    },
    {
      "epoch": 0.2844,
      "grad_norm": 1.4788411855697632,
      "learning_rate": 3.651020408163266e-05,
      "loss": 0.0738,
      "step": 1422
    },
    {
      "epoch": 0.2846,
      "grad_norm": 4.117801666259766,
      "learning_rate": 3.65e-05,
      "loss": 0.7203,
      "step": 1423
    },
    {
      "epoch": 0.2848,
      "grad_norm": 6.469876766204834,
      "learning_rate": 3.648979591836735e-05,
      "loss": 1.1329,
      "step": 1424
    },
    {
      "epoch": 0.285,
      "grad_norm": 5.63407039642334,
      "learning_rate": 3.64795918367347e-05,
      "loss": 1.4169,
      "step": 1425
    },
    {
      "epoch": 0.2852,
      "grad_norm": 1.5634034872055054,
      "learning_rate": 3.646938775510204e-05,
      "loss": 0.0892,
      "step": 1426
    },
    {
      "epoch": 0.2854,
      "grad_norm": 13.888527870178223,
      "learning_rate": 3.645918367346939e-05,
      "loss": 2.2912,
      "step": 1427
    },
    {
      "epoch": 0.2856,
      "grad_norm": 4.702810764312744,
      "learning_rate": 3.644897959183673e-05,
      "loss": 1.0284,
      "step": 1428
    },
    {
      "epoch": 0.2858,
      "grad_norm": 18.649410247802734,
      "learning_rate": 3.643877551020408e-05,
      "loss": 4.8366,
      "step": 1429
    },
    {
      "epoch": 0.286,
      "grad_norm": 36.403133392333984,
      "learning_rate": 3.642857142857143e-05,
      "loss": 5.4918,
      "step": 1430
    },
    {
      "epoch": 0.2862,
      "grad_norm": 10.971161842346191,
      "learning_rate": 3.641836734693877e-05,
      "loss": 4.3258,
      "step": 1431
    },
    {
      "epoch": 0.2864,
      "grad_norm": 8.168219566345215,
      "learning_rate": 3.640816326530612e-05,
      "loss": 1.0876,
      "step": 1432
    },
    {
      "epoch": 0.2866,
      "grad_norm": 39.15559768676758,
      "learning_rate": 3.639795918367347e-05,
      "loss": 5.2519,
      "step": 1433
    },
    {
      "epoch": 0.2868,
      "grad_norm": 5.92279052734375,
      "learning_rate": 3.638775510204081e-05,
      "loss": 1.0407,
      "step": 1434
    },
    {
      "epoch": 0.287,
      "grad_norm": 1.6077378988265991,
      "learning_rate": 3.637755102040816e-05,
      "loss": 0.1247,
      "step": 1435
    },
    {
      "epoch": 0.2872,
      "grad_norm": 1.6817654371261597,
      "learning_rate": 3.636734693877551e-05,
      "loss": 0.1037,
      "step": 1436
    },
    {
      "epoch": 0.2874,
      "grad_norm": 1.44280207157135,
      "learning_rate": 3.6357142857142854e-05,
      "loss": 0.0773,
      "step": 1437
    },
    {
      "epoch": 0.2876,
      "grad_norm": 1.7035020589828491,
      "learning_rate": 3.63469387755102e-05,
      "loss": 0.1836,
      "step": 1438
    },
    {
      "epoch": 0.2878,
      "grad_norm": 1.8268977403640747,
      "learning_rate": 3.633673469387755e-05,
      "loss": 0.164,
      "step": 1439
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.641530990600586,
      "learning_rate": 3.63265306122449e-05,
      "loss": 0.1704,
      "step": 1440
    },
    {
      "epoch": 0.2882,
      "grad_norm": 1.2546417713165283,
      "learning_rate": 3.631632653061225e-05,
      "loss": 0.0956,
      "step": 1441
    },
    {
      "epoch": 0.2884,
      "grad_norm": 1.6430646181106567,
      "learning_rate": 3.630612244897959e-05,
      "loss": 0.1431,
      "step": 1442
    },
    {
      "epoch": 0.2886,
      "grad_norm": 1.5561422109603882,
      "learning_rate": 3.629591836734694e-05,
      "loss": 0.1346,
      "step": 1443
    },
    {
      "epoch": 0.2888,
      "grad_norm": 1.5593615770339966,
      "learning_rate": 3.628571428571429e-05,
      "loss": 0.0824,
      "step": 1444
    },
    {
      "epoch": 0.289,
      "grad_norm": 1.296095848083496,
      "learning_rate": 3.6275510204081634e-05,
      "loss": 0.0708,
      "step": 1445
    },
    {
      "epoch": 0.2892,
      "grad_norm": 1.6430326700210571,
      "learning_rate": 3.626530612244898e-05,
      "loss": 0.181,
      "step": 1446
    },
    {
      "epoch": 0.2894,
      "grad_norm": 75.7375259399414,
      "learning_rate": 3.625510204081633e-05,
      "loss": 5.6181,
      "step": 1447
    },
    {
      "epoch": 0.2896,
      "grad_norm": 1.7995988130569458,
      "learning_rate": 3.6244897959183675e-05,
      "loss": 0.1908,
      "step": 1448
    },
    {
      "epoch": 0.2898,
      "grad_norm": 1.6368839740753174,
      "learning_rate": 3.6234693877551024e-05,
      "loss": 0.1367,
      "step": 1449
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4290481805801392,
      "learning_rate": 3.622448979591837e-05,
      "loss": 0.1843,
      "step": 1450
    },
    {
      "epoch": 0.2902,
      "grad_norm": 30.49867057800293,
      "learning_rate": 3.6214285714285716e-05,
      "loss": 4.8434,
      "step": 1451
    },
    {
      "epoch": 0.2904,
      "grad_norm": 1.2859183549880981,
      "learning_rate": 3.6204081632653065e-05,
      "loss": 0.1097,
      "step": 1452
    },
    {
      "epoch": 0.2906,
      "grad_norm": 20.768646240234375,
      "learning_rate": 3.6193877551020414e-05,
      "loss": 4.2763,
      "step": 1453
    },
    {
      "epoch": 0.2908,
      "grad_norm": 1.3919540643692017,
      "learning_rate": 3.6183673469387756e-05,
      "loss": 0.1183,
      "step": 1454
    },
    {
      "epoch": 0.291,
      "grad_norm": 4.011405944824219,
      "learning_rate": 3.6173469387755106e-05,
      "loss": 0.6685,
      "step": 1455
    },
    {
      "epoch": 0.2912,
      "grad_norm": 1.4617115259170532,
      "learning_rate": 3.6163265306122455e-05,
      "loss": 0.1463,
      "step": 1456
    },
    {
      "epoch": 0.2914,
      "grad_norm": 1.5064582824707031,
      "learning_rate": 3.61530612244898e-05,
      "loss": 0.0668,
      "step": 1457
    },
    {
      "epoch": 0.2916,
      "grad_norm": 1.6540746688842773,
      "learning_rate": 3.6142857142857146e-05,
      "loss": 0.1695,
      "step": 1458
    },
    {
      "epoch": 0.2918,
      "grad_norm": 1.683559536933899,
      "learning_rate": 3.613265306122449e-05,
      "loss": 0.0959,
      "step": 1459
    },
    {
      "epoch": 0.292,
      "grad_norm": 1.4306719303131104,
      "learning_rate": 3.612244897959184e-05,
      "loss": 0.0992,
      "step": 1460
    },
    {
      "epoch": 0.2922,
      "grad_norm": 6.173213481903076,
      "learning_rate": 3.611224489795919e-05,
      "loss": 2.266,
      "step": 1461
    },
    {
      "epoch": 0.2924,
      "grad_norm": 5.689418315887451,
      "learning_rate": 3.610204081632653e-05,
      "loss": 1.2446,
      "step": 1462
    },
    {
      "epoch": 0.2926,
      "grad_norm": 1.2365597486495972,
      "learning_rate": 3.609183673469388e-05,
      "loss": 0.1461,
      "step": 1463
    },
    {
      "epoch": 0.2928,
      "grad_norm": 1.5029525756835938,
      "learning_rate": 3.608163265306123e-05,
      "loss": 0.0894,
      "step": 1464
    },
    {
      "epoch": 0.293,
      "grad_norm": 12.302665710449219,
      "learning_rate": 3.607142857142857e-05,
      "loss": 2.1756,
      "step": 1465
    },
    {
      "epoch": 0.2932,
      "grad_norm": 1.846310019493103,
      "learning_rate": 3.606122448979592e-05,
      "loss": 0.166,
      "step": 1466
    },
    {
      "epoch": 0.2934,
      "grad_norm": 29.496362686157227,
      "learning_rate": 3.605102040816327e-05,
      "loss": 4.3579,
      "step": 1467
    },
    {
      "epoch": 0.2936,
      "grad_norm": 1.7736221551895142,
      "learning_rate": 3.604081632653061e-05,
      "loss": 0.1465,
      "step": 1468
    },
    {
      "epoch": 0.2938,
      "grad_norm": 8.128005981445312,
      "learning_rate": 3.603061224489796e-05,
      "loss": 2.0517,
      "step": 1469
    },
    {
      "epoch": 0.294,
      "grad_norm": 1.6332521438598633,
      "learning_rate": 3.602040816326531e-05,
      "loss": 0.0818,
      "step": 1470
    },
    {
      "epoch": 0.2942,
      "grad_norm": 1.4304476976394653,
      "learning_rate": 3.601020408163265e-05,
      "loss": 0.1144,
      "step": 1471
    },
    {
      "epoch": 0.2944,
      "grad_norm": 6.116512298583984,
      "learning_rate": 3.6e-05,
      "loss": 1.1423,
      "step": 1472
    },
    {
      "epoch": 0.2946,
      "grad_norm": 1.713259220123291,
      "learning_rate": 3.598979591836735e-05,
      "loss": 0.2148,
      "step": 1473
    },
    {
      "epoch": 0.2948,
      "grad_norm": 4.142031669616699,
      "learning_rate": 3.597959183673469e-05,
      "loss": 0.6924,
      "step": 1474
    },
    {
      "epoch": 0.295,
      "grad_norm": 1.4502673149108887,
      "learning_rate": 3.596938775510204e-05,
      "loss": 0.0688,
      "step": 1475
    },
    {
      "epoch": 0.2952,
      "grad_norm": 1.564724326133728,
      "learning_rate": 3.595918367346939e-05,
      "loss": 0.1236,
      "step": 1476
    },
    {
      "epoch": 0.2954,
      "grad_norm": 1.7544804811477661,
      "learning_rate": 3.594897959183673e-05,
      "loss": 0.083,
      "step": 1477
    },
    {
      "epoch": 0.2956,
      "grad_norm": 13.240604400634766,
      "learning_rate": 3.593877551020408e-05,
      "loss": 2.213,
      "step": 1478
    },
    {
      "epoch": 0.2958,
      "grad_norm": 1.6230263710021973,
      "learning_rate": 3.5928571428571425e-05,
      "loss": 0.0949,
      "step": 1479
    },
    {
      "epoch": 0.296,
      "grad_norm": 1.812746524810791,
      "learning_rate": 3.5918367346938774e-05,
      "loss": 0.1133,
      "step": 1480
    },
    {
      "epoch": 0.2962,
      "grad_norm": 5.771973609924316,
      "learning_rate": 3.590816326530612e-05,
      "loss": 1.136,
      "step": 1481
    },
    {
      "epoch": 0.2964,
      "grad_norm": 1.594550609588623,
      "learning_rate": 3.5897959183673466e-05,
      "loss": 0.1483,
      "step": 1482
    },
    {
      "epoch": 0.2966,
      "grad_norm": 17.456066131591797,
      "learning_rate": 3.5887755102040815e-05,
      "loss": 4.0837,
      "step": 1483
    },
    {
      "epoch": 0.2968,
      "grad_norm": 1.3515254259109497,
      "learning_rate": 3.587755102040817e-05,
      "loss": 0.06,
      "step": 1484
    },
    {
      "epoch": 0.297,
      "grad_norm": 1.5891889333724976,
      "learning_rate": 3.586734693877551e-05,
      "loss": 0.1569,
      "step": 1485
    },
    {
      "epoch": 0.2972,
      "grad_norm": 1.8010159730911255,
      "learning_rate": 3.585714285714286e-05,
      "loss": 0.1797,
      "step": 1486
    },
    {
      "epoch": 0.2974,
      "grad_norm": 6.173261642456055,
      "learning_rate": 3.584693877551021e-05,
      "loss": 2.3401,
      "step": 1487
    },
    {
      "epoch": 0.2976,
      "grad_norm": 6.937226295471191,
      "learning_rate": 3.5836734693877554e-05,
      "loss": 1.1559,
      "step": 1488
    },
    {
      "epoch": 0.2978,
      "grad_norm": 1.6864588260650635,
      "learning_rate": 3.58265306122449e-05,
      "loss": 0.1606,
      "step": 1489
    },
    {
      "epoch": 0.298,
      "grad_norm": 2.268903970718384,
      "learning_rate": 3.5816326530612245e-05,
      "loss": 0.0849,
      "step": 1490
    },
    {
      "epoch": 0.2982,
      "grad_norm": 2.1504127979278564,
      "learning_rate": 3.5806122448979595e-05,
      "loss": 0.1037,
      "step": 1491
    },
    {
      "epoch": 0.2984,
      "grad_norm": 1.5428582429885864,
      "learning_rate": 3.5795918367346944e-05,
      "loss": 0.1281,
      "step": 1492
    },
    {
      "epoch": 0.2986,
      "grad_norm": 1.5299909114837646,
      "learning_rate": 3.5785714285714286e-05,
      "loss": 0.0599,
      "step": 1493
    },
    {
      "epoch": 0.2988,
      "grad_norm": 6.747359752655029,
      "learning_rate": 3.5775510204081635e-05,
      "loss": 2.2304,
      "step": 1494
    },
    {
      "epoch": 0.299,
      "grad_norm": 1.5359011888504028,
      "learning_rate": 3.5765306122448985e-05,
      "loss": 0.1195,
      "step": 1495
    },
    {
      "epoch": 0.2992,
      "grad_norm": 1.6429435014724731,
      "learning_rate": 3.575510204081633e-05,
      "loss": 0.1849,
      "step": 1496
    },
    {
      "epoch": 0.2994,
      "grad_norm": 1.3016010522842407,
      "learning_rate": 3.5744897959183676e-05,
      "loss": 0.1018,
      "step": 1497
    },
    {
      "epoch": 0.2996,
      "grad_norm": 1.7025047540664673,
      "learning_rate": 3.5734693877551025e-05,
      "loss": 0.1944,
      "step": 1498
    },
    {
      "epoch": 0.2998,
      "grad_norm": 1.5438400506973267,
      "learning_rate": 3.572448979591837e-05,
      "loss": 0.1396,
      "step": 1499
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3732315301895142,
      "learning_rate": 3.571428571428572e-05,
      "loss": 0.1188,
      "step": 1500
    },
    {
      "epoch": 0.3002,
      "grad_norm": 22.419750213623047,
      "learning_rate": 3.5704081632653066e-05,
      "loss": 5.4139,
      "step": 1501
    },
    {
      "epoch": 0.3004,
      "grad_norm": 4.706147193908691,
      "learning_rate": 3.569387755102041e-05,
      "loss": 0.8861,
      "step": 1502
    },
    {
      "epoch": 0.3006,
      "grad_norm": 1.3744498491287231,
      "learning_rate": 3.568367346938776e-05,
      "loss": 0.18,
      "step": 1503
    },
    {
      "epoch": 0.3008,
      "grad_norm": 6.645823001861572,
      "learning_rate": 3.567346938775511e-05,
      "loss": 1.2648,
      "step": 1504
    },
    {
      "epoch": 0.301,
      "grad_norm": 6.157020092010498,
      "learning_rate": 3.566326530612245e-05,
      "loss": 1.0222,
      "step": 1505
    },
    {
      "epoch": 0.3012,
      "grad_norm": 1.47323739528656,
      "learning_rate": 3.56530612244898e-05,
      "loss": 0.1314,
      "step": 1506
    },
    {
      "epoch": 0.3014,
      "grad_norm": 1.8938440084457397,
      "learning_rate": 3.564285714285715e-05,
      "loss": 0.1721,
      "step": 1507
    },
    {
      "epoch": 0.3016,
      "grad_norm": 6.2161030769348145,
      "learning_rate": 3.563265306122449e-05,
      "loss": 1.1228,
      "step": 1508
    },
    {
      "epoch": 0.3018,
      "grad_norm": 2.3820409774780273,
      "learning_rate": 3.562244897959184e-05,
      "loss": 0.1067,
      "step": 1509
    },
    {
      "epoch": 0.302,
      "grad_norm": 1.7744817733764648,
      "learning_rate": 3.561224489795918e-05,
      "loss": 0.2073,
      "step": 1510
    },
    {
      "epoch": 0.3022,
      "grad_norm": 6.334085464477539,
      "learning_rate": 3.560204081632653e-05,
      "loss": 1.4715,
      "step": 1511
    },
    {
      "epoch": 0.3024,
      "grad_norm": 1.767067551612854,
      "learning_rate": 3.559183673469388e-05,
      "loss": 0.1587,
      "step": 1512
    },
    {
      "epoch": 0.3026,
      "grad_norm": 2.8065779209136963,
      "learning_rate": 3.558163265306122e-05,
      "loss": 0.1068,
      "step": 1513
    },
    {
      "epoch": 0.3028,
      "grad_norm": 1.4385590553283691,
      "learning_rate": 3.557142857142857e-05,
      "loss": 0.1007,
      "step": 1514
    },
    {
      "epoch": 0.303,
      "grad_norm": 1.512893557548523,
      "learning_rate": 3.556122448979592e-05,
      "loss": 0.1425,
      "step": 1515
    },
    {
      "epoch": 0.3032,
      "grad_norm": 1.9610958099365234,
      "learning_rate": 3.555102040816326e-05,
      "loss": 0.1041,
      "step": 1516
    },
    {
      "epoch": 0.3034,
      "grad_norm": 18.42789649963379,
      "learning_rate": 3.554081632653061e-05,
      "loss": 3.6185,
      "step": 1517
    },
    {
      "epoch": 0.3036,
      "grad_norm": 6.517113208770752,
      "learning_rate": 3.553061224489796e-05,
      "loss": 1.0958,
      "step": 1518
    },
    {
      "epoch": 0.3038,
      "grad_norm": 20.108379364013672,
      "learning_rate": 3.5520408163265304e-05,
      "loss": 4.195,
      "step": 1519
    },
    {
      "epoch": 0.304,
      "grad_norm": 5.463354587554932,
      "learning_rate": 3.551020408163265e-05,
      "loss": 1.0336,
      "step": 1520
    },
    {
      "epoch": 0.3042,
      "grad_norm": 1.7834296226501465,
      "learning_rate": 3.55e-05,
      "loss": 0.1257,
      "step": 1521
    },
    {
      "epoch": 0.3044,
      "grad_norm": 30.76213836669922,
      "learning_rate": 3.5489795918367345e-05,
      "loss": 5.2445,
      "step": 1522
    },
    {
      "epoch": 0.3046,
      "grad_norm": 17.685440063476562,
      "learning_rate": 3.5479591836734694e-05,
      "loss": 3.5709,
      "step": 1523
    },
    {
      "epoch": 0.3048,
      "grad_norm": 7.154433250427246,
      "learning_rate": 3.546938775510204e-05,
      "loss": 1.2881,
      "step": 1524
    },
    {
      "epoch": 0.305,
      "grad_norm": 6.08471155166626,
      "learning_rate": 3.5459183673469385e-05,
      "loss": 1.1282,
      "step": 1525
    },
    {
      "epoch": 0.3052,
      "grad_norm": 1.3227505683898926,
      "learning_rate": 3.5448979591836735e-05,
      "loss": 0.1083,
      "step": 1526
    },
    {
      "epoch": 0.3054,
      "grad_norm": 5.789578914642334,
      "learning_rate": 3.5438775510204084e-05,
      "loss": 2.3024,
      "step": 1527
    },
    {
      "epoch": 0.3056,
      "grad_norm": 1.3606492280960083,
      "learning_rate": 3.5428571428571426e-05,
      "loss": 0.0869,
      "step": 1528
    },
    {
      "epoch": 0.3058,
      "grad_norm": 7.820291996002197,
      "learning_rate": 3.5418367346938775e-05,
      "loss": 1.2844,
      "step": 1529
    },
    {
      "epoch": 0.306,
      "grad_norm": 1.8580995798110962,
      "learning_rate": 3.5408163265306125e-05,
      "loss": 0.1626,
      "step": 1530
    },
    {
      "epoch": 0.3062,
      "grad_norm": 9.6140718460083,
      "learning_rate": 3.5397959183673474e-05,
      "loss": 3.9964,
      "step": 1531
    },
    {
      "epoch": 0.3064,
      "grad_norm": 1.9866973161697388,
      "learning_rate": 3.538775510204082e-05,
      "loss": 0.2138,
      "step": 1532
    },
    {
      "epoch": 0.3066,
      "grad_norm": 1.411624789237976,
      "learning_rate": 3.5377551020408165e-05,
      "loss": 0.1443,
      "step": 1533
    },
    {
      "epoch": 0.3068,
      "grad_norm": 1.9197992086410522,
      "learning_rate": 3.5367346938775515e-05,
      "loss": 0.1875,
      "step": 1534
    },
    {
      "epoch": 0.307,
      "grad_norm": 40.50486373901367,
      "learning_rate": 3.5357142857142864e-05,
      "loss": 5.3346,
      "step": 1535
    },
    {
      "epoch": 0.3072,
      "grad_norm": 1.443135380744934,
      "learning_rate": 3.5346938775510206e-05,
      "loss": 0.1313,
      "step": 1536
    },
    {
      "epoch": 0.3074,
      "grad_norm": 4.470092296600342,
      "learning_rate": 3.5336734693877555e-05,
      "loss": 1.0869,
      "step": 1537
    },
    {
      "epoch": 0.3076,
      "grad_norm": 1.6345386505126953,
      "learning_rate": 3.5326530612244904e-05,
      "loss": 0.1208,
      "step": 1538
    },
    {
      "epoch": 0.3078,
      "grad_norm": 4.804007530212402,
      "learning_rate": 3.531632653061225e-05,
      "loss": 0.8006,
      "step": 1539
    },
    {
      "epoch": 0.308,
      "grad_norm": 1.4564540386199951,
      "learning_rate": 3.5306122448979596e-05,
      "loss": 0.1651,
      "step": 1540
    },
    {
      "epoch": 0.3082,
      "grad_norm": 1.2920421361923218,
      "learning_rate": 3.529591836734694e-05,
      "loss": 0.1073,
      "step": 1541
    },
    {
      "epoch": 0.3084,
      "grad_norm": 4.00438928604126,
      "learning_rate": 3.528571428571429e-05,
      "loss": 0.7153,
      "step": 1542
    },
    {
      "epoch": 0.3086,
      "grad_norm": 1.3874545097351074,
      "learning_rate": 3.527551020408164e-05,
      "loss": 0.0839,
      "step": 1543
    },
    {
      "epoch": 0.3088,
      "grad_norm": 6.14603853225708,
      "learning_rate": 3.526530612244898e-05,
      "loss": 2.2278,
      "step": 1544
    },
    {
      "epoch": 0.309,
      "grad_norm": 1.3765803575515747,
      "learning_rate": 3.525510204081633e-05,
      "loss": 0.0968,
      "step": 1545
    },
    {
      "epoch": 0.3092,
      "grad_norm": 4.640695571899414,
      "learning_rate": 3.524489795918368e-05,
      "loss": 1.1675,
      "step": 1546
    },
    {
      "epoch": 0.3094,
      "grad_norm": 4.243666172027588,
      "learning_rate": 3.523469387755102e-05,
      "loss": 0.7011,
      "step": 1547
    },
    {
      "epoch": 0.3096,
      "grad_norm": 5.763128280639648,
      "learning_rate": 3.522448979591837e-05,
      "loss": 1.3464,
      "step": 1548
    },
    {
      "epoch": 0.3098,
      "grad_norm": 1.5605864524841309,
      "learning_rate": 3.521428571428572e-05,
      "loss": 0.1439,
      "step": 1549
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6594971418380737,
      "learning_rate": 3.520408163265306e-05,
      "loss": 0.1253,
      "step": 1550
    },
    {
      "epoch": 0.3102,
      "grad_norm": 1.8500962257385254,
      "learning_rate": 3.519387755102041e-05,
      "loss": 0.1942,
      "step": 1551
    },
    {
      "epoch": 0.3104,
      "grad_norm": 5.679625511169434,
      "learning_rate": 3.518367346938776e-05,
      "loss": 1.3988,
      "step": 1552
    },
    {
      "epoch": 0.3106,
      "grad_norm": 7.082156181335449,
      "learning_rate": 3.51734693877551e-05,
      "loss": 1.1313,
      "step": 1553
    },
    {
      "epoch": 0.3108,
      "grad_norm": 10.332091331481934,
      "learning_rate": 3.516326530612245e-05,
      "loss": 1.809,
      "step": 1554
    },
    {
      "epoch": 0.311,
      "grad_norm": 1.4994076490402222,
      "learning_rate": 3.51530612244898e-05,
      "loss": 0.1259,
      "step": 1555
    },
    {
      "epoch": 0.3112,
      "grad_norm": 1.9050859212875366,
      "learning_rate": 3.514285714285714e-05,
      "loss": 0.1662,
      "step": 1556
    },
    {
      "epoch": 0.3114,
      "grad_norm": 1.4891732931137085,
      "learning_rate": 3.513265306122449e-05,
      "loss": 0.117,
      "step": 1557
    },
    {
      "epoch": 0.3116,
      "grad_norm": 1.6584659814834595,
      "learning_rate": 3.512244897959184e-05,
      "loss": 0.1396,
      "step": 1558
    },
    {
      "epoch": 0.3118,
      "grad_norm": 4.019477844238281,
      "learning_rate": 3.511224489795918e-05,
      "loss": 1.0247,
      "step": 1559
    },
    {
      "epoch": 0.312,
      "grad_norm": 4.059731483459473,
      "learning_rate": 3.510204081632653e-05,
      "loss": 0.9791,
      "step": 1560
    },
    {
      "epoch": 0.3122,
      "grad_norm": 9.344042778015137,
      "learning_rate": 3.509183673469388e-05,
      "loss": 1.0842,
      "step": 1561
    },
    {
      "epoch": 0.3124,
      "grad_norm": 2.1318531036376953,
      "learning_rate": 3.5081632653061224e-05,
      "loss": 0.2142,
      "step": 1562
    },
    {
      "epoch": 0.3126,
      "grad_norm": 1.78006911277771,
      "learning_rate": 3.507142857142857e-05,
      "loss": 0.1534,
      "step": 1563
    },
    {
      "epoch": 0.3128,
      "grad_norm": 1.7451825141906738,
      "learning_rate": 3.5061224489795915e-05,
      "loss": 0.1632,
      "step": 1564
    },
    {
      "epoch": 0.313,
      "grad_norm": 1.4965324401855469,
      "learning_rate": 3.5051020408163265e-05,
      "loss": 0.1359,
      "step": 1565
    },
    {
      "epoch": 0.3132,
      "grad_norm": 1.3011749982833862,
      "learning_rate": 3.5040816326530614e-05,
      "loss": 0.0872,
      "step": 1566
    },
    {
      "epoch": 0.3134,
      "grad_norm": 1.8529915809631348,
      "learning_rate": 3.5030612244897956e-05,
      "loss": 0.1895,
      "step": 1567
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.5333659648895264,
      "learning_rate": 3.5020408163265305e-05,
      "loss": 0.1305,
      "step": 1568
    },
    {
      "epoch": 0.3138,
      "grad_norm": 4.501252174377441,
      "learning_rate": 3.5010204081632655e-05,
      "loss": 0.7046,
      "step": 1569
    },
    {
      "epoch": 0.314,
      "grad_norm": 5.666863441467285,
      "learning_rate": 3.5e-05,
      "loss": 1.0643,
      "step": 1570
    },
    {
      "epoch": 0.3142,
      "grad_norm": 1.5210033655166626,
      "learning_rate": 3.4989795918367346e-05,
      "loss": 0.1249,
      "step": 1571
    },
    {
      "epoch": 0.3144,
      "grad_norm": 1.483896255493164,
      "learning_rate": 3.4979591836734695e-05,
      "loss": 0.117,
      "step": 1572
    },
    {
      "epoch": 0.3146,
      "grad_norm": 14.870137214660645,
      "learning_rate": 3.496938775510204e-05,
      "loss": 4.297,
      "step": 1573
    },
    {
      "epoch": 0.3148,
      "grad_norm": 1.2303497791290283,
      "learning_rate": 3.495918367346939e-05,
      "loss": 0.0996,
      "step": 1574
    },
    {
      "epoch": 0.315,
      "grad_norm": 3.4915120601654053,
      "learning_rate": 3.4948979591836736e-05,
      "loss": 0.8037,
      "step": 1575
    },
    {
      "epoch": 0.3152,
      "grad_norm": 1.8560577630996704,
      "learning_rate": 3.4938775510204085e-05,
      "loss": 0.1712,
      "step": 1576
    },
    {
      "epoch": 0.3154,
      "grad_norm": 1.430997610092163,
      "learning_rate": 3.4928571428571434e-05,
      "loss": 0.0985,
      "step": 1577
    },
    {
      "epoch": 0.3156,
      "grad_norm": 1.5378358364105225,
      "learning_rate": 3.491836734693878e-05,
      "loss": 0.163,
      "step": 1578
    },
    {
      "epoch": 0.3158,
      "grad_norm": 1.7097241878509521,
      "learning_rate": 3.4908163265306126e-05,
      "loss": 0.1624,
      "step": 1579
    },
    {
      "epoch": 0.316,
      "grad_norm": 7.346837043762207,
      "learning_rate": 3.4897959183673475e-05,
      "loss": 2.2195,
      "step": 1580
    },
    {
      "epoch": 0.3162,
      "grad_norm": 5.800686359405518,
      "learning_rate": 3.488775510204082e-05,
      "loss": 1.1779,
      "step": 1581
    },
    {
      "epoch": 0.3164,
      "grad_norm": 3.9609789848327637,
      "learning_rate": 3.487755102040817e-05,
      "loss": 1.0466,
      "step": 1582
    },
    {
      "epoch": 0.3166,
      "grad_norm": 1.1616785526275635,
      "learning_rate": 3.4867346938775516e-05,
      "loss": 0.098,
      "step": 1583
    },
    {
      "epoch": 0.3168,
      "grad_norm": 1.6771498918533325,
      "learning_rate": 3.485714285714286e-05,
      "loss": 0.1413,
      "step": 1584
    },
    {
      "epoch": 0.317,
      "grad_norm": 1.4764608144760132,
      "learning_rate": 3.484693877551021e-05,
      "loss": 0.1243,
      "step": 1585
    },
    {
      "epoch": 0.3172,
      "grad_norm": 1.3374871015548706,
      "learning_rate": 3.483673469387756e-05,
      "loss": 0.1146,
      "step": 1586
    },
    {
      "epoch": 0.3174,
      "grad_norm": 3.5561840534210205,
      "learning_rate": 3.48265306122449e-05,
      "loss": 0.8987,
      "step": 1587
    },
    {
      "epoch": 0.3176,
      "grad_norm": 1.4515995979309082,
      "learning_rate": 3.481632653061225e-05,
      "loss": 0.125,
      "step": 1588
    },
    {
      "epoch": 0.3178,
      "grad_norm": 25.665393829345703,
      "learning_rate": 3.48061224489796e-05,
      "loss": 5.1746,
      "step": 1589
    },
    {
      "epoch": 0.318,
      "grad_norm": 1.3759853839874268,
      "learning_rate": 3.479591836734694e-05,
      "loss": 0.16,
      "step": 1590
    },
    {
      "epoch": 0.3182,
      "grad_norm": 7.616851806640625,
      "learning_rate": 3.478571428571429e-05,
      "loss": 3.5554,
      "step": 1591
    },
    {
      "epoch": 0.3184,
      "grad_norm": 3.7456676959991455,
      "learning_rate": 3.477551020408164e-05,
      "loss": 0.9717,
      "step": 1592
    },
    {
      "epoch": 0.3186,
      "grad_norm": 1.3557777404785156,
      "learning_rate": 3.476530612244898e-05,
      "loss": 0.1145,
      "step": 1593
    },
    {
      "epoch": 0.3188,
      "grad_norm": 4.116989612579346,
      "learning_rate": 3.475510204081633e-05,
      "loss": 0.6787,
      "step": 1594
    },
    {
      "epoch": 0.319,
      "grad_norm": 3.567126750946045,
      "learning_rate": 3.474489795918367e-05,
      "loss": 0.8604,
      "step": 1595
    },
    {
      "epoch": 0.3192,
      "grad_norm": 3.4982337951660156,
      "learning_rate": 3.473469387755102e-05,
      "loss": 0.9317,
      "step": 1596
    },
    {
      "epoch": 0.3194,
      "grad_norm": 3.7683043479919434,
      "learning_rate": 3.472448979591837e-05,
      "loss": 0.6925,
      "step": 1597
    },
    {
      "epoch": 0.3196,
      "grad_norm": 4.050169944763184,
      "learning_rate": 3.471428571428571e-05,
      "loss": 0.7471,
      "step": 1598
    },
    {
      "epoch": 0.3198,
      "grad_norm": 1.2799166440963745,
      "learning_rate": 3.470408163265306e-05,
      "loss": 0.1135,
      "step": 1599
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3863531351089478,
      "learning_rate": 3.469387755102041e-05,
      "loss": 0.1476,
      "step": 1600
    },
    {
      "epoch": 0.3202,
      "grad_norm": 8.264937400817871,
      "learning_rate": 3.4683673469387754e-05,
      "loss": 1.4777,
      "step": 1601
    },
    {
      "epoch": 0.3204,
      "grad_norm": 1.742661714553833,
      "learning_rate": 3.46734693877551e-05,
      "loss": 0.1704,
      "step": 1602
    },
    {
      "epoch": 0.3206,
      "grad_norm": 4.8802266120910645,
      "learning_rate": 3.466326530612245e-05,
      "loss": 1.318,
      "step": 1603
    },
    {
      "epoch": 0.3208,
      "grad_norm": 4.035736083984375,
      "learning_rate": 3.4653061224489795e-05,
      "loss": 1.0556,
      "step": 1604
    },
    {
      "epoch": 0.321,
      "grad_norm": 6.3922343254089355,
      "learning_rate": 3.4642857142857144e-05,
      "loss": 1.1756,
      "step": 1605
    },
    {
      "epoch": 0.3212,
      "grad_norm": 8.551189422607422,
      "learning_rate": 3.463265306122449e-05,
      "loss": 3.5888,
      "step": 1606
    },
    {
      "epoch": 0.3214,
      "grad_norm": 54.71131896972656,
      "learning_rate": 3.4622448979591835e-05,
      "loss": 5.4349,
      "step": 1607
    },
    {
      "epoch": 0.3216,
      "grad_norm": 37.75558090209961,
      "learning_rate": 3.4612244897959184e-05,
      "loss": 5.3459,
      "step": 1608
    },
    {
      "epoch": 0.3218,
      "grad_norm": 1.561134696006775,
      "learning_rate": 3.4602040816326534e-05,
      "loss": 0.1272,
      "step": 1609
    },
    {
      "epoch": 0.322,
      "grad_norm": 6.835804462432861,
      "learning_rate": 3.4591836734693876e-05,
      "loss": 1.4204,
      "step": 1610
    },
    {
      "epoch": 0.3222,
      "grad_norm": 1.407302975654602,
      "learning_rate": 3.4581632653061225e-05,
      "loss": 0.1035,
      "step": 1611
    },
    {
      "epoch": 0.3224,
      "grad_norm": 1.1173968315124512,
      "learning_rate": 3.4571428571428574e-05,
      "loss": 0.0943,
      "step": 1612
    },
    {
      "epoch": 0.3226,
      "grad_norm": 1.4095560312271118,
      "learning_rate": 3.456122448979592e-05,
      "loss": 0.1168,
      "step": 1613
    },
    {
      "epoch": 0.3228,
      "grad_norm": 28.86614990234375,
      "learning_rate": 3.4551020408163266e-05,
      "loss": 5.2445,
      "step": 1614
    },
    {
      "epoch": 0.323,
      "grad_norm": 1.3415619134902954,
      "learning_rate": 3.454081632653061e-05,
      "loss": 0.0933,
      "step": 1615
    },
    {
      "epoch": 0.3232,
      "grad_norm": 10.243077278137207,
      "learning_rate": 3.453061224489796e-05,
      "loss": 1.2456,
      "step": 1616
    },
    {
      "epoch": 0.3234,
      "grad_norm": 6.853682994842529,
      "learning_rate": 3.452040816326531e-05,
      "loss": 1.2804,
      "step": 1617
    },
    {
      "epoch": 0.3236,
      "grad_norm": 1.4520295858383179,
      "learning_rate": 3.451020408163265e-05,
      "loss": 0.1343,
      "step": 1618
    },
    {
      "epoch": 0.3238,
      "grad_norm": 1.510238528251648,
      "learning_rate": 3.45e-05,
      "loss": 0.1346,
      "step": 1619
    },
    {
      "epoch": 0.324,
      "grad_norm": 1.762850046157837,
      "learning_rate": 3.4489795918367354e-05,
      "loss": 0.1664,
      "step": 1620
    },
    {
      "epoch": 0.3242,
      "grad_norm": 5.683584690093994,
      "learning_rate": 3.44795918367347e-05,
      "loss": 2.1875,
      "step": 1621
    },
    {
      "epoch": 0.3244,
      "grad_norm": 3.9202487468719482,
      "learning_rate": 3.4469387755102046e-05,
      "loss": 0.986,
      "step": 1622
    },
    {
      "epoch": 0.3246,
      "grad_norm": 1.4274541139602661,
      "learning_rate": 3.4459183673469395e-05,
      "loss": 0.1336,
      "step": 1623
    },
    {
      "epoch": 0.3248,
      "grad_norm": 1.455907940864563,
      "learning_rate": 3.444897959183674e-05,
      "loss": 0.1164,
      "step": 1624
    },
    {
      "epoch": 0.325,
      "grad_norm": 1.8156874179840088,
      "learning_rate": 3.443877551020409e-05,
      "loss": 0.1452,
      "step": 1625
    },
    {
      "epoch": 0.3252,
      "grad_norm": 7.767525672912598,
      "learning_rate": 3.442857142857143e-05,
      "loss": 3.2278,
      "step": 1626
    },
    {
      "epoch": 0.3254,
      "grad_norm": 6.3254170417785645,
      "learning_rate": 3.441836734693878e-05,
      "loss": 1.2977,
      "step": 1627
    },
    {
      "epoch": 0.3256,
      "grad_norm": 1.2721400260925293,
      "learning_rate": 3.440816326530613e-05,
      "loss": 0.1012,
      "step": 1628
    },
    {
      "epoch": 0.3258,
      "grad_norm": 1.3631278276443481,
      "learning_rate": 3.439795918367347e-05,
      "loss": 0.1276,
      "step": 1629
    },
    {
      "epoch": 0.326,
      "grad_norm": 1.3635259866714478,
      "learning_rate": 3.438775510204082e-05,
      "loss": 0.0953,
      "step": 1630
    },
    {
      "epoch": 0.3262,
      "grad_norm": 1.0951149463653564,
      "learning_rate": 3.437755102040817e-05,
      "loss": 0.0772,
      "step": 1631
    },
    {
      "epoch": 0.3264,
      "grad_norm": 5.393881320953369,
      "learning_rate": 3.436734693877551e-05,
      "loss": 0.8863,
      "step": 1632
    },
    {
      "epoch": 0.3266,
      "grad_norm": 3.9565374851226807,
      "learning_rate": 3.435714285714286e-05,
      "loss": 1.0638,
      "step": 1633
    },
    {
      "epoch": 0.3268,
      "grad_norm": 1.1917364597320557,
      "learning_rate": 3.434693877551021e-05,
      "loss": 0.1068,
      "step": 1634
    },
    {
      "epoch": 0.327,
      "grad_norm": 6.110203266143799,
      "learning_rate": 3.433673469387755e-05,
      "loss": 1.1439,
      "step": 1635
    },
    {
      "epoch": 0.3272,
      "grad_norm": 1.4958648681640625,
      "learning_rate": 3.43265306122449e-05,
      "loss": 0.1234,
      "step": 1636
    },
    {
      "epoch": 0.3274,
      "grad_norm": 5.4393534660339355,
      "learning_rate": 3.431632653061225e-05,
      "loss": 1.1319,
      "step": 1637
    },
    {
      "epoch": 0.3276,
      "grad_norm": 1.2154313325881958,
      "learning_rate": 3.430612244897959e-05,
      "loss": 0.0805,
      "step": 1638
    },
    {
      "epoch": 0.3278,
      "grad_norm": 1.614425539970398,
      "learning_rate": 3.429591836734694e-05,
      "loss": 0.1391,
      "step": 1639
    },
    {
      "epoch": 0.328,
      "grad_norm": 1.4102503061294556,
      "learning_rate": 3.428571428571429e-05,
      "loss": 0.1195,
      "step": 1640
    },
    {
      "epoch": 0.3282,
      "grad_norm": 1.0450716018676758,
      "learning_rate": 3.427551020408163e-05,
      "loss": 0.0539,
      "step": 1641
    },
    {
      "epoch": 0.3284,
      "grad_norm": 1.5549017190933228,
      "learning_rate": 3.426530612244898e-05,
      "loss": 0.1209,
      "step": 1642
    },
    {
      "epoch": 0.3286,
      "grad_norm": 1.6046826839447021,
      "learning_rate": 3.425510204081633e-05,
      "loss": 0.1655,
      "step": 1643
    },
    {
      "epoch": 0.3288,
      "grad_norm": 1.2701417207717896,
      "learning_rate": 3.4244897959183674e-05,
      "loss": 0.1058,
      "step": 1644
    },
    {
      "epoch": 0.329,
      "grad_norm": 1.348042368888855,
      "learning_rate": 3.423469387755102e-05,
      "loss": 0.0948,
      "step": 1645
    },
    {
      "epoch": 0.3292,
      "grad_norm": 1.8712621927261353,
      "learning_rate": 3.4224489795918365e-05,
      "loss": 0.1133,
      "step": 1646
    },
    {
      "epoch": 0.3294,
      "grad_norm": 3.555974006652832,
      "learning_rate": 3.4214285714285714e-05,
      "loss": 0.8684,
      "step": 1647
    },
    {
      "epoch": 0.3296,
      "grad_norm": 1.5502355098724365,
      "learning_rate": 3.4204081632653064e-05,
      "loss": 0.1568,
      "step": 1648
    },
    {
      "epoch": 0.3298,
      "grad_norm": 3.769658088684082,
      "learning_rate": 3.4193877551020406e-05,
      "loss": 0.9487,
      "step": 1649
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6740633249282837,
      "learning_rate": 3.4183673469387755e-05,
      "loss": 0.1423,
      "step": 1650
    },
    {
      "epoch": 0.3302,
      "grad_norm": 1.3294124603271484,
      "learning_rate": 3.4173469387755104e-05,
      "loss": 0.0909,
      "step": 1651
    },
    {
      "epoch": 0.3304,
      "grad_norm": 1.5210949182510376,
      "learning_rate": 3.416326530612245e-05,
      "loss": 0.1255,
      "step": 1652
    },
    {
      "epoch": 0.3306,
      "grad_norm": 1.7085466384887695,
      "learning_rate": 3.4153061224489796e-05,
      "loss": 0.1542,
      "step": 1653
    },
    {
      "epoch": 0.3308,
      "grad_norm": 1.6548869609832764,
      "learning_rate": 3.4142857142857145e-05,
      "loss": 0.144,
      "step": 1654
    },
    {
      "epoch": 0.331,
      "grad_norm": 1.3550704717636108,
      "learning_rate": 3.413265306122449e-05,
      "loss": 0.129,
      "step": 1655
    },
    {
      "epoch": 0.3312,
      "grad_norm": 4.087093353271484,
      "learning_rate": 3.412244897959184e-05,
      "loss": 0.7072,
      "step": 1656
    },
    {
      "epoch": 0.3314,
      "grad_norm": 10.05302906036377,
      "learning_rate": 3.4112244897959186e-05,
      "loss": 3.2497,
      "step": 1657
    },
    {
      "epoch": 0.3316,
      "grad_norm": 1.5289970636367798,
      "learning_rate": 3.410204081632653e-05,
      "loss": 0.1442,
      "step": 1658
    },
    {
      "epoch": 0.3318,
      "grad_norm": 3.952613115310669,
      "learning_rate": 3.409183673469388e-05,
      "loss": 1.0591,
      "step": 1659
    },
    {
      "epoch": 0.332,
      "grad_norm": 1.4845707416534424,
      "learning_rate": 3.408163265306123e-05,
      "loss": 0.1287,
      "step": 1660
    },
    {
      "epoch": 0.3322,
      "grad_norm": 1.3955657482147217,
      "learning_rate": 3.407142857142857e-05,
      "loss": 0.1285,
      "step": 1661
    },
    {
      "epoch": 0.3324,
      "grad_norm": 1.4724160432815552,
      "learning_rate": 3.406122448979592e-05,
      "loss": 0.132,
      "step": 1662
    },
    {
      "epoch": 0.3326,
      "grad_norm": 1.2959953546524048,
      "learning_rate": 3.405102040816327e-05,
      "loss": 0.1044,
      "step": 1663
    },
    {
      "epoch": 0.3328,
      "grad_norm": 1.2368966341018677,
      "learning_rate": 3.404081632653061e-05,
      "loss": 0.085,
      "step": 1664
    },
    {
      "epoch": 0.333,
      "grad_norm": 1.550514817237854,
      "learning_rate": 3.403061224489796e-05,
      "loss": 0.1256,
      "step": 1665
    },
    {
      "epoch": 0.3332,
      "grad_norm": 3.974156618118286,
      "learning_rate": 3.402040816326531e-05,
      "loss": 0.9368,
      "step": 1666
    },
    {
      "epoch": 0.3334,
      "grad_norm": 3.496253490447998,
      "learning_rate": 3.401020408163266e-05,
      "loss": 1.0436,
      "step": 1667
    },
    {
      "epoch": 0.3336,
      "grad_norm": 3.534266471862793,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 1.0133,
      "step": 1668
    },
    {
      "epoch": 0.3338,
      "grad_norm": 6.372529029846191,
      "learning_rate": 3.398979591836735e-05,
      "loss": 2.2158,
      "step": 1669
    },
    {
      "epoch": 0.334,
      "grad_norm": 1.311484932899475,
      "learning_rate": 3.39795918367347e-05,
      "loss": 0.0882,
      "step": 1670
    },
    {
      "epoch": 0.3342,
      "grad_norm": 3.420142412185669,
      "learning_rate": 3.396938775510205e-05,
      "loss": 0.8702,
      "step": 1671
    },
    {
      "epoch": 0.3344,
      "grad_norm": 8.62233829498291,
      "learning_rate": 3.395918367346939e-05,
      "loss": 1.4761,
      "step": 1672
    },
    {
      "epoch": 0.3346,
      "grad_norm": 1.5052247047424316,
      "learning_rate": 3.394897959183674e-05,
      "loss": 0.1379,
      "step": 1673
    },
    {
      "epoch": 0.3348,
      "grad_norm": 1.2894690036773682,
      "learning_rate": 3.393877551020409e-05,
      "loss": 0.0978,
      "step": 1674
    },
    {
      "epoch": 0.335,
      "grad_norm": 1.4060401916503906,
      "learning_rate": 3.392857142857143e-05,
      "loss": 0.0982,
      "step": 1675
    },
    {
      "epoch": 0.3352,
      "grad_norm": 5.907136917114258,
      "learning_rate": 3.391836734693878e-05,
      "loss": 2.2422,
      "step": 1676
    },
    {
      "epoch": 0.3354,
      "grad_norm": 1.534026026725769,
      "learning_rate": 3.390816326530613e-05,
      "loss": 0.1418,
      "step": 1677
    },
    {
      "epoch": 0.3356,
      "grad_norm": 8.832184791564941,
      "learning_rate": 3.389795918367347e-05,
      "loss": 1.2558,
      "step": 1678
    },
    {
      "epoch": 0.3358,
      "grad_norm": 18.30379867553711,
      "learning_rate": 3.388775510204082e-05,
      "loss": 3.42,
      "step": 1679
    },
    {
      "epoch": 0.336,
      "grad_norm": 7.9011077880859375,
      "learning_rate": 3.387755102040816e-05,
      "loss": 1.2031,
      "step": 1680
    },
    {
      "epoch": 0.3362,
      "grad_norm": 1.344774842262268,
      "learning_rate": 3.386734693877551e-05,
      "loss": 0.118,
      "step": 1681
    },
    {
      "epoch": 0.3364,
      "grad_norm": 7.134794235229492,
      "learning_rate": 3.385714285714286e-05,
      "loss": 1.2171,
      "step": 1682
    },
    {
      "epoch": 0.3366,
      "grad_norm": 1.4707818031311035,
      "learning_rate": 3.3846938775510204e-05,
      "loss": 0.1157,
      "step": 1683
    },
    {
      "epoch": 0.3368,
      "grad_norm": 1.6916249990463257,
      "learning_rate": 3.383673469387755e-05,
      "loss": 0.1132,
      "step": 1684
    },
    {
      "epoch": 0.337,
      "grad_norm": 11.05109977722168,
      "learning_rate": 3.38265306122449e-05,
      "loss": 1.1971,
      "step": 1685
    },
    {
      "epoch": 0.3372,
      "grad_norm": 5.596589088439941,
      "learning_rate": 3.3816326530612244e-05,
      "loss": 1.1734,
      "step": 1686
    },
    {
      "epoch": 0.3374,
      "grad_norm": 3.7325363159179688,
      "learning_rate": 3.3806122448979593e-05,
      "loss": 0.8977,
      "step": 1687
    },
    {
      "epoch": 0.3376,
      "grad_norm": 1.798915147781372,
      "learning_rate": 3.379591836734694e-05,
      "loss": 0.1708,
      "step": 1688
    },
    {
      "epoch": 0.3378,
      "grad_norm": 1.2309712171554565,
      "learning_rate": 3.3785714285714285e-05,
      "loss": 0.1188,
      "step": 1689
    },
    {
      "epoch": 0.338,
      "grad_norm": 1.7836970090866089,
      "learning_rate": 3.3775510204081634e-05,
      "loss": 0.1625,
      "step": 1690
    },
    {
      "epoch": 0.3382,
      "grad_norm": 1.6843546628952026,
      "learning_rate": 3.3765306122448983e-05,
      "loss": 0.1446,
      "step": 1691
    },
    {
      "epoch": 0.3384,
      "grad_norm": 1.4100183248519897,
      "learning_rate": 3.3755102040816326e-05,
      "loss": 0.1293,
      "step": 1692
    },
    {
      "epoch": 0.3386,
      "grad_norm": 1.7047845125198364,
      "learning_rate": 3.3744897959183675e-05,
      "loss": 0.2804,
      "step": 1693
    },
    {
      "epoch": 0.3388,
      "grad_norm": 4.1712260246276855,
      "learning_rate": 3.3734693877551024e-05,
      "loss": 0.6373,
      "step": 1694
    },
    {
      "epoch": 0.339,
      "grad_norm": 1.97983717918396,
      "learning_rate": 3.3724489795918367e-05,
      "loss": 0.1642,
      "step": 1695
    },
    {
      "epoch": 0.3392,
      "grad_norm": 1.362574577331543,
      "learning_rate": 3.3714285714285716e-05,
      "loss": 0.1343,
      "step": 1696
    },
    {
      "epoch": 0.3394,
      "grad_norm": 1.8942652940750122,
      "learning_rate": 3.3704081632653065e-05,
      "loss": 0.1933,
      "step": 1697
    },
    {
      "epoch": 0.3396,
      "grad_norm": 5.709615230560303,
      "learning_rate": 3.369387755102041e-05,
      "loss": 2.2058,
      "step": 1698
    },
    {
      "epoch": 0.3398,
      "grad_norm": 4.429149627685547,
      "learning_rate": 3.3683673469387757e-05,
      "loss": 0.962,
      "step": 1699
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3776390552520752,
      "learning_rate": 3.36734693877551e-05,
      "loss": 0.1654,
      "step": 1700
    },
    {
      "epoch": 0.3402,
      "grad_norm": 1.5495575666427612,
      "learning_rate": 3.366326530612245e-05,
      "loss": 0.1326,
      "step": 1701
    },
    {
      "epoch": 0.3404,
      "grad_norm": 1.2185381650924683,
      "learning_rate": 3.36530612244898e-05,
      "loss": 0.159,
      "step": 1702
    },
    {
      "epoch": 0.3406,
      "grad_norm": 1.8954464197158813,
      "learning_rate": 3.364285714285714e-05,
      "loss": 0.3059,
      "step": 1703
    },
    {
      "epoch": 0.3408,
      "grad_norm": 4.421078205108643,
      "learning_rate": 3.363265306122449e-05,
      "loss": 1.0207,
      "step": 1704
    },
    {
      "epoch": 0.341,
      "grad_norm": 1.355732798576355,
      "learning_rate": 3.362244897959184e-05,
      "loss": 0.1387,
      "step": 1705
    },
    {
      "epoch": 0.3412,
      "grad_norm": 7.854058742523193,
      "learning_rate": 3.361224489795918e-05,
      "loss": 1.0543,
      "step": 1706
    },
    {
      "epoch": 0.3414,
      "grad_norm": 1.3484553098678589,
      "learning_rate": 3.360204081632653e-05,
      "loss": 0.1473,
      "step": 1707
    },
    {
      "epoch": 0.3416,
      "grad_norm": 22.08734130859375,
      "learning_rate": 3.359183673469388e-05,
      "loss": 4.1627,
      "step": 1708
    },
    {
      "epoch": 0.3418,
      "grad_norm": 18.592674255371094,
      "learning_rate": 3.358163265306122e-05,
      "loss": 4.042,
      "step": 1709
    },
    {
      "epoch": 0.342,
      "grad_norm": 1.5277100801467896,
      "learning_rate": 3.357142857142857e-05,
      "loss": 0.166,
      "step": 1710
    },
    {
      "epoch": 0.3422,
      "grad_norm": 3.34956955909729,
      "learning_rate": 3.356122448979592e-05,
      "loss": 0.8024,
      "step": 1711
    },
    {
      "epoch": 0.3424,
      "grad_norm": 1.134932279586792,
      "learning_rate": 3.355102040816327e-05,
      "loss": 0.0938,
      "step": 1712
    },
    {
      "epoch": 0.3426,
      "grad_norm": 1.764827847480774,
      "learning_rate": 3.354081632653062e-05,
      "loss": 0.2044,
      "step": 1713
    },
    {
      "epoch": 0.3428,
      "grad_norm": 3.3121185302734375,
      "learning_rate": 3.353061224489796e-05,
      "loss": 0.8107,
      "step": 1714
    },
    {
      "epoch": 0.343,
      "grad_norm": 1.1516286134719849,
      "learning_rate": 3.352040816326531e-05,
      "loss": 0.0882,
      "step": 1715
    },
    {
      "epoch": 0.3432,
      "grad_norm": 16.110246658325195,
      "learning_rate": 3.351020408163266e-05,
      "loss": 4.0145,
      "step": 1716
    },
    {
      "epoch": 0.3434,
      "grad_norm": 1.7402606010437012,
      "learning_rate": 3.35e-05,
      "loss": 0.1657,
      "step": 1717
    },
    {
      "epoch": 0.3436,
      "grad_norm": 3.8523192405700684,
      "learning_rate": 3.348979591836735e-05,
      "loss": 1.0029,
      "step": 1718
    },
    {
      "epoch": 0.3438,
      "grad_norm": 6.852540016174316,
      "learning_rate": 3.34795918367347e-05,
      "loss": 1.1104,
      "step": 1719
    },
    {
      "epoch": 0.344,
      "grad_norm": 3.598719358444214,
      "learning_rate": 3.346938775510204e-05,
      "loss": 0.9867,
      "step": 1720
    },
    {
      "epoch": 0.3442,
      "grad_norm": 1.3935763835906982,
      "learning_rate": 3.345918367346939e-05,
      "loss": 0.1097,
      "step": 1721
    },
    {
      "epoch": 0.3444,
      "grad_norm": 1.6227840185165405,
      "learning_rate": 3.344897959183674e-05,
      "loss": 0.1254,
      "step": 1722
    },
    {
      "epoch": 0.3446,
      "grad_norm": 1.2965258359909058,
      "learning_rate": 3.343877551020408e-05,
      "loss": 0.1083,
      "step": 1723
    },
    {
      "epoch": 0.3448,
      "grad_norm": 1.3589463233947754,
      "learning_rate": 3.342857142857143e-05,
      "loss": 0.1148,
      "step": 1724
    },
    {
      "epoch": 0.345,
      "grad_norm": 5.295959949493408,
      "learning_rate": 3.341836734693878e-05,
      "loss": 1.0539,
      "step": 1725
    },
    {
      "epoch": 0.3452,
      "grad_norm": 3.4420156478881836,
      "learning_rate": 3.3408163265306123e-05,
      "loss": 0.7745,
      "step": 1726
    },
    {
      "epoch": 0.3454,
      "grad_norm": 16.860403060913086,
      "learning_rate": 3.339795918367347e-05,
      "loss": 3.1186,
      "step": 1727
    },
    {
      "epoch": 0.3456,
      "grad_norm": 15.784538269042969,
      "learning_rate": 3.338775510204082e-05,
      "loss": 1.4251,
      "step": 1728
    },
    {
      "epoch": 0.3458,
      "grad_norm": 4.1750874519348145,
      "learning_rate": 3.3377551020408164e-05,
      "loss": 0.6993,
      "step": 1729
    },
    {
      "epoch": 0.346,
      "grad_norm": 3.424389362335205,
      "learning_rate": 3.336734693877551e-05,
      "loss": 0.8525,
      "step": 1730
    },
    {
      "epoch": 0.3462,
      "grad_norm": 6.516100883483887,
      "learning_rate": 3.3357142857142856e-05,
      "loss": 2.2779,
      "step": 1731
    },
    {
      "epoch": 0.3464,
      "grad_norm": 4.072001934051514,
      "learning_rate": 3.3346938775510205e-05,
      "loss": 0.6746,
      "step": 1732
    },
    {
      "epoch": 0.3466,
      "grad_norm": 8.418037414550781,
      "learning_rate": 3.3336734693877554e-05,
      "loss": 1.1869,
      "step": 1733
    },
    {
      "epoch": 0.3468,
      "grad_norm": 3.950887441635132,
      "learning_rate": 3.3326530612244897e-05,
      "loss": 0.7377,
      "step": 1734
    },
    {
      "epoch": 0.347,
      "grad_norm": 3.9305362701416016,
      "learning_rate": 3.3316326530612246e-05,
      "loss": 0.6518,
      "step": 1735
    },
    {
      "epoch": 0.3472,
      "grad_norm": 4.105065822601318,
      "learning_rate": 3.3306122448979595e-05,
      "loss": 0.7078,
      "step": 1736
    },
    {
      "epoch": 0.3474,
      "grad_norm": 3.4574952125549316,
      "learning_rate": 3.329591836734694e-05,
      "loss": 0.9431,
      "step": 1737
    },
    {
      "epoch": 0.3476,
      "grad_norm": 1.2512776851654053,
      "learning_rate": 3.3285714285714286e-05,
      "loss": 0.1185,
      "step": 1738
    },
    {
      "epoch": 0.3478,
      "grad_norm": 10.10742473602295,
      "learning_rate": 3.3275510204081636e-05,
      "loss": 1.3306,
      "step": 1739
    },
    {
      "epoch": 0.348,
      "grad_norm": 4.465425968170166,
      "learning_rate": 3.326530612244898e-05,
      "loss": 0.6967,
      "step": 1740
    },
    {
      "epoch": 0.3482,
      "grad_norm": 6.026064872741699,
      "learning_rate": 3.325510204081633e-05,
      "loss": 1.0078,
      "step": 1741
    },
    {
      "epoch": 0.3484,
      "grad_norm": 1.0465497970581055,
      "learning_rate": 3.3244897959183676e-05,
      "loss": 0.0829,
      "step": 1742
    },
    {
      "epoch": 0.3486,
      "grad_norm": 3.51488995552063,
      "learning_rate": 3.323469387755102e-05,
      "loss": 1.0144,
      "step": 1743
    },
    {
      "epoch": 0.3488,
      "grad_norm": 10.781816482543945,
      "learning_rate": 3.322448979591837e-05,
      "loss": 1.3457,
      "step": 1744
    },
    {
      "epoch": 0.349,
      "grad_norm": 3.6929171085357666,
      "learning_rate": 3.321428571428572e-05,
      "loss": 0.8511,
      "step": 1745
    },
    {
      "epoch": 0.3492,
      "grad_norm": 1.1213587522506714,
      "learning_rate": 3.320408163265306e-05,
      "loss": 0.0851,
      "step": 1746
    },
    {
      "epoch": 0.3494,
      "grad_norm": 9.472102165222168,
      "learning_rate": 3.319387755102041e-05,
      "loss": 2.6352,
      "step": 1747
    },
    {
      "epoch": 0.3496,
      "grad_norm": 1.7029060125350952,
      "learning_rate": 3.318367346938776e-05,
      "loss": 0.155,
      "step": 1748
    },
    {
      "epoch": 0.3498,
      "grad_norm": 1.4820170402526855,
      "learning_rate": 3.31734693877551e-05,
      "loss": 0.1119,
      "step": 1749
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.579092502593994,
      "learning_rate": 3.316326530612245e-05,
      "loss": 0.7513,
      "step": 1750
    },
    {
      "epoch": 0.3502,
      "grad_norm": 1.3974214792251587,
      "learning_rate": 3.31530612244898e-05,
      "loss": 0.1146,
      "step": 1751
    },
    {
      "epoch": 0.3504,
      "grad_norm": 4.359090805053711,
      "learning_rate": 3.314285714285714e-05,
      "loss": 0.6645,
      "step": 1752
    },
    {
      "epoch": 0.3506,
      "grad_norm": 4.183372497558594,
      "learning_rate": 3.313265306122449e-05,
      "loss": 0.747,
      "step": 1753
    },
    {
      "epoch": 0.3508,
      "grad_norm": 6.204871654510498,
      "learning_rate": 3.312244897959183e-05,
      "loss": 1.3034,
      "step": 1754
    },
    {
      "epoch": 0.351,
      "grad_norm": 3.895108938217163,
      "learning_rate": 3.311224489795918e-05,
      "loss": 0.9852,
      "step": 1755
    },
    {
      "epoch": 0.3512,
      "grad_norm": 1.3820586204528809,
      "learning_rate": 3.310204081632653e-05,
      "loss": 0.1083,
      "step": 1756
    },
    {
      "epoch": 0.3514,
      "grad_norm": 8.781912803649902,
      "learning_rate": 3.309183673469388e-05,
      "loss": 3.3185,
      "step": 1757
    },
    {
      "epoch": 0.3516,
      "grad_norm": 7.169617176055908,
      "learning_rate": 3.308163265306123e-05,
      "loss": 1.1307,
      "step": 1758
    },
    {
      "epoch": 0.3518,
      "grad_norm": 5.979995250701904,
      "learning_rate": 3.307142857142858e-05,
      "loss": 2.2011,
      "step": 1759
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.4406485557556152,
      "learning_rate": 3.306122448979592e-05,
      "loss": 0.1699,
      "step": 1760
    },
    {
      "epoch": 0.3522,
      "grad_norm": 1.7303863763809204,
      "learning_rate": 3.305102040816327e-05,
      "loss": 0.1256,
      "step": 1761
    },
    {
      "epoch": 0.3524,
      "grad_norm": 1.4316778182983398,
      "learning_rate": 3.304081632653061e-05,
      "loss": 0.132,
      "step": 1762
    },
    {
      "epoch": 0.3526,
      "grad_norm": 6.620378494262695,
      "learning_rate": 3.303061224489796e-05,
      "loss": 3.2379,
      "step": 1763
    },
    {
      "epoch": 0.3528,
      "grad_norm": 3.6853671073913574,
      "learning_rate": 3.302040816326531e-05,
      "loss": 1.0066,
      "step": 1764
    },
    {
      "epoch": 0.353,
      "grad_norm": 5.730157375335693,
      "learning_rate": 3.301020408163265e-05,
      "loss": 0.7325,
      "step": 1765
    },
    {
      "epoch": 0.3532,
      "grad_norm": 1.6502915620803833,
      "learning_rate": 3.3e-05,
      "loss": 0.1872,
      "step": 1766
    },
    {
      "epoch": 0.3534,
      "grad_norm": 1.535474419593811,
      "learning_rate": 3.298979591836735e-05,
      "loss": 0.1412,
      "step": 1767
    },
    {
      "epoch": 0.3536,
      "grad_norm": 7.426815509796143,
      "learning_rate": 3.2979591836734694e-05,
      "loss": 1.1631,
      "step": 1768
    },
    {
      "epoch": 0.3538,
      "grad_norm": 1.3126943111419678,
      "learning_rate": 3.296938775510204e-05,
      "loss": 0.1495,
      "step": 1769
    },
    {
      "epoch": 0.354,
      "grad_norm": 3.619706153869629,
      "learning_rate": 3.295918367346939e-05,
      "loss": 1.0407,
      "step": 1770
    },
    {
      "epoch": 0.3542,
      "grad_norm": 1.4426507949829102,
      "learning_rate": 3.2948979591836735e-05,
      "loss": 0.1323,
      "step": 1771
    },
    {
      "epoch": 0.3544,
      "grad_norm": 1.4123917818069458,
      "learning_rate": 3.2938775510204084e-05,
      "loss": 0.1208,
      "step": 1772
    },
    {
      "epoch": 0.3546,
      "grad_norm": 1.712772250175476,
      "learning_rate": 3.292857142857143e-05,
      "loss": 0.1838,
      "step": 1773
    },
    {
      "epoch": 0.3548,
      "grad_norm": 1.4630318880081177,
      "learning_rate": 3.2918367346938776e-05,
      "loss": 0.1842,
      "step": 1774
    },
    {
      "epoch": 0.355,
      "grad_norm": 1.5501052141189575,
      "learning_rate": 3.2908163265306125e-05,
      "loss": 0.1932,
      "step": 1775
    },
    {
      "epoch": 0.3552,
      "grad_norm": 8.72156810760498,
      "learning_rate": 3.2897959183673474e-05,
      "loss": 3.2597,
      "step": 1776
    },
    {
      "epoch": 0.3554,
      "grad_norm": 3.496630907058716,
      "learning_rate": 3.2887755102040816e-05,
      "loss": 0.9562,
      "step": 1777
    },
    {
      "epoch": 0.3556,
      "grad_norm": 1.2298327684402466,
      "learning_rate": 3.2877551020408166e-05,
      "loss": 0.1136,
      "step": 1778
    },
    {
      "epoch": 0.3558,
      "grad_norm": 5.842172622680664,
      "learning_rate": 3.2867346938775515e-05,
      "loss": 2.3213,
      "step": 1779
    },
    {
      "epoch": 0.356,
      "grad_norm": 1.2177481651306152,
      "learning_rate": 3.285714285714286e-05,
      "loss": 0.0954,
      "step": 1780
    },
    {
      "epoch": 0.3562,
      "grad_norm": 3.6284985542297363,
      "learning_rate": 3.2846938775510206e-05,
      "loss": 0.8903,
      "step": 1781
    },
    {
      "epoch": 0.3564,
      "grad_norm": 6.315433502197266,
      "learning_rate": 3.2836734693877556e-05,
      "loss": 1.3574,
      "step": 1782
    },
    {
      "epoch": 0.3566,
      "grad_norm": 5.381503582000732,
      "learning_rate": 3.28265306122449e-05,
      "loss": 1.0625,
      "step": 1783
    },
    {
      "epoch": 0.3568,
      "grad_norm": 3.6157219409942627,
      "learning_rate": 3.281632653061225e-05,
      "loss": 1.0214,
      "step": 1784
    },
    {
      "epoch": 0.357,
      "grad_norm": 3.836179494857788,
      "learning_rate": 3.280612244897959e-05,
      "loss": 0.6615,
      "step": 1785
    },
    {
      "epoch": 0.3572,
      "grad_norm": 3.564310312271118,
      "learning_rate": 3.279591836734694e-05,
      "loss": 0.8603,
      "step": 1786
    },
    {
      "epoch": 0.3574,
      "grad_norm": 10.237130165100098,
      "learning_rate": 3.278571428571429e-05,
      "loss": 3.5878,
      "step": 1787
    },
    {
      "epoch": 0.3576,
      "grad_norm": 3.611448287963867,
      "learning_rate": 3.277551020408163e-05,
      "loss": 0.9539,
      "step": 1788
    },
    {
      "epoch": 0.3578,
      "grad_norm": 5.697876453399658,
      "learning_rate": 3.276530612244898e-05,
      "loss": 1.0622,
      "step": 1789
    },
    {
      "epoch": 0.358,
      "grad_norm": 6.092684268951416,
      "learning_rate": 3.275510204081633e-05,
      "loss": 1.3316,
      "step": 1790
    },
    {
      "epoch": 0.3582,
      "grad_norm": 3.4067060947418213,
      "learning_rate": 3.274489795918367e-05,
      "loss": 0.8333,
      "step": 1791
    },
    {
      "epoch": 0.3584,
      "grad_norm": 5.643919944763184,
      "learning_rate": 3.273469387755102e-05,
      "loss": 1.0826,
      "step": 1792
    },
    {
      "epoch": 0.3586,
      "grad_norm": 5.771454334259033,
      "learning_rate": 3.272448979591837e-05,
      "loss": 1.1095,
      "step": 1793
    },
    {
      "epoch": 0.3588,
      "grad_norm": 5.640023231506348,
      "learning_rate": 3.271428571428571e-05,
      "loss": 1.1269,
      "step": 1794
    },
    {
      "epoch": 0.359,
      "grad_norm": 4.021417617797852,
      "learning_rate": 3.270408163265306e-05,
      "loss": 0.6279,
      "step": 1795
    },
    {
      "epoch": 0.3592,
      "grad_norm": 10.706412315368652,
      "learning_rate": 3.269387755102041e-05,
      "loss": 3.4476,
      "step": 1796
    },
    {
      "epoch": 0.3594,
      "grad_norm": 9.943094253540039,
      "learning_rate": 3.268367346938775e-05,
      "loss": 3.3975,
      "step": 1797
    },
    {
      "epoch": 0.3596,
      "grad_norm": 3.9147348403930664,
      "learning_rate": 3.26734693877551e-05,
      "loss": 0.6569,
      "step": 1798
    },
    {
      "epoch": 0.3598,
      "grad_norm": 6.527913570404053,
      "learning_rate": 3.266326530612245e-05,
      "loss": 0.9437,
      "step": 1799
    },
    {
      "epoch": 0.36,
      "grad_norm": 6.382730484008789,
      "learning_rate": 3.265306122448979e-05,
      "loss": 1.0618,
      "step": 1800
    },
    {
      "epoch": 0.3602,
      "grad_norm": 5.790674209594727,
      "learning_rate": 3.264285714285714e-05,
      "loss": 2.3089,
      "step": 1801
    },
    {
      "epoch": 0.3604,
      "grad_norm": 5.195087432861328,
      "learning_rate": 3.263265306122449e-05,
      "loss": 1.0027,
      "step": 1802
    },
    {
      "epoch": 0.3606,
      "grad_norm": 4.964000701904297,
      "learning_rate": 3.262244897959184e-05,
      "loss": 1.2752,
      "step": 1803
    },
    {
      "epoch": 0.3608,
      "grad_norm": 5.07620906829834,
      "learning_rate": 3.261224489795919e-05,
      "loss": 1.007,
      "step": 1804
    },
    {
      "epoch": 0.361,
      "grad_norm": 5.06422233581543,
      "learning_rate": 3.260204081632653e-05,
      "loss": 2.202,
      "step": 1805
    },
    {
      "epoch": 0.3612,
      "grad_norm": 5.387059688568115,
      "learning_rate": 3.259183673469388e-05,
      "loss": 1.014,
      "step": 1806
    },
    {
      "epoch": 0.3614,
      "grad_norm": 6.059737205505371,
      "learning_rate": 3.258163265306123e-05,
      "loss": 1.2625,
      "step": 1807
    },
    {
      "epoch": 0.3616,
      "grad_norm": 5.065815448760986,
      "learning_rate": 3.257142857142857e-05,
      "loss": 0.9486,
      "step": 1808
    },
    {
      "epoch": 0.3618,
      "grad_norm": 33.03364181518555,
      "learning_rate": 3.256122448979592e-05,
      "loss": 3.7953,
      "step": 1809
    },
    {
      "epoch": 0.362,
      "grad_norm": 7.243016242980957,
      "learning_rate": 3.255102040816327e-05,
      "loss": 1.1437,
      "step": 1810
    },
    {
      "epoch": 0.3622,
      "grad_norm": 4.155308246612549,
      "learning_rate": 3.2540816326530614e-05,
      "loss": 0.5773,
      "step": 1811
    },
    {
      "epoch": 0.3624,
      "grad_norm": 21.944828033447266,
      "learning_rate": 3.253061224489796e-05,
      "loss": 1.3502,
      "step": 1812
    },
    {
      "epoch": 0.3626,
      "grad_norm": 4.042191505432129,
      "learning_rate": 3.252040816326531e-05,
      "loss": 0.6203,
      "step": 1813
    },
    {
      "epoch": 0.3628,
      "grad_norm": 30.818618774414062,
      "learning_rate": 3.2510204081632655e-05,
      "loss": 1.3049,
      "step": 1814
    },
    {
      "epoch": 0.363,
      "grad_norm": 16.38467025756836,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 3.4577,
      "step": 1815
    },
    {
      "epoch": 0.3632,
      "grad_norm": 11.60643482208252,
      "learning_rate": 3.2489795918367346e-05,
      "loss": 3.4165,
      "step": 1816
    },
    {
      "epoch": 0.3634,
      "grad_norm": 6.943589210510254,
      "learning_rate": 3.2479591836734696e-05,
      "loss": 1.152,
      "step": 1817
    },
    {
      "epoch": 0.3636,
      "grad_norm": 8.237154960632324,
      "learning_rate": 3.2469387755102045e-05,
      "loss": 3.092,
      "step": 1818
    },
    {
      "epoch": 0.3638,
      "grad_norm": 10.214845657348633,
      "learning_rate": 3.245918367346939e-05,
      "loss": 3.4323,
      "step": 1819
    },
    {
      "epoch": 0.364,
      "grad_norm": 5.321285247802734,
      "learning_rate": 3.2448979591836736e-05,
      "loss": 0.8715,
      "step": 1820
    },
    {
      "epoch": 0.3642,
      "grad_norm": 5.701889991760254,
      "learning_rate": 3.2438775510204085e-05,
      "loss": 1.0432,
      "step": 1821
    },
    {
      "epoch": 0.3644,
      "grad_norm": 9.685300827026367,
      "learning_rate": 3.242857142857143e-05,
      "loss": 3.4456,
      "step": 1822
    },
    {
      "epoch": 0.3646,
      "grad_norm": 5.464596271514893,
      "learning_rate": 3.241836734693878e-05,
      "loss": 1.0736,
      "step": 1823
    },
    {
      "epoch": 0.3648,
      "grad_norm": 17.551986694335938,
      "learning_rate": 3.2408163265306126e-05,
      "loss": 1.3006,
      "step": 1824
    },
    {
      "epoch": 0.365,
      "grad_norm": 4.076687812805176,
      "learning_rate": 3.239795918367347e-05,
      "loss": 0.894,
      "step": 1825
    },
    {
      "epoch": 0.3652,
      "grad_norm": 9.390284538269043,
      "learning_rate": 3.238775510204082e-05,
      "loss": 3.2777,
      "step": 1826
    },
    {
      "epoch": 0.3654,
      "grad_norm": 3.8985793590545654,
      "learning_rate": 3.237755102040817e-05,
      "loss": 0.9889,
      "step": 1827
    },
    {
      "epoch": 0.3656,
      "grad_norm": 4.8796234130859375,
      "learning_rate": 3.236734693877551e-05,
      "loss": 0.6503,
      "step": 1828
    },
    {
      "epoch": 0.3658,
      "grad_norm": 4.104645729064941,
      "learning_rate": 3.235714285714286e-05,
      "loss": 0.5622,
      "step": 1829
    },
    {
      "epoch": 0.366,
      "grad_norm": 13.190925598144531,
      "learning_rate": 3.234693877551021e-05,
      "loss": 1.3046,
      "step": 1830
    },
    {
      "epoch": 0.3662,
      "grad_norm": 9.27966022491455,
      "learning_rate": 3.233673469387755e-05,
      "loss": 2.9469,
      "step": 1831
    },
    {
      "epoch": 0.3664,
      "grad_norm": 7.0232319831848145,
      "learning_rate": 3.23265306122449e-05,
      "loss": 1.1279,
      "step": 1832
    },
    {
      "epoch": 0.3666,
      "grad_norm": 4.636045455932617,
      "learning_rate": 3.231632653061225e-05,
      "loss": 0.9341,
      "step": 1833
    },
    {
      "epoch": 0.3668,
      "grad_norm": 4.049642562866211,
      "learning_rate": 3.230612244897959e-05,
      "loss": 0.6906,
      "step": 1834
    },
    {
      "epoch": 0.367,
      "grad_norm": 3.690493583679199,
      "learning_rate": 3.229591836734694e-05,
      "loss": 1.0225,
      "step": 1835
    },
    {
      "epoch": 0.3672,
      "grad_norm": 9.44584846496582,
      "learning_rate": 3.228571428571428e-05,
      "loss": 1.2827,
      "step": 1836
    },
    {
      "epoch": 0.3674,
      "grad_norm": 7.8593668937683105,
      "learning_rate": 3.227551020408163e-05,
      "loss": 1.1259,
      "step": 1837
    },
    {
      "epoch": 0.3676,
      "grad_norm": 12.65119743347168,
      "learning_rate": 3.226530612244898e-05,
      "loss": 1.3475,
      "step": 1838
    },
    {
      "epoch": 0.3678,
      "grad_norm": 14.825552940368652,
      "learning_rate": 3.225510204081632e-05,
      "loss": 1.2212,
      "step": 1839
    },
    {
      "epoch": 0.368,
      "grad_norm": 10.312542915344238,
      "learning_rate": 3.224489795918367e-05,
      "loss": 2.8721,
      "step": 1840
    },
    {
      "epoch": 0.3682,
      "grad_norm": 8.35904598236084,
      "learning_rate": 3.223469387755102e-05,
      "loss": 1.1117,
      "step": 1841
    },
    {
      "epoch": 0.3684,
      "grad_norm": 21.328636169433594,
      "learning_rate": 3.2224489795918364e-05,
      "loss": 1.2698,
      "step": 1842
    },
    {
      "epoch": 0.3686,
      "grad_norm": 11.115662574768066,
      "learning_rate": 3.221428571428571e-05,
      "loss": 1.2428,
      "step": 1843
    },
    {
      "epoch": 0.3688,
      "grad_norm": 8.736370086669922,
      "learning_rate": 3.220408163265306e-05,
      "loss": 2.8465,
      "step": 1844
    },
    {
      "epoch": 0.369,
      "grad_norm": 4.694275379180908,
      "learning_rate": 3.2193877551020405e-05,
      "loss": 0.6521,
      "step": 1845
    },
    {
      "epoch": 0.3692,
      "grad_norm": 20.076202392578125,
      "learning_rate": 3.2183673469387754e-05,
      "loss": 1.2942,
      "step": 1846
    },
    {
      "epoch": 0.3694,
      "grad_norm": 3.7784104347229004,
      "learning_rate": 3.21734693877551e-05,
      "loss": 0.8574,
      "step": 1847
    },
    {
      "epoch": 0.3696,
      "grad_norm": 3.9411227703094482,
      "learning_rate": 3.216326530612245e-05,
      "loss": 0.6632,
      "step": 1848
    },
    {
      "epoch": 0.3698,
      "grad_norm": 3.7799296379089355,
      "learning_rate": 3.21530612244898e-05,
      "loss": 0.8616,
      "step": 1849
    },
    {
      "epoch": 0.37,
      "grad_norm": 5.967972278594971,
      "learning_rate": 3.2142857142857144e-05,
      "loss": 0.9687,
      "step": 1850
    },
    {
      "epoch": 0.3702,
      "grad_norm": 6.866252899169922,
      "learning_rate": 3.213265306122449e-05,
      "loss": 2.7619,
      "step": 1851
    },
    {
      "epoch": 0.3704,
      "grad_norm": 3.4781665802001953,
      "learning_rate": 3.212244897959184e-05,
      "loss": 0.8378,
      "step": 1852
    },
    {
      "epoch": 0.3706,
      "grad_norm": 12.557558059692383,
      "learning_rate": 3.2112244897959185e-05,
      "loss": 1.2461,
      "step": 1853
    },
    {
      "epoch": 0.3708,
      "grad_norm": 3.999380111694336,
      "learning_rate": 3.2102040816326534e-05,
      "loss": 0.9952,
      "step": 1854
    },
    {
      "epoch": 0.371,
      "grad_norm": 6.250697612762451,
      "learning_rate": 3.209183673469388e-05,
      "loss": 1.1547,
      "step": 1855
    },
    {
      "epoch": 0.3712,
      "grad_norm": 3.5486440658569336,
      "learning_rate": 3.2081632653061225e-05,
      "loss": 0.8971,
      "step": 1856
    },
    {
      "epoch": 0.3714,
      "grad_norm": 4.229161262512207,
      "learning_rate": 3.2071428571428575e-05,
      "loss": 0.6361,
      "step": 1857
    },
    {
      "epoch": 0.3716,
      "grad_norm": 8.195026397705078,
      "learning_rate": 3.2061224489795924e-05,
      "loss": 3.408,
      "step": 1858
    },
    {
      "epoch": 0.3718,
      "grad_norm": 8.057538986206055,
      "learning_rate": 3.2051020408163266e-05,
      "loss": 1.2027,
      "step": 1859
    },
    {
      "epoch": 0.372,
      "grad_norm": 3.8987393379211426,
      "learning_rate": 3.2040816326530615e-05,
      "loss": 0.9229,
      "step": 1860
    },
    {
      "epoch": 0.3722,
      "grad_norm": 10.013763427734375,
      "learning_rate": 3.2030612244897965e-05,
      "loss": 3.5451,
      "step": 1861
    },
    {
      "epoch": 0.3724,
      "grad_norm": 6.321132183074951,
      "learning_rate": 3.202040816326531e-05,
      "loss": 1.0868,
      "step": 1862
    },
    {
      "epoch": 0.3726,
      "grad_norm": 3.6911380290985107,
      "learning_rate": 3.2010204081632656e-05,
      "loss": 0.9743,
      "step": 1863
    },
    {
      "epoch": 0.3728,
      "grad_norm": 6.979063510894775,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.0945,
      "step": 1864
    },
    {
      "epoch": 0.373,
      "grad_norm": 6.380610942840576,
      "learning_rate": 3.198979591836735e-05,
      "loss": 1.0856,
      "step": 1865
    },
    {
      "epoch": 0.3732,
      "grad_norm": 4.044541835784912,
      "learning_rate": 3.19795918367347e-05,
      "loss": 0.6641,
      "step": 1866
    },
    {
      "epoch": 0.3734,
      "grad_norm": 4.208895206451416,
      "learning_rate": 3.196938775510204e-05,
      "loss": 0.8649,
      "step": 1867
    },
    {
      "epoch": 0.3736,
      "grad_norm": 3.6989431381225586,
      "learning_rate": 3.195918367346939e-05,
      "loss": 0.9247,
      "step": 1868
    },
    {
      "epoch": 0.3738,
      "grad_norm": 6.997406005859375,
      "learning_rate": 3.194897959183674e-05,
      "loss": 1.1066,
      "step": 1869
    },
    {
      "epoch": 0.374,
      "grad_norm": 10.595443725585938,
      "learning_rate": 3.193877551020408e-05,
      "loss": 3.605,
      "step": 1870
    },
    {
      "epoch": 0.3742,
      "grad_norm": 5.664329528808594,
      "learning_rate": 3.192857142857143e-05,
      "loss": 1.0672,
      "step": 1871
    },
    {
      "epoch": 0.3744,
      "grad_norm": 3.9504380226135254,
      "learning_rate": 3.191836734693878e-05,
      "loss": 1.0203,
      "step": 1872
    },
    {
      "epoch": 0.3746,
      "grad_norm": 4.3827996253967285,
      "learning_rate": 3.190816326530612e-05,
      "loss": 0.7952,
      "step": 1873
    },
    {
      "epoch": 0.3748,
      "grad_norm": 3.58048152923584,
      "learning_rate": 3.189795918367347e-05,
      "loss": 0.8945,
      "step": 1874
    },
    {
      "epoch": 0.375,
      "grad_norm": 7.100594520568848,
      "learning_rate": 3.188775510204082e-05,
      "loss": 1.0029,
      "step": 1875
    },
    {
      "epoch": 0.3752,
      "grad_norm": 5.567595481872559,
      "learning_rate": 3.187755102040816e-05,
      "loss": 0.9961,
      "step": 1876
    },
    {
      "epoch": 0.3754,
      "grad_norm": 6.2788405418396,
      "learning_rate": 3.186734693877551e-05,
      "loss": 1.2381,
      "step": 1877
    },
    {
      "epoch": 0.3756,
      "grad_norm": 10.337995529174805,
      "learning_rate": 3.185714285714286e-05,
      "loss": 3.6624,
      "step": 1878
    },
    {
      "epoch": 0.3758,
      "grad_norm": 3.7018885612487793,
      "learning_rate": 3.18469387755102e-05,
      "loss": 1.0117,
      "step": 1879
    },
    {
      "epoch": 0.376,
      "grad_norm": 3.6972994804382324,
      "learning_rate": 3.183673469387755e-05,
      "loss": 0.803,
      "step": 1880
    },
    {
      "epoch": 0.3762,
      "grad_norm": 3.676743268966675,
      "learning_rate": 3.18265306122449e-05,
      "loss": 0.8658,
      "step": 1881
    },
    {
      "epoch": 0.3764,
      "grad_norm": 3.5221686363220215,
      "learning_rate": 3.181632653061224e-05,
      "loss": 0.5166,
      "step": 1882
    },
    {
      "epoch": 0.3766,
      "grad_norm": 9.288930892944336,
      "learning_rate": 3.180612244897959e-05,
      "loss": 3.1491,
      "step": 1883
    },
    {
      "epoch": 0.3768,
      "grad_norm": 9.832517623901367,
      "learning_rate": 3.179591836734694e-05,
      "loss": 1.2434,
      "step": 1884
    },
    {
      "epoch": 0.377,
      "grad_norm": 7.8202223777771,
      "learning_rate": 3.1785714285714284e-05,
      "loss": 1.105,
      "step": 1885
    },
    {
      "epoch": 0.3772,
      "grad_norm": 5.503623962402344,
      "learning_rate": 3.177551020408163e-05,
      "loss": 0.9261,
      "step": 1886
    },
    {
      "epoch": 0.3774,
      "grad_norm": 3.5174360275268555,
      "learning_rate": 3.176530612244898e-05,
      "loss": 0.5461,
      "step": 1887
    },
    {
      "epoch": 0.3776,
      "grad_norm": 9.028364181518555,
      "learning_rate": 3.1755102040816325e-05,
      "loss": 3.6403,
      "step": 1888
    },
    {
      "epoch": 0.3778,
      "grad_norm": 9.071781158447266,
      "learning_rate": 3.1744897959183674e-05,
      "loss": 3.6504,
      "step": 1889
    },
    {
      "epoch": 0.378,
      "grad_norm": 8.815235137939453,
      "learning_rate": 3.1734693877551016e-05,
      "loss": 3.2857,
      "step": 1890
    },
    {
      "epoch": 0.3782,
      "grad_norm": 3.7210514545440674,
      "learning_rate": 3.1724489795918365e-05,
      "loss": 0.9614,
      "step": 1891
    },
    {
      "epoch": 0.3784,
      "grad_norm": 10.739581108093262,
      "learning_rate": 3.1714285714285715e-05,
      "loss": 3.3892,
      "step": 1892
    },
    {
      "epoch": 0.3786,
      "grad_norm": 9.99426555633545,
      "learning_rate": 3.1704081632653064e-05,
      "loss": 3.0964,
      "step": 1893
    },
    {
      "epoch": 0.3788,
      "grad_norm": 3.962157964706421,
      "learning_rate": 3.169387755102041e-05,
      "loss": 0.9553,
      "step": 1894
    },
    {
      "epoch": 0.379,
      "grad_norm": 5.60536003112793,
      "learning_rate": 3.168367346938776e-05,
      "loss": 1.024,
      "step": 1895
    },
    {
      "epoch": 0.3792,
      "grad_norm": 8.856130599975586,
      "learning_rate": 3.1673469387755105e-05,
      "loss": 1.236,
      "step": 1896
    },
    {
      "epoch": 0.3794,
      "grad_norm": 5.325572490692139,
      "learning_rate": 3.1663265306122454e-05,
      "loss": 1.1324,
      "step": 1897
    },
    {
      "epoch": 0.3796,
      "grad_norm": 3.662156343460083,
      "learning_rate": 3.1653061224489796e-05,
      "loss": 0.8391,
      "step": 1898
    },
    {
      "epoch": 0.3798,
      "grad_norm": 6.837795734405518,
      "learning_rate": 3.1642857142857145e-05,
      "loss": 1.0977,
      "step": 1899
    },
    {
      "epoch": 0.38,
      "grad_norm": 6.636470317840576,
      "learning_rate": 3.1632653061224494e-05,
      "loss": 1.1679,
      "step": 1900
    },
    {
      "epoch": 0.3802,
      "grad_norm": 5.536612033843994,
      "learning_rate": 3.162244897959184e-05,
      "loss": 0.9724,
      "step": 1901
    },
    {
      "epoch": 0.3804,
      "grad_norm": 5.437753200531006,
      "learning_rate": 3.1612244897959186e-05,
      "loss": 0.9618,
      "step": 1902
    },
    {
      "epoch": 0.3806,
      "grad_norm": 5.53395938873291,
      "learning_rate": 3.1602040816326535e-05,
      "loss": 0.9768,
      "step": 1903
    },
    {
      "epoch": 0.3808,
      "grad_norm": 26.427989959716797,
      "learning_rate": 3.159183673469388e-05,
      "loss": 3.7384,
      "step": 1904
    },
    {
      "epoch": 0.381,
      "grad_norm": 3.5214948654174805,
      "learning_rate": 3.158163265306123e-05,
      "loss": 0.5824,
      "step": 1905
    },
    {
      "epoch": 0.3812,
      "grad_norm": 4.140063762664795,
      "learning_rate": 3.1571428571428576e-05,
      "loss": 0.6823,
      "step": 1906
    },
    {
      "epoch": 0.3814,
      "grad_norm": 5.837836265563965,
      "learning_rate": 3.156122448979592e-05,
      "loss": 1.1179,
      "step": 1907
    },
    {
      "epoch": 0.3816,
      "grad_norm": 5.40811824798584,
      "learning_rate": 3.155102040816327e-05,
      "loss": 1.0316,
      "step": 1908
    },
    {
      "epoch": 0.3818,
      "grad_norm": 4.175589084625244,
      "learning_rate": 3.154081632653062e-05,
      "loss": 0.9813,
      "step": 1909
    },
    {
      "epoch": 0.382,
      "grad_norm": 5.406588554382324,
      "learning_rate": 3.153061224489796e-05,
      "loss": 0.8905,
      "step": 1910
    },
    {
      "epoch": 0.3822,
      "grad_norm": 3.529309034347534,
      "learning_rate": 3.152040816326531e-05,
      "loss": 0.9562,
      "step": 1911
    },
    {
      "epoch": 0.3824,
      "grad_norm": 11.12568187713623,
      "learning_rate": 3.151020408163266e-05,
      "loss": 1.2959,
      "step": 1912
    },
    {
      "epoch": 0.3826,
      "grad_norm": 7.0006184577941895,
      "learning_rate": 3.15e-05,
      "loss": 1.0464,
      "step": 1913
    },
    {
      "epoch": 0.3828,
      "grad_norm": 3.590198040008545,
      "learning_rate": 3.148979591836735e-05,
      "loss": 0.9875,
      "step": 1914
    },
    {
      "epoch": 0.383,
      "grad_norm": 5.859086513519287,
      "learning_rate": 3.14795918367347e-05,
      "loss": 1.0345,
      "step": 1915
    },
    {
      "epoch": 0.3832,
      "grad_norm": 3.9727370738983154,
      "learning_rate": 3.146938775510204e-05,
      "loss": 0.6396,
      "step": 1916
    },
    {
      "epoch": 0.3834,
      "grad_norm": 3.9990367889404297,
      "learning_rate": 3.145918367346939e-05,
      "loss": 0.6313,
      "step": 1917
    },
    {
      "epoch": 0.3836,
      "grad_norm": 4.18431282043457,
      "learning_rate": 3.144897959183674e-05,
      "loss": 0.8224,
      "step": 1918
    },
    {
      "epoch": 0.3838,
      "grad_norm": 5.706707954406738,
      "learning_rate": 3.143877551020408e-05,
      "loss": 1.0675,
      "step": 1919
    },
    {
      "epoch": 0.384,
      "grad_norm": 3.66581392288208,
      "learning_rate": 3.142857142857143e-05,
      "loss": 0.9332,
      "step": 1920
    },
    {
      "epoch": 0.3842,
      "grad_norm": 3.9625906944274902,
      "learning_rate": 3.141836734693877e-05,
      "loss": 0.6328,
      "step": 1921
    },
    {
      "epoch": 0.3844,
      "grad_norm": 17.030656814575195,
      "learning_rate": 3.140816326530612e-05,
      "loss": 3.9366,
      "step": 1922
    },
    {
      "epoch": 0.3846,
      "grad_norm": 5.616252899169922,
      "learning_rate": 3.139795918367347e-05,
      "loss": 0.9011,
      "step": 1923
    },
    {
      "epoch": 0.3848,
      "grad_norm": 6.17678689956665,
      "learning_rate": 3.1387755102040814e-05,
      "loss": 1.0409,
      "step": 1924
    },
    {
      "epoch": 0.385,
      "grad_norm": 4.09219217300415,
      "learning_rate": 3.137755102040816e-05,
      "loss": 0.6887,
      "step": 1925
    },
    {
      "epoch": 0.3852,
      "grad_norm": 8.822908401489258,
      "learning_rate": 3.136734693877551e-05,
      "loss": 3.6639,
      "step": 1926
    },
    {
      "epoch": 0.3854,
      "grad_norm": 5.703184604644775,
      "learning_rate": 3.1357142857142855e-05,
      "loss": 0.9716,
      "step": 1927
    },
    {
      "epoch": 0.3856,
      "grad_norm": 5.323953151702881,
      "learning_rate": 3.1346938775510204e-05,
      "loss": 0.9321,
      "step": 1928
    },
    {
      "epoch": 0.3858,
      "grad_norm": 3.9743220806121826,
      "learning_rate": 3.133673469387755e-05,
      "loss": 0.809,
      "step": 1929
    },
    {
      "epoch": 0.386,
      "grad_norm": 3.9991021156311035,
      "learning_rate": 3.1326530612244895e-05,
      "loss": 0.5908,
      "step": 1930
    },
    {
      "epoch": 0.3862,
      "grad_norm": 3.7616851329803467,
      "learning_rate": 3.1316326530612245e-05,
      "loss": 0.906,
      "step": 1931
    },
    {
      "epoch": 0.3864,
      "grad_norm": 11.455646514892578,
      "learning_rate": 3.1306122448979594e-05,
      "loss": 1.278,
      "step": 1932
    },
    {
      "epoch": 0.3866,
      "grad_norm": 5.890268802642822,
      "learning_rate": 3.1295918367346936e-05,
      "loss": 1.075,
      "step": 1933
    },
    {
      "epoch": 0.3868,
      "grad_norm": 10.48533821105957,
      "learning_rate": 3.1285714285714285e-05,
      "loss": 3.1911,
      "step": 1934
    },
    {
      "epoch": 0.387,
      "grad_norm": 3.556635856628418,
      "learning_rate": 3.1275510204081634e-05,
      "loss": 0.9989,
      "step": 1935
    },
    {
      "epoch": 0.3872,
      "grad_norm": 9.573206901550293,
      "learning_rate": 3.126530612244898e-05,
      "loss": 2.9485,
      "step": 1936
    },
    {
      "epoch": 0.3874,
      "grad_norm": 5.702581882476807,
      "learning_rate": 3.1255102040816326e-05,
      "loss": 1.0445,
      "step": 1937
    },
    {
      "epoch": 0.3876,
      "grad_norm": 5.90617561340332,
      "learning_rate": 3.1244897959183675e-05,
      "loss": 1.0362,
      "step": 1938
    },
    {
      "epoch": 0.3878,
      "grad_norm": 5.865963459014893,
      "learning_rate": 3.1234693877551024e-05,
      "loss": 1.0315,
      "step": 1939
    },
    {
      "epoch": 0.388,
      "grad_norm": 6.144934177398682,
      "learning_rate": 3.1224489795918374e-05,
      "loss": 1.0469,
      "step": 1940
    },
    {
      "epoch": 0.3882,
      "grad_norm": 5.400904178619385,
      "learning_rate": 3.1214285714285716e-05,
      "loss": 0.9472,
      "step": 1941
    },
    {
      "epoch": 0.3884,
      "grad_norm": 5.631744384765625,
      "learning_rate": 3.1204081632653065e-05,
      "loss": 1.0641,
      "step": 1942
    },
    {
      "epoch": 0.3886,
      "grad_norm": 9.171345710754395,
      "learning_rate": 3.1193877551020414e-05,
      "loss": 1.2373,
      "step": 1943
    },
    {
      "epoch": 0.3888,
      "grad_norm": 4.0366530418396,
      "learning_rate": 3.118367346938776e-05,
      "loss": 0.5937,
      "step": 1944
    },
    {
      "epoch": 0.389,
      "grad_norm": 3.8050081729888916,
      "learning_rate": 3.1173469387755106e-05,
      "loss": 0.8319,
      "step": 1945
    },
    {
      "epoch": 0.3892,
      "grad_norm": 5.817963123321533,
      "learning_rate": 3.1163265306122455e-05,
      "loss": 1.0481,
      "step": 1946
    },
    {
      "epoch": 0.3894,
      "grad_norm": 6.208949089050293,
      "learning_rate": 3.11530612244898e-05,
      "loss": 1.0638,
      "step": 1947
    },
    {
      "epoch": 0.3896,
      "grad_norm": 3.700920343399048,
      "learning_rate": 3.114285714285715e-05,
      "loss": 0.6374,
      "step": 1948
    },
    {
      "epoch": 0.3898,
      "grad_norm": 5.019497871398926,
      "learning_rate": 3.1132653061224496e-05,
      "loss": 0.9131,
      "step": 1949
    },
    {
      "epoch": 0.39,
      "grad_norm": 21.337099075317383,
      "learning_rate": 3.112244897959184e-05,
      "loss": 3.1491,
      "step": 1950
    },
    {
      "epoch": 0.3902,
      "grad_norm": 18.03508758544922,
      "learning_rate": 3.111224489795919e-05,
      "loss": 3.3816,
      "step": 1951
    },
    {
      "epoch": 0.3904,
      "grad_norm": 4.190047740936279,
      "learning_rate": 3.110204081632653e-05,
      "loss": 0.7675,
      "step": 1952
    },
    {
      "epoch": 0.3906,
      "grad_norm": 14.109248161315918,
      "learning_rate": 3.109183673469388e-05,
      "loss": 1.2913,
      "step": 1953
    },
    {
      "epoch": 0.3908,
      "grad_norm": 4.065311431884766,
      "learning_rate": 3.108163265306123e-05,
      "loss": 0.6556,
      "step": 1954
    },
    {
      "epoch": 0.391,
      "grad_norm": 3.806204080581665,
      "learning_rate": 3.107142857142857e-05,
      "loss": 0.8599,
      "step": 1955
    },
    {
      "epoch": 0.3912,
      "grad_norm": 14.704195976257324,
      "learning_rate": 3.106122448979592e-05,
      "loss": 1.2845,
      "step": 1956
    },
    {
      "epoch": 0.3914,
      "grad_norm": 17.34805679321289,
      "learning_rate": 3.105102040816327e-05,
      "loss": 1.3075,
      "step": 1957
    },
    {
      "epoch": 0.3916,
      "grad_norm": 5.221651077270508,
      "learning_rate": 3.104081632653061e-05,
      "loss": 1.0788,
      "step": 1958
    },
    {
      "epoch": 0.3918,
      "grad_norm": 20.793325424194336,
      "learning_rate": 3.103061224489796e-05,
      "loss": 1.2658,
      "step": 1959
    },
    {
      "epoch": 0.392,
      "grad_norm": 3.698373556137085,
      "learning_rate": 3.102040816326531e-05,
      "loss": 0.9564,
      "step": 1960
    },
    {
      "epoch": 0.3922,
      "grad_norm": 6.547873020172119,
      "learning_rate": 3.101020408163265e-05,
      "loss": 1.096,
      "step": 1961
    },
    {
      "epoch": 0.3924,
      "grad_norm": 3.6642978191375732,
      "learning_rate": 3.1e-05,
      "loss": 0.9244,
      "step": 1962
    },
    {
      "epoch": 0.3926,
      "grad_norm": 4.003145694732666,
      "learning_rate": 3.098979591836735e-05,
      "loss": 0.6351,
      "step": 1963
    },
    {
      "epoch": 0.3928,
      "grad_norm": 37.91643524169922,
      "learning_rate": 3.097959183673469e-05,
      "loss": 4.2161,
      "step": 1964
    },
    {
      "epoch": 0.393,
      "grad_norm": 5.784246444702148,
      "learning_rate": 3.096938775510204e-05,
      "loss": 0.9189,
      "step": 1965
    },
    {
      "epoch": 0.3932,
      "grad_norm": 3.4148221015930176,
      "learning_rate": 3.095918367346939e-05,
      "loss": 0.7949,
      "step": 1966
    },
    {
      "epoch": 0.3934,
      "grad_norm": 3.614133834838867,
      "learning_rate": 3.0948979591836734e-05,
      "loss": 0.5784,
      "step": 1967
    },
    {
      "epoch": 0.3936,
      "grad_norm": 8.935428619384766,
      "learning_rate": 3.093877551020408e-05,
      "loss": 1.1826,
      "step": 1968
    },
    {
      "epoch": 0.3938,
      "grad_norm": 5.664531230926514,
      "learning_rate": 3.092857142857143e-05,
      "loss": 0.9772,
      "step": 1969
    },
    {
      "epoch": 0.394,
      "grad_norm": 3.752865791320801,
      "learning_rate": 3.0918367346938774e-05,
      "loss": 0.613,
      "step": 1970
    },
    {
      "epoch": 0.3942,
      "grad_norm": 15.220873832702637,
      "learning_rate": 3.0908163265306124e-05,
      "loss": 3.7921,
      "step": 1971
    },
    {
      "epoch": 0.3944,
      "grad_norm": 3.931366443634033,
      "learning_rate": 3.0897959183673466e-05,
      "loss": 0.6618,
      "step": 1972
    },
    {
      "epoch": 0.3946,
      "grad_norm": 3.9736874103546143,
      "learning_rate": 3.0887755102040815e-05,
      "loss": 0.6155,
      "step": 1973
    },
    {
      "epoch": 0.3948,
      "grad_norm": 6.704489707946777,
      "learning_rate": 3.0877551020408164e-05,
      "loss": 1.1973,
      "step": 1974
    },
    {
      "epoch": 0.395,
      "grad_norm": 8.21762466430664,
      "learning_rate": 3.086734693877551e-05,
      "loss": 1.3387,
      "step": 1975
    },
    {
      "epoch": 0.3952,
      "grad_norm": 5.635308742523193,
      "learning_rate": 3.0857142857142856e-05,
      "loss": 1.0915,
      "step": 1976
    },
    {
      "epoch": 0.3954,
      "grad_norm": 3.429318428039551,
      "learning_rate": 3.0846938775510205e-05,
      "loss": 0.8407,
      "step": 1977
    },
    {
      "epoch": 0.3956,
      "grad_norm": 5.429908275604248,
      "learning_rate": 3.083673469387755e-05,
      "loss": 0.9866,
      "step": 1978
    },
    {
      "epoch": 0.3958,
      "grad_norm": 3.98653507232666,
      "learning_rate": 3.08265306122449e-05,
      "loss": 1.0904,
      "step": 1979
    },
    {
      "epoch": 0.396,
      "grad_norm": 3.7446930408477783,
      "learning_rate": 3.0816326530612246e-05,
      "loss": 0.9527,
      "step": 1980
    },
    {
      "epoch": 0.3962,
      "grad_norm": 3.985771417617798,
      "learning_rate": 3.080612244897959e-05,
      "loss": 0.5949,
      "step": 1981
    },
    {
      "epoch": 0.3964,
      "grad_norm": 5.2019453048706055,
      "learning_rate": 3.079591836734694e-05,
      "loss": 0.9761,
      "step": 1982
    },
    {
      "epoch": 0.3966,
      "grad_norm": 3.969365119934082,
      "learning_rate": 3.078571428571429e-05,
      "loss": 0.6185,
      "step": 1983
    },
    {
      "epoch": 0.3968,
      "grad_norm": 3.3323633670806885,
      "learning_rate": 3.0775510204081636e-05,
      "loss": 0.9703,
      "step": 1984
    },
    {
      "epoch": 0.397,
      "grad_norm": 19.202472686767578,
      "learning_rate": 3.0765306122448985e-05,
      "loss": 3.8001,
      "step": 1985
    },
    {
      "epoch": 0.3972,
      "grad_norm": 3.3859543800354004,
      "learning_rate": 3.075510204081633e-05,
      "loss": 0.959,
      "step": 1986
    },
    {
      "epoch": 0.3974,
      "grad_norm": 3.364506244659424,
      "learning_rate": 3.074489795918368e-05,
      "loss": 0.9518,
      "step": 1987
    },
    {
      "epoch": 0.3976,
      "grad_norm": 11.729973793029785,
      "learning_rate": 3.0734693877551026e-05,
      "loss": 4.0912,
      "step": 1988
    },
    {
      "epoch": 0.3978,
      "grad_norm": 9.226360321044922,
      "learning_rate": 3.072448979591837e-05,
      "loss": 4.0417,
      "step": 1989
    },
    {
      "epoch": 0.398,
      "grad_norm": 8.097306251525879,
      "learning_rate": 3.071428571428572e-05,
      "loss": 1.2364,
      "step": 1990
    },
    {
      "epoch": 0.3982,
      "grad_norm": 5.448502063751221,
      "learning_rate": 3.0704081632653067e-05,
      "loss": 1.0877,
      "step": 1991
    },
    {
      "epoch": 0.3984,
      "grad_norm": 3.436793327331543,
      "learning_rate": 3.069387755102041e-05,
      "loss": 0.8769,
      "step": 1992
    },
    {
      "epoch": 0.3986,
      "grad_norm": 3.4772262573242188,
      "learning_rate": 3.068367346938776e-05,
      "loss": 0.9147,
      "step": 1993
    },
    {
      "epoch": 0.3988,
      "grad_norm": 10.667865753173828,
      "learning_rate": 3.067346938775511e-05,
      "loss": 1.2431,
      "step": 1994
    },
    {
      "epoch": 0.399,
      "grad_norm": 3.464256763458252,
      "learning_rate": 3.066326530612245e-05,
      "loss": 0.9025,
      "step": 1995
    },
    {
      "epoch": 0.3992,
      "grad_norm": 4.23721170425415,
      "learning_rate": 3.06530612244898e-05,
      "loss": 0.6652,
      "step": 1996
    },
    {
      "epoch": 0.3994,
      "grad_norm": 4.40722131729126,
      "learning_rate": 3.064285714285715e-05,
      "loss": 0.5962,
      "step": 1997
    },
    {
      "epoch": 0.3996,
      "grad_norm": 5.087538719177246,
      "learning_rate": 3.063265306122449e-05,
      "loss": 1.0236,
      "step": 1998
    },
    {
      "epoch": 0.3998,
      "grad_norm": 3.472217321395874,
      "learning_rate": 3.062244897959184e-05,
      "loss": 0.8398,
      "step": 1999
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.690279960632324,
      "learning_rate": 3.061224489795919e-05,
      "loss": 0.9392,
      "step": 2000
    },
    {
      "epoch": 0.4002,
      "grad_norm": 6.091668605804443,
      "learning_rate": 3.060204081632653e-05,
      "loss": 1.065,
      "step": 2001
    },
    {
      "epoch": 0.4004,
      "grad_norm": 5.0249924659729,
      "learning_rate": 3.059183673469388e-05,
      "loss": 1.0291,
      "step": 2002
    },
    {
      "epoch": 0.4006,
      "grad_norm": 4.939494609832764,
      "learning_rate": 3.058163265306123e-05,
      "loss": 0.8855,
      "step": 2003
    },
    {
      "epoch": 0.4008,
      "grad_norm": 4.587623119354248,
      "learning_rate": 3.057142857142857e-05,
      "loss": 0.9673,
      "step": 2004
    },
    {
      "epoch": 0.401,
      "grad_norm": 7.077375411987305,
      "learning_rate": 3.056122448979592e-05,
      "loss": 0.9796,
      "step": 2005
    },
    {
      "epoch": 0.4012,
      "grad_norm": 5.092260837554932,
      "learning_rate": 3.0551020408163264e-05,
      "loss": 0.9718,
      "step": 2006
    },
    {
      "epoch": 0.4014,
      "grad_norm": 3.5044186115264893,
      "learning_rate": 3.054081632653061e-05,
      "loss": 0.8064,
      "step": 2007
    },
    {
      "epoch": 0.4016,
      "grad_norm": 6.15052604675293,
      "learning_rate": 3.053061224489796e-05,
      "loss": 0.9817,
      "step": 2008
    },
    {
      "epoch": 0.4018,
      "grad_norm": 9.287921905517578,
      "learning_rate": 3.0520408163265304e-05,
      "loss": 1.2837,
      "step": 2009
    },
    {
      "epoch": 0.402,
      "grad_norm": 3.6782681941986084,
      "learning_rate": 3.0510204081632654e-05,
      "loss": 0.859,
      "step": 2010
    },
    {
      "epoch": 0.4022,
      "grad_norm": 5.276793479919434,
      "learning_rate": 3.05e-05,
      "loss": 0.6698,
      "step": 2011
    },
    {
      "epoch": 0.4024,
      "grad_norm": 4.4981560707092285,
      "learning_rate": 3.048979591836735e-05,
      "loss": 0.6247,
      "step": 2012
    },
    {
      "epoch": 0.4026,
      "grad_norm": 4.182480335235596,
      "learning_rate": 3.0479591836734694e-05,
      "loss": 0.9111,
      "step": 2013
    },
    {
      "epoch": 0.4028,
      "grad_norm": 4.58298397064209,
      "learning_rate": 3.046938775510204e-05,
      "loss": 0.9857,
      "step": 2014
    },
    {
      "epoch": 0.403,
      "grad_norm": 7.412697792053223,
      "learning_rate": 3.045918367346939e-05,
      "loss": 1.0283,
      "step": 2015
    },
    {
      "epoch": 0.4032,
      "grad_norm": 4.1168694496154785,
      "learning_rate": 3.0448979591836735e-05,
      "loss": 0.6491,
      "step": 2016
    },
    {
      "epoch": 0.4034,
      "grad_norm": 3.5667405128479004,
      "learning_rate": 3.043877551020408e-05,
      "loss": 0.9392,
      "step": 2017
    },
    {
      "epoch": 0.4036,
      "grad_norm": 6.736431121826172,
      "learning_rate": 3.042857142857143e-05,
      "loss": 1.0647,
      "step": 2018
    },
    {
      "epoch": 0.4038,
      "grad_norm": 3.888291120529175,
      "learning_rate": 3.0418367346938776e-05,
      "loss": 0.8569,
      "step": 2019
    },
    {
      "epoch": 0.404,
      "grad_norm": 6.160373687744141,
      "learning_rate": 3.040816326530612e-05,
      "loss": 1.0167,
      "step": 2020
    },
    {
      "epoch": 0.4042,
      "grad_norm": 4.123867511749268,
      "learning_rate": 3.0397959183673467e-05,
      "loss": 0.8689,
      "step": 2021
    },
    {
      "epoch": 0.4044,
      "grad_norm": 13.053849220275879,
      "learning_rate": 3.0387755102040817e-05,
      "loss": 1.236,
      "step": 2022
    },
    {
      "epoch": 0.4046,
      "grad_norm": 9.465170860290527,
      "learning_rate": 3.0377551020408162e-05,
      "loss": 1.2494,
      "step": 2023
    },
    {
      "epoch": 0.4048,
      "grad_norm": 7.582245349884033,
      "learning_rate": 3.0367346938775508e-05,
      "loss": 1.214,
      "step": 2024
    },
    {
      "epoch": 0.405,
      "grad_norm": 7.706305980682373,
      "learning_rate": 3.0357142857142857e-05,
      "loss": 1.0576,
      "step": 2025
    },
    {
      "epoch": 0.4052,
      "grad_norm": 14.02505874633789,
      "learning_rate": 3.0346938775510203e-05,
      "loss": 1.0928,
      "step": 2026
    },
    {
      "epoch": 0.4054,
      "grad_norm": 7.376889705657959,
      "learning_rate": 3.033673469387755e-05,
      "loss": 1.2182,
      "step": 2027
    },
    {
      "epoch": 0.4056,
      "grad_norm": 3.411616086959839,
      "learning_rate": 3.0326530612244898e-05,
      "loss": 0.8649,
      "step": 2028
    },
    {
      "epoch": 0.4058,
      "grad_norm": 11.657422065734863,
      "learning_rate": 3.031632653061225e-05,
      "loss": 1.2291,
      "step": 2029
    },
    {
      "epoch": 0.406,
      "grad_norm": 4.260434150695801,
      "learning_rate": 3.0306122448979597e-05,
      "loss": 0.5884,
      "step": 2030
    },
    {
      "epoch": 0.4062,
      "grad_norm": 8.48438835144043,
      "learning_rate": 3.0295918367346942e-05,
      "loss": 1.062,
      "step": 2031
    },
    {
      "epoch": 0.4064,
      "grad_norm": 3.534970998764038,
      "learning_rate": 3.0285714285714288e-05,
      "loss": 0.889,
      "step": 2032
    },
    {
      "epoch": 0.4066,
      "grad_norm": 3.441049337387085,
      "learning_rate": 3.0275510204081637e-05,
      "loss": 0.958,
      "step": 2033
    },
    {
      "epoch": 0.4068,
      "grad_norm": 3.2886886596679688,
      "learning_rate": 3.0265306122448983e-05,
      "loss": 0.786,
      "step": 2034
    },
    {
      "epoch": 0.407,
      "grad_norm": 3.797276020050049,
      "learning_rate": 3.025510204081633e-05,
      "loss": 0.5687,
      "step": 2035
    },
    {
      "epoch": 0.4072,
      "grad_norm": 3.5251455307006836,
      "learning_rate": 3.0244897959183678e-05,
      "loss": 0.7998,
      "step": 2036
    },
    {
      "epoch": 0.4074,
      "grad_norm": 3.2015249729156494,
      "learning_rate": 3.0234693877551024e-05,
      "loss": 0.743,
      "step": 2037
    },
    {
      "epoch": 0.4076,
      "grad_norm": 3.5168328285217285,
      "learning_rate": 3.022448979591837e-05,
      "loss": 0.8678,
      "step": 2038
    },
    {
      "epoch": 0.4078,
      "grad_norm": 3.90354061126709,
      "learning_rate": 3.021428571428572e-05,
      "loss": 0.6304,
      "step": 2039
    },
    {
      "epoch": 0.408,
      "grad_norm": 3.3159353733062744,
      "learning_rate": 3.0204081632653065e-05,
      "loss": 0.7902,
      "step": 2040
    },
    {
      "epoch": 0.4082,
      "grad_norm": 3.9887642860412598,
      "learning_rate": 3.019387755102041e-05,
      "loss": 0.545,
      "step": 2041
    },
    {
      "epoch": 0.4084,
      "grad_norm": 3.527355670928955,
      "learning_rate": 3.0183673469387756e-05,
      "loss": 0.7923,
      "step": 2042
    },
    {
      "epoch": 0.4086,
      "grad_norm": 3.8808462619781494,
      "learning_rate": 3.0173469387755105e-05,
      "loss": 0.9774,
      "step": 2043
    },
    {
      "epoch": 0.4088,
      "grad_norm": 3.7496337890625,
      "learning_rate": 3.016326530612245e-05,
      "loss": 0.7932,
      "step": 2044
    },
    {
      "epoch": 0.409,
      "grad_norm": 3.646125316619873,
      "learning_rate": 3.0153061224489797e-05,
      "loss": 0.9224,
      "step": 2045
    },
    {
      "epoch": 0.4092,
      "grad_norm": 3.7936694622039795,
      "learning_rate": 3.0142857142857146e-05,
      "loss": 0.9792,
      "step": 2046
    },
    {
      "epoch": 0.4094,
      "grad_norm": 4.863035202026367,
      "learning_rate": 3.0132653061224492e-05,
      "loss": 0.7245,
      "step": 2047
    },
    {
      "epoch": 0.4096,
      "grad_norm": 4.080334186553955,
      "learning_rate": 3.0122448979591838e-05,
      "loss": 0.5945,
      "step": 2048
    },
    {
      "epoch": 0.4098,
      "grad_norm": 3.421778440475464,
      "learning_rate": 3.0112244897959187e-05,
      "loss": 0.9618,
      "step": 2049
    },
    {
      "epoch": 0.41,
      "grad_norm": 3.776970624923706,
      "learning_rate": 3.0102040816326533e-05,
      "loss": 0.9939,
      "step": 2050
    },
    {
      "epoch": 0.4102,
      "grad_norm": 11.517478942871094,
      "learning_rate": 3.009183673469388e-05,
      "loss": 1.2231,
      "step": 2051
    },
    {
      "epoch": 0.4104,
      "grad_norm": 3.593852996826172,
      "learning_rate": 3.0081632653061224e-05,
      "loss": 0.8188,
      "step": 2052
    },
    {
      "epoch": 0.4106,
      "grad_norm": 3.9655921459198,
      "learning_rate": 3.0071428571428573e-05,
      "loss": 0.5983,
      "step": 2053
    },
    {
      "epoch": 0.4108,
      "grad_norm": 7.757854461669922,
      "learning_rate": 3.006122448979592e-05,
      "loss": 1.2068,
      "step": 2054
    },
    {
      "epoch": 0.411,
      "grad_norm": 3.8167548179626465,
      "learning_rate": 3.0051020408163265e-05,
      "loss": 0.8267,
      "step": 2055
    },
    {
      "epoch": 0.4112,
      "grad_norm": 3.506699323654175,
      "learning_rate": 3.0040816326530614e-05,
      "loss": 0.8534,
      "step": 2056
    },
    {
      "epoch": 0.4114,
      "grad_norm": 6.759268760681152,
      "learning_rate": 3.003061224489796e-05,
      "loss": 1.2464,
      "step": 2057
    },
    {
      "epoch": 0.4116,
      "grad_norm": 3.3585410118103027,
      "learning_rate": 3.0020408163265306e-05,
      "loss": 0.8959,
      "step": 2058
    },
    {
      "epoch": 0.4118,
      "grad_norm": 3.979999303817749,
      "learning_rate": 3.0010204081632655e-05,
      "loss": 0.5908,
      "step": 2059
    },
    {
      "epoch": 0.412,
      "grad_norm": 3.32076096534729,
      "learning_rate": 3e-05,
      "loss": 0.8333,
      "step": 2060
    },
    {
      "epoch": 0.4122,
      "grad_norm": 3.9741761684417725,
      "learning_rate": 2.9989795918367347e-05,
      "loss": 0.603,
      "step": 2061
    },
    {
      "epoch": 0.4124,
      "grad_norm": 3.654820442199707,
      "learning_rate": 2.9979591836734692e-05,
      "loss": 0.5471,
      "step": 2062
    },
    {
      "epoch": 0.4126,
      "grad_norm": 3.9359383583068848,
      "learning_rate": 2.996938775510204e-05,
      "loss": 0.6082,
      "step": 2063
    },
    {
      "epoch": 0.4128,
      "grad_norm": 3.843722105026245,
      "learning_rate": 2.9959183673469387e-05,
      "loss": 0.597,
      "step": 2064
    },
    {
      "epoch": 0.413,
      "grad_norm": 8.112070083618164,
      "learning_rate": 2.9948979591836733e-05,
      "loss": 1.2302,
      "step": 2065
    },
    {
      "epoch": 0.4132,
      "grad_norm": 7.104030132293701,
      "learning_rate": 2.9938775510204082e-05,
      "loss": 1.2015,
      "step": 2066
    },
    {
      "epoch": 0.4134,
      "grad_norm": 10.215073585510254,
      "learning_rate": 2.9928571428571428e-05,
      "loss": 1.1589,
      "step": 2067
    },
    {
      "epoch": 0.4136,
      "grad_norm": 3.378031015396118,
      "learning_rate": 2.9918367346938774e-05,
      "loss": 0.9331,
      "step": 2068
    },
    {
      "epoch": 0.4138,
      "grad_norm": 7.290038585662842,
      "learning_rate": 2.9908163265306123e-05,
      "loss": 1.196,
      "step": 2069
    },
    {
      "epoch": 0.414,
      "grad_norm": 4.649845600128174,
      "learning_rate": 2.989795918367347e-05,
      "loss": 0.6947,
      "step": 2070
    },
    {
      "epoch": 0.4142,
      "grad_norm": 6.744777679443359,
      "learning_rate": 2.9887755102040815e-05,
      "loss": 1.2755,
      "step": 2071
    },
    {
      "epoch": 0.4144,
      "grad_norm": 4.500767707824707,
      "learning_rate": 2.987755102040816e-05,
      "loss": 0.6236,
      "step": 2072
    },
    {
      "epoch": 0.4146,
      "grad_norm": 4.191307067871094,
      "learning_rate": 2.986734693877551e-05,
      "loss": 0.6218,
      "step": 2073
    },
    {
      "epoch": 0.4148,
      "grad_norm": 4.284450531005859,
      "learning_rate": 2.9857142857142862e-05,
      "loss": 0.6073,
      "step": 2074
    },
    {
      "epoch": 0.415,
      "grad_norm": 3.9721145629882812,
      "learning_rate": 2.9846938775510208e-05,
      "loss": 0.6127,
      "step": 2075
    },
    {
      "epoch": 0.4152,
      "grad_norm": 4.18171501159668,
      "learning_rate": 2.9836734693877554e-05,
      "loss": 0.6471,
      "step": 2076
    },
    {
      "epoch": 0.4154,
      "grad_norm": 3.992933511734009,
      "learning_rate": 2.9826530612244903e-05,
      "loss": 0.5906,
      "step": 2077
    },
    {
      "epoch": 0.4156,
      "grad_norm": 9.78303337097168,
      "learning_rate": 2.981632653061225e-05,
      "loss": 1.2285,
      "step": 2078
    },
    {
      "epoch": 0.4158,
      "grad_norm": 7.465812683105469,
      "learning_rate": 2.9806122448979595e-05,
      "loss": 1.2131,
      "step": 2079
    },
    {
      "epoch": 0.416,
      "grad_norm": 7.9334940910339355,
      "learning_rate": 2.9795918367346944e-05,
      "loss": 1.0955,
      "step": 2080
    },
    {
      "epoch": 0.4162,
      "grad_norm": 8.485240936279297,
      "learning_rate": 2.978571428571429e-05,
      "loss": 1.3146,
      "step": 2081
    },
    {
      "epoch": 0.4164,
      "grad_norm": 6.9409637451171875,
      "learning_rate": 2.9775510204081635e-05,
      "loss": 1.143,
      "step": 2082
    },
    {
      "epoch": 0.4166,
      "grad_norm": 7.945091247558594,
      "learning_rate": 2.976530612244898e-05,
      "loss": 1.1643,
      "step": 2083
    },
    {
      "epoch": 0.4168,
      "grad_norm": 18.780855178833008,
      "learning_rate": 2.975510204081633e-05,
      "loss": 0.8423,
      "step": 2084
    },
    {
      "epoch": 1.0002,
      "grad_norm": 15.242094039916992,
      "learning_rate": 2.9744897959183676e-05,
      "loss": 1.1254,
      "step": 2085
    },
    {
      "epoch": 1.0004,
      "grad_norm": 51.17671585083008,
      "learning_rate": 2.9734693877551022e-05,
      "loss": 3.1271,
      "step": 2086
    },
    {
      "epoch": 1.0006,
      "grad_norm": 2.6297574043273926,
      "learning_rate": 2.972448979591837e-05,
      "loss": 0.336,
      "step": 2087
    },
    {
      "epoch": 1.0008,
      "grad_norm": 8.47588062286377,
      "learning_rate": 2.9714285714285717e-05,
      "loss": 0.5913,
      "step": 2088
    },
    {
      "epoch": 1.001,
      "grad_norm": 7.058892726898193,
      "learning_rate": 2.9704081632653063e-05,
      "loss": 0.4798,
      "step": 2089
    },
    {
      "epoch": 1.0012,
      "grad_norm": 1.6835764646530151,
      "learning_rate": 2.9693877551020412e-05,
      "loss": 0.0532,
      "step": 2090
    },
    {
      "epoch": 1.0014,
      "grad_norm": 2.0876736640930176,
      "learning_rate": 2.9683673469387758e-05,
      "loss": 0.26,
      "step": 2091
    },
    {
      "epoch": 1.0016,
      "grad_norm": 2.948742628097534,
      "learning_rate": 2.9673469387755103e-05,
      "loss": 0.5332,
      "step": 2092
    },
    {
      "epoch": 1.0018,
      "grad_norm": 7.258228302001953,
      "learning_rate": 2.966326530612245e-05,
      "loss": 1.5851,
      "step": 2093
    },
    {
      "epoch": 1.002,
      "grad_norm": 44.64124298095703,
      "learning_rate": 2.96530612244898e-05,
      "loss": 0.4057,
      "step": 2094
    },
    {
      "epoch": 1.0022,
      "grad_norm": 90.02332305908203,
      "learning_rate": 2.9642857142857144e-05,
      "loss": 0.2384,
      "step": 2095
    },
    {
      "epoch": 1.0024,
      "grad_norm": 2.4125702381134033,
      "learning_rate": 2.963265306122449e-05,
      "loss": 0.4608,
      "step": 2096
    },
    {
      "epoch": 1.0026,
      "grad_norm": 3.5664608478546143,
      "learning_rate": 2.962244897959184e-05,
      "loss": 1.5708,
      "step": 2097
    },
    {
      "epoch": 1.0028,
      "grad_norm": 2.176363945007324,
      "learning_rate": 2.9612244897959185e-05,
      "loss": 0.232,
      "step": 2098
    },
    {
      "epoch": 1.003,
      "grad_norm": 4.480269908905029,
      "learning_rate": 2.960204081632653e-05,
      "loss": 0.2929,
      "step": 2099
    },
    {
      "epoch": 1.0032,
      "grad_norm": 4.131799697875977,
      "learning_rate": 2.959183673469388e-05,
      "loss": 1.0002,
      "step": 2100
    },
    {
      "epoch": 1.0034,
      "grad_norm": 15.502717018127441,
      "learning_rate": 2.9581632653061226e-05,
      "loss": 3.1208,
      "step": 2101
    },
    {
      "epoch": 1.0036,
      "grad_norm": 21.303125381469727,
      "learning_rate": 2.957142857142857e-05,
      "loss": 3.2585,
      "step": 2102
    },
    {
      "epoch": 1.0038,
      "grad_norm": 2.70643949508667,
      "learning_rate": 2.956122448979592e-05,
      "loss": 0.389,
      "step": 2103
    },
    {
      "epoch": 1.004,
      "grad_norm": 7.2120466232299805,
      "learning_rate": 2.9551020408163266e-05,
      "loss": 0.9416,
      "step": 2104
    },
    {
      "epoch": 1.0042,
      "grad_norm": 33.94477462768555,
      "learning_rate": 2.9540816326530612e-05,
      "loss": 1.8284,
      "step": 2105
    },
    {
      "epoch": 1.0044,
      "grad_norm": 98.7635498046875,
      "learning_rate": 2.9530612244897958e-05,
      "loss": 5.7999,
      "step": 2106
    },
    {
      "epoch": 1.0046,
      "grad_norm": 1.7753891944885254,
      "learning_rate": 2.9520408163265307e-05,
      "loss": 0.0865,
      "step": 2107
    },
    {
      "epoch": 1.0048,
      "grad_norm": 2.728239059448242,
      "learning_rate": 2.9510204081632653e-05,
      "loss": 0.2501,
      "step": 2108
    },
    {
      "epoch": 1.005,
      "grad_norm": 4.694189548492432,
      "learning_rate": 2.95e-05,
      "loss": 0.4039,
      "step": 2109
    },
    {
      "epoch": 1.0052,
      "grad_norm": 9.07700252532959,
      "learning_rate": 2.9489795918367348e-05,
      "loss": 0.85,
      "step": 2110
    },
    {
      "epoch": 1.0054,
      "grad_norm": 11.796422958374023,
      "learning_rate": 2.9479591836734694e-05,
      "loss": 1.6915,
      "step": 2111
    },
    {
      "epoch": 1.0056,
      "grad_norm": 15.555567741394043,
      "learning_rate": 2.946938775510204e-05,
      "loss": 5.2889,
      "step": 2112
    },
    {
      "epoch": 1.0058,
      "grad_norm": 3.4670357704162598,
      "learning_rate": 2.945918367346939e-05,
      "loss": 0.4462,
      "step": 2113
    },
    {
      "epoch": 1.006,
      "grad_norm": 3.21122407913208,
      "learning_rate": 2.9448979591836735e-05,
      "loss": 0.4042,
      "step": 2114
    },
    {
      "epoch": 1.0062,
      "grad_norm": 2.0091428756713867,
      "learning_rate": 2.943877551020408e-05,
      "loss": 0.2797,
      "step": 2115
    },
    {
      "epoch": 1.0064,
      "grad_norm": 3.6005043983459473,
      "learning_rate": 2.9428571428571426e-05,
      "loss": 0.2846,
      "step": 2116
    },
    {
      "epoch": 1.0066,
      "grad_norm": 8.382363319396973,
      "learning_rate": 2.9418367346938775e-05,
      "loss": 1.2789,
      "step": 2117
    },
    {
      "epoch": 1.0068,
      "grad_norm": 4.641507148742676,
      "learning_rate": 2.940816326530612e-05,
      "loss": 0.1561,
      "step": 2118
    },
    {
      "epoch": 1.007,
      "grad_norm": 4.312943458557129,
      "learning_rate": 2.9397959183673474e-05,
      "loss": 0.3616,
      "step": 2119
    },
    {
      "epoch": 1.0072,
      "grad_norm": 7.527416706085205,
      "learning_rate": 2.938775510204082e-05,
      "loss": 1.0976,
      "step": 2120
    },
    {
      "epoch": 1.0074,
      "grad_norm": 5.3701019287109375,
      "learning_rate": 2.937755102040817e-05,
      "loss": 1.6283,
      "step": 2121
    },
    {
      "epoch": 1.0076,
      "grad_norm": 2.4299004077911377,
      "learning_rate": 2.9367346938775514e-05,
      "loss": 0.1998,
      "step": 2122
    },
    {
      "epoch": 1.0078,
      "grad_norm": 3.094940662384033,
      "learning_rate": 2.935714285714286e-05,
      "loss": 0.9545,
      "step": 2123
    },
    {
      "epoch": 1.008,
      "grad_norm": 7.588383197784424,
      "learning_rate": 2.9346938775510206e-05,
      "loss": 2.3132,
      "step": 2124
    },
    {
      "epoch": 1.0082,
      "grad_norm": 14.091776847839355,
      "learning_rate": 2.9336734693877555e-05,
      "loss": 2.5728,
      "step": 2125
    },
    {
      "epoch": 1.0084,
      "grad_norm": 1.9190328121185303,
      "learning_rate": 2.93265306122449e-05,
      "loss": 0.2431,
      "step": 2126
    },
    {
      "epoch": 1.0086,
      "grad_norm": 5.02833366394043,
      "learning_rate": 2.9316326530612247e-05,
      "loss": 0.5762,
      "step": 2127
    },
    {
      "epoch": 1.0088,
      "grad_norm": 8.859189987182617,
      "learning_rate": 2.9306122448979596e-05,
      "loss": 1.2586,
      "step": 2128
    },
    {
      "epoch": 1.009,
      "grad_norm": 2.1577792167663574,
      "learning_rate": 2.9295918367346942e-05,
      "loss": 0.2873,
      "step": 2129
    },
    {
      "epoch": 1.0092,
      "grad_norm": 1.858289122581482,
      "learning_rate": 2.9285714285714288e-05,
      "loss": 0.2176,
      "step": 2130
    },
    {
      "epoch": 1.0094,
      "grad_norm": 11.388729095458984,
      "learning_rate": 2.9275510204081637e-05,
      "loss": 1.218,
      "step": 2131
    },
    {
      "epoch": 1.0096,
      "grad_norm": 18.279762268066406,
      "learning_rate": 2.9265306122448982e-05,
      "loss": 2.053,
      "step": 2132
    },
    {
      "epoch": 1.0098,
      "grad_norm": 4.1785783767700195,
      "learning_rate": 2.9255102040816328e-05,
      "loss": 0.7075,
      "step": 2133
    },
    {
      "epoch": 1.01,
      "grad_norm": 5.734426975250244,
      "learning_rate": 2.9244897959183677e-05,
      "loss": 1.6235,
      "step": 2134
    },
    {
      "epoch": 1.0102,
      "grad_norm": 7.153301239013672,
      "learning_rate": 2.9234693877551023e-05,
      "loss": 0.9697,
      "step": 2135
    },
    {
      "epoch": 1.0104,
      "grad_norm": 7.6792426109313965,
      "learning_rate": 2.922448979591837e-05,
      "loss": 0.8291,
      "step": 2136
    },
    {
      "epoch": 1.0106,
      "grad_norm": 2.6842920780181885,
      "learning_rate": 2.9214285714285715e-05,
      "loss": 0.753,
      "step": 2137
    },
    {
      "epoch": 1.0108,
      "grad_norm": 3.8635406494140625,
      "learning_rate": 2.9204081632653064e-05,
      "loss": 1.7667,
      "step": 2138
    },
    {
      "epoch": 1.011,
      "grad_norm": 2.733811140060425,
      "learning_rate": 2.919387755102041e-05,
      "loss": 0.3324,
      "step": 2139
    },
    {
      "epoch": 1.0112,
      "grad_norm": 2.376089334487915,
      "learning_rate": 2.9183673469387756e-05,
      "loss": 0.2683,
      "step": 2140
    },
    {
      "epoch": 1.0114,
      "grad_norm": 4.434942722320557,
      "learning_rate": 2.9173469387755105e-05,
      "loss": 0.7013,
      "step": 2141
    },
    {
      "epoch": 1.0116,
      "grad_norm": 4.3742356300354,
      "learning_rate": 2.916326530612245e-05,
      "loss": 1.9174,
      "step": 2142
    },
    {
      "epoch": 1.0118,
      "grad_norm": 9.535046577453613,
      "learning_rate": 2.9153061224489796e-05,
      "loss": 2.0534,
      "step": 2143
    },
    {
      "epoch": 1.012,
      "grad_norm": 12.418166160583496,
      "learning_rate": 2.9142857142857146e-05,
      "loss": 1.9329,
      "step": 2144
    },
    {
      "epoch": 1.0122,
      "grad_norm": 3.672943353652954,
      "learning_rate": 2.913265306122449e-05,
      "loss": 1.7427,
      "step": 2145
    },
    {
      "epoch": 1.0124,
      "grad_norm": 3.344444751739502,
      "learning_rate": 2.9122448979591837e-05,
      "loss": 0.6165,
      "step": 2146
    },
    {
      "epoch": 1.0126,
      "grad_norm": 3.1958985328674316,
      "learning_rate": 2.9112244897959183e-05,
      "loss": 1.4177,
      "step": 2147
    },
    {
      "epoch": 1.0128,
      "grad_norm": 2.36991810798645,
      "learning_rate": 2.9102040816326532e-05,
      "loss": 0.6834,
      "step": 2148
    },
    {
      "epoch": 1.013,
      "grad_norm": 8.034588813781738,
      "learning_rate": 2.9091836734693878e-05,
      "loss": 2.3334,
      "step": 2149
    },
    {
      "epoch": 1.0132,
      "grad_norm": 7.280432224273682,
      "learning_rate": 2.9081632653061224e-05,
      "loss": 2.5005,
      "step": 2150
    },
    {
      "epoch": 1.0134,
      "grad_norm": 5.838300704956055,
      "learning_rate": 2.9071428571428573e-05,
      "loss": 1.8218,
      "step": 2151
    },
    {
      "epoch": 1.0136,
      "grad_norm": 10.983949661254883,
      "learning_rate": 2.906122448979592e-05,
      "loss": 2.1662,
      "step": 2152
    },
    {
      "epoch": 1.0138,
      "grad_norm": 20.635364532470703,
      "learning_rate": 2.9051020408163264e-05,
      "loss": 4.3268,
      "step": 2153
    },
    {
      "epoch": 1.014,
      "grad_norm": 13.782269477844238,
      "learning_rate": 2.9040816326530614e-05,
      "loss": 2.6215,
      "step": 2154
    },
    {
      "epoch": 1.0142,
      "grad_norm": 5.808487415313721,
      "learning_rate": 2.903061224489796e-05,
      "loss": 1.1143,
      "step": 2155
    },
    {
      "epoch": 1.0144,
      "grad_norm": 9.513835906982422,
      "learning_rate": 2.9020408163265305e-05,
      "loss": 2.3511,
      "step": 2156
    },
    {
      "epoch": 1.0146,
      "grad_norm": 1.9956474304199219,
      "learning_rate": 2.901020408163265e-05,
      "loss": 0.2929,
      "step": 2157
    },
    {
      "epoch": 1.0148,
      "grad_norm": 5.999602317810059,
      "learning_rate": 2.9e-05,
      "loss": 1.599,
      "step": 2158
    },
    {
      "epoch": 1.015,
      "grad_norm": 14.991220474243164,
      "learning_rate": 2.8989795918367346e-05,
      "loss": 4.129,
      "step": 2159
    },
    {
      "epoch": 1.0152,
      "grad_norm": 16.006671905517578,
      "learning_rate": 2.8979591836734692e-05,
      "loss": 2.3101,
      "step": 2160
    },
    {
      "epoch": 1.0154,
      "grad_norm": 10.431235313415527,
      "learning_rate": 2.896938775510204e-05,
      "loss": 1.7555,
      "step": 2161
    },
    {
      "epoch": 1.0156,
      "grad_norm": 5.553926467895508,
      "learning_rate": 2.8959183673469387e-05,
      "loss": 2.2064,
      "step": 2162
    },
    {
      "epoch": 1.0158,
      "grad_norm": 6.1060590744018555,
      "learning_rate": 2.8948979591836733e-05,
      "loss": 1.9016,
      "step": 2163
    },
    {
      "epoch": 1.016,
      "grad_norm": 2.589744806289673,
      "learning_rate": 2.8938775510204082e-05,
      "loss": 0.679,
      "step": 2164
    },
    {
      "epoch": 1.0162,
      "grad_norm": 7.968225479125977,
      "learning_rate": 2.8928571428571434e-05,
      "loss": 2.2678,
      "step": 2165
    },
    {
      "epoch": 1.0164,
      "grad_norm": 20.811412811279297,
      "learning_rate": 2.891836734693878e-05,
      "loss": 2.487,
      "step": 2166
    },
    {
      "epoch": 1.0166,
      "grad_norm": 2.4238314628601074,
      "learning_rate": 2.8908163265306126e-05,
      "loss": 0.3541,
      "step": 2167
    },
    {
      "epoch": 1.0168,
      "grad_norm": 16.10127067565918,
      "learning_rate": 2.889795918367347e-05,
      "loss": 1.8777,
      "step": 2168
    },
    {
      "epoch": 1.017,
      "grad_norm": 36.814212799072266,
      "learning_rate": 2.888775510204082e-05,
      "loss": 3.9616,
      "step": 2169
    },
    {
      "epoch": 1.0172,
      "grad_norm": 7.001033306121826,
      "learning_rate": 2.8877551020408167e-05,
      "loss": 0.4239,
      "step": 2170
    },
    {
      "epoch": 1.0174,
      "grad_norm": 2.069876194000244,
      "learning_rate": 2.8867346938775512e-05,
      "loss": 0.2434,
      "step": 2171
    },
    {
      "epoch": 1.0176,
      "grad_norm": 3.3106343746185303,
      "learning_rate": 2.885714285714286e-05,
      "loss": 0.2273,
      "step": 2172
    },
    {
      "epoch": 1.0178,
      "grad_norm": 3.6341891288757324,
      "learning_rate": 2.8846938775510207e-05,
      "loss": 0.5256,
      "step": 2173
    },
    {
      "epoch": 1.018,
      "grad_norm": 5.140353679656982,
      "learning_rate": 2.8836734693877553e-05,
      "loss": 0.8109,
      "step": 2174
    },
    {
      "epoch": 1.0182,
      "grad_norm": 3.3280324935913086,
      "learning_rate": 2.8826530612244902e-05,
      "loss": 0.4409,
      "step": 2175
    },
    {
      "epoch": 1.0184,
      "grad_norm": 3.295837163925171,
      "learning_rate": 2.8816326530612248e-05,
      "loss": 0.3681,
      "step": 2176
    },
    {
      "epoch": 1.0186,
      "grad_norm": 3.2037105560302734,
      "learning_rate": 2.8806122448979594e-05,
      "loss": 0.1784,
      "step": 2177
    },
    {
      "epoch": 1.0188,
      "grad_norm": 4.193380832672119,
      "learning_rate": 2.879591836734694e-05,
      "loss": 0.3336,
      "step": 2178
    },
    {
      "epoch": 1.019,
      "grad_norm": 7.2576985359191895,
      "learning_rate": 2.878571428571429e-05,
      "loss": 1.288,
      "step": 2179
    },
    {
      "epoch": 1.0192,
      "grad_norm": 7.04248571395874,
      "learning_rate": 2.8775510204081635e-05,
      "loss": 2.1678,
      "step": 2180
    },
    {
      "epoch": 1.0194,
      "grad_norm": 16.95452308654785,
      "learning_rate": 2.876530612244898e-05,
      "loss": 4.9449,
      "step": 2181
    },
    {
      "epoch": 1.0196,
      "grad_norm": 6.738562107086182,
      "learning_rate": 2.875510204081633e-05,
      "loss": 0.7365,
      "step": 2182
    },
    {
      "epoch": 1.0198,
      "grad_norm": 7.8314666748046875,
      "learning_rate": 2.8744897959183675e-05,
      "loss": 1.8211,
      "step": 2183
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.4186290502548218,
      "learning_rate": 2.873469387755102e-05,
      "loss": 0.0832,
      "step": 2184
    },
    {
      "epoch": 1.0202,
      "grad_norm": 1.4884039163589478,
      "learning_rate": 2.872448979591837e-05,
      "loss": 0.0918,
      "step": 2185
    },
    {
      "epoch": 1.0204,
      "grad_norm": 4.299226760864258,
      "learning_rate": 2.8714285714285716e-05,
      "loss": 0.374,
      "step": 2186
    },
    {
      "epoch": 1.0206,
      "grad_norm": 2.9364874362945557,
      "learning_rate": 2.8704081632653062e-05,
      "loss": 0.2143,
      "step": 2187
    },
    {
      "epoch": 1.0208,
      "grad_norm": 0.9889697432518005,
      "learning_rate": 2.8693877551020408e-05,
      "loss": 0.0406,
      "step": 2188
    },
    {
      "epoch": 1.021,
      "grad_norm": 1.8314415216445923,
      "learning_rate": 2.8683673469387757e-05,
      "loss": 0.232,
      "step": 2189
    },
    {
      "epoch": 1.0212,
      "grad_norm": 2.237391471862793,
      "learning_rate": 2.8673469387755103e-05,
      "loss": 0.2606,
      "step": 2190
    },
    {
      "epoch": 1.0214,
      "grad_norm": 4.936470985412598,
      "learning_rate": 2.866326530612245e-05,
      "loss": 1.1271,
      "step": 2191
    },
    {
      "epoch": 1.0216,
      "grad_norm": 10.1080961227417,
      "learning_rate": 2.8653061224489798e-05,
      "loss": 1.7872,
      "step": 2192
    },
    {
      "epoch": 1.0218,
      "grad_norm": 8.04566764831543,
      "learning_rate": 2.8642857142857144e-05,
      "loss": 0.4089,
      "step": 2193
    },
    {
      "epoch": 1.022,
      "grad_norm": 2.041060209274292,
      "learning_rate": 2.863265306122449e-05,
      "loss": 0.2214,
      "step": 2194
    },
    {
      "epoch": 1.0222,
      "grad_norm": 4.039699554443359,
      "learning_rate": 2.862244897959184e-05,
      "loss": 0.3278,
      "step": 2195
    },
    {
      "epoch": 1.0224,
      "grad_norm": 7.945012092590332,
      "learning_rate": 2.8612244897959184e-05,
      "loss": 1.0167,
      "step": 2196
    },
    {
      "epoch": 1.0226,
      "grad_norm": 8.669960975646973,
      "learning_rate": 2.860204081632653e-05,
      "loss": 1.5223,
      "step": 2197
    },
    {
      "epoch": 1.0228,
      "grad_norm": 3.9770047664642334,
      "learning_rate": 2.8591836734693876e-05,
      "loss": 1.5937,
      "step": 2198
    },
    {
      "epoch": 1.023,
      "grad_norm": 4.4404144287109375,
      "learning_rate": 2.8581632653061225e-05,
      "loss": 0.7484,
      "step": 2199
    },
    {
      "epoch": 1.0232,
      "grad_norm": 2.1829888820648193,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.2419,
      "step": 2200
    },
    {
      "epoch": 1.0234,
      "grad_norm": 1.9747893810272217,
      "learning_rate": 2.8561224489795917e-05,
      "loss": 0.2527,
      "step": 2201
    },
    {
      "epoch": 1.0236,
      "grad_norm": 4.661055564880371,
      "learning_rate": 2.8551020408163266e-05,
      "loss": 0.7782,
      "step": 2202
    },
    {
      "epoch": 1.0238,
      "grad_norm": 11.441447257995605,
      "learning_rate": 2.854081632653061e-05,
      "loss": 2.4732,
      "step": 2203
    },
    {
      "epoch": 1.024,
      "grad_norm": 4.55333137512207,
      "learning_rate": 2.8530612244897957e-05,
      "loss": 1.3935,
      "step": 2204
    },
    {
      "epoch": 1.0242,
      "grad_norm": 4.206186771392822,
      "learning_rate": 2.8520408163265307e-05,
      "loss": 1.894,
      "step": 2205
    },
    {
      "epoch": 1.0244,
      "grad_norm": 11.083173751831055,
      "learning_rate": 2.8510204081632652e-05,
      "loss": 2.4657,
      "step": 2206
    },
    {
      "epoch": 1.0246,
      "grad_norm": 23.377429962158203,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 2.9622,
      "step": 2207
    },
    {
      "epoch": 1.0248,
      "grad_norm": 11.613309860229492,
      "learning_rate": 2.8489795918367347e-05,
      "loss": 3.649,
      "step": 2208
    },
    {
      "epoch": 1.025,
      "grad_norm": 15.78592300415039,
      "learning_rate": 2.8479591836734693e-05,
      "loss": 5.3943,
      "step": 2209
    },
    {
      "epoch": 1.0252,
      "grad_norm": 26.071020126342773,
      "learning_rate": 2.8469387755102046e-05,
      "loss": 2.2782,
      "step": 2210
    },
    {
      "epoch": 1.0254,
      "grad_norm": 14.612674713134766,
      "learning_rate": 2.845918367346939e-05,
      "loss": 4.8628,
      "step": 2211
    },
    {
      "epoch": 1.0256,
      "grad_norm": 3.795036554336548,
      "learning_rate": 2.8448979591836737e-05,
      "loss": 0.4615,
      "step": 2212
    },
    {
      "epoch": 1.0258,
      "grad_norm": 4.833165168762207,
      "learning_rate": 2.8438775510204086e-05,
      "loss": 1.0494,
      "step": 2213
    },
    {
      "epoch": 1.026,
      "grad_norm": 3.8589091300964355,
      "learning_rate": 2.8428571428571432e-05,
      "loss": 1.6959,
      "step": 2214
    },
    {
      "epoch": 1.0262,
      "grad_norm": 3.95047664642334,
      "learning_rate": 2.8418367346938778e-05,
      "loss": 0.7726,
      "step": 2215
    },
    {
      "epoch": 1.0264,
      "grad_norm": 1.8891721963882446,
      "learning_rate": 2.8408163265306127e-05,
      "loss": 0.5153,
      "step": 2216
    },
    {
      "epoch": 1.0266,
      "grad_norm": 4.715999603271484,
      "learning_rate": 2.8397959183673473e-05,
      "loss": 2.9044,
      "step": 2217
    },
    {
      "epoch": 1.0268,
      "grad_norm": 7.398796558380127,
      "learning_rate": 2.838775510204082e-05,
      "loss": 5.2357,
      "step": 2218
    },
    {
      "epoch": 1.027,
      "grad_norm": 16.269620895385742,
      "learning_rate": 2.8377551020408165e-05,
      "loss": 2.6862,
      "step": 2219
    },
    {
      "epoch": 1.0272,
      "grad_norm": 69.98822021484375,
      "learning_rate": 2.8367346938775514e-05,
      "loss": 4.26,
      "step": 2220
    },
    {
      "epoch": 1.0274,
      "grad_norm": 2.350796937942505,
      "learning_rate": 2.835714285714286e-05,
      "loss": 0.1944,
      "step": 2221
    },
    {
      "epoch": 1.0276,
      "grad_norm": 4.043328285217285,
      "learning_rate": 2.8346938775510205e-05,
      "loss": 0.664,
      "step": 2222
    },
    {
      "epoch": 1.0278,
      "grad_norm": 9.004681587219238,
      "learning_rate": 2.8336734693877555e-05,
      "loss": 0.7774,
      "step": 2223
    },
    {
      "epoch": 1.028,
      "grad_norm": 16.564464569091797,
      "learning_rate": 2.83265306122449e-05,
      "loss": 2.6385,
      "step": 2224
    },
    {
      "epoch": 1.0282,
      "grad_norm": 5.071820259094238,
      "learning_rate": 2.8316326530612246e-05,
      "loss": 0.903,
      "step": 2225
    },
    {
      "epoch": 1.0284,
      "grad_norm": 3.974750518798828,
      "learning_rate": 2.8306122448979595e-05,
      "loss": 0.6421,
      "step": 2226
    },
    {
      "epoch": 1.0286,
      "grad_norm": 2.1501924991607666,
      "learning_rate": 2.829591836734694e-05,
      "loss": 0.52,
      "step": 2227
    },
    {
      "epoch": 1.0288,
      "grad_norm": 3.4971842765808105,
      "learning_rate": 2.8285714285714287e-05,
      "loss": 1.5575,
      "step": 2228
    },
    {
      "epoch": 1.029,
      "grad_norm": 5.169280052185059,
      "learning_rate": 2.8275510204081636e-05,
      "loss": 0.1911,
      "step": 2229
    },
    {
      "epoch": 1.0292,
      "grad_norm": 5.614638805389404,
      "learning_rate": 2.8265306122448982e-05,
      "loss": 1.5052,
      "step": 2230
    },
    {
      "epoch": 1.0294,
      "grad_norm": 10.963964462280273,
      "learning_rate": 2.8255102040816328e-05,
      "loss": 5.0517,
      "step": 2231
    },
    {
      "epoch": 1.0296,
      "grad_norm": 7.341145038604736,
      "learning_rate": 2.8244897959183673e-05,
      "loss": 2.3456,
      "step": 2232
    },
    {
      "epoch": 1.0298,
      "grad_norm": 10.776498794555664,
      "learning_rate": 2.8234693877551023e-05,
      "loss": 3.0459,
      "step": 2233
    },
    {
      "epoch": 1.03,
      "grad_norm": 12.878778457641602,
      "learning_rate": 2.822448979591837e-05,
      "loss": 3.6761,
      "step": 2234
    },
    {
      "epoch": 1.0302,
      "grad_norm": 1.7765483856201172,
      "learning_rate": 2.8214285714285714e-05,
      "loss": 0.2764,
      "step": 2235
    },
    {
      "epoch": 1.0304,
      "grad_norm": 3.567537784576416,
      "learning_rate": 2.8204081632653063e-05,
      "loss": 0.4874,
      "step": 2236
    },
    {
      "epoch": 1.0306,
      "grad_norm": 11.235349655151367,
      "learning_rate": 2.819387755102041e-05,
      "loss": 2.031,
      "step": 2237
    },
    {
      "epoch": 1.0308,
      "grad_norm": 11.563838005065918,
      "learning_rate": 2.8183673469387755e-05,
      "loss": 1.9193,
      "step": 2238
    },
    {
      "epoch": 1.031,
      "grad_norm": 4.753085136413574,
      "learning_rate": 2.8173469387755104e-05,
      "loss": 2.6111,
      "step": 2239
    },
    {
      "epoch": 1.0312,
      "grad_norm": 12.879108428955078,
      "learning_rate": 2.816326530612245e-05,
      "loss": 2.582,
      "step": 2240
    },
    {
      "epoch": 1.0314,
      "grad_norm": 3.6239147186279297,
      "learning_rate": 2.8153061224489796e-05,
      "loss": 1.5615,
      "step": 2241
    },
    {
      "epoch": 1.0316,
      "grad_norm": 1.8712190389633179,
      "learning_rate": 2.814285714285714e-05,
      "loss": 0.2572,
      "step": 2242
    },
    {
      "epoch": 1.0318,
      "grad_norm": 5.439129829406738,
      "learning_rate": 2.813265306122449e-05,
      "loss": 0.6856,
      "step": 2243
    },
    {
      "epoch": 1.032,
      "grad_norm": 14.700053215026855,
      "learning_rate": 2.8122448979591837e-05,
      "loss": 1.5368,
      "step": 2244
    },
    {
      "epoch": 1.0322,
      "grad_norm": 1.8174935579299927,
      "learning_rate": 2.8112244897959182e-05,
      "loss": 0.1421,
      "step": 2245
    },
    {
      "epoch": 1.0324,
      "grad_norm": 11.371535301208496,
      "learning_rate": 2.810204081632653e-05,
      "loss": 1.4882,
      "step": 2246
    },
    {
      "epoch": 1.0326,
      "grad_norm": 34.32445526123047,
      "learning_rate": 2.8091836734693877e-05,
      "loss": 3.6614,
      "step": 2247
    },
    {
      "epoch": 1.0328,
      "grad_norm": 1.8845922946929932,
      "learning_rate": 2.8081632653061223e-05,
      "loss": 0.0578,
      "step": 2248
    },
    {
      "epoch": 1.033,
      "grad_norm": 1.6151620149612427,
      "learning_rate": 2.8071428571428572e-05,
      "loss": 0.1738,
      "step": 2249
    },
    {
      "epoch": 1.0332,
      "grad_norm": 4.2584452629089355,
      "learning_rate": 2.8061224489795918e-05,
      "loss": 0.4365,
      "step": 2250
    },
    {
      "epoch": 1.0334,
      "grad_norm": 7.917820453643799,
      "learning_rate": 2.8051020408163264e-05,
      "loss": 1.2255,
      "step": 2251
    },
    {
      "epoch": 1.0336,
      "grad_norm": 3.721857786178589,
      "learning_rate": 2.804081632653061e-05,
      "loss": 0.2738,
      "step": 2252
    },
    {
      "epoch": 1.0338,
      "grad_norm": 8.463653564453125,
      "learning_rate": 2.803061224489796e-05,
      "loss": 1.2317,
      "step": 2253
    },
    {
      "epoch": 1.034,
      "grad_norm": 4.422754287719727,
      "learning_rate": 2.8020408163265305e-05,
      "loss": 0.7327,
      "step": 2254
    },
    {
      "epoch": 1.0342,
      "grad_norm": 2.656949758529663,
      "learning_rate": 2.801020408163265e-05,
      "loss": 0.5733,
      "step": 2255
    },
    {
      "epoch": 1.0344,
      "grad_norm": 2.132868766784668,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.163,
      "step": 2256
    },
    {
      "epoch": 1.0346,
      "grad_norm": 11.262079238891602,
      "learning_rate": 2.7989795918367352e-05,
      "loss": 2.2782,
      "step": 2257
    },
    {
      "epoch": 1.0348,
      "grad_norm": 26.79599952697754,
      "learning_rate": 2.7979591836734698e-05,
      "loss": 5.6807,
      "step": 2258
    },
    {
      "epoch": 1.035,
      "grad_norm": 9.496004104614258,
      "learning_rate": 2.7969387755102044e-05,
      "loss": 2.106,
      "step": 2259
    },
    {
      "epoch": 1.0352,
      "grad_norm": 1.3949973583221436,
      "learning_rate": 2.7959183673469393e-05,
      "loss": 0.0548,
      "step": 2260
    },
    {
      "epoch": 1.0354,
      "grad_norm": 2.057509183883667,
      "learning_rate": 2.794897959183674e-05,
      "loss": 0.2224,
      "step": 2261
    },
    {
      "epoch": 1.0356,
      "grad_norm": 1.8621536493301392,
      "learning_rate": 2.7938775510204084e-05,
      "loss": 0.1989,
      "step": 2262
    },
    {
      "epoch": 1.0358,
      "grad_norm": 1.6629083156585693,
      "learning_rate": 2.792857142857143e-05,
      "loss": 0.1907,
      "step": 2263
    },
    {
      "epoch": 1.036,
      "grad_norm": 2.4138829708099365,
      "learning_rate": 2.791836734693878e-05,
      "loss": 0.3385,
      "step": 2264
    },
    {
      "epoch": 1.0362,
      "grad_norm": 5.232615947723389,
      "learning_rate": 2.7908163265306125e-05,
      "loss": 0.9645,
      "step": 2265
    },
    {
      "epoch": 1.0364,
      "grad_norm": 7.121108531951904,
      "learning_rate": 2.789795918367347e-05,
      "loss": 1.0515,
      "step": 2266
    },
    {
      "epoch": 1.0366,
      "grad_norm": 2.2986886501312256,
      "learning_rate": 2.788775510204082e-05,
      "loss": 0.6423,
      "step": 2267
    },
    {
      "epoch": 1.0368,
      "grad_norm": 3.717360496520996,
      "learning_rate": 2.7877551020408166e-05,
      "loss": 1.6657,
      "step": 2268
    },
    {
      "epoch": 1.037,
      "grad_norm": 3.4986190795898438,
      "learning_rate": 2.7867346938775512e-05,
      "loss": 0.1756,
      "step": 2269
    },
    {
      "epoch": 1.0372,
      "grad_norm": 6.572997093200684,
      "learning_rate": 2.785714285714286e-05,
      "loss": 0.3995,
      "step": 2270
    },
    {
      "epoch": 1.0374,
      "grad_norm": 1.7881145477294922,
      "learning_rate": 2.7846938775510207e-05,
      "loss": 0.215,
      "step": 2271
    },
    {
      "epoch": 1.0376,
      "grad_norm": 2.556283712387085,
      "learning_rate": 2.7836734693877553e-05,
      "loss": 0.2269,
      "step": 2272
    },
    {
      "epoch": 1.0378,
      "grad_norm": 1.8592932224273682,
      "learning_rate": 2.78265306122449e-05,
      "loss": 0.1474,
      "step": 2273
    },
    {
      "epoch": 1.038,
      "grad_norm": 4.909307956695557,
      "learning_rate": 2.7816326530612248e-05,
      "loss": 0.8709,
      "step": 2274
    },
    {
      "epoch": 1.0382,
      "grad_norm": 10.424930572509766,
      "learning_rate": 2.7806122448979593e-05,
      "loss": 2.0107,
      "step": 2275
    },
    {
      "epoch": 1.0384,
      "grad_norm": 3.212265968322754,
      "learning_rate": 2.779591836734694e-05,
      "loss": 0.3066,
      "step": 2276
    },
    {
      "epoch": 1.0386,
      "grad_norm": 3.5865800380706787,
      "learning_rate": 2.778571428571429e-05,
      "loss": 0.5172,
      "step": 2277
    },
    {
      "epoch": 1.0388,
      "grad_norm": 4.890392780303955,
      "learning_rate": 2.7775510204081634e-05,
      "loss": 0.8389,
      "step": 2278
    },
    {
      "epoch": 1.039,
      "grad_norm": 4.13357400894165,
      "learning_rate": 2.776530612244898e-05,
      "loss": 0.8173,
      "step": 2279
    },
    {
      "epoch": 1.0392,
      "grad_norm": 2.429373264312744,
      "learning_rate": 2.775510204081633e-05,
      "loss": 0.2955,
      "step": 2280
    },
    {
      "epoch": 1.0394,
      "grad_norm": 4.9131622314453125,
      "learning_rate": 2.7744897959183675e-05,
      "loss": 0.7203,
      "step": 2281
    },
    {
      "epoch": 1.0396,
      "grad_norm": 1.8296152353286743,
      "learning_rate": 2.773469387755102e-05,
      "loss": 0.2563,
      "step": 2282
    },
    {
      "epoch": 1.0398,
      "grad_norm": 2.247852087020874,
      "learning_rate": 2.7724489795918366e-05,
      "loss": 0.1898,
      "step": 2283
    },
    {
      "epoch": 1.04,
      "grad_norm": 4.431122303009033,
      "learning_rate": 2.7714285714285716e-05,
      "loss": 0.859,
      "step": 2284
    },
    {
      "epoch": 1.0402,
      "grad_norm": 7.977433204650879,
      "learning_rate": 2.770408163265306e-05,
      "loss": 1.6311,
      "step": 2285
    },
    {
      "epoch": 1.0404,
      "grad_norm": 11.255778312683105,
      "learning_rate": 2.7693877551020407e-05,
      "loss": 0.903,
      "step": 2286
    },
    {
      "epoch": 1.0406,
      "grad_norm": 6.330371379852295,
      "learning_rate": 2.7683673469387756e-05,
      "loss": 0.8694,
      "step": 2287
    },
    {
      "epoch": 1.0408,
      "grad_norm": 5.011168479919434,
      "learning_rate": 2.7673469387755102e-05,
      "loss": 1.5897,
      "step": 2288
    },
    {
      "epoch": 1.041,
      "grad_norm": 4.034277439117432,
      "learning_rate": 2.7663265306122448e-05,
      "loss": 0.5642,
      "step": 2289
    },
    {
      "epoch": 1.0412,
      "grad_norm": 12.039741516113281,
      "learning_rate": 2.7653061224489797e-05,
      "loss": 1.9871,
      "step": 2290
    },
    {
      "epoch": 1.0414,
      "grad_norm": 4.279134273529053,
      "learning_rate": 2.7642857142857143e-05,
      "loss": 1.0659,
      "step": 2291
    },
    {
      "epoch": 1.0416,
      "grad_norm": 3.9430532455444336,
      "learning_rate": 2.763265306122449e-05,
      "loss": 1.552,
      "step": 2292
    },
    {
      "epoch": 1.0418,
      "grad_norm": 4.80662727355957,
      "learning_rate": 2.7622448979591835e-05,
      "loss": 0.675,
      "step": 2293
    },
    {
      "epoch": 1.042,
      "grad_norm": 1.777544379234314,
      "learning_rate": 2.7612244897959184e-05,
      "loss": 0.2428,
      "step": 2294
    },
    {
      "epoch": 1.0422,
      "grad_norm": 1.8538479804992676,
      "learning_rate": 2.760204081632653e-05,
      "loss": 0.2281,
      "step": 2295
    },
    {
      "epoch": 1.0424,
      "grad_norm": 2.8823657035827637,
      "learning_rate": 2.7591836734693875e-05,
      "loss": 0.5083,
      "step": 2296
    },
    {
      "epoch": 1.0426,
      "grad_norm": 6.437721252441406,
      "learning_rate": 2.7581632653061224e-05,
      "loss": 1.7292,
      "step": 2297
    },
    {
      "epoch": 1.0428,
      "grad_norm": 6.733942985534668,
      "learning_rate": 2.757142857142857e-05,
      "loss": 1.5915,
      "step": 2298
    },
    {
      "epoch": 1.043,
      "grad_norm": 1.15110445022583,
      "learning_rate": 2.7561224489795916e-05,
      "loss": 0.0507,
      "step": 2299
    },
    {
      "epoch": 1.0432,
      "grad_norm": 3.2653064727783203,
      "learning_rate": 2.7551020408163265e-05,
      "loss": 0.2169,
      "step": 2300
    },
    {
      "epoch": 1.0434,
      "grad_norm": 5.073391914367676,
      "learning_rate": 2.7540816326530618e-05,
      "loss": 1.0899,
      "step": 2301
    },
    {
      "epoch": 1.0436,
      "grad_norm": 3.506643056869507,
      "learning_rate": 2.7530612244897964e-05,
      "loss": 1.5375,
      "step": 2302
    },
    {
      "epoch": 1.0438,
      "grad_norm": 5.033652305603027,
      "learning_rate": 2.752040816326531e-05,
      "loss": 1.0176,
      "step": 2303
    },
    {
      "epoch": 1.044,
      "grad_norm": 9.596454620361328,
      "learning_rate": 2.7510204081632655e-05,
      "loss": 1.6695,
      "step": 2304
    },
    {
      "epoch": 1.0442,
      "grad_norm": 7.5560736656188965,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.9778,
      "step": 2305
    },
    {
      "epoch": 1.0444,
      "grad_norm": 17.96355628967285,
      "learning_rate": 2.748979591836735e-05,
      "loss": 3.7537,
      "step": 2306
    },
    {
      "epoch": 1.0446,
      "grad_norm": 6.530677795410156,
      "learning_rate": 2.7479591836734696e-05,
      "loss": 1.1766,
      "step": 2307
    },
    {
      "epoch": 1.0448,
      "grad_norm": 11.135425567626953,
      "learning_rate": 2.7469387755102045e-05,
      "loss": 1.8137,
      "step": 2308
    },
    {
      "epoch": 1.045,
      "grad_norm": 9.50981616973877,
      "learning_rate": 2.745918367346939e-05,
      "loss": 2.141,
      "step": 2309
    },
    {
      "epoch": 1.0452,
      "grad_norm": 9.981623649597168,
      "learning_rate": 2.7448979591836737e-05,
      "loss": 2.3876,
      "step": 2310
    },
    {
      "epoch": 1.0454,
      "grad_norm": 1.8994221687316895,
      "learning_rate": 2.7438775510204086e-05,
      "loss": 0.4994,
      "step": 2311
    },
    {
      "epoch": 1.0456,
      "grad_norm": 2.903367042541504,
      "learning_rate": 2.742857142857143e-05,
      "loss": 1.3922,
      "step": 2312
    },
    {
      "epoch": 1.0458,
      "grad_norm": 1.6302586793899536,
      "learning_rate": 2.7418367346938777e-05,
      "loss": 0.064,
      "step": 2313
    },
    {
      "epoch": 1.046,
      "grad_norm": 2.146597146987915,
      "learning_rate": 2.7408163265306123e-05,
      "loss": 0.2463,
      "step": 2314
    },
    {
      "epoch": 1.0462,
      "grad_norm": 14.185137748718262,
      "learning_rate": 2.7397959183673472e-05,
      "loss": 0.1766,
      "step": 2315
    },
    {
      "epoch": 1.0464,
      "grad_norm": 5.094812393188477,
      "learning_rate": 2.7387755102040818e-05,
      "loss": 0.3866,
      "step": 2316
    },
    {
      "epoch": 1.0466,
      "grad_norm": 10.57387924194336,
      "learning_rate": 2.7377551020408164e-05,
      "loss": 1.2765,
      "step": 2317
    },
    {
      "epoch": 1.0468,
      "grad_norm": 15.01219654083252,
      "learning_rate": 2.7367346938775513e-05,
      "loss": 1.539,
      "step": 2318
    },
    {
      "epoch": 1.047,
      "grad_norm": 2.043815851211548,
      "learning_rate": 2.735714285714286e-05,
      "loss": 0.1832,
      "step": 2319
    },
    {
      "epoch": 1.0472,
      "grad_norm": 2.302161931991577,
      "learning_rate": 2.7346938775510205e-05,
      "loss": 0.3694,
      "step": 2320
    },
    {
      "epoch": 1.0474,
      "grad_norm": 3.4057183265686035,
      "learning_rate": 2.7336734693877554e-05,
      "loss": 0.7633,
      "step": 2321
    },
    {
      "epoch": 1.0476,
      "grad_norm": 4.351888179779053,
      "learning_rate": 2.73265306122449e-05,
      "loss": 0.7517,
      "step": 2322
    },
    {
      "epoch": 1.0478,
      "grad_norm": 2.020519495010376,
      "learning_rate": 2.7316326530612246e-05,
      "loss": 0.2264,
      "step": 2323
    },
    {
      "epoch": 1.048,
      "grad_norm": 6.416701316833496,
      "learning_rate": 2.730612244897959e-05,
      "loss": 1.2137,
      "step": 2324
    },
    {
      "epoch": 1.0482,
      "grad_norm": 8.749337196350098,
      "learning_rate": 2.729591836734694e-05,
      "loss": 2.1079,
      "step": 2325
    },
    {
      "epoch": 1.0484,
      "grad_norm": 4.0751471519470215,
      "learning_rate": 2.7285714285714286e-05,
      "loss": 0.582,
      "step": 2326
    },
    {
      "epoch": 1.0486,
      "grad_norm": 6.795112133026123,
      "learning_rate": 2.7275510204081632e-05,
      "loss": 2.4138,
      "step": 2327
    },
    {
      "epoch": 1.0488,
      "grad_norm": 3.4426896572113037,
      "learning_rate": 2.726530612244898e-05,
      "loss": 1.6688,
      "step": 2328
    },
    {
      "epoch": 1.049,
      "grad_norm": 1.9708757400512695,
      "learning_rate": 2.7255102040816327e-05,
      "loss": 0.2078,
      "step": 2329
    },
    {
      "epoch": 1.0492,
      "grad_norm": 3.8974573612213135,
      "learning_rate": 2.7244897959183673e-05,
      "loss": 0.3048,
      "step": 2330
    },
    {
      "epoch": 1.0493999999999999,
      "grad_norm": 4.25654411315918,
      "learning_rate": 2.7234693877551022e-05,
      "loss": 0.9219,
      "step": 2331
    },
    {
      "epoch": 1.0496,
      "grad_norm": 2.8432178497314453,
      "learning_rate": 2.7224489795918368e-05,
      "loss": 0.6184,
      "step": 2332
    },
    {
      "epoch": 1.0498,
      "grad_norm": 2.5125699043273926,
      "learning_rate": 2.7214285714285714e-05,
      "loss": 0.3014,
      "step": 2333
    },
    {
      "epoch": 1.05,
      "grad_norm": 5.020925998687744,
      "learning_rate": 2.7204081632653063e-05,
      "loss": 0.2826,
      "step": 2334
    },
    {
      "epoch": 1.0502,
      "grad_norm": 6.0717082023620605,
      "learning_rate": 2.719387755102041e-05,
      "loss": 1.4458,
      "step": 2335
    },
    {
      "epoch": 1.0504,
      "grad_norm": 33.236656188964844,
      "learning_rate": 2.7183673469387754e-05,
      "loss": 3.6638,
      "step": 2336
    },
    {
      "epoch": 1.0506,
      "grad_norm": 3.4322261810302734,
      "learning_rate": 2.71734693877551e-05,
      "loss": 1.5605,
      "step": 2337
    },
    {
      "epoch": 1.0508,
      "grad_norm": 3.385988235473633,
      "learning_rate": 2.716326530612245e-05,
      "loss": 0.4192,
      "step": 2338
    },
    {
      "epoch": 1.051,
      "grad_norm": 3.708045482635498,
      "learning_rate": 2.7153061224489795e-05,
      "loss": 0.4939,
      "step": 2339
    },
    {
      "epoch": 1.0512,
      "grad_norm": 6.477840423583984,
      "learning_rate": 2.714285714285714e-05,
      "loss": 1.4888,
      "step": 2340
    },
    {
      "epoch": 1.0514000000000001,
      "grad_norm": 1.9212522506713867,
      "learning_rate": 2.713265306122449e-05,
      "loss": 0.1784,
      "step": 2341
    },
    {
      "epoch": 1.0516,
      "grad_norm": 5.865115165710449,
      "learning_rate": 2.7122448979591836e-05,
      "loss": 1.7386,
      "step": 2342
    },
    {
      "epoch": 1.0518,
      "grad_norm": 19.44985580444336,
      "learning_rate": 2.7112244897959182e-05,
      "loss": 2.1967,
      "step": 2343
    },
    {
      "epoch": 1.052,
      "grad_norm": 3.969221353530884,
      "learning_rate": 2.710204081632653e-05,
      "loss": 0.5162,
      "step": 2344
    },
    {
      "epoch": 1.0522,
      "grad_norm": 18.67195701599121,
      "learning_rate": 2.7091836734693877e-05,
      "loss": 2.5846,
      "step": 2345
    },
    {
      "epoch": 1.0524,
      "grad_norm": 35.398231506347656,
      "learning_rate": 2.708163265306123e-05,
      "loss": 5.0282,
      "step": 2346
    },
    {
      "epoch": 1.0526,
      "grad_norm": 144.9508056640625,
      "learning_rate": 2.7071428571428575e-05,
      "loss": 6.2646,
      "step": 2347
    },
    {
      "epoch": 1.0528,
      "grad_norm": 63.41022491455078,
      "learning_rate": 2.706122448979592e-05,
      "loss": 2.7425,
      "step": 2348
    },
    {
      "epoch": 1.053,
      "grad_norm": 23.23649787902832,
      "learning_rate": 2.705102040816327e-05,
      "loss": 4.6877,
      "step": 2349
    },
    {
      "epoch": 1.0532,
      "grad_norm": 3.7437732219696045,
      "learning_rate": 2.7040816326530616e-05,
      "loss": 1.4819,
      "step": 2350
    },
    {
      "epoch": 1.0534,
      "grad_norm": 2.6368460655212402,
      "learning_rate": 2.703061224489796e-05,
      "loss": 0.1497,
      "step": 2351
    },
    {
      "epoch": 1.0536,
      "grad_norm": 4.356257438659668,
      "learning_rate": 2.702040816326531e-05,
      "loss": 0.6518,
      "step": 2352
    },
    {
      "epoch": 1.0538,
      "grad_norm": 1.6669964790344238,
      "learning_rate": 2.7010204081632657e-05,
      "loss": 0.1701,
      "step": 2353
    },
    {
      "epoch": 1.054,
      "grad_norm": 34.541290283203125,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 1.8341,
      "step": 2354
    },
    {
      "epoch": 1.0542,
      "grad_norm": 82.706787109375,
      "learning_rate": 2.698979591836735e-05,
      "loss": 5.5993,
      "step": 2355
    },
    {
      "epoch": 1.0544,
      "grad_norm": 6.800225734710693,
      "learning_rate": 2.6979591836734697e-05,
      "loss": 0.7273,
      "step": 2356
    },
    {
      "epoch": 1.0546,
      "grad_norm": 3.865412473678589,
      "learning_rate": 2.6969387755102043e-05,
      "loss": 0.9293,
      "step": 2357
    },
    {
      "epoch": 1.0548,
      "grad_norm": 3.1924800872802734,
      "learning_rate": 2.695918367346939e-05,
      "loss": 1.0414,
      "step": 2358
    },
    {
      "epoch": 1.055,
      "grad_norm": 12.705007553100586,
      "learning_rate": 2.6948979591836738e-05,
      "loss": 2.6006,
      "step": 2359
    },
    {
      "epoch": 1.0552,
      "grad_norm": 30.791608810424805,
      "learning_rate": 2.6938775510204084e-05,
      "loss": 4.2604,
      "step": 2360
    },
    {
      "epoch": 1.0554,
      "grad_norm": 3.631416082382202,
      "learning_rate": 2.692857142857143e-05,
      "loss": 0.1501,
      "step": 2361
    },
    {
      "epoch": 1.0556,
      "grad_norm": 5.945212364196777,
      "learning_rate": 2.691836734693878e-05,
      "loss": 0.5055,
      "step": 2362
    },
    {
      "epoch": 1.0558,
      "grad_norm": 22.41519546508789,
      "learning_rate": 2.6908163265306125e-05,
      "loss": 1.3106,
      "step": 2363
    },
    {
      "epoch": 1.056,
      "grad_norm": 49.526512145996094,
      "learning_rate": 2.689795918367347e-05,
      "loss": 4.4305,
      "step": 2364
    },
    {
      "epoch": 1.0562,
      "grad_norm": 3.019853353500366,
      "learning_rate": 2.688775510204082e-05,
      "loss": 0.2182,
      "step": 2365
    },
    {
      "epoch": 1.0564,
      "grad_norm": 5.977735996246338,
      "learning_rate": 2.6877551020408165e-05,
      "loss": 0.3451,
      "step": 2366
    },
    {
      "epoch": 1.0566,
      "grad_norm": 1.628416895866394,
      "learning_rate": 2.686734693877551e-05,
      "loss": 0.1811,
      "step": 2367
    },
    {
      "epoch": 1.0568,
      "grad_norm": 1.9607353210449219,
      "learning_rate": 2.6857142857142857e-05,
      "loss": 0.219,
      "step": 2368
    },
    {
      "epoch": 1.057,
      "grad_norm": 3.1588733196258545,
      "learning_rate": 2.6846938775510206e-05,
      "loss": 0.0474,
      "step": 2369
    },
    {
      "epoch": 1.0572,
      "grad_norm": 1.7066285610198975,
      "learning_rate": 2.6836734693877552e-05,
      "loss": 0.2578,
      "step": 2370
    },
    {
      "epoch": 1.0574,
      "grad_norm": 2.4532039165496826,
      "learning_rate": 2.6826530612244898e-05,
      "loss": 0.6573,
      "step": 2371
    },
    {
      "epoch": 1.0576,
      "grad_norm": 3.317470073699951,
      "learning_rate": 2.6816326530612247e-05,
      "loss": 1.6299,
      "step": 2372
    },
    {
      "epoch": 1.0578,
      "grad_norm": 4.707334518432617,
      "learning_rate": 2.6806122448979593e-05,
      "loss": 0.5495,
      "step": 2373
    },
    {
      "epoch": 1.058,
      "grad_norm": 10.748777389526367,
      "learning_rate": 2.679591836734694e-05,
      "loss": 1.7668,
      "step": 2374
    },
    {
      "epoch": 1.0582,
      "grad_norm": 4.457859516143799,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 0.6171,
      "step": 2375
    },
    {
      "epoch": 1.0584,
      "grad_norm": 2.8171865940093994,
      "learning_rate": 2.6775510204081634e-05,
      "loss": 0.2549,
      "step": 2376
    },
    {
      "epoch": 1.0586,
      "grad_norm": 4.005404949188232,
      "learning_rate": 2.676530612244898e-05,
      "loss": 0.2262,
      "step": 2377
    },
    {
      "epoch": 1.0588,
      "grad_norm": 2.286299467086792,
      "learning_rate": 2.6755102040816325e-05,
      "loss": 0.4013,
      "step": 2378
    },
    {
      "epoch": 1.059,
      "grad_norm": 3.72717547416687,
      "learning_rate": 2.6744897959183674e-05,
      "loss": 1.6463,
      "step": 2379
    },
    {
      "epoch": 1.0592,
      "grad_norm": 6.668277740478516,
      "learning_rate": 2.673469387755102e-05,
      "loss": 0.5391,
      "step": 2380
    },
    {
      "epoch": 1.0594,
      "grad_norm": 9.107120513916016,
      "learning_rate": 2.6724489795918366e-05,
      "loss": 1.5799,
      "step": 2381
    },
    {
      "epoch": 1.0596,
      "grad_norm": 4.307889938354492,
      "learning_rate": 2.6714285714285715e-05,
      "loss": 0.2707,
      "step": 2382
    },
    {
      "epoch": 1.0598,
      "grad_norm": 3.1701438426971436,
      "learning_rate": 2.670408163265306e-05,
      "loss": 0.6578,
      "step": 2383
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.8328818082809448,
      "learning_rate": 2.6693877551020407e-05,
      "loss": 0.0478,
      "step": 2384
    },
    {
      "epoch": 1.0602,
      "grad_norm": 11.048552513122559,
      "learning_rate": 2.6683673469387756e-05,
      "loss": 1.4496,
      "step": 2385
    },
    {
      "epoch": 1.0604,
      "grad_norm": 17.557401657104492,
      "learning_rate": 2.66734693877551e-05,
      "loss": 4.7514,
      "step": 2386
    },
    {
      "epoch": 1.0606,
      "grad_norm": 2.2788662910461426,
      "learning_rate": 2.6663265306122447e-05,
      "loss": 0.0864,
      "step": 2387
    },
    {
      "epoch": 1.0608,
      "grad_norm": 1.4921162128448486,
      "learning_rate": 2.6653061224489793e-05,
      "loss": 0.0922,
      "step": 2388
    },
    {
      "epoch": 1.061,
      "grad_norm": 2.2919957637786865,
      "learning_rate": 2.6642857142857142e-05,
      "loss": 0.161,
      "step": 2389
    },
    {
      "epoch": 1.0612,
      "grad_norm": 21.927486419677734,
      "learning_rate": 2.6632653061224488e-05,
      "loss": 2.0904,
      "step": 2390
    },
    {
      "epoch": 1.0614,
      "grad_norm": 6.614194393157959,
      "learning_rate": 2.6622448979591834e-05,
      "loss": 4.6801,
      "step": 2391
    },
    {
      "epoch": 1.0616,
      "grad_norm": 2.3331987857818604,
      "learning_rate": 2.6612244897959187e-05,
      "loss": 0.5976,
      "step": 2392
    },
    {
      "epoch": 1.0618,
      "grad_norm": 3.3811841011047363,
      "learning_rate": 2.6602040816326536e-05,
      "loss": 1.6392,
      "step": 2393
    },
    {
      "epoch": 1.062,
      "grad_norm": 2.3435702323913574,
      "learning_rate": 2.659183673469388e-05,
      "loss": 0.3477,
      "step": 2394
    },
    {
      "epoch": 1.0622,
      "grad_norm": 6.553333282470703,
      "learning_rate": 2.6581632653061227e-05,
      "loss": 1.6144,
      "step": 2395
    },
    {
      "epoch": 1.0624,
      "grad_norm": 10.608148574829102,
      "learning_rate": 2.6571428571428576e-05,
      "loss": 2.163,
      "step": 2396
    },
    {
      "epoch": 1.0626,
      "grad_norm": 2.376661539077759,
      "learning_rate": 2.6561224489795922e-05,
      "loss": 0.2497,
      "step": 2397
    },
    {
      "epoch": 1.0628,
      "grad_norm": 5.275718688964844,
      "learning_rate": 2.6551020408163268e-05,
      "loss": 0.707,
      "step": 2398
    },
    {
      "epoch": 1.063,
      "grad_norm": 4.847681999206543,
      "learning_rate": 2.6540816326530614e-05,
      "loss": 0.5223,
      "step": 2399
    },
    {
      "epoch": 1.0632,
      "grad_norm": 5.3941650390625,
      "learning_rate": 2.6530612244897963e-05,
      "loss": 1.6102,
      "step": 2400
    },
    {
      "epoch": 1.0634,
      "grad_norm": 6.420630931854248,
      "learning_rate": 2.652040816326531e-05,
      "loss": 0.8017,
      "step": 2401
    },
    {
      "epoch": 1.0636,
      "grad_norm": 5.283441543579102,
      "learning_rate": 2.6510204081632655e-05,
      "loss": 0.4855,
      "step": 2402
    },
    {
      "epoch": 1.0638,
      "grad_norm": 13.458806037902832,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.7235,
      "step": 2403
    },
    {
      "epoch": 1.064,
      "grad_norm": 28.971445083618164,
      "learning_rate": 2.648979591836735e-05,
      "loss": 2.4488,
      "step": 2404
    },
    {
      "epoch": 1.0642,
      "grad_norm": 3.099283456802368,
      "learning_rate": 2.6479591836734695e-05,
      "loss": 0.374,
      "step": 2405
    },
    {
      "epoch": 1.0644,
      "grad_norm": 4.976246356964111,
      "learning_rate": 2.6469387755102045e-05,
      "loss": 1.3718,
      "step": 2406
    },
    {
      "epoch": 1.0646,
      "grad_norm": 0.9986436367034912,
      "learning_rate": 2.645918367346939e-05,
      "loss": 0.041,
      "step": 2407
    },
    {
      "epoch": 1.0648,
      "grad_norm": 2.3621811866760254,
      "learning_rate": 2.6448979591836736e-05,
      "loss": 0.6865,
      "step": 2408
    },
    {
      "epoch": 1.065,
      "grad_norm": 3.2157063484191895,
      "learning_rate": 2.6438775510204082e-05,
      "loss": 1.4494,
      "step": 2409
    },
    {
      "epoch": 1.0652,
      "grad_norm": 1.8791629076004028,
      "learning_rate": 2.642857142857143e-05,
      "loss": 0.255,
      "step": 2410
    },
    {
      "epoch": 1.0654,
      "grad_norm": 1.9395337104797363,
      "learning_rate": 2.6418367346938777e-05,
      "loss": 0.2685,
      "step": 2411
    },
    {
      "epoch": 1.0656,
      "grad_norm": 1.4387263059616089,
      "learning_rate": 2.6408163265306123e-05,
      "loss": 0.1137,
      "step": 2412
    },
    {
      "epoch": 1.0658,
      "grad_norm": 3.260917901992798,
      "learning_rate": 2.6397959183673472e-05,
      "loss": 0.9726,
      "step": 2413
    },
    {
      "epoch": 1.066,
      "grad_norm": 3.520772695541382,
      "learning_rate": 2.6387755102040818e-05,
      "loss": 1.5219,
      "step": 2414
    },
    {
      "epoch": 1.0662,
      "grad_norm": 7.0468316078186035,
      "learning_rate": 2.6377551020408163e-05,
      "loss": 0.5509,
      "step": 2415
    },
    {
      "epoch": 1.0664,
      "grad_norm": 4.3371663093566895,
      "learning_rate": 2.6367346938775513e-05,
      "loss": 0.8301,
      "step": 2416
    },
    {
      "epoch": 1.0666,
      "grad_norm": 4.128448486328125,
      "learning_rate": 2.635714285714286e-05,
      "loss": 1.2042,
      "step": 2417
    },
    {
      "epoch": 1.0668,
      "grad_norm": 3.1050233840942383,
      "learning_rate": 2.6346938775510204e-05,
      "loss": 1.4974,
      "step": 2418
    },
    {
      "epoch": 1.067,
      "grad_norm": 1.9743566513061523,
      "learning_rate": 2.633673469387755e-05,
      "loss": 0.138,
      "step": 2419
    },
    {
      "epoch": 1.0672,
      "grad_norm": 1.6516227722167969,
      "learning_rate": 2.63265306122449e-05,
      "loss": 0.1792,
      "step": 2420
    },
    {
      "epoch": 1.0674,
      "grad_norm": 2.782998561859131,
      "learning_rate": 2.6316326530612245e-05,
      "loss": 0.3191,
      "step": 2421
    },
    {
      "epoch": 1.0676,
      "grad_norm": 3.995413303375244,
      "learning_rate": 2.630612244897959e-05,
      "loss": 0.5936,
      "step": 2422
    },
    {
      "epoch": 1.0678,
      "grad_norm": 3.0147314071655273,
      "learning_rate": 2.629591836734694e-05,
      "loss": 0.2437,
      "step": 2423
    },
    {
      "epoch": 1.068,
      "grad_norm": 9.054986000061035,
      "learning_rate": 2.6285714285714286e-05,
      "loss": 1.7769,
      "step": 2424
    },
    {
      "epoch": 1.0682,
      "grad_norm": 8.601308822631836,
      "learning_rate": 2.627551020408163e-05,
      "loss": 1.6046,
      "step": 2425
    },
    {
      "epoch": 1.0684,
      "grad_norm": 4.747802257537842,
      "learning_rate": 2.626530612244898e-05,
      "loss": 0.7729,
      "step": 2426
    },
    {
      "epoch": 1.0686,
      "grad_norm": 3.94885516166687,
      "learning_rate": 2.6255102040816326e-05,
      "loss": 1.7273,
      "step": 2427
    },
    {
      "epoch": 1.0688,
      "grad_norm": 2.5352084636688232,
      "learning_rate": 2.6244897959183672e-05,
      "loss": 0.6568,
      "step": 2428
    },
    {
      "epoch": 1.069,
      "grad_norm": 6.1487507820129395,
      "learning_rate": 2.623469387755102e-05,
      "loss": 2.8582,
      "step": 2429
    },
    {
      "epoch": 1.0692,
      "grad_norm": 10.868823051452637,
      "learning_rate": 2.6224489795918367e-05,
      "loss": 4.6807,
      "step": 2430
    },
    {
      "epoch": 1.0694,
      "grad_norm": 2.121119260787964,
      "learning_rate": 2.6214285714285713e-05,
      "loss": 0.2144,
      "step": 2431
    },
    {
      "epoch": 1.0695999999999999,
      "grad_norm": 5.630823612213135,
      "learning_rate": 2.620408163265306e-05,
      "loss": 1.0378,
      "step": 2432
    },
    {
      "epoch": 1.0698,
      "grad_norm": 14.371367454528809,
      "learning_rate": 2.6193877551020408e-05,
      "loss": 3.6745,
      "step": 2433
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.156108856201172,
      "learning_rate": 2.6183673469387754e-05,
      "loss": 4.7997,
      "step": 2434
    },
    {
      "epoch": 1.0702,
      "grad_norm": 4.25835657119751,
      "learning_rate": 2.61734693877551e-05,
      "loss": 0.5825,
      "step": 2435
    },
    {
      "epoch": 1.0704,
      "grad_norm": 3.216127872467041,
      "learning_rate": 2.616326530612245e-05,
      "loss": 0.366,
      "step": 2436
    },
    {
      "epoch": 1.0706,
      "grad_norm": 1.3122063875198364,
      "learning_rate": 2.61530612244898e-05,
      "loss": 0.0502,
      "step": 2437
    },
    {
      "epoch": 1.0708,
      "grad_norm": 1.7881473302841187,
      "learning_rate": 2.6142857142857147e-05,
      "loss": 0.0359,
      "step": 2438
    },
    {
      "epoch": 1.071,
      "grad_norm": 1.5982885360717773,
      "learning_rate": 2.6132653061224493e-05,
      "loss": 0.1776,
      "step": 2439
    },
    {
      "epoch": 1.0712,
      "grad_norm": 1.4499366283416748,
      "learning_rate": 2.612244897959184e-05,
      "loss": 0.0721,
      "step": 2440
    },
    {
      "epoch": 1.0714,
      "grad_norm": 2.7761549949645996,
      "learning_rate": 2.6112244897959188e-05,
      "loss": 0.2619,
      "step": 2441
    },
    {
      "epoch": 1.0716,
      "grad_norm": 3.3720993995666504,
      "learning_rate": 2.6102040816326534e-05,
      "loss": 0.1568,
      "step": 2442
    },
    {
      "epoch": 1.0718,
      "grad_norm": 1.3414360284805298,
      "learning_rate": 2.609183673469388e-05,
      "loss": 0.0378,
      "step": 2443
    },
    {
      "epoch": 1.072,
      "grad_norm": 1.1581441164016724,
      "learning_rate": 2.608163265306123e-05,
      "loss": 0.0442,
      "step": 2444
    },
    {
      "epoch": 1.0722,
      "grad_norm": 1.699424386024475,
      "learning_rate": 2.6071428571428574e-05,
      "loss": 0.2574,
      "step": 2445
    },
    {
      "epoch": 1.0724,
      "grad_norm": 3.8419086933135986,
      "learning_rate": 2.606122448979592e-05,
      "loss": 0.2357,
      "step": 2446
    },
    {
      "epoch": 1.0726,
      "grad_norm": 8.273889541625977,
      "learning_rate": 2.605102040816327e-05,
      "loss": 1.3546,
      "step": 2447
    },
    {
      "epoch": 1.0728,
      "grad_norm": 2.8728175163269043,
      "learning_rate": 2.6040816326530615e-05,
      "loss": 0.64,
      "step": 2448
    },
    {
      "epoch": 1.073,
      "grad_norm": 6.3875579833984375,
      "learning_rate": 2.603061224489796e-05,
      "loss": 1.9487,
      "step": 2449
    },
    {
      "epoch": 1.0732,
      "grad_norm": 9.781810760498047,
      "learning_rate": 2.6020408163265307e-05,
      "loss": 1.4451,
      "step": 2450
    },
    {
      "epoch": 1.0734,
      "grad_norm": 13.755586624145508,
      "learning_rate": 2.6010204081632656e-05,
      "loss": 0.9554,
      "step": 2451
    },
    {
      "epoch": 1.0735999999999999,
      "grad_norm": 4.130563259124756,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.173,
      "step": 2452
    },
    {
      "epoch": 1.0738,
      "grad_norm": 1.6912357807159424,
      "learning_rate": 2.5989795918367348e-05,
      "loss": 0.1261,
      "step": 2453
    },
    {
      "epoch": 1.074,
      "grad_norm": 4.4955925941467285,
      "learning_rate": 2.5979591836734697e-05,
      "loss": 0.2242,
      "step": 2454
    },
    {
      "epoch": 1.0742,
      "grad_norm": 2.2355611324310303,
      "learning_rate": 2.5969387755102043e-05,
      "loss": 0.1369,
      "step": 2455
    },
    {
      "epoch": 1.0744,
      "grad_norm": 1.3106834888458252,
      "learning_rate": 2.595918367346939e-05,
      "loss": 0.0492,
      "step": 2456
    },
    {
      "epoch": 1.0746,
      "grad_norm": 7.732717990875244,
      "learning_rate": 2.5948979591836738e-05,
      "loss": 0.6055,
      "step": 2457
    },
    {
      "epoch": 1.0748,
      "grad_norm": 13.904815673828125,
      "learning_rate": 2.5938775510204083e-05,
      "loss": 1.3689,
      "step": 2458
    },
    {
      "epoch": 1.075,
      "grad_norm": 1.706587314605713,
      "learning_rate": 2.592857142857143e-05,
      "loss": 0.0827,
      "step": 2459
    },
    {
      "epoch": 1.0752,
      "grad_norm": 7.618509292602539,
      "learning_rate": 2.5918367346938778e-05,
      "loss": 1.062,
      "step": 2460
    },
    {
      "epoch": 1.0754,
      "grad_norm": 14.990194320678711,
      "learning_rate": 2.5908163265306124e-05,
      "loss": 2.4455,
      "step": 2461
    },
    {
      "epoch": 1.0756000000000001,
      "grad_norm": 7.136456489562988,
      "learning_rate": 2.589795918367347e-05,
      "loss": 0.6104,
      "step": 2462
    },
    {
      "epoch": 1.0758,
      "grad_norm": 19.634891510009766,
      "learning_rate": 2.5887755102040816e-05,
      "loss": 1.4636,
      "step": 2463
    },
    {
      "epoch": 1.076,
      "grad_norm": 2.359050750732422,
      "learning_rate": 2.5877551020408165e-05,
      "loss": 0.4074,
      "step": 2464
    },
    {
      "epoch": 1.0762,
      "grad_norm": 5.584263801574707,
      "learning_rate": 2.586734693877551e-05,
      "loss": 2.4626,
      "step": 2465
    },
    {
      "epoch": 1.0764,
      "grad_norm": 4.744572162628174,
      "learning_rate": 2.5857142857142856e-05,
      "loss": 1.764,
      "step": 2466
    },
    {
      "epoch": 1.0766,
      "grad_norm": 4.963972568511963,
      "learning_rate": 2.5846938775510206e-05,
      "loss": 0.3838,
      "step": 2467
    },
    {
      "epoch": 1.0768,
      "grad_norm": 12.521373748779297,
      "learning_rate": 2.583673469387755e-05,
      "loss": 1.2165,
      "step": 2468
    },
    {
      "epoch": 1.077,
      "grad_norm": 33.27798843383789,
      "learning_rate": 2.5826530612244897e-05,
      "loss": 4.0389,
      "step": 2469
    },
    {
      "epoch": 1.0772,
      "grad_norm": 4.73541259765625,
      "learning_rate": 2.5816326530612246e-05,
      "loss": 0.9516,
      "step": 2470
    },
    {
      "epoch": 1.0774,
      "grad_norm": 7.456980228424072,
      "learning_rate": 2.5806122448979592e-05,
      "loss": 1.5232,
      "step": 2471
    },
    {
      "epoch": 1.0776,
      "grad_norm": 3.147059202194214,
      "learning_rate": 2.5795918367346938e-05,
      "loss": 0.2818,
      "step": 2472
    },
    {
      "epoch": 1.0778,
      "grad_norm": 6.3879008293151855,
      "learning_rate": 2.5785714285714284e-05,
      "loss": 0.7583,
      "step": 2473
    },
    {
      "epoch": 1.078,
      "grad_norm": 2.2574191093444824,
      "learning_rate": 2.5775510204081633e-05,
      "loss": 0.4802,
      "step": 2474
    },
    {
      "epoch": 1.0782,
      "grad_norm": 6.5756378173828125,
      "learning_rate": 2.576530612244898e-05,
      "loss": 2.5003,
      "step": 2475
    },
    {
      "epoch": 1.0784,
      "grad_norm": 11.439837455749512,
      "learning_rate": 2.5755102040816325e-05,
      "loss": 3.7775,
      "step": 2476
    },
    {
      "epoch": 1.0786,
      "grad_norm": 23.936866760253906,
      "learning_rate": 2.5744897959183674e-05,
      "loss": 2.1445,
      "step": 2477
    },
    {
      "epoch": 1.0788,
      "grad_norm": 23.30131721496582,
      "learning_rate": 2.573469387755102e-05,
      "loss": 3.8271,
      "step": 2478
    },
    {
      "epoch": 1.079,
      "grad_norm": 2.3670778274536133,
      "learning_rate": 2.5724489795918365e-05,
      "loss": 0.3152,
      "step": 2479
    },
    {
      "epoch": 1.0792,
      "grad_norm": 3.0792884826660156,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.5886,
      "step": 2480
    },
    {
      "epoch": 1.0794,
      "grad_norm": 14.284750938415527,
      "learning_rate": 2.570408163265306e-05,
      "loss": 1.2113,
      "step": 2481
    },
    {
      "epoch": 1.0796000000000001,
      "grad_norm": 29.00316047668457,
      "learning_rate": 2.5693877551020413e-05,
      "loss": 3.9031,
      "step": 2482
    },
    {
      "epoch": 1.0798,
      "grad_norm": 2.106374502182007,
      "learning_rate": 2.568367346938776e-05,
      "loss": 0.2462,
      "step": 2483
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.4427016973495483,
      "learning_rate": 2.5673469387755104e-05,
      "loss": 0.0533,
      "step": 2484
    },
    {
      "epoch": 1.0802,
      "grad_norm": 2.1444180011749268,
      "learning_rate": 2.5663265306122454e-05,
      "loss": 0.2781,
      "step": 2485
    },
    {
      "epoch": 1.0804,
      "grad_norm": 6.793755054473877,
      "learning_rate": 2.56530612244898e-05,
      "loss": 1.5963,
      "step": 2486
    },
    {
      "epoch": 1.0806,
      "grad_norm": 15.605535507202148,
      "learning_rate": 2.5642857142857145e-05,
      "loss": 5.3253,
      "step": 2487
    },
    {
      "epoch": 1.0808,
      "grad_norm": 6.925433158874512,
      "learning_rate": 2.5632653061224494e-05,
      "loss": 0.827,
      "step": 2488
    },
    {
      "epoch": 1.081,
      "grad_norm": 7.272058963775635,
      "learning_rate": 2.562244897959184e-05,
      "loss": 1.4671,
      "step": 2489
    },
    {
      "epoch": 1.0812,
      "grad_norm": 3.7872958183288574,
      "learning_rate": 2.5612244897959186e-05,
      "loss": 0.5281,
      "step": 2490
    },
    {
      "epoch": 1.0814,
      "grad_norm": 1.976417899131775,
      "learning_rate": 2.5602040816326535e-05,
      "loss": 0.2451,
      "step": 2491
    },
    {
      "epoch": 1.0816,
      "grad_norm": 2.881429433822632,
      "learning_rate": 2.559183673469388e-05,
      "loss": 0.2317,
      "step": 2492
    },
    {
      "epoch": 1.0818,
      "grad_norm": 3.65335750579834,
      "learning_rate": 2.5581632653061227e-05,
      "loss": 0.3894,
      "step": 2493
    },
    {
      "epoch": 1.082,
      "grad_norm": 2.544227123260498,
      "learning_rate": 2.5571428571428572e-05,
      "loss": 0.3121,
      "step": 2494
    },
    {
      "epoch": 1.0822,
      "grad_norm": 7.100771903991699,
      "learning_rate": 2.556122448979592e-05,
      "loss": 1.753,
      "step": 2495
    },
    {
      "epoch": 1.0824,
      "grad_norm": 12.864798545837402,
      "learning_rate": 2.5551020408163267e-05,
      "loss": 2.8146,
      "step": 2496
    },
    {
      "epoch": 1.0826,
      "grad_norm": 22.558868408203125,
      "learning_rate": 2.5540816326530613e-05,
      "loss": 1.1523,
      "step": 2497
    },
    {
      "epoch": 1.0828,
      "grad_norm": 73.28495025634766,
      "learning_rate": 2.5530612244897962e-05,
      "loss": 5.9876,
      "step": 2498
    },
    {
      "epoch": 1.083,
      "grad_norm": 80.7002182006836,
      "learning_rate": 2.5520408163265308e-05,
      "loss": 5.6509,
      "step": 2499
    },
    {
      "epoch": 1.0832,
      "grad_norm": 2.3751392364501953,
      "learning_rate": 2.5510204081632654e-05,
      "loss": 0.3095,
      "step": 2500
    },
    {
      "epoch": 1.0834,
      "grad_norm": 5.686920642852783,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.7626,
      "step": 2501
    },
    {
      "epoch": 1.0836,
      "grad_norm": 4.439648628234863,
      "learning_rate": 2.548979591836735e-05,
      "loss": 0.4574,
      "step": 2502
    },
    {
      "epoch": 1.0838,
      "grad_norm": 4.587183475494385,
      "learning_rate": 2.5479591836734695e-05,
      "loss": 0.6623,
      "step": 2503
    },
    {
      "epoch": 1.084,
      "grad_norm": 2.004228353500366,
      "learning_rate": 2.546938775510204e-05,
      "loss": 0.0682,
      "step": 2504
    },
    {
      "epoch": 1.0842,
      "grad_norm": 4.348333358764648,
      "learning_rate": 2.545918367346939e-05,
      "loss": 0.1526,
      "step": 2505
    },
    {
      "epoch": 1.0844,
      "grad_norm": 2.2945642471313477,
      "learning_rate": 2.5448979591836736e-05,
      "loss": 0.6309,
      "step": 2506
    },
    {
      "epoch": 1.0846,
      "grad_norm": 3.4506020545959473,
      "learning_rate": 2.543877551020408e-05,
      "loss": 1.6623,
      "step": 2507
    },
    {
      "epoch": 1.0848,
      "grad_norm": 3.223053216934204,
      "learning_rate": 2.542857142857143e-05,
      "loss": 0.3624,
      "step": 2508
    },
    {
      "epoch": 1.085,
      "grad_norm": 2.0197436809539795,
      "learning_rate": 2.5418367346938776e-05,
      "loss": 0.4646,
      "step": 2509
    },
    {
      "epoch": 1.0852,
      "grad_norm": 3.190613031387329,
      "learning_rate": 2.5408163265306122e-05,
      "loss": 1.6204,
      "step": 2510
    },
    {
      "epoch": 1.0854,
      "grad_norm": 1.880837082862854,
      "learning_rate": 2.539795918367347e-05,
      "loss": 0.1061,
      "step": 2511
    },
    {
      "epoch": 1.0856,
      "grad_norm": 3.710780143737793,
      "learning_rate": 2.5387755102040817e-05,
      "loss": 0.4916,
      "step": 2512
    },
    {
      "epoch": 1.0858,
      "grad_norm": 4.7563886642456055,
      "learning_rate": 2.5377551020408163e-05,
      "loss": 0.6781,
      "step": 2513
    },
    {
      "epoch": 1.086,
      "grad_norm": 6.385908603668213,
      "learning_rate": 2.536734693877551e-05,
      "loss": 0.4621,
      "step": 2514
    },
    {
      "epoch": 1.0862,
      "grad_norm": 9.943044662475586,
      "learning_rate": 2.5357142857142858e-05,
      "loss": 1.7608,
      "step": 2515
    },
    {
      "epoch": 1.0864,
      "grad_norm": 7.606964588165283,
      "learning_rate": 2.5346938775510204e-05,
      "loss": 2.5999,
      "step": 2516
    },
    {
      "epoch": 1.0866,
      "grad_norm": 8.824433326721191,
      "learning_rate": 2.533673469387755e-05,
      "loss": 3.2743,
      "step": 2517
    },
    {
      "epoch": 1.0868,
      "grad_norm": 3.378680467605591,
      "learning_rate": 2.53265306122449e-05,
      "loss": 1.7249,
      "step": 2518
    },
    {
      "epoch": 1.087,
      "grad_norm": 7.698119163513184,
      "learning_rate": 2.5316326530612244e-05,
      "loss": 1.4085,
      "step": 2519
    },
    {
      "epoch": 1.0872,
      "grad_norm": 16.662708282470703,
      "learning_rate": 2.530612244897959e-05,
      "loss": 3.6055,
      "step": 2520
    },
    {
      "epoch": 1.0874,
      "grad_norm": 2.6420581340789795,
      "learning_rate": 2.529591836734694e-05,
      "loss": 0.2005,
      "step": 2521
    },
    {
      "epoch": 1.0876,
      "grad_norm": 3.360490560531616,
      "learning_rate": 2.5285714285714285e-05,
      "loss": 0.862,
      "step": 2522
    },
    {
      "epoch": 1.0878,
      "grad_norm": 3.689589500427246,
      "learning_rate": 2.527551020408163e-05,
      "loss": 2.2184,
      "step": 2523
    },
    {
      "epoch": 1.088,
      "grad_norm": 3.124295711517334,
      "learning_rate": 2.526530612244898e-05,
      "loss": 1.5167,
      "step": 2524
    },
    {
      "epoch": 1.0882,
      "grad_norm": 2.8372583389282227,
      "learning_rate": 2.5255102040816326e-05,
      "loss": 0.5618,
      "step": 2525
    },
    {
      "epoch": 1.0884,
      "grad_norm": 4.048417091369629,
      "learning_rate": 2.5244897959183672e-05,
      "loss": 0.5793,
      "step": 2526
    },
    {
      "epoch": 1.0886,
      "grad_norm": 2.170830249786377,
      "learning_rate": 2.5234693877551017e-05,
      "loss": 0.2466,
      "step": 2527
    },
    {
      "epoch": 1.0888,
      "grad_norm": 1.8162426948547363,
      "learning_rate": 2.522448979591837e-05,
      "loss": 0.1953,
      "step": 2528
    },
    {
      "epoch": 1.089,
      "grad_norm": 2.095097303390503,
      "learning_rate": 2.521428571428572e-05,
      "loss": 0.1488,
      "step": 2529
    },
    {
      "epoch": 1.0892,
      "grad_norm": 3.6445815563201904,
      "learning_rate": 2.5204081632653065e-05,
      "loss": 0.5365,
      "step": 2530
    },
    {
      "epoch": 1.0894,
      "grad_norm": 2.005758047103882,
      "learning_rate": 2.519387755102041e-05,
      "loss": 0.2343,
      "step": 2531
    },
    {
      "epoch": 1.0896,
      "grad_norm": 2.506690502166748,
      "learning_rate": 2.518367346938776e-05,
      "loss": 0.276,
      "step": 2532
    },
    {
      "epoch": 1.0898,
      "grad_norm": 4.830741882324219,
      "learning_rate": 2.5173469387755106e-05,
      "loss": 0.7821,
      "step": 2533
    },
    {
      "epoch": 1.09,
      "grad_norm": 16.803346633911133,
      "learning_rate": 2.516326530612245e-05,
      "loss": 1.7534,
      "step": 2534
    },
    {
      "epoch": 1.0902,
      "grad_norm": 20.17479705810547,
      "learning_rate": 2.5153061224489797e-05,
      "loss": 1.6485,
      "step": 2535
    },
    {
      "epoch": 1.0904,
      "grad_norm": 6.046016216278076,
      "learning_rate": 2.5142857142857147e-05,
      "loss": 1.2223,
      "step": 2536
    },
    {
      "epoch": 1.0906,
      "grad_norm": 11.118535995483398,
      "learning_rate": 2.5132653061224492e-05,
      "loss": 2.9827,
      "step": 2537
    },
    {
      "epoch": 1.0908,
      "grad_norm": 4.158987045288086,
      "learning_rate": 2.5122448979591838e-05,
      "loss": 0.8138,
      "step": 2538
    },
    {
      "epoch": 1.091,
      "grad_norm": 5.505331516265869,
      "learning_rate": 2.5112244897959187e-05,
      "loss": 0.3746,
      "step": 2539
    },
    {
      "epoch": 1.0912,
      "grad_norm": 17.089290618896484,
      "learning_rate": 2.5102040816326533e-05,
      "loss": 0.8031,
      "step": 2540
    },
    {
      "epoch": 1.0914,
      "grad_norm": 1.8025397062301636,
      "learning_rate": 2.509183673469388e-05,
      "loss": 0.2397,
      "step": 2541
    },
    {
      "epoch": 1.0916,
      "grad_norm": 1.887250542640686,
      "learning_rate": 2.5081632653061228e-05,
      "loss": 0.1461,
      "step": 2542
    },
    {
      "epoch": 1.0918,
      "grad_norm": 1.9648981094360352,
      "learning_rate": 2.5071428571428574e-05,
      "loss": 0.0805,
      "step": 2543
    },
    {
      "epoch": 1.092,
      "grad_norm": 4.035311698913574,
      "learning_rate": 2.506122448979592e-05,
      "loss": 0.5857,
      "step": 2544
    },
    {
      "epoch": 1.0922,
      "grad_norm": 6.021888732910156,
      "learning_rate": 2.5051020408163265e-05,
      "loss": 1.2061,
      "step": 2545
    },
    {
      "epoch": 1.0924,
      "grad_norm": 1.1482367515563965,
      "learning_rate": 2.5040816326530615e-05,
      "loss": 0.0516,
      "step": 2546
    },
    {
      "epoch": 1.0926,
      "grad_norm": 1.3609832525253296,
      "learning_rate": 2.503061224489796e-05,
      "loss": 0.0763,
      "step": 2547
    },
    {
      "epoch": 1.0928,
      "grad_norm": 1.8674981594085693,
      "learning_rate": 2.5020408163265306e-05,
      "loss": 0.207,
      "step": 2548
    },
    {
      "epoch": 1.093,
      "grad_norm": 2.574284791946411,
      "learning_rate": 2.5010204081632655e-05,
      "loss": 0.4801,
      "step": 2549
    },
    {
      "epoch": 1.0932,
      "grad_norm": 5.7880940437316895,
      "learning_rate": 2.5e-05,
      "loss": 1.706,
      "step": 2550
    },
    {
      "epoch": 1.0934,
      "grad_norm": 8.400511741638184,
      "learning_rate": 2.4989795918367347e-05,
      "loss": 0.9191,
      "step": 2551
    },
    {
      "epoch": 1.0936,
      "grad_norm": 6.415307998657227,
      "learning_rate": 2.4979591836734696e-05,
      "loss": 0.9075,
      "step": 2552
    },
    {
      "epoch": 1.0937999999999999,
      "grad_norm": 4.819631099700928,
      "learning_rate": 2.4969387755102042e-05,
      "loss": 1.2393,
      "step": 2553
    },
    {
      "epoch": 1.094,
      "grad_norm": 4.074408531188965,
      "learning_rate": 2.4959183673469388e-05,
      "loss": 0.7136,
      "step": 2554
    },
    {
      "epoch": 1.0942,
      "grad_norm": 1.512075662612915,
      "learning_rate": 2.4948979591836737e-05,
      "loss": 0.1217,
      "step": 2555
    },
    {
      "epoch": 1.0944,
      "grad_norm": 1.1646981239318848,
      "learning_rate": 2.4938775510204083e-05,
      "loss": 0.0507,
      "step": 2556
    },
    {
      "epoch": 1.0946,
      "grad_norm": 1.680336594581604,
      "learning_rate": 2.492857142857143e-05,
      "loss": 0.1584,
      "step": 2557
    },
    {
      "epoch": 1.0948,
      "grad_norm": 2.7351956367492676,
      "learning_rate": 2.4918367346938774e-05,
      "loss": 0.37,
      "step": 2558
    },
    {
      "epoch": 1.095,
      "grad_norm": 6.237983703613281,
      "learning_rate": 2.4908163265306123e-05,
      "loss": 1.5575,
      "step": 2559
    },
    {
      "epoch": 1.0952,
      "grad_norm": 2.0216691493988037,
      "learning_rate": 2.489795918367347e-05,
      "loss": 0.1551,
      "step": 2560
    },
    {
      "epoch": 1.0954,
      "grad_norm": 7.477294921875,
      "learning_rate": 2.4887755102040815e-05,
      "loss": 0.5022,
      "step": 2561
    },
    {
      "epoch": 1.0956,
      "grad_norm": 17.65044403076172,
      "learning_rate": 2.4877551020408164e-05,
      "loss": 2.1472,
      "step": 2562
    },
    {
      "epoch": 1.0958,
      "grad_norm": 1.9280164241790771,
      "learning_rate": 2.4867346938775513e-05,
      "loss": 0.1854,
      "step": 2563
    },
    {
      "epoch": 1.096,
      "grad_norm": 3.616777181625366,
      "learning_rate": 2.485714285714286e-05,
      "loss": 0.2811,
      "step": 2564
    },
    {
      "epoch": 1.0962,
      "grad_norm": 3.648794651031494,
      "learning_rate": 2.4846938775510205e-05,
      "loss": 0.496,
      "step": 2565
    },
    {
      "epoch": 1.0964,
      "grad_norm": 4.593664646148682,
      "learning_rate": 2.4836734693877554e-05,
      "loss": 0.6411,
      "step": 2566
    },
    {
      "epoch": 1.0966,
      "grad_norm": 2.9092133045196533,
      "learning_rate": 2.48265306122449e-05,
      "loss": 1.4583,
      "step": 2567
    },
    {
      "epoch": 1.0968,
      "grad_norm": 5.004642486572266,
      "learning_rate": 2.4816326530612246e-05,
      "loss": 5.1528,
      "step": 2568
    },
    {
      "epoch": 1.097,
      "grad_norm": 3.7020583152770996,
      "learning_rate": 2.4806122448979595e-05,
      "loss": 0.7462,
      "step": 2569
    },
    {
      "epoch": 1.0972,
      "grad_norm": 4.8899736404418945,
      "learning_rate": 2.479591836734694e-05,
      "loss": 0.7615,
      "step": 2570
    },
    {
      "epoch": 1.0974,
      "grad_norm": 12.54307746887207,
      "learning_rate": 2.4785714285714287e-05,
      "loss": 1.0085,
      "step": 2571
    },
    {
      "epoch": 1.0976,
      "grad_norm": 17.654830932617188,
      "learning_rate": 2.4775510204081632e-05,
      "loss": 2.3768,
      "step": 2572
    },
    {
      "epoch": 1.0977999999999999,
      "grad_norm": 3.0789108276367188,
      "learning_rate": 2.476530612244898e-05,
      "loss": 0.6546,
      "step": 2573
    },
    {
      "epoch": 1.098,
      "grad_norm": 2.571528911590576,
      "learning_rate": 2.4755102040816327e-05,
      "loss": 0.6165,
      "step": 2574
    },
    {
      "epoch": 1.0982,
      "grad_norm": 4.755832672119141,
      "learning_rate": 2.4744897959183673e-05,
      "loss": 2.2378,
      "step": 2575
    },
    {
      "epoch": 1.0984,
      "grad_norm": 4.332126617431641,
      "learning_rate": 2.4734693877551022e-05,
      "loss": 2.0653,
      "step": 2576
    },
    {
      "epoch": 1.0986,
      "grad_norm": 3.1858487129211426,
      "learning_rate": 2.4724489795918368e-05,
      "loss": 1.5829,
      "step": 2577
    },
    {
      "epoch": 1.0988,
      "grad_norm": 12.445663452148438,
      "learning_rate": 2.4714285714285714e-05,
      "loss": 1.4157,
      "step": 2578
    },
    {
      "epoch": 1.099,
      "grad_norm": 25.608064651489258,
      "learning_rate": 2.4704081632653063e-05,
      "loss": 3.2528,
      "step": 2579
    },
    {
      "epoch": 1.0992,
      "grad_norm": 1.2519673109054565,
      "learning_rate": 2.469387755102041e-05,
      "loss": 0.0214,
      "step": 2580
    },
    {
      "epoch": 1.0994,
      "grad_norm": 2.4915757179260254,
      "learning_rate": 2.4683673469387755e-05,
      "loss": 0.3216,
      "step": 2581
    },
    {
      "epoch": 1.0996,
      "grad_norm": 3.9431369304656982,
      "learning_rate": 2.46734693877551e-05,
      "loss": 0.5245,
      "step": 2582
    },
    {
      "epoch": 1.0998,
      "grad_norm": 1.9523743391036987,
      "learning_rate": 2.466326530612245e-05,
      "loss": 0.2015,
      "step": 2583
    },
    {
      "epoch": 1.1,
      "grad_norm": 2.3328518867492676,
      "learning_rate": 2.4653061224489795e-05,
      "loss": 0.2947,
      "step": 2584
    },
    {
      "epoch": 1.1002,
      "grad_norm": 4.120367050170898,
      "learning_rate": 2.4642857142857145e-05,
      "loss": 0.5547,
      "step": 2585
    },
    {
      "epoch": 1.1004,
      "grad_norm": 1.4598771333694458,
      "learning_rate": 2.4632653061224494e-05,
      "loss": 0.055,
      "step": 2586
    },
    {
      "epoch": 1.1006,
      "grad_norm": 2.297677755355835,
      "learning_rate": 2.462244897959184e-05,
      "loss": 0.262,
      "step": 2587
    },
    {
      "epoch": 1.1008,
      "grad_norm": 4.969046115875244,
      "learning_rate": 2.4612244897959185e-05,
      "loss": 0.6594,
      "step": 2588
    },
    {
      "epoch": 1.101,
      "grad_norm": 5.883965492248535,
      "learning_rate": 2.460204081632653e-05,
      "loss": 0.2786,
      "step": 2589
    },
    {
      "epoch": 1.1012,
      "grad_norm": 0.9458068609237671,
      "learning_rate": 2.459183673469388e-05,
      "loss": 0.02,
      "step": 2590
    },
    {
      "epoch": 1.1014,
      "grad_norm": 2.026034355163574,
      "learning_rate": 2.4581632653061226e-05,
      "loss": 0.1354,
      "step": 2591
    },
    {
      "epoch": 1.1016,
      "grad_norm": 5.589249134063721,
      "learning_rate": 2.4571428571428572e-05,
      "loss": 0.9564,
      "step": 2592
    },
    {
      "epoch": 1.1018,
      "grad_norm": 14.085370063781738,
      "learning_rate": 2.456122448979592e-05,
      "loss": 1.7646,
      "step": 2593
    },
    {
      "epoch": 1.102,
      "grad_norm": 28.355955123901367,
      "learning_rate": 2.4551020408163267e-05,
      "loss": 3.7831,
      "step": 2594
    },
    {
      "epoch": 1.1022,
      "grad_norm": 4.164278507232666,
      "learning_rate": 2.4540816326530613e-05,
      "loss": 0.0793,
      "step": 2595
    },
    {
      "epoch": 1.1024,
      "grad_norm": 5.802898406982422,
      "learning_rate": 2.4530612244897962e-05,
      "loss": 0.2683,
      "step": 2596
    },
    {
      "epoch": 1.1026,
      "grad_norm": 10.879895210266113,
      "learning_rate": 2.4520408163265308e-05,
      "loss": 1.8366,
      "step": 2597
    },
    {
      "epoch": 1.1028,
      "grad_norm": 10.431557655334473,
      "learning_rate": 2.4510204081632653e-05,
      "loss": 3.1627,
      "step": 2598
    },
    {
      "epoch": 1.103,
      "grad_norm": 1.2934068441390991,
      "learning_rate": 2.45e-05,
      "loss": 0.0489,
      "step": 2599
    },
    {
      "epoch": 1.1032,
      "grad_norm": 2.8279800415039062,
      "learning_rate": 2.448979591836735e-05,
      "loss": 0.134,
      "step": 2600
    },
    {
      "epoch": 1.1034,
      "grad_norm": 1.7171381711959839,
      "learning_rate": 2.4479591836734694e-05,
      "loss": 0.1996,
      "step": 2601
    },
    {
      "epoch": 1.1036,
      "grad_norm": 1.8562647104263306,
      "learning_rate": 2.446938775510204e-05,
      "loss": 0.1764,
      "step": 2602
    },
    {
      "epoch": 1.1038000000000001,
      "grad_norm": 1.6449317932128906,
      "learning_rate": 2.445918367346939e-05,
      "loss": 0.1458,
      "step": 2603
    },
    {
      "epoch": 1.104,
      "grad_norm": 1.1087433099746704,
      "learning_rate": 2.4448979591836735e-05,
      "loss": 0.0387,
      "step": 2604
    },
    {
      "epoch": 1.1042,
      "grad_norm": 1.929565191268921,
      "learning_rate": 2.443877551020408e-05,
      "loss": 0.2624,
      "step": 2605
    },
    {
      "epoch": 1.1044,
      "grad_norm": 5.939025402069092,
      "learning_rate": 2.442857142857143e-05,
      "loss": 0.7206,
      "step": 2606
    },
    {
      "epoch": 1.1046,
      "grad_norm": 14.030128479003906,
      "learning_rate": 2.441836734693878e-05,
      "loss": 1.3375,
      "step": 2607
    },
    {
      "epoch": 1.1048,
      "grad_norm": 2.9569661617279053,
      "learning_rate": 2.4408163265306125e-05,
      "loss": 0.6407,
      "step": 2608
    },
    {
      "epoch": 1.105,
      "grad_norm": 6.541789531707764,
      "learning_rate": 2.439795918367347e-05,
      "loss": 1.7242,
      "step": 2609
    },
    {
      "epoch": 1.1052,
      "grad_norm": 1.908613681793213,
      "learning_rate": 2.438775510204082e-05,
      "loss": 0.1024,
      "step": 2610
    },
    {
      "epoch": 1.1054,
      "grad_norm": 4.430586338043213,
      "learning_rate": 2.4377551020408166e-05,
      "loss": 0.9539,
      "step": 2611
    },
    {
      "epoch": 1.1056,
      "grad_norm": 10.48686408996582,
      "learning_rate": 2.436734693877551e-05,
      "loss": 3.0465,
      "step": 2612
    },
    {
      "epoch": 1.1058,
      "grad_norm": 2.074552297592163,
      "learning_rate": 2.4357142857142857e-05,
      "loss": 0.2299,
      "step": 2613
    },
    {
      "epoch": 1.106,
      "grad_norm": 3.4246790409088135,
      "learning_rate": 2.4346938775510206e-05,
      "loss": 0.4109,
      "step": 2614
    },
    {
      "epoch": 1.1062,
      "grad_norm": 11.392557144165039,
      "learning_rate": 2.4336734693877552e-05,
      "loss": 2.621,
      "step": 2615
    },
    {
      "epoch": 1.1064,
      "grad_norm": 17.0239200592041,
      "learning_rate": 2.4326530612244898e-05,
      "loss": 5.0606,
      "step": 2616
    },
    {
      "epoch": 1.1066,
      "grad_norm": 2.5136921405792236,
      "learning_rate": 2.4316326530612247e-05,
      "loss": 0.303,
      "step": 2617
    },
    {
      "epoch": 1.1068,
      "grad_norm": 4.009968280792236,
      "learning_rate": 2.4306122448979593e-05,
      "loss": 0.5713,
      "step": 2618
    },
    {
      "epoch": 1.107,
      "grad_norm": 2.3244709968566895,
      "learning_rate": 2.429591836734694e-05,
      "loss": 0.0982,
      "step": 2619
    },
    {
      "epoch": 1.1072,
      "grad_norm": 3.1784238815307617,
      "learning_rate": 2.4285714285714288e-05,
      "loss": 0.1382,
      "step": 2620
    },
    {
      "epoch": 1.1074,
      "grad_norm": 1.188054084777832,
      "learning_rate": 2.4275510204081634e-05,
      "loss": 0.042,
      "step": 2621
    },
    {
      "epoch": 1.1076,
      "grad_norm": 6.872524261474609,
      "learning_rate": 2.426530612244898e-05,
      "loss": 0.6016,
      "step": 2622
    },
    {
      "epoch": 1.1078000000000001,
      "grad_norm": 24.797069549560547,
      "learning_rate": 2.425510204081633e-05,
      "loss": 2.0936,
      "step": 2623
    },
    {
      "epoch": 1.108,
      "grad_norm": 1.7459396123886108,
      "learning_rate": 2.4244897959183674e-05,
      "loss": 0.1488,
      "step": 2624
    },
    {
      "epoch": 1.1082,
      "grad_norm": 1.2388994693756104,
      "learning_rate": 2.423469387755102e-05,
      "loss": 0.0466,
      "step": 2625
    },
    {
      "epoch": 1.1084,
      "grad_norm": 7.152106761932373,
      "learning_rate": 2.4224489795918366e-05,
      "loss": 1.1114,
      "step": 2626
    },
    {
      "epoch": 1.1086,
      "grad_norm": 14.875006675720215,
      "learning_rate": 2.4214285714285715e-05,
      "loss": 3.5813,
      "step": 2627
    },
    {
      "epoch": 1.1088,
      "grad_norm": 4.775994300842285,
      "learning_rate": 2.420408163265306e-05,
      "loss": 0.3872,
      "step": 2628
    },
    {
      "epoch": 1.109,
      "grad_norm": 12.121844291687012,
      "learning_rate": 2.4193877551020407e-05,
      "loss": 1.6924,
      "step": 2629
    },
    {
      "epoch": 1.1092,
      "grad_norm": 1.738283395767212,
      "learning_rate": 2.4183673469387756e-05,
      "loss": 0.1797,
      "step": 2630
    },
    {
      "epoch": 1.1094,
      "grad_norm": 1.5898929834365845,
      "learning_rate": 2.4173469387755105e-05,
      "loss": 0.1573,
      "step": 2631
    },
    {
      "epoch": 1.1096,
      "grad_norm": 1.715936541557312,
      "learning_rate": 2.416326530612245e-05,
      "loss": 0.1822,
      "step": 2632
    },
    {
      "epoch": 1.1098,
      "grad_norm": 1.7776638269424438,
      "learning_rate": 2.4153061224489797e-05,
      "loss": 0.1873,
      "step": 2633
    },
    {
      "epoch": 1.11,
      "grad_norm": 2.2790379524230957,
      "learning_rate": 2.4142857142857146e-05,
      "loss": 0.3086,
      "step": 2634
    },
    {
      "epoch": 1.1102,
      "grad_norm": 3.2239675521850586,
      "learning_rate": 2.4132653061224492e-05,
      "loss": 0.6393,
      "step": 2635
    },
    {
      "epoch": 1.1104,
      "grad_norm": 3.7233240604400635,
      "learning_rate": 2.4122448979591838e-05,
      "loss": 0.3281,
      "step": 2636
    },
    {
      "epoch": 1.1106,
      "grad_norm": 11.728107452392578,
      "learning_rate": 2.4112244897959187e-05,
      "loss": 1.6738,
      "step": 2637
    },
    {
      "epoch": 1.1108,
      "grad_norm": 12.237937927246094,
      "learning_rate": 2.4102040816326533e-05,
      "loss": 3.5454,
      "step": 2638
    },
    {
      "epoch": 1.111,
      "grad_norm": 1.3609619140625,
      "learning_rate": 2.409183673469388e-05,
      "loss": 0.0409,
      "step": 2639
    },
    {
      "epoch": 1.1112,
      "grad_norm": 2.414645195007324,
      "learning_rate": 2.4081632653061224e-05,
      "loss": 0.5846,
      "step": 2640
    },
    {
      "epoch": 1.1114,
      "grad_norm": 4.060914039611816,
      "learning_rate": 2.4071428571428573e-05,
      "loss": 1.8276,
      "step": 2641
    },
    {
      "epoch": 1.1116,
      "grad_norm": 3.0793774127960205,
      "learning_rate": 2.406122448979592e-05,
      "loss": 0.6324,
      "step": 2642
    },
    {
      "epoch": 1.1118,
      "grad_norm": 1.6483436822891235,
      "learning_rate": 2.4051020408163265e-05,
      "loss": 0.1275,
      "step": 2643
    },
    {
      "epoch": 1.112,
      "grad_norm": 1.8608285188674927,
      "learning_rate": 2.4040816326530614e-05,
      "loss": 0.2238,
      "step": 2644
    },
    {
      "epoch": 1.1122,
      "grad_norm": 3.8672969341278076,
      "learning_rate": 2.403061224489796e-05,
      "loss": 0.8668,
      "step": 2645
    },
    {
      "epoch": 1.1124,
      "grad_norm": 7.139001369476318,
      "learning_rate": 2.4020408163265306e-05,
      "loss": 2.4751,
      "step": 2646
    },
    {
      "epoch": 1.1126,
      "grad_norm": 2.5560009479522705,
      "learning_rate": 2.4010204081632655e-05,
      "loss": 0.2781,
      "step": 2647
    },
    {
      "epoch": 1.1128,
      "grad_norm": 13.595659255981445,
      "learning_rate": 2.4e-05,
      "loss": 1.1682,
      "step": 2648
    },
    {
      "epoch": 1.113,
      "grad_norm": 16.68004035949707,
      "learning_rate": 2.3989795918367346e-05,
      "loss": 2.1596,
      "step": 2649
    },
    {
      "epoch": 1.1132,
      "grad_norm": 1.8965356349945068,
      "learning_rate": 2.3979591836734696e-05,
      "loss": 0.2235,
      "step": 2650
    },
    {
      "epoch": 1.1134,
      "grad_norm": 1.8255671262741089,
      "learning_rate": 2.396938775510204e-05,
      "loss": 0.1737,
      "step": 2651
    },
    {
      "epoch": 1.1136,
      "grad_norm": 5.417263031005859,
      "learning_rate": 2.3959183673469387e-05,
      "loss": 0.3718,
      "step": 2652
    },
    {
      "epoch": 1.1138,
      "grad_norm": 7.282905578613281,
      "learning_rate": 2.3948979591836736e-05,
      "loss": 1.4317,
      "step": 2653
    },
    {
      "epoch": 1.114,
      "grad_norm": 1.0953459739685059,
      "learning_rate": 2.3938775510204086e-05,
      "loss": 0.0445,
      "step": 2654
    },
    {
      "epoch": 1.1142,
      "grad_norm": 1.5249688625335693,
      "learning_rate": 2.392857142857143e-05,
      "loss": 0.0532,
      "step": 2655
    },
    {
      "epoch": 1.1144,
      "grad_norm": 3.1003592014312744,
      "learning_rate": 2.3918367346938777e-05,
      "loss": 0.3299,
      "step": 2656
    },
    {
      "epoch": 1.1146,
      "grad_norm": 7.614220142364502,
      "learning_rate": 2.3908163265306123e-05,
      "loss": 2.4333,
      "step": 2657
    },
    {
      "epoch": 1.1148,
      "grad_norm": 23.154939651489258,
      "learning_rate": 2.3897959183673472e-05,
      "loss": 3.3009,
      "step": 2658
    },
    {
      "epoch": 1.115,
      "grad_norm": 6.271807670593262,
      "learning_rate": 2.3887755102040818e-05,
      "loss": 0.8004,
      "step": 2659
    },
    {
      "epoch": 1.1152,
      "grad_norm": 8.839851379394531,
      "learning_rate": 2.3877551020408164e-05,
      "loss": 0.7816,
      "step": 2660
    },
    {
      "epoch": 1.1154,
      "grad_norm": 2.4665706157684326,
      "learning_rate": 2.3867346938775513e-05,
      "loss": 0.1041,
      "step": 2661
    },
    {
      "epoch": 1.1156,
      "grad_norm": 1.2891920804977417,
      "learning_rate": 2.385714285714286e-05,
      "loss": 0.053,
      "step": 2662
    },
    {
      "epoch": 1.1158,
      "grad_norm": 5.995043754577637,
      "learning_rate": 2.3846938775510204e-05,
      "loss": 1.0304,
      "step": 2663
    },
    {
      "epoch": 1.116,
      "grad_norm": 9.764463424682617,
      "learning_rate": 2.3836734693877554e-05,
      "loss": 2.591,
      "step": 2664
    },
    {
      "epoch": 1.1162,
      "grad_norm": 1.9951735734939575,
      "learning_rate": 2.38265306122449e-05,
      "loss": 0.2677,
      "step": 2665
    },
    {
      "epoch": 1.1164,
      "grad_norm": 3.084277629852295,
      "learning_rate": 2.3816326530612245e-05,
      "loss": 0.6408,
      "step": 2666
    },
    {
      "epoch": 1.1166,
      "grad_norm": 1.1833219528198242,
      "learning_rate": 2.380612244897959e-05,
      "loss": 0.0451,
      "step": 2667
    },
    {
      "epoch": 1.1168,
      "grad_norm": 2.448613405227661,
      "learning_rate": 2.379591836734694e-05,
      "loss": 0.6766,
      "step": 2668
    },
    {
      "epoch": 1.117,
      "grad_norm": 3.4040493965148926,
      "learning_rate": 2.3785714285714286e-05,
      "loss": 1.5933,
      "step": 2669
    },
    {
      "epoch": 1.1172,
      "grad_norm": 14.958999633789062,
      "learning_rate": 2.3775510204081632e-05,
      "loss": 0.8691,
      "step": 2670
    },
    {
      "epoch": 1.1174,
      "grad_norm": 3.2945313453674316,
      "learning_rate": 2.376530612244898e-05,
      "loss": 0.3583,
      "step": 2671
    },
    {
      "epoch": 1.1176,
      "grad_norm": 5.522057056427002,
      "learning_rate": 2.3755102040816327e-05,
      "loss": 1.0304,
      "step": 2672
    },
    {
      "epoch": 1.1178,
      "grad_norm": 8.773022651672363,
      "learning_rate": 2.3744897959183673e-05,
      "loss": 2.6604,
      "step": 2673
    },
    {
      "epoch": 1.1179999999999999,
      "grad_norm": 1.739574670791626,
      "learning_rate": 2.373469387755102e-05,
      "loss": 0.2143,
      "step": 2674
    },
    {
      "epoch": 1.1182,
      "grad_norm": 2.3119235038757324,
      "learning_rate": 2.372448979591837e-05,
      "loss": 0.3143,
      "step": 2675
    },
    {
      "epoch": 1.1184,
      "grad_norm": 3.644256591796875,
      "learning_rate": 2.3714285714285717e-05,
      "loss": 0.1712,
      "step": 2676
    },
    {
      "epoch": 1.1186,
      "grad_norm": 1.1288031339645386,
      "learning_rate": 2.3704081632653062e-05,
      "loss": 0.0345,
      "step": 2677
    },
    {
      "epoch": 1.1188,
      "grad_norm": 1.857331395149231,
      "learning_rate": 2.369387755102041e-05,
      "loss": 0.2174,
      "step": 2678
    },
    {
      "epoch": 1.119,
      "grad_norm": 2.1009767055511475,
      "learning_rate": 2.3683673469387757e-05,
      "loss": 0.2082,
      "step": 2679
    },
    {
      "epoch": 1.1192,
      "grad_norm": 2.547800302505493,
      "learning_rate": 2.3673469387755103e-05,
      "loss": 0.2913,
      "step": 2680
    },
    {
      "epoch": 1.1194,
      "grad_norm": 4.963531494140625,
      "learning_rate": 2.3663265306122452e-05,
      "loss": 0.7662,
      "step": 2681
    },
    {
      "epoch": 1.1196,
      "grad_norm": 4.640702247619629,
      "learning_rate": 2.3653061224489798e-05,
      "loss": 0.77,
      "step": 2682
    },
    {
      "epoch": 1.1198,
      "grad_norm": 3.045327663421631,
      "learning_rate": 2.3642857142857144e-05,
      "loss": 1.4742,
      "step": 2683
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.779814720153809,
      "learning_rate": 2.363265306122449e-05,
      "loss": 0.9135,
      "step": 2684
    },
    {
      "epoch": 1.1202,
      "grad_norm": 17.087533950805664,
      "learning_rate": 2.362244897959184e-05,
      "loss": 3.5725,
      "step": 2685
    },
    {
      "epoch": 1.1204,
      "grad_norm": 1.8048440217971802,
      "learning_rate": 2.3612244897959185e-05,
      "loss": 0.1151,
      "step": 2686
    },
    {
      "epoch": 1.1206,
      "grad_norm": 4.408398151397705,
      "learning_rate": 2.360204081632653e-05,
      "loss": 0.4616,
      "step": 2687
    },
    {
      "epoch": 1.1208,
      "grad_norm": 4.695296287536621,
      "learning_rate": 2.359183673469388e-05,
      "loss": 1.5793,
      "step": 2688
    },
    {
      "epoch": 1.121,
      "grad_norm": 12.745363235473633,
      "learning_rate": 2.3581632653061226e-05,
      "loss": 3.0204,
      "step": 2689
    },
    {
      "epoch": 1.1212,
      "grad_norm": 10.47659969329834,
      "learning_rate": 2.357142857142857e-05,
      "loss": 0.5482,
      "step": 2690
    },
    {
      "epoch": 1.1214,
      "grad_norm": 9.649550437927246,
      "learning_rate": 2.356122448979592e-05,
      "loss": 1.2656,
      "step": 2691
    },
    {
      "epoch": 1.1216,
      "grad_norm": 2.9947545528411865,
      "learning_rate": 2.3551020408163266e-05,
      "loss": 0.9884,
      "step": 2692
    },
    {
      "epoch": 1.1218,
      "grad_norm": 8.209904670715332,
      "learning_rate": 2.3540816326530612e-05,
      "loss": 3.1175,
      "step": 2693
    },
    {
      "epoch": 1.1219999999999999,
      "grad_norm": 5.573894023895264,
      "learning_rate": 2.3530612244897958e-05,
      "loss": 0.7264,
      "step": 2694
    },
    {
      "epoch": 1.1222,
      "grad_norm": 3.8499183654785156,
      "learning_rate": 2.3520408163265307e-05,
      "loss": 0.1304,
      "step": 2695
    },
    {
      "epoch": 1.1224,
      "grad_norm": 10.379673957824707,
      "learning_rate": 2.3510204081632653e-05,
      "loss": 1.1817,
      "step": 2696
    },
    {
      "epoch": 1.1226,
      "grad_norm": 10.155116081237793,
      "learning_rate": 2.35e-05,
      "loss": 1.4249,
      "step": 2697
    },
    {
      "epoch": 1.1228,
      "grad_norm": 4.0061163902282715,
      "learning_rate": 2.3489795918367348e-05,
      "loss": 0.3103,
      "step": 2698
    },
    {
      "epoch": 1.123,
      "grad_norm": 4.94691801071167,
      "learning_rate": 2.3479591836734697e-05,
      "loss": 0.797,
      "step": 2699
    },
    {
      "epoch": 1.1232,
      "grad_norm": 1.813581943511963,
      "learning_rate": 2.3469387755102043e-05,
      "loss": 0.2344,
      "step": 2700
    },
    {
      "epoch": 1.1234,
      "grad_norm": 1.988603949546814,
      "learning_rate": 2.345918367346939e-05,
      "loss": 0.228,
      "step": 2701
    },
    {
      "epoch": 1.1236,
      "grad_norm": 3.084177255630493,
      "learning_rate": 2.3448979591836738e-05,
      "loss": 0.3087,
      "step": 2702
    },
    {
      "epoch": 1.1238,
      "grad_norm": 3.0632259845733643,
      "learning_rate": 2.3438775510204084e-05,
      "loss": 0.5855,
      "step": 2703
    },
    {
      "epoch": 1.124,
      "grad_norm": 2.8814468383789062,
      "learning_rate": 2.342857142857143e-05,
      "loss": 0.2033,
      "step": 2704
    },
    {
      "epoch": 1.1242,
      "grad_norm": 3.1632778644561768,
      "learning_rate": 2.341836734693878e-05,
      "loss": 0.5382,
      "step": 2705
    },
    {
      "epoch": 1.1244,
      "grad_norm": 1.307254672050476,
      "learning_rate": 2.3408163265306124e-05,
      "loss": 0.0582,
      "step": 2706
    },
    {
      "epoch": 1.1246,
      "grad_norm": 3.011153221130371,
      "learning_rate": 2.339795918367347e-05,
      "loss": 0.1373,
      "step": 2707
    },
    {
      "epoch": 1.1248,
      "grad_norm": 2.2650563716888428,
      "learning_rate": 2.3387755102040816e-05,
      "loss": 0.286,
      "step": 2708
    },
    {
      "epoch": 1.125,
      "grad_norm": 6.4699296951293945,
      "learning_rate": 2.3377551020408165e-05,
      "loss": 0.7095,
      "step": 2709
    },
    {
      "epoch": 1.1252,
      "grad_norm": 7.103350639343262,
      "learning_rate": 2.336734693877551e-05,
      "loss": 1.4044,
      "step": 2710
    },
    {
      "epoch": 1.1254,
      "grad_norm": 2.83284330368042,
      "learning_rate": 2.3357142857142857e-05,
      "loss": 0.1647,
      "step": 2711
    },
    {
      "epoch": 1.1256,
      "grad_norm": 6.117814064025879,
      "learning_rate": 2.3346938775510206e-05,
      "loss": 0.9263,
      "step": 2712
    },
    {
      "epoch": 1.1258,
      "grad_norm": 11.084607124328613,
      "learning_rate": 2.333673469387755e-05,
      "loss": 1.686,
      "step": 2713
    },
    {
      "epoch": 1.126,
      "grad_norm": 27.052371978759766,
      "learning_rate": 2.3326530612244897e-05,
      "loss": 4.1773,
      "step": 2714
    },
    {
      "epoch": 1.1262,
      "grad_norm": 10.01918888092041,
      "learning_rate": 2.3316326530612247e-05,
      "loss": 1.4462,
      "step": 2715
    },
    {
      "epoch": 1.1264,
      "grad_norm": 2.8444294929504395,
      "learning_rate": 2.3306122448979592e-05,
      "loss": 0.5133,
      "step": 2716
    },
    {
      "epoch": 1.1266,
      "grad_norm": 3.0026142597198486,
      "learning_rate": 2.3295918367346938e-05,
      "loss": 0.286,
      "step": 2717
    },
    {
      "epoch": 1.1268,
      "grad_norm": 5.488697528839111,
      "learning_rate": 2.3285714285714287e-05,
      "loss": 0.8064,
      "step": 2718
    },
    {
      "epoch": 1.127,
      "grad_norm": 2.136697292327881,
      "learning_rate": 2.3275510204081633e-05,
      "loss": 0.1877,
      "step": 2719
    },
    {
      "epoch": 1.1272,
      "grad_norm": 3.03145432472229,
      "learning_rate": 2.326530612244898e-05,
      "loss": 0.3336,
      "step": 2720
    },
    {
      "epoch": 1.1274,
      "grad_norm": 1.7051206827163696,
      "learning_rate": 2.3255102040816328e-05,
      "loss": 0.22,
      "step": 2721
    },
    {
      "epoch": 1.1276,
      "grad_norm": 1.7277824878692627,
      "learning_rate": 2.3244897959183677e-05,
      "loss": 0.1903,
      "step": 2722
    },
    {
      "epoch": 1.1278,
      "grad_norm": 2.966791868209839,
      "learning_rate": 2.3234693877551023e-05,
      "loss": 0.5597,
      "step": 2723
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 1.6014727354049683,
      "learning_rate": 2.322448979591837e-05,
      "loss": 0.0624,
      "step": 2724
    },
    {
      "epoch": 1.1282,
      "grad_norm": 2.1757612228393555,
      "learning_rate": 2.3214285714285715e-05,
      "loss": 0.1369,
      "step": 2725
    },
    {
      "epoch": 1.1284,
      "grad_norm": 1.3579509258270264,
      "learning_rate": 2.3204081632653064e-05,
      "loss": 0.0576,
      "step": 2726
    },
    {
      "epoch": 1.1286,
      "grad_norm": 1.526038408279419,
      "learning_rate": 2.319387755102041e-05,
      "loss": 0.134,
      "step": 2727
    },
    {
      "epoch": 1.1288,
      "grad_norm": 1.9145957231521606,
      "learning_rate": 2.3183673469387755e-05,
      "loss": 0.2593,
      "step": 2728
    },
    {
      "epoch": 1.129,
      "grad_norm": 9.260457992553711,
      "learning_rate": 2.3173469387755105e-05,
      "loss": 1.0714,
      "step": 2729
    },
    {
      "epoch": 1.1292,
      "grad_norm": 29.887434005737305,
      "learning_rate": 2.316326530612245e-05,
      "loss": 3.3454,
      "step": 2730
    },
    {
      "epoch": 1.1294,
      "grad_norm": 13.774479866027832,
      "learning_rate": 2.3153061224489796e-05,
      "loss": 3.6583,
      "step": 2731
    },
    {
      "epoch": 1.1296,
      "grad_norm": 6.605763912200928,
      "learning_rate": 2.3142857142857145e-05,
      "loss": 0.8855,
      "step": 2732
    },
    {
      "epoch": 1.1298,
      "grad_norm": 4.741060733795166,
      "learning_rate": 2.313265306122449e-05,
      "loss": 0.7902,
      "step": 2733
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.5933618545532227,
      "learning_rate": 2.3122448979591837e-05,
      "loss": 0.1623,
      "step": 2734
    },
    {
      "epoch": 1.1302,
      "grad_norm": 1.6804769039154053,
      "learning_rate": 2.3112244897959183e-05,
      "loss": 0.1537,
      "step": 2735
    },
    {
      "epoch": 1.1304,
      "grad_norm": 2.6978795528411865,
      "learning_rate": 2.3102040816326532e-05,
      "loss": 0.4404,
      "step": 2736
    },
    {
      "epoch": 1.1306,
      "grad_norm": 5.578824043273926,
      "learning_rate": 2.3091836734693878e-05,
      "loss": 2.2101,
      "step": 2737
    },
    {
      "epoch": 1.1308,
      "grad_norm": 5.527901649475098,
      "learning_rate": 2.3081632653061224e-05,
      "loss": 1.5192,
      "step": 2738
    },
    {
      "epoch": 1.131,
      "grad_norm": 3.8173131942749023,
      "learning_rate": 2.3071428571428573e-05,
      "loss": 0.7459,
      "step": 2739
    },
    {
      "epoch": 1.1312,
      "grad_norm": 5.779356479644775,
      "learning_rate": 2.306122448979592e-05,
      "loss": 0.1954,
      "step": 2740
    },
    {
      "epoch": 1.1314,
      "grad_norm": 7.4463701248168945,
      "learning_rate": 2.3051020408163264e-05,
      "loss": 1.1728,
      "step": 2741
    },
    {
      "epoch": 1.1316,
      "grad_norm": 5.37211275100708,
      "learning_rate": 2.3040816326530613e-05,
      "loss": 0.6996,
      "step": 2742
    },
    {
      "epoch": 1.1318,
      "grad_norm": 2.5998642444610596,
      "learning_rate": 2.3030612244897963e-05,
      "loss": 0.3492,
      "step": 2743
    },
    {
      "epoch": 1.1320000000000001,
      "grad_norm": 5.17749547958374,
      "learning_rate": 2.302040816326531e-05,
      "loss": 1.4707,
      "step": 2744
    },
    {
      "epoch": 1.1322,
      "grad_norm": 1.720008134841919,
      "learning_rate": 2.3010204081632654e-05,
      "loss": 0.1492,
      "step": 2745
    },
    {
      "epoch": 1.1324,
      "grad_norm": 4.889779567718506,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.9116,
      "step": 2746
    },
    {
      "epoch": 1.1326,
      "grad_norm": 9.019208908081055,
      "learning_rate": 2.298979591836735e-05,
      "loss": 3.042,
      "step": 2747
    },
    {
      "epoch": 1.1328,
      "grad_norm": 3.348464250564575,
      "learning_rate": 2.2979591836734695e-05,
      "loss": 1.6472,
      "step": 2748
    },
    {
      "epoch": 1.133,
      "grad_norm": 4.644619941711426,
      "learning_rate": 2.2969387755102044e-05,
      "loss": 0.3318,
      "step": 2749
    },
    {
      "epoch": 1.1332,
      "grad_norm": 6.385184288024902,
      "learning_rate": 2.295918367346939e-05,
      "loss": 0.817,
      "step": 2750
    },
    {
      "epoch": 1.1334,
      "grad_norm": 3.095552682876587,
      "learning_rate": 2.2948979591836736e-05,
      "loss": 0.5922,
      "step": 2751
    },
    {
      "epoch": 1.1336,
      "grad_norm": 1.7094886302947998,
      "learning_rate": 2.293877551020408e-05,
      "loss": 0.2121,
      "step": 2752
    },
    {
      "epoch": 1.1338,
      "grad_norm": 3.2444093227386475,
      "learning_rate": 2.292857142857143e-05,
      "loss": 0.5553,
      "step": 2753
    },
    {
      "epoch": 1.134,
      "grad_norm": 1.6598012447357178,
      "learning_rate": 2.2918367346938777e-05,
      "loss": 0.0689,
      "step": 2754
    },
    {
      "epoch": 1.1342,
      "grad_norm": 1.6716395616531372,
      "learning_rate": 2.2908163265306122e-05,
      "loss": 0.1675,
      "step": 2755
    },
    {
      "epoch": 1.1344,
      "grad_norm": 1.6261208057403564,
      "learning_rate": 2.289795918367347e-05,
      "loss": 0.1817,
      "step": 2756
    },
    {
      "epoch": 1.1346,
      "grad_norm": 2.096707582473755,
      "learning_rate": 2.2887755102040817e-05,
      "loss": 0.1242,
      "step": 2757
    },
    {
      "epoch": 1.1348,
      "grad_norm": 5.705234527587891,
      "learning_rate": 2.2877551020408163e-05,
      "loss": 0.7205,
      "step": 2758
    },
    {
      "epoch": 1.135,
      "grad_norm": 2.8146116733551025,
      "learning_rate": 2.2867346938775512e-05,
      "loss": 0.3474,
      "step": 2759
    },
    {
      "epoch": 1.1352,
      "grad_norm": 3.8745510578155518,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 0.3032,
      "step": 2760
    },
    {
      "epoch": 1.1354,
      "grad_norm": 1.6446934938430786,
      "learning_rate": 2.2846938775510204e-05,
      "loss": 0.1739,
      "step": 2761
    },
    {
      "epoch": 1.1356,
      "grad_norm": 1.8880884647369385,
      "learning_rate": 2.283673469387755e-05,
      "loss": 0.0904,
      "step": 2762
    },
    {
      "epoch": 1.1358,
      "grad_norm": 2.831763744354248,
      "learning_rate": 2.28265306122449e-05,
      "loss": 0.3695,
      "step": 2763
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 2.0684707164764404,
      "learning_rate": 2.2816326530612245e-05,
      "loss": 0.147,
      "step": 2764
    },
    {
      "epoch": 1.1362,
      "grad_norm": 20.502220153808594,
      "learning_rate": 2.280612244897959e-05,
      "loss": 1.2422,
      "step": 2765
    },
    {
      "epoch": 1.1364,
      "grad_norm": 5.634237289428711,
      "learning_rate": 2.279591836734694e-05,
      "loss": 0.7513,
      "step": 2766
    },
    {
      "epoch": 1.1366,
      "grad_norm": 1.6980818510055542,
      "learning_rate": 2.278571428571429e-05,
      "loss": 0.1942,
      "step": 2767
    },
    {
      "epoch": 1.1368,
      "grad_norm": 1.4760496616363525,
      "learning_rate": 2.2775510204081635e-05,
      "loss": 0.0931,
      "step": 2768
    },
    {
      "epoch": 1.137,
      "grad_norm": 2.79160475730896,
      "learning_rate": 2.276530612244898e-05,
      "loss": 0.5251,
      "step": 2769
    },
    {
      "epoch": 1.1372,
      "grad_norm": 2.9664740562438965,
      "learning_rate": 2.275510204081633e-05,
      "loss": 0.164,
      "step": 2770
    },
    {
      "epoch": 1.1374,
      "grad_norm": 1.9510027170181274,
      "learning_rate": 2.2744897959183675e-05,
      "loss": 0.219,
      "step": 2771
    },
    {
      "epoch": 1.1376,
      "grad_norm": 1.5220974683761597,
      "learning_rate": 2.273469387755102e-05,
      "loss": 0.1522,
      "step": 2772
    },
    {
      "epoch": 1.1378,
      "grad_norm": 0.9969020485877991,
      "learning_rate": 2.272448979591837e-05,
      "loss": 0.0287,
      "step": 2773
    },
    {
      "epoch": 1.138,
      "grad_norm": 1.9152190685272217,
      "learning_rate": 2.2714285714285716e-05,
      "loss": 0.2731,
      "step": 2774
    },
    {
      "epoch": 1.1381999999999999,
      "grad_norm": 1.6108945608139038,
      "learning_rate": 2.2704081632653062e-05,
      "loss": 0.1778,
      "step": 2775
    },
    {
      "epoch": 1.1384,
      "grad_norm": 2.6869864463806152,
      "learning_rate": 2.269387755102041e-05,
      "loss": 0.306,
      "step": 2776
    },
    {
      "epoch": 1.1386,
      "grad_norm": 5.675705909729004,
      "learning_rate": 2.2683673469387757e-05,
      "loss": 0.3161,
      "step": 2777
    },
    {
      "epoch": 1.1388,
      "grad_norm": 0.9622205495834351,
      "learning_rate": 2.2673469387755103e-05,
      "loss": 0.0312,
      "step": 2778
    },
    {
      "epoch": 1.139,
      "grad_norm": 1.7289884090423584,
      "learning_rate": 2.266326530612245e-05,
      "loss": 0.2262,
      "step": 2779
    },
    {
      "epoch": 1.1392,
      "grad_norm": 2.9984047412872314,
      "learning_rate": 2.2653061224489798e-05,
      "loss": 0.4141,
      "step": 2780
    },
    {
      "epoch": 1.1394,
      "grad_norm": 6.064243793487549,
      "learning_rate": 2.2642857142857143e-05,
      "loss": 0.4489,
      "step": 2781
    },
    {
      "epoch": 1.1396,
      "grad_norm": 6.7196364402771,
      "learning_rate": 2.263265306122449e-05,
      "loss": 0.786,
      "step": 2782
    },
    {
      "epoch": 1.1398,
      "grad_norm": 7.996720790863037,
      "learning_rate": 2.262244897959184e-05,
      "loss": 1.0503,
      "step": 2783
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 20.268320083618164,
      "learning_rate": 2.2612244897959184e-05,
      "loss": 3.538,
      "step": 2784
    },
    {
      "epoch": 1.1402,
      "grad_norm": 1.163985013961792,
      "learning_rate": 2.260204081632653e-05,
      "loss": 0.0461,
      "step": 2785
    },
    {
      "epoch": 1.1404,
      "grad_norm": 0.9983113408088684,
      "learning_rate": 2.259183673469388e-05,
      "loss": 0.0153,
      "step": 2786
    },
    {
      "epoch": 1.1406,
      "grad_norm": 1.6703228950500488,
      "learning_rate": 2.2581632653061225e-05,
      "loss": 0.1554,
      "step": 2787
    },
    {
      "epoch": 1.1408,
      "grad_norm": 1.7943878173828125,
      "learning_rate": 2.257142857142857e-05,
      "loss": 0.1643,
      "step": 2788
    },
    {
      "epoch": 1.141,
      "grad_norm": 1.7003551721572876,
      "learning_rate": 2.256122448979592e-05,
      "loss": 0.0332,
      "step": 2789
    },
    {
      "epoch": 1.1412,
      "grad_norm": 2.893758773803711,
      "learning_rate": 2.255102040816327e-05,
      "loss": 0.08,
      "step": 2790
    },
    {
      "epoch": 1.1414,
      "grad_norm": 10.646772384643555,
      "learning_rate": 2.2540816326530615e-05,
      "loss": 0.3478,
      "step": 2791
    },
    {
      "epoch": 1.1416,
      "grad_norm": 4.976737022399902,
      "learning_rate": 2.253061224489796e-05,
      "loss": 0.7227,
      "step": 2792
    },
    {
      "epoch": 1.1418,
      "grad_norm": 1.5799471139907837,
      "learning_rate": 2.2520408163265306e-05,
      "loss": 0.0398,
      "step": 2793
    },
    {
      "epoch": 1.142,
      "grad_norm": 4.041989803314209,
      "learning_rate": 2.2510204081632656e-05,
      "loss": 0.3589,
      "step": 2794
    },
    {
      "epoch": 1.1421999999999999,
      "grad_norm": 6.398179054260254,
      "learning_rate": 2.25e-05,
      "loss": 1.5622,
      "step": 2795
    },
    {
      "epoch": 1.1424,
      "grad_norm": 3.265444040298462,
      "learning_rate": 2.2489795918367347e-05,
      "loss": 0.3875,
      "step": 2796
    },
    {
      "epoch": 1.1426,
      "grad_norm": 1.83922278881073,
      "learning_rate": 2.2479591836734696e-05,
      "loss": 0.1368,
      "step": 2797
    },
    {
      "epoch": 1.1428,
      "grad_norm": 4.628585338592529,
      "learning_rate": 2.2469387755102042e-05,
      "loss": 0.7204,
      "step": 2798
    },
    {
      "epoch": 1.143,
      "grad_norm": 7.918242454528809,
      "learning_rate": 2.2459183673469388e-05,
      "loss": 0.4492,
      "step": 2799
    },
    {
      "epoch": 1.1432,
      "grad_norm": 1.9701615571975708,
      "learning_rate": 2.2448979591836737e-05,
      "loss": 0.2359,
      "step": 2800
    },
    {
      "epoch": 1.1434,
      "grad_norm": 5.108238220214844,
      "learning_rate": 2.2438775510204083e-05,
      "loss": 0.9741,
      "step": 2801
    },
    {
      "epoch": 1.1436,
      "grad_norm": 7.751604080200195,
      "learning_rate": 2.242857142857143e-05,
      "loss": 2.4672,
      "step": 2802
    },
    {
      "epoch": 1.1438,
      "grad_norm": 6.921794414520264,
      "learning_rate": 2.2418367346938775e-05,
      "loss": 0.8374,
      "step": 2803
    },
    {
      "epoch": 1.144,
      "grad_norm": 4.900245189666748,
      "learning_rate": 2.2408163265306124e-05,
      "loss": 0.6811,
      "step": 2804
    },
    {
      "epoch": 1.1442,
      "grad_norm": 2.111574172973633,
      "learning_rate": 2.239795918367347e-05,
      "loss": 0.2138,
      "step": 2805
    },
    {
      "epoch": 1.1444,
      "grad_norm": 2.1236510276794434,
      "learning_rate": 2.2387755102040815e-05,
      "loss": 0.527,
      "step": 2806
    },
    {
      "epoch": 1.1446,
      "grad_norm": 1.9605672359466553,
      "learning_rate": 2.2377551020408164e-05,
      "loss": 0.3985,
      "step": 2807
    },
    {
      "epoch": 1.1448,
      "grad_norm": 7.262815475463867,
      "learning_rate": 2.236734693877551e-05,
      "loss": 2.0073,
      "step": 2808
    },
    {
      "epoch": 1.145,
      "grad_norm": 10.495835304260254,
      "learning_rate": 2.2357142857142856e-05,
      "loss": 1.4147,
      "step": 2809
    },
    {
      "epoch": 1.1452,
      "grad_norm": 2.1620757579803467,
      "learning_rate": 2.2346938775510205e-05,
      "loss": 0.5296,
      "step": 2810
    },
    {
      "epoch": 1.1454,
      "grad_norm": 1.7605044841766357,
      "learning_rate": 2.2336734693877554e-05,
      "loss": 0.1249,
      "step": 2811
    },
    {
      "epoch": 1.1456,
      "grad_norm": 2.030941963195801,
      "learning_rate": 2.23265306122449e-05,
      "loss": 0.1877,
      "step": 2812
    },
    {
      "epoch": 1.1458,
      "grad_norm": 1.8107985258102417,
      "learning_rate": 2.2316326530612246e-05,
      "loss": 0.1866,
      "step": 2813
    },
    {
      "epoch": 1.146,
      "grad_norm": 2.911255121231079,
      "learning_rate": 2.2306122448979595e-05,
      "loss": 0.0739,
      "step": 2814
    },
    {
      "epoch": 1.1461999999999999,
      "grad_norm": 1.6085364818572998,
      "learning_rate": 2.229591836734694e-05,
      "loss": 0.0729,
      "step": 2815
    },
    {
      "epoch": 1.1464,
      "grad_norm": 7.5536699295043945,
      "learning_rate": 2.2285714285714287e-05,
      "loss": 0.4134,
      "step": 2816
    },
    {
      "epoch": 1.1466,
      "grad_norm": 24.68170928955078,
      "learning_rate": 2.2275510204081636e-05,
      "loss": 2.1009,
      "step": 2817
    },
    {
      "epoch": 1.1468,
      "grad_norm": 15.094805717468262,
      "learning_rate": 2.2265306122448982e-05,
      "loss": 1.6492,
      "step": 2818
    },
    {
      "epoch": 1.147,
      "grad_norm": 4.507059097290039,
      "learning_rate": 2.2255102040816328e-05,
      "loss": 0.3622,
      "step": 2819
    },
    {
      "epoch": 1.1472,
      "grad_norm": 2.311285972595215,
      "learning_rate": 2.2244897959183673e-05,
      "loss": 0.3124,
      "step": 2820
    },
    {
      "epoch": 1.1474,
      "grad_norm": 4.111201286315918,
      "learning_rate": 2.2234693877551022e-05,
      "loss": 0.9661,
      "step": 2821
    },
    {
      "epoch": 1.1476,
      "grad_norm": 2.7623486518859863,
      "learning_rate": 2.2224489795918368e-05,
      "loss": 0.5781,
      "step": 2822
    },
    {
      "epoch": 1.1478,
      "grad_norm": 2.8520424365997314,
      "learning_rate": 2.2214285714285714e-05,
      "loss": 0.3171,
      "step": 2823
    },
    {
      "epoch": 1.148,
      "grad_norm": 1.5777324438095093,
      "learning_rate": 2.2204081632653063e-05,
      "loss": 0.1292,
      "step": 2824
    },
    {
      "epoch": 1.1482,
      "grad_norm": 1.6323140859603882,
      "learning_rate": 2.219387755102041e-05,
      "loss": 0.1628,
      "step": 2825
    },
    {
      "epoch": 1.1484,
      "grad_norm": 0.8256015777587891,
      "learning_rate": 2.2183673469387755e-05,
      "loss": 0.0198,
      "step": 2826
    },
    {
      "epoch": 1.1486,
      "grad_norm": 1.7223306894302368,
      "learning_rate": 2.2173469387755104e-05,
      "loss": 0.1819,
      "step": 2827
    },
    {
      "epoch": 1.1488,
      "grad_norm": 2.447789430618286,
      "learning_rate": 2.216326530612245e-05,
      "loss": 0.3057,
      "step": 2828
    },
    {
      "epoch": 1.149,
      "grad_norm": 5.144397735595703,
      "learning_rate": 2.2153061224489796e-05,
      "loss": 0.7765,
      "step": 2829
    },
    {
      "epoch": 1.1492,
      "grad_norm": 4.504195690155029,
      "learning_rate": 2.214285714285714e-05,
      "loss": 0.251,
      "step": 2830
    },
    {
      "epoch": 1.1494,
      "grad_norm": 7.414604663848877,
      "learning_rate": 2.213265306122449e-05,
      "loss": 1.3508,
      "step": 2831
    },
    {
      "epoch": 1.1496,
      "grad_norm": 5.837427139282227,
      "learning_rate": 2.2122448979591836e-05,
      "loss": 0.196,
      "step": 2832
    },
    {
      "epoch": 1.1498,
      "grad_norm": 7.726154327392578,
      "learning_rate": 2.2112244897959182e-05,
      "loss": 0.3343,
      "step": 2833
    },
    {
      "epoch": 1.15,
      "grad_norm": 2.2711503505706787,
      "learning_rate": 2.210204081632653e-05,
      "loss": 0.1605,
      "step": 2834
    },
    {
      "epoch": 1.1502,
      "grad_norm": 2.346430778503418,
      "learning_rate": 2.209183673469388e-05,
      "loss": 0.2132,
      "step": 2835
    },
    {
      "epoch": 1.1504,
      "grad_norm": 4.680408477783203,
      "learning_rate": 2.2081632653061226e-05,
      "loss": 0.7817,
      "step": 2836
    },
    {
      "epoch": 1.1506,
      "grad_norm": 2.951364755630493,
      "learning_rate": 2.2071428571428572e-05,
      "loss": 0.5717,
      "step": 2837
    },
    {
      "epoch": 1.1508,
      "grad_norm": 1.5978069305419922,
      "learning_rate": 2.206122448979592e-05,
      "loss": 0.161,
      "step": 2838
    },
    {
      "epoch": 1.151,
      "grad_norm": 2.3276004791259766,
      "learning_rate": 2.2051020408163267e-05,
      "loss": 0.3314,
      "step": 2839
    },
    {
      "epoch": 1.1512,
      "grad_norm": 3.6032259464263916,
      "learning_rate": 2.2040816326530613e-05,
      "loss": 0.9733,
      "step": 2840
    },
    {
      "epoch": 1.1514,
      "grad_norm": 6.305304527282715,
      "learning_rate": 2.2030612244897962e-05,
      "loss": 0.6498,
      "step": 2841
    },
    {
      "epoch": 1.1516,
      "grad_norm": 2.920128345489502,
      "learning_rate": 2.2020408163265308e-05,
      "loss": 0.3438,
      "step": 2842
    },
    {
      "epoch": 1.1518,
      "grad_norm": 0.9799144268035889,
      "learning_rate": 2.2010204081632654e-05,
      "loss": 0.0364,
      "step": 2843
    },
    {
      "epoch": 1.152,
      "grad_norm": 1.5188324451446533,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.1618,
      "step": 2844
    },
    {
      "epoch": 1.1522000000000001,
      "grad_norm": 3.6000258922576904,
      "learning_rate": 2.198979591836735e-05,
      "loss": 0.4784,
      "step": 2845
    },
    {
      "epoch": 1.1524,
      "grad_norm": 7.604703903198242,
      "learning_rate": 2.1979591836734694e-05,
      "loss": 2.1591,
      "step": 2846
    },
    {
      "epoch": 1.1526,
      "grad_norm": 6.9416728019714355,
      "learning_rate": 2.196938775510204e-05,
      "loss": 1.4784,
      "step": 2847
    },
    {
      "epoch": 1.1528,
      "grad_norm": 1.5903897285461426,
      "learning_rate": 2.195918367346939e-05,
      "loss": 0.0872,
      "step": 2848
    },
    {
      "epoch": 1.153,
      "grad_norm": 1.9738491773605347,
      "learning_rate": 2.1948979591836735e-05,
      "loss": 0.1707,
      "step": 2849
    },
    {
      "epoch": 1.1532,
      "grad_norm": 4.482814788818359,
      "learning_rate": 2.193877551020408e-05,
      "loss": 0.8933,
      "step": 2850
    },
    {
      "epoch": 1.1534,
      "grad_norm": 5.364544868469238,
      "learning_rate": 2.192857142857143e-05,
      "loss": 0.7052,
      "step": 2851
    },
    {
      "epoch": 1.1536,
      "grad_norm": 11.5966157913208,
      "learning_rate": 2.1918367346938776e-05,
      "loss": 1.5005,
      "step": 2852
    },
    {
      "epoch": 1.1538,
      "grad_norm": 34.515357971191406,
      "learning_rate": 2.1908163265306122e-05,
      "loss": 4.0731,
      "step": 2853
    },
    {
      "epoch": 1.154,
      "grad_norm": 11.817828178405762,
      "learning_rate": 2.189795918367347e-05,
      "loss": 1.3178,
      "step": 2854
    },
    {
      "epoch": 1.1542,
      "grad_norm": 1.158492922782898,
      "learning_rate": 2.1887755102040817e-05,
      "loss": 0.045,
      "step": 2855
    },
    {
      "epoch": 1.1544,
      "grad_norm": 2.2836451530456543,
      "learning_rate": 2.1877551020408162e-05,
      "loss": 0.1339,
      "step": 2856
    },
    {
      "epoch": 1.1546,
      "grad_norm": 4.321943759918213,
      "learning_rate": 2.186734693877551e-05,
      "loss": 0.6489,
      "step": 2857
    },
    {
      "epoch": 1.1548,
      "grad_norm": 1.2813799381256104,
      "learning_rate": 2.185714285714286e-05,
      "loss": 0.0529,
      "step": 2858
    },
    {
      "epoch": 1.155,
      "grad_norm": 1.838179588317871,
      "learning_rate": 2.1846938775510207e-05,
      "loss": 0.2322,
      "step": 2859
    },
    {
      "epoch": 1.1552,
      "grad_norm": 2.256678819656372,
      "learning_rate": 2.1836734693877552e-05,
      "loss": 0.3033,
      "step": 2860
    },
    {
      "epoch": 1.1554,
      "grad_norm": 3.089646100997925,
      "learning_rate": 2.1826530612244898e-05,
      "loss": 0.3594,
      "step": 2861
    },
    {
      "epoch": 1.1556,
      "grad_norm": 1.164261817932129,
      "learning_rate": 2.1816326530612247e-05,
      "loss": 0.0458,
      "step": 2862
    },
    {
      "epoch": 1.1558,
      "grad_norm": 1.6352413892745972,
      "learning_rate": 2.1806122448979593e-05,
      "loss": 0.1352,
      "step": 2863
    },
    {
      "epoch": 1.156,
      "grad_norm": 1.7097831964492798,
      "learning_rate": 2.179591836734694e-05,
      "loss": 0.2099,
      "step": 2864
    },
    {
      "epoch": 1.1562000000000001,
      "grad_norm": 1.9434983730316162,
      "learning_rate": 2.1785714285714288e-05,
      "loss": 0.2527,
      "step": 2865
    },
    {
      "epoch": 1.1564,
      "grad_norm": 2.9280757904052734,
      "learning_rate": 2.1775510204081634e-05,
      "loss": 0.5403,
      "step": 2866
    },
    {
      "epoch": 1.1566,
      "grad_norm": 2.071007490158081,
      "learning_rate": 2.176530612244898e-05,
      "loss": 0.3943,
      "step": 2867
    },
    {
      "epoch": 1.1568,
      "grad_norm": 3.7801718711853027,
      "learning_rate": 2.175510204081633e-05,
      "loss": 1.6641,
      "step": 2868
    },
    {
      "epoch": 1.157,
      "grad_norm": 5.573559284210205,
      "learning_rate": 2.1744897959183675e-05,
      "loss": 0.6697,
      "step": 2869
    },
    {
      "epoch": 1.1572,
      "grad_norm": 5.357125759124756,
      "learning_rate": 2.173469387755102e-05,
      "loss": 0.9319,
      "step": 2870
    },
    {
      "epoch": 1.1574,
      "grad_norm": 11.930858612060547,
      "learning_rate": 2.1724489795918366e-05,
      "loss": 3.4857,
      "step": 2871
    },
    {
      "epoch": 1.1576,
      "grad_norm": 21.174842834472656,
      "learning_rate": 2.1714285714285715e-05,
      "loss": 3.564,
      "step": 2872
    },
    {
      "epoch": 1.1578,
      "grad_norm": 6.427459716796875,
      "learning_rate": 2.170408163265306e-05,
      "loss": 1.6017,
      "step": 2873
    },
    {
      "epoch": 1.158,
      "grad_norm": 2.9862544536590576,
      "learning_rate": 2.1693877551020407e-05,
      "loss": 0.3427,
      "step": 2874
    },
    {
      "epoch": 1.1582,
      "grad_norm": 1.7005441188812256,
      "learning_rate": 2.1683673469387756e-05,
      "loss": 0.029,
      "step": 2875
    },
    {
      "epoch": 1.1584,
      "grad_norm": 1.1053560972213745,
      "learning_rate": 2.1673469387755102e-05,
      "loss": 0.0429,
      "step": 2876
    },
    {
      "epoch": 1.1586,
      "grad_norm": 2.360154628753662,
      "learning_rate": 2.1663265306122448e-05,
      "loss": 0.5475,
      "step": 2877
    },
    {
      "epoch": 1.1588,
      "grad_norm": 3.2414939403533936,
      "learning_rate": 2.1653061224489797e-05,
      "loss": 1.5426,
      "step": 2878
    },
    {
      "epoch": 1.159,
      "grad_norm": 2.4206106662750244,
      "learning_rate": 2.1642857142857146e-05,
      "loss": 0.1732,
      "step": 2879
    },
    {
      "epoch": 1.1592,
      "grad_norm": 4.766385555267334,
      "learning_rate": 2.1632653061224492e-05,
      "loss": 0.6206,
      "step": 2880
    },
    {
      "epoch": 1.1594,
      "grad_norm": 1.7571815252304077,
      "learning_rate": 2.1622448979591838e-05,
      "loss": 0.1629,
      "step": 2881
    },
    {
      "epoch": 1.1596,
      "grad_norm": 2.1201984882354736,
      "learning_rate": 2.1612244897959187e-05,
      "loss": 0.2275,
      "step": 2882
    },
    {
      "epoch": 1.1598,
      "grad_norm": 1.9399231672286987,
      "learning_rate": 2.1602040816326533e-05,
      "loss": 0.1805,
      "step": 2883
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.9238301515579224,
      "learning_rate": 2.159183673469388e-05,
      "loss": 0.4092,
      "step": 2884
    },
    {
      "epoch": 1.1602000000000001,
      "grad_norm": 3.6335549354553223,
      "learning_rate": 2.1581632653061228e-05,
      "loss": 1.6746,
      "step": 2885
    },
    {
      "epoch": 1.1604,
      "grad_norm": 1.0903916358947754,
      "learning_rate": 2.1571428571428574e-05,
      "loss": 0.0294,
      "step": 2886
    },
    {
      "epoch": 1.1606,
      "grad_norm": 1.3453727960586548,
      "learning_rate": 2.156122448979592e-05,
      "loss": 0.027,
      "step": 2887
    },
    {
      "epoch": 1.1608,
      "grad_norm": 1.4440014362335205,
      "learning_rate": 2.1551020408163265e-05,
      "loss": 0.1507,
      "step": 2888
    },
    {
      "epoch": 1.161,
      "grad_norm": 1.1431560516357422,
      "learning_rate": 2.1540816326530614e-05,
      "loss": 0.0303,
      "step": 2889
    },
    {
      "epoch": 1.1612,
      "grad_norm": 1.644680142402649,
      "learning_rate": 2.153061224489796e-05,
      "loss": 0.1439,
      "step": 2890
    },
    {
      "epoch": 1.1614,
      "grad_norm": 2.0987346172332764,
      "learning_rate": 2.1520408163265306e-05,
      "loss": 0.5029,
      "step": 2891
    },
    {
      "epoch": 1.1616,
      "grad_norm": 4.1892409324646,
      "learning_rate": 2.1510204081632655e-05,
      "loss": 1.8529,
      "step": 2892
    },
    {
      "epoch": 1.1618,
      "grad_norm": 7.047264575958252,
      "learning_rate": 2.15e-05,
      "loss": 1.5588,
      "step": 2893
    },
    {
      "epoch": 1.162,
      "grad_norm": 4.856655597686768,
      "learning_rate": 2.1489795918367347e-05,
      "loss": 0.6655,
      "step": 2894
    },
    {
      "epoch": 1.1622,
      "grad_norm": 3.8010571002960205,
      "learning_rate": 2.1479591836734696e-05,
      "loss": 2.1848,
      "step": 2895
    },
    {
      "epoch": 1.1623999999999999,
      "grad_norm": 3.943370819091797,
      "learning_rate": 2.146938775510204e-05,
      "loss": 2.3074,
      "step": 2896
    },
    {
      "epoch": 1.1626,
      "grad_norm": 8.42847728729248,
      "learning_rate": 2.1459183673469387e-05,
      "loss": 2.3577,
      "step": 2897
    },
    {
      "epoch": 1.1628,
      "grad_norm": 15.801399230957031,
      "learning_rate": 2.1448979591836733e-05,
      "loss": 3.7016,
      "step": 2898
    },
    {
      "epoch": 1.163,
      "grad_norm": 6.718034267425537,
      "learning_rate": 2.1438775510204082e-05,
      "loss": 0.6954,
      "step": 2899
    },
    {
      "epoch": 1.1632,
      "grad_norm": 1.4650180339813232,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.0764,
      "step": 2900
    },
    {
      "epoch": 1.1634,
      "grad_norm": 2.4374194145202637,
      "learning_rate": 2.1418367346938774e-05,
      "loss": 0.3817,
      "step": 2901
    },
    {
      "epoch": 1.1636,
      "grad_norm": 3.6774742603302,
      "learning_rate": 2.1408163265306127e-05,
      "loss": 0.9914,
      "step": 2902
    },
    {
      "epoch": 1.1638,
      "grad_norm": 4.494981288909912,
      "learning_rate": 2.1397959183673472e-05,
      "loss": 1.9728,
      "step": 2903
    },
    {
      "epoch": 1.164,
      "grad_norm": 6.808770179748535,
      "learning_rate": 2.1387755102040818e-05,
      "loss": 2.264,
      "step": 2904
    },
    {
      "epoch": 1.1642000000000001,
      "grad_norm": 10.066740036010742,
      "learning_rate": 2.1377551020408164e-05,
      "loss": 2.7497,
      "step": 2905
    },
    {
      "epoch": 1.1644,
      "grad_norm": 6.652487277984619,
      "learning_rate": 2.1367346938775513e-05,
      "loss": 2.5345,
      "step": 2906
    },
    {
      "epoch": 1.1646,
      "grad_norm": 7.7479472160339355,
      "learning_rate": 2.135714285714286e-05,
      "loss": 2.5017,
      "step": 2907
    },
    {
      "epoch": 1.1648,
      "grad_norm": 2.5272955894470215,
      "learning_rate": 2.1346938775510205e-05,
      "loss": 0.1819,
      "step": 2908
    },
    {
      "epoch": 1.165,
      "grad_norm": 5.05612325668335,
      "learning_rate": 2.1336734693877554e-05,
      "loss": 0.6378,
      "step": 2909
    },
    {
      "epoch": 1.1652,
      "grad_norm": 1.5177295207977295,
      "learning_rate": 2.13265306122449e-05,
      "loss": 0.1516,
      "step": 2910
    },
    {
      "epoch": 1.1654,
      "grad_norm": 1.9821985960006714,
      "learning_rate": 2.1316326530612245e-05,
      "loss": 0.4778,
      "step": 2911
    },
    {
      "epoch": 1.1656,
      "grad_norm": 1.3456918001174927,
      "learning_rate": 2.1306122448979595e-05,
      "loss": 0.06,
      "step": 2912
    },
    {
      "epoch": 1.1658,
      "grad_norm": 10.177454948425293,
      "learning_rate": 2.129591836734694e-05,
      "loss": 0.9334,
      "step": 2913
    },
    {
      "epoch": 1.166,
      "grad_norm": 22.78434181213379,
      "learning_rate": 2.1285714285714286e-05,
      "loss": 3.5745,
      "step": 2914
    },
    {
      "epoch": 1.1662,
      "grad_norm": 1.3361148834228516,
      "learning_rate": 2.1275510204081632e-05,
      "loss": 0.0313,
      "step": 2915
    },
    {
      "epoch": 1.1663999999999999,
      "grad_norm": 1.5880099534988403,
      "learning_rate": 2.126530612244898e-05,
      "loss": 0.1572,
      "step": 2916
    },
    {
      "epoch": 1.1666,
      "grad_norm": 1.2747163772583008,
      "learning_rate": 2.1255102040816327e-05,
      "loss": 0.0516,
      "step": 2917
    },
    {
      "epoch": 1.1668,
      "grad_norm": 2.1888797283172607,
      "learning_rate": 2.1244897959183673e-05,
      "loss": 0.2343,
      "step": 2918
    },
    {
      "epoch": 1.167,
      "grad_norm": 3.0382328033447266,
      "learning_rate": 2.1234693877551022e-05,
      "loss": 0.5813,
      "step": 2919
    },
    {
      "epoch": 1.1672,
      "grad_norm": 1.4096482992172241,
      "learning_rate": 2.1224489795918368e-05,
      "loss": 0.0571,
      "step": 2920
    },
    {
      "epoch": 1.1674,
      "grad_norm": 1.185367226600647,
      "learning_rate": 2.1214285714285713e-05,
      "loss": 0.0407,
      "step": 2921
    },
    {
      "epoch": 1.1676,
      "grad_norm": 1.6025385856628418,
      "learning_rate": 2.1204081632653063e-05,
      "loss": 0.1827,
      "step": 2922
    },
    {
      "epoch": 1.1678,
      "grad_norm": 1.0219346284866333,
      "learning_rate": 2.119387755102041e-05,
      "loss": 0.0378,
      "step": 2923
    },
    {
      "epoch": 1.168,
      "grad_norm": 1.3154290914535522,
      "learning_rate": 2.1183673469387754e-05,
      "loss": 0.1056,
      "step": 2924
    },
    {
      "epoch": 1.1682,
      "grad_norm": 1.2452430725097656,
      "learning_rate": 2.1173469387755103e-05,
      "loss": 0.1115,
      "step": 2925
    },
    {
      "epoch": 1.1684,
      "grad_norm": 1.9842383861541748,
      "learning_rate": 2.1163265306122453e-05,
      "loss": 0.2558,
      "step": 2926
    },
    {
      "epoch": 1.1686,
      "grad_norm": 3.636225938796997,
      "learning_rate": 2.11530612244898e-05,
      "loss": 1.0264,
      "step": 2927
    },
    {
      "epoch": 1.1688,
      "grad_norm": 3.466745138168335,
      "learning_rate": 2.1142857142857144e-05,
      "loss": 1.7151,
      "step": 2928
    },
    {
      "epoch": 1.169,
      "grad_norm": 2.697654962539673,
      "learning_rate": 2.113265306122449e-05,
      "loss": 0.6576,
      "step": 2929
    },
    {
      "epoch": 1.1692,
      "grad_norm": 3.2328388690948486,
      "learning_rate": 2.112244897959184e-05,
      "loss": 0.459,
      "step": 2930
    },
    {
      "epoch": 1.1694,
      "grad_norm": 1.6146492958068848,
      "learning_rate": 2.1112244897959185e-05,
      "loss": 0.1081,
      "step": 2931
    },
    {
      "epoch": 1.1696,
      "grad_norm": 2.285536766052246,
      "learning_rate": 2.110204081632653e-05,
      "loss": 0.2701,
      "step": 2932
    },
    {
      "epoch": 1.1698,
      "grad_norm": 6.2433905601501465,
      "learning_rate": 2.109183673469388e-05,
      "loss": 0.8872,
      "step": 2933
    },
    {
      "epoch": 1.17,
      "grad_norm": 8.182297706604004,
      "learning_rate": 2.1081632653061226e-05,
      "loss": 0.8029,
      "step": 2934
    },
    {
      "epoch": 1.1702,
      "grad_norm": 2.641026020050049,
      "learning_rate": 2.107142857142857e-05,
      "loss": 0.1031,
      "step": 2935
    },
    {
      "epoch": 1.1703999999999999,
      "grad_norm": 3.0824596881866455,
      "learning_rate": 2.106122448979592e-05,
      "loss": 0.4613,
      "step": 2936
    },
    {
      "epoch": 1.1706,
      "grad_norm": 4.887169361114502,
      "learning_rate": 2.1051020408163266e-05,
      "loss": 1.4961,
      "step": 2937
    },
    {
      "epoch": 1.1708,
      "grad_norm": 3.300015449523926,
      "learning_rate": 2.1040816326530612e-05,
      "loss": 0.3345,
      "step": 2938
    },
    {
      "epoch": 1.171,
      "grad_norm": 7.756126403808594,
      "learning_rate": 2.103061224489796e-05,
      "loss": 1.3592,
      "step": 2939
    },
    {
      "epoch": 1.1712,
      "grad_norm": 1.2789355516433716,
      "learning_rate": 2.1020408163265307e-05,
      "loss": 0.033,
      "step": 2940
    },
    {
      "epoch": 1.1714,
      "grad_norm": 2.148930311203003,
      "learning_rate": 2.1010204081632653e-05,
      "loss": 0.1399,
      "step": 2941
    },
    {
      "epoch": 1.1716,
      "grad_norm": 5.470130443572998,
      "learning_rate": 2.1e-05,
      "loss": 0.8179,
      "step": 2942
    },
    {
      "epoch": 1.1718,
      "grad_norm": 2.832042932510376,
      "learning_rate": 2.0989795918367348e-05,
      "loss": 0.3213,
      "step": 2943
    },
    {
      "epoch": 1.172,
      "grad_norm": 1.6176263093948364,
      "learning_rate": 2.0979591836734694e-05,
      "loss": 0.08,
      "step": 2944
    },
    {
      "epoch": 1.1722,
      "grad_norm": 8.061407089233398,
      "learning_rate": 2.096938775510204e-05,
      "loss": 0.4205,
      "step": 2945
    },
    {
      "epoch": 1.1724,
      "grad_norm": 21.434839248657227,
      "learning_rate": 2.095918367346939e-05,
      "loss": 2.1196,
      "step": 2946
    },
    {
      "epoch": 1.1726,
      "grad_norm": 5.325974941253662,
      "learning_rate": 2.0948979591836738e-05,
      "loss": 1.4808,
      "step": 2947
    },
    {
      "epoch": 1.1728,
      "grad_norm": 6.673322677612305,
      "learning_rate": 2.0938775510204084e-05,
      "loss": 0.9208,
      "step": 2948
    },
    {
      "epoch": 1.173,
      "grad_norm": 22.24078941345215,
      "learning_rate": 2.092857142857143e-05,
      "loss": 2.9982,
      "step": 2949
    },
    {
      "epoch": 1.1732,
      "grad_norm": 2.5346529483795166,
      "learning_rate": 2.091836734693878e-05,
      "loss": 0.6132,
      "step": 2950
    },
    {
      "epoch": 1.1734,
      "grad_norm": 6.196047782897949,
      "learning_rate": 2.0908163265306125e-05,
      "loss": 1.766,
      "step": 2951
    },
    {
      "epoch": 1.1736,
      "grad_norm": 9.942660331726074,
      "learning_rate": 2.089795918367347e-05,
      "loss": 1.4465,
      "step": 2952
    },
    {
      "epoch": 1.1738,
      "grad_norm": 2.346346139907837,
      "learning_rate": 2.088775510204082e-05,
      "loss": 0.2552,
      "step": 2953
    },
    {
      "epoch": 1.174,
      "grad_norm": 4.952291011810303,
      "learning_rate": 2.0877551020408165e-05,
      "loss": 1.0868,
      "step": 2954
    },
    {
      "epoch": 1.1742,
      "grad_norm": 3.304058790206909,
      "learning_rate": 2.086734693877551e-05,
      "loss": 1.5127,
      "step": 2955
    },
    {
      "epoch": 1.1743999999999999,
      "grad_norm": 1.2883349657058716,
      "learning_rate": 2.0857142857142857e-05,
      "loss": 0.0566,
      "step": 2956
    },
    {
      "epoch": 1.1746,
      "grad_norm": 3.0064966678619385,
      "learning_rate": 2.0846938775510206e-05,
      "loss": 0.1192,
      "step": 2957
    },
    {
      "epoch": 1.1748,
      "grad_norm": 5.3768768310546875,
      "learning_rate": 2.0836734693877552e-05,
      "loss": 0.3003,
      "step": 2958
    },
    {
      "epoch": 1.175,
      "grad_norm": 1.6847586631774902,
      "learning_rate": 2.0826530612244898e-05,
      "loss": 0.1338,
      "step": 2959
    },
    {
      "epoch": 1.1752,
      "grad_norm": 1.9989336729049683,
      "learning_rate": 2.0816326530612247e-05,
      "loss": 0.4576,
      "step": 2960
    },
    {
      "epoch": 1.1754,
      "grad_norm": 3.390285015106201,
      "learning_rate": 2.0806122448979593e-05,
      "loss": 1.6451,
      "step": 2961
    },
    {
      "epoch": 1.1756,
      "grad_norm": 1.4990715980529785,
      "learning_rate": 2.079591836734694e-05,
      "loss": 0.1426,
      "step": 2962
    },
    {
      "epoch": 1.1758,
      "grad_norm": 1.853268027305603,
      "learning_rate": 2.0785714285714288e-05,
      "loss": 0.2308,
      "step": 2963
    },
    {
      "epoch": 1.176,
      "grad_norm": 1.5178651809692383,
      "learning_rate": 2.0775510204081633e-05,
      "loss": 0.1052,
      "step": 2964
    },
    {
      "epoch": 1.1762,
      "grad_norm": 1.68022882938385,
      "learning_rate": 2.076530612244898e-05,
      "loss": 0.1792,
      "step": 2965
    },
    {
      "epoch": 1.1764000000000001,
      "grad_norm": 1.5579159259796143,
      "learning_rate": 2.0755102040816325e-05,
      "loss": 0.0838,
      "step": 2966
    },
    {
      "epoch": 1.1766,
      "grad_norm": 4.088912010192871,
      "learning_rate": 2.0744897959183674e-05,
      "loss": 0.5077,
      "step": 2967
    },
    {
      "epoch": 1.1768,
      "grad_norm": 7.16737699508667,
      "learning_rate": 2.073469387755102e-05,
      "loss": 1.4126,
      "step": 2968
    },
    {
      "epoch": 1.177,
      "grad_norm": 5.091823101043701,
      "learning_rate": 2.0724489795918366e-05,
      "loss": 0.2374,
      "step": 2969
    },
    {
      "epoch": 1.1772,
      "grad_norm": 6.514739990234375,
      "learning_rate": 2.0714285714285718e-05,
      "loss": 0.3321,
      "step": 2970
    },
    {
      "epoch": 1.1774,
      "grad_norm": 10.223052978515625,
      "learning_rate": 2.0704081632653064e-05,
      "loss": 0.7807,
      "step": 2971
    },
    {
      "epoch": 1.1776,
      "grad_norm": 20.88498878479004,
      "learning_rate": 2.069387755102041e-05,
      "loss": 2.1198,
      "step": 2972
    },
    {
      "epoch": 1.1778,
      "grad_norm": 8.11330795288086,
      "learning_rate": 2.0683673469387756e-05,
      "loss": 0.7407,
      "step": 2973
    },
    {
      "epoch": 1.178,
      "grad_norm": 7.516994476318359,
      "learning_rate": 2.0673469387755105e-05,
      "loss": 0.9794,
      "step": 2974
    },
    {
      "epoch": 1.1782,
      "grad_norm": 16.233720779418945,
      "learning_rate": 2.066326530612245e-05,
      "loss": 3.877,
      "step": 2975
    },
    {
      "epoch": 1.1784,
      "grad_norm": 3.6696817874908447,
      "learning_rate": 2.0653061224489796e-05,
      "loss": 1.5323,
      "step": 2976
    },
    {
      "epoch": 1.1786,
      "grad_norm": 2.3854358196258545,
      "learning_rate": 2.0642857142857146e-05,
      "loss": 0.242,
      "step": 2977
    },
    {
      "epoch": 1.1788,
      "grad_norm": 4.538883686065674,
      "learning_rate": 2.063265306122449e-05,
      "loss": 0.781,
      "step": 2978
    },
    {
      "epoch": 1.179,
      "grad_norm": 2.7626986503601074,
      "learning_rate": 2.0622448979591837e-05,
      "loss": 0.3337,
      "step": 2979
    },
    {
      "epoch": 1.1792,
      "grad_norm": 1.7479041814804077,
      "learning_rate": 2.0612244897959186e-05,
      "loss": 0.1616,
      "step": 2980
    },
    {
      "epoch": 1.1794,
      "grad_norm": 1.5093389749526978,
      "learning_rate": 2.0602040816326532e-05,
      "loss": 0.145,
      "step": 2981
    },
    {
      "epoch": 1.1796,
      "grad_norm": 2.2258784770965576,
      "learning_rate": 2.0591836734693878e-05,
      "loss": 0.1724,
      "step": 2982
    },
    {
      "epoch": 1.1798,
      "grad_norm": 4.365813255310059,
      "learning_rate": 2.0581632653061224e-05,
      "loss": 0.3035,
      "step": 2983
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.0965051651000977,
      "learning_rate": 2.0571428571428573e-05,
      "loss": 0.0122,
      "step": 2984
    },
    {
      "epoch": 1.1802,
      "grad_norm": 1.604235291481018,
      "learning_rate": 2.056122448979592e-05,
      "loss": 0.1682,
      "step": 2985
    },
    {
      "epoch": 1.1804000000000001,
      "grad_norm": 2.189631223678589,
      "learning_rate": 2.0551020408163265e-05,
      "loss": 0.4954,
      "step": 2986
    },
    {
      "epoch": 1.1806,
      "grad_norm": 1.478208065032959,
      "learning_rate": 2.0540816326530614e-05,
      "loss": 0.097,
      "step": 2987
    },
    {
      "epoch": 1.1808,
      "grad_norm": 1.4729284048080444,
      "learning_rate": 2.053061224489796e-05,
      "loss": 0.088,
      "step": 2988
    },
    {
      "epoch": 1.181,
      "grad_norm": 1.6487613916397095,
      "learning_rate": 2.0520408163265305e-05,
      "loss": 0.116,
      "step": 2989
    },
    {
      "epoch": 1.1812,
      "grad_norm": 3.5400404930114746,
      "learning_rate": 2.0510204081632654e-05,
      "loss": 0.629,
      "step": 2990
    },
    {
      "epoch": 1.1814,
      "grad_norm": 6.919580459594727,
      "learning_rate": 2.05e-05,
      "loss": 0.7759,
      "step": 2991
    },
    {
      "epoch": 1.1816,
      "grad_norm": 2.9006593227386475,
      "learning_rate": 2.0489795918367346e-05,
      "loss": 0.2965,
      "step": 2992
    },
    {
      "epoch": 1.1818,
      "grad_norm": 11.443617820739746,
      "learning_rate": 2.0479591836734695e-05,
      "loss": 2.0913,
      "step": 2993
    },
    {
      "epoch": 1.182,
      "grad_norm": 16.701017379760742,
      "learning_rate": 2.0469387755102044e-05,
      "loss": 2.2104,
      "step": 2994
    },
    {
      "epoch": 1.1822,
      "grad_norm": 5.281843185424805,
      "learning_rate": 2.045918367346939e-05,
      "loss": 0.8237,
      "step": 2995
    },
    {
      "epoch": 1.1824,
      "grad_norm": 4.46371603012085,
      "learning_rate": 2.0448979591836736e-05,
      "loss": 0.6812,
      "step": 2996
    },
    {
      "epoch": 1.1826,
      "grad_norm": 1.2886015176773071,
      "learning_rate": 2.0438775510204082e-05,
      "loss": 0.0697,
      "step": 2997
    },
    {
      "epoch": 1.1828,
      "grad_norm": 3.352893114089966,
      "learning_rate": 2.042857142857143e-05,
      "loss": 0.3906,
      "step": 2998
    },
    {
      "epoch": 1.183,
      "grad_norm": 5.506088733673096,
      "learning_rate": 2.0418367346938777e-05,
      "loss": 1.8157,
      "step": 2999
    },
    {
      "epoch": 1.1832,
      "grad_norm": 3.538123846054077,
      "learning_rate": 2.0408163265306123e-05,
      "loss": 1.6393,
      "step": 3000
    },
    {
      "epoch": 1.1834,
      "grad_norm": 0.9296956062316895,
      "learning_rate": 2.0397959183673472e-05,
      "loss": 0.0733,
      "step": 3001
    },
    {
      "epoch": 1.1836,
      "grad_norm": 1.3598219156265259,
      "learning_rate": 2.0387755102040817e-05,
      "loss": 0.0486,
      "step": 3002
    },
    {
      "epoch": 1.1838,
      "grad_norm": 2.0247433185577393,
      "learning_rate": 2.0377551020408163e-05,
      "loss": 0.1865,
      "step": 3003
    },
    {
      "epoch": 1.184,
      "grad_norm": 5.148133277893066,
      "learning_rate": 2.0367346938775512e-05,
      "loss": 0.8735,
      "step": 3004
    },
    {
      "epoch": 1.1842,
      "grad_norm": 3.1829111576080322,
      "learning_rate": 2.0357142857142858e-05,
      "loss": 0.8626,
      "step": 3005
    },
    {
      "epoch": 1.1844000000000001,
      "grad_norm": 3.3566997051239014,
      "learning_rate": 2.0346938775510204e-05,
      "loss": 1.5234,
      "step": 3006
    },
    {
      "epoch": 1.1846,
      "grad_norm": 1.5200315713882446,
      "learning_rate": 2.0336734693877553e-05,
      "loss": 0.1196,
      "step": 3007
    },
    {
      "epoch": 1.1848,
      "grad_norm": 8.360516548156738,
      "learning_rate": 2.03265306122449e-05,
      "loss": 1.1216,
      "step": 3008
    },
    {
      "epoch": 1.185,
      "grad_norm": 16.254709243774414,
      "learning_rate": 2.0316326530612245e-05,
      "loss": 3.6941,
      "step": 3009
    },
    {
      "epoch": 1.1852,
      "grad_norm": 2.454667091369629,
      "learning_rate": 2.030612244897959e-05,
      "loss": 0.1772,
      "step": 3010
    },
    {
      "epoch": 1.1854,
      "grad_norm": 1.7685350179672241,
      "learning_rate": 2.029591836734694e-05,
      "loss": 0.2342,
      "step": 3011
    },
    {
      "epoch": 1.1856,
      "grad_norm": 4.389267444610596,
      "learning_rate": 2.0285714285714286e-05,
      "loss": 0.7286,
      "step": 3012
    },
    {
      "epoch": 1.1858,
      "grad_norm": 13.457568168640137,
      "learning_rate": 2.027551020408163e-05,
      "loss": 3.6436,
      "step": 3013
    },
    {
      "epoch": 1.186,
      "grad_norm": 14.3905611038208,
      "learning_rate": 2.026530612244898e-05,
      "loss": 3.3189,
      "step": 3014
    },
    {
      "epoch": 1.1862,
      "grad_norm": 7.432793140411377,
      "learning_rate": 2.0255102040816326e-05,
      "loss": 0.8565,
      "step": 3015
    },
    {
      "epoch": 1.1864,
      "grad_norm": 2.9249179363250732,
      "learning_rate": 2.0244897959183676e-05,
      "loss": 0.5177,
      "step": 3016
    },
    {
      "epoch": 1.1865999999999999,
      "grad_norm": 2.024263620376587,
      "learning_rate": 2.023469387755102e-05,
      "loss": 0.4307,
      "step": 3017
    },
    {
      "epoch": 1.1868,
      "grad_norm": 3.8027567863464355,
      "learning_rate": 2.022448979591837e-05,
      "loss": 1.5573,
      "step": 3018
    },
    {
      "epoch": 1.187,
      "grad_norm": 3.958740234375,
      "learning_rate": 2.0214285714285716e-05,
      "loss": 0.6836,
      "step": 3019
    },
    {
      "epoch": 1.1872,
      "grad_norm": 1.5115995407104492,
      "learning_rate": 2.0204081632653062e-05,
      "loss": 0.1058,
      "step": 3020
    },
    {
      "epoch": 1.1874,
      "grad_norm": 6.807767868041992,
      "learning_rate": 2.019387755102041e-05,
      "loss": 1.069,
      "step": 3021
    },
    {
      "epoch": 1.1876,
      "grad_norm": 10.496637344360352,
      "learning_rate": 2.0183673469387757e-05,
      "loss": 3.1418,
      "step": 3022
    },
    {
      "epoch": 1.1878,
      "grad_norm": 3.6764719486236572,
      "learning_rate": 2.0173469387755103e-05,
      "loss": 0.6877,
      "step": 3023
    },
    {
      "epoch": 1.188,
      "grad_norm": 4.998138904571533,
      "learning_rate": 2.016326530612245e-05,
      "loss": 0.7316,
      "step": 3024
    },
    {
      "epoch": 1.1882,
      "grad_norm": 3.1024022102355957,
      "learning_rate": 2.0153061224489798e-05,
      "loss": 0.5964,
      "step": 3025
    },
    {
      "epoch": 1.1884000000000001,
      "grad_norm": 5.601747512817383,
      "learning_rate": 2.0142857142857144e-05,
      "loss": 0.4643,
      "step": 3026
    },
    {
      "epoch": 1.1886,
      "grad_norm": 15.740729331970215,
      "learning_rate": 2.013265306122449e-05,
      "loss": 0.979,
      "step": 3027
    },
    {
      "epoch": 1.1888,
      "grad_norm": 6.91904878616333,
      "learning_rate": 2.012244897959184e-05,
      "loss": 0.9966,
      "step": 3028
    },
    {
      "epoch": 1.189,
      "grad_norm": 5.038677215576172,
      "learning_rate": 2.0112244897959184e-05,
      "loss": 1.5011,
      "step": 3029
    },
    {
      "epoch": 1.1892,
      "grad_norm": 1.3942502737045288,
      "learning_rate": 2.010204081632653e-05,
      "loss": 0.1064,
      "step": 3030
    },
    {
      "epoch": 1.1894,
      "grad_norm": 3.739537000656128,
      "learning_rate": 2.009183673469388e-05,
      "loss": 0.3063,
      "step": 3031
    },
    {
      "epoch": 1.1896,
      "grad_norm": 4.991539478302002,
      "learning_rate": 2.0081632653061225e-05,
      "loss": 0.6707,
      "step": 3032
    },
    {
      "epoch": 1.1898,
      "grad_norm": 1.6191712617874146,
      "learning_rate": 2.007142857142857e-05,
      "loss": 0.1287,
      "step": 3033
    },
    {
      "epoch": 1.19,
      "grad_norm": 3.295368194580078,
      "learning_rate": 2.0061224489795917e-05,
      "loss": 0.125,
      "step": 3034
    },
    {
      "epoch": 1.1902,
      "grad_norm": 8.2177734375,
      "learning_rate": 2.0051020408163266e-05,
      "loss": 0.7098,
      "step": 3035
    },
    {
      "epoch": 1.1904,
      "grad_norm": 10.9459228515625,
      "learning_rate": 2.004081632653061e-05,
      "loss": 2.0341,
      "step": 3036
    },
    {
      "epoch": 1.1905999999999999,
      "grad_norm": 10.621405601501465,
      "learning_rate": 2.0030612244897957e-05,
      "loss": 1.8083,
      "step": 3037
    },
    {
      "epoch": 1.1908,
      "grad_norm": 6.680018901824951,
      "learning_rate": 2.002040816326531e-05,
      "loss": 1.4806,
      "step": 3038
    },
    {
      "epoch": 1.191,
      "grad_norm": 1.1898123025894165,
      "learning_rate": 2.0010204081632656e-05,
      "loss": 0.1244,
      "step": 3039
    },
    {
      "epoch": 1.1912,
      "grad_norm": 1.4971431493759155,
      "learning_rate": 2e-05,
      "loss": 0.1304,
      "step": 3040
    },
    {
      "epoch": 1.1914,
      "grad_norm": 1.1757605075836182,
      "learning_rate": 1.9989795918367347e-05,
      "loss": 0.0622,
      "step": 3041
    },
    {
      "epoch": 1.1916,
      "grad_norm": 1.6284112930297852,
      "learning_rate": 1.9979591836734697e-05,
      "loss": 0.0736,
      "step": 3042
    },
    {
      "epoch": 1.1918,
      "grad_norm": 1.1026115417480469,
      "learning_rate": 1.9969387755102042e-05,
      "loss": 0.042,
      "step": 3043
    },
    {
      "epoch": 1.192,
      "grad_norm": 1.9990594387054443,
      "learning_rate": 1.9959183673469388e-05,
      "loss": 0.1913,
      "step": 3044
    },
    {
      "epoch": 1.1922,
      "grad_norm": 3.820249319076538,
      "learning_rate": 1.9948979591836737e-05,
      "loss": 0.6869,
      "step": 3045
    },
    {
      "epoch": 1.1924,
      "grad_norm": 1.741510033607483,
      "learning_rate": 1.9938775510204083e-05,
      "loss": 0.0858,
      "step": 3046
    },
    {
      "epoch": 1.1926,
      "grad_norm": 3.434689998626709,
      "learning_rate": 1.992857142857143e-05,
      "loss": 0.3755,
      "step": 3047
    },
    {
      "epoch": 1.1928,
      "grad_norm": 5.762261390686035,
      "learning_rate": 1.9918367346938778e-05,
      "loss": 1.5206,
      "step": 3048
    },
    {
      "epoch": 1.193,
      "grad_norm": 8.78773021697998,
      "learning_rate": 1.9908163265306124e-05,
      "loss": 0.86,
      "step": 3049
    },
    {
      "epoch": 1.1932,
      "grad_norm": 7.344252109527588,
      "learning_rate": 1.989795918367347e-05,
      "loss": 0.3056,
      "step": 3050
    },
    {
      "epoch": 1.1934,
      "grad_norm": 1.1593364477157593,
      "learning_rate": 1.9887755102040816e-05,
      "loss": 0.0734,
      "step": 3051
    },
    {
      "epoch": 1.1936,
      "grad_norm": 1.4719830751419067,
      "learning_rate": 1.9877551020408165e-05,
      "loss": 0.0712,
      "step": 3052
    },
    {
      "epoch": 1.1938,
      "grad_norm": 2.024540662765503,
      "learning_rate": 1.986734693877551e-05,
      "loss": 0.4275,
      "step": 3053
    },
    {
      "epoch": 1.194,
      "grad_norm": 4.2963056564331055,
      "learning_rate": 1.9857142857142856e-05,
      "loss": 1.6131,
      "step": 3054
    },
    {
      "epoch": 1.1942,
      "grad_norm": 8.733445167541504,
      "learning_rate": 1.9846938775510205e-05,
      "loss": 0.4647,
      "step": 3055
    },
    {
      "epoch": 1.1944,
      "grad_norm": 3.532928466796875,
      "learning_rate": 1.983673469387755e-05,
      "loss": 0.7477,
      "step": 3056
    },
    {
      "epoch": 1.1945999999999999,
      "grad_norm": 1.657246470451355,
      "learning_rate": 1.9826530612244897e-05,
      "loss": 0.1621,
      "step": 3057
    },
    {
      "epoch": 1.1948,
      "grad_norm": 2.976757049560547,
      "learning_rate": 1.9816326530612246e-05,
      "loss": 0.6082,
      "step": 3058
    },
    {
      "epoch": 1.195,
      "grad_norm": 1.20772385597229,
      "learning_rate": 1.9806122448979592e-05,
      "loss": 0.1211,
      "step": 3059
    },
    {
      "epoch": 1.1952,
      "grad_norm": 2.3030781745910645,
      "learning_rate": 1.9795918367346938e-05,
      "loss": 0.4913,
      "step": 3060
    },
    {
      "epoch": 1.1954,
      "grad_norm": 1.7301090955734253,
      "learning_rate": 1.9785714285714287e-05,
      "loss": 0.1146,
      "step": 3061
    },
    {
      "epoch": 1.1956,
      "grad_norm": 1.2910375595092773,
      "learning_rate": 1.9775510204081636e-05,
      "loss": 0.0515,
      "step": 3062
    },
    {
      "epoch": 1.1958,
      "grad_norm": 1.4276667833328247,
      "learning_rate": 1.9765306122448982e-05,
      "loss": 0.0557,
      "step": 3063
    },
    {
      "epoch": 1.196,
      "grad_norm": 5.736569881439209,
      "learning_rate": 1.9755102040816328e-05,
      "loss": 0.1938,
      "step": 3064
    },
    {
      "epoch": 1.1962,
      "grad_norm": 12.777549743652344,
      "learning_rate": 1.9744897959183677e-05,
      "loss": 1.8733,
      "step": 3065
    },
    {
      "epoch": 1.1964,
      "grad_norm": 17.816858291625977,
      "learning_rate": 1.9734693877551023e-05,
      "loss": 3.5921,
      "step": 3066
    },
    {
      "epoch": 1.1966,
      "grad_norm": 1.7031697034835815,
      "learning_rate": 1.972448979591837e-05,
      "loss": 0.1164,
      "step": 3067
    },
    {
      "epoch": 1.1968,
      "grad_norm": 4.166122913360596,
      "learning_rate": 1.9714285714285714e-05,
      "loss": 0.1969,
      "step": 3068
    },
    {
      "epoch": 1.197,
      "grad_norm": 9.597853660583496,
      "learning_rate": 1.9704081632653063e-05,
      "loss": 0.8492,
      "step": 3069
    },
    {
      "epoch": 1.1972,
      "grad_norm": 1.3580381870269775,
      "learning_rate": 1.969387755102041e-05,
      "loss": 0.0847,
      "step": 3070
    },
    {
      "epoch": 1.1974,
      "grad_norm": 6.079249858856201,
      "learning_rate": 1.9683673469387755e-05,
      "loss": 0.7113,
      "step": 3071
    },
    {
      "epoch": 1.1976,
      "grad_norm": 9.949956893920898,
      "learning_rate": 1.9673469387755104e-05,
      "loss": 1.831,
      "step": 3072
    },
    {
      "epoch": 1.1978,
      "grad_norm": 3.1551754474639893,
      "learning_rate": 1.966326530612245e-05,
      "loss": 0.3292,
      "step": 3073
    },
    {
      "epoch": 1.198,
      "grad_norm": 1.6016051769256592,
      "learning_rate": 1.9653061224489796e-05,
      "loss": 0.0865,
      "step": 3074
    },
    {
      "epoch": 1.1982,
      "grad_norm": 1.842907428741455,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 0.0748,
      "step": 3075
    },
    {
      "epoch": 1.1984,
      "grad_norm": 1.17067289352417,
      "learning_rate": 1.963265306122449e-05,
      "loss": 0.0344,
      "step": 3076
    },
    {
      "epoch": 1.1985999999999999,
      "grad_norm": 2.7815639972686768,
      "learning_rate": 1.9622448979591837e-05,
      "loss": 0.2074,
      "step": 3077
    },
    {
      "epoch": 1.1988,
      "grad_norm": 4.956741809844971,
      "learning_rate": 1.9612244897959182e-05,
      "loss": 0.6889,
      "step": 3078
    },
    {
      "epoch": 1.199,
      "grad_norm": 2.6895835399627686,
      "learning_rate": 1.960204081632653e-05,
      "loss": 0.1869,
      "step": 3079
    },
    {
      "epoch": 1.1992,
      "grad_norm": 4.918212890625,
      "learning_rate": 1.9591836734693877e-05,
      "loss": 0.2545,
      "step": 3080
    },
    {
      "epoch": 1.1994,
      "grad_norm": 1.636683702468872,
      "learning_rate": 1.9581632653061223e-05,
      "loss": 0.2151,
      "step": 3081
    },
    {
      "epoch": 1.1996,
      "grad_norm": 4.025855541229248,
      "learning_rate": 1.9571428571428572e-05,
      "loss": 0.2909,
      "step": 3082
    },
    {
      "epoch": 1.1998,
      "grad_norm": 9.657797813415527,
      "learning_rate": 1.9561224489795918e-05,
      "loss": 1.3987,
      "step": 3083
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.4951329231262207,
      "learning_rate": 1.9551020408163267e-05,
      "loss": 0.1309,
      "step": 3084
    },
    {
      "epoch": 1.2002,
      "grad_norm": 5.286871910095215,
      "learning_rate": 1.9540816326530613e-05,
      "loss": 0.9136,
      "step": 3085
    },
    {
      "epoch": 1.2004,
      "grad_norm": 7.972781658172607,
      "learning_rate": 1.9530612244897962e-05,
      "loss": 2.049,
      "step": 3086
    },
    {
      "epoch": 1.2006000000000001,
      "grad_norm": 7.049349308013916,
      "learning_rate": 1.9520408163265308e-05,
      "loss": 0.9573,
      "step": 3087
    },
    {
      "epoch": 1.2008,
      "grad_norm": 20.929615020751953,
      "learning_rate": 1.9510204081632654e-05,
      "loss": 3.7053,
      "step": 3088
    },
    {
      "epoch": 1.201,
      "grad_norm": 4.493311882019043,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.48,
      "step": 3089
    },
    {
      "epoch": 1.2012,
      "grad_norm": 4.102753639221191,
      "learning_rate": 1.948979591836735e-05,
      "loss": 0.8753,
      "step": 3090
    },
    {
      "epoch": 1.2014,
      "grad_norm": 5.101694583892822,
      "learning_rate": 1.9479591836734695e-05,
      "loss": 1.4985,
      "step": 3091
    },
    {
      "epoch": 1.2016,
      "grad_norm": 4.904029846191406,
      "learning_rate": 1.946938775510204e-05,
      "loss": 0.6939,
      "step": 3092
    },
    {
      "epoch": 1.2018,
      "grad_norm": 3.5857245922088623,
      "learning_rate": 1.945918367346939e-05,
      "loss": 0.2649,
      "step": 3093
    },
    {
      "epoch": 1.202,
      "grad_norm": 1.2739484310150146,
      "learning_rate": 1.9448979591836735e-05,
      "loss": 0.0412,
      "step": 3094
    },
    {
      "epoch": 1.2022,
      "grad_norm": 1.253515362739563,
      "learning_rate": 1.943877551020408e-05,
      "loss": 0.0718,
      "step": 3095
    },
    {
      "epoch": 1.2024,
      "grad_norm": 7.14802885055542,
      "learning_rate": 1.942857142857143e-05,
      "loss": 0.5822,
      "step": 3096
    },
    {
      "epoch": 1.2026,
      "grad_norm": 15.709341049194336,
      "learning_rate": 1.9418367346938776e-05,
      "loss": 2.2194,
      "step": 3097
    },
    {
      "epoch": 1.2028,
      "grad_norm": 7.323596477508545,
      "learning_rate": 1.9408163265306122e-05,
      "loss": 0.9511,
      "step": 3098
    },
    {
      "epoch": 1.203,
      "grad_norm": 23.892974853515625,
      "learning_rate": 1.939795918367347e-05,
      "loss": 3.5908,
      "step": 3099
    },
    {
      "epoch": 1.2032,
      "grad_norm": 1.673535704612732,
      "learning_rate": 1.9387755102040817e-05,
      "loss": 0.0876,
      "step": 3100
    },
    {
      "epoch": 1.2034,
      "grad_norm": 1.2727751731872559,
      "learning_rate": 1.9377551020408163e-05,
      "loss": 0.0544,
      "step": 3101
    },
    {
      "epoch": 1.2036,
      "grad_norm": 1.1253530979156494,
      "learning_rate": 1.9367346938775512e-05,
      "loss": 0.0778,
      "step": 3102
    },
    {
      "epoch": 1.2038,
      "grad_norm": 1.170177698135376,
      "learning_rate": 1.9357142857142858e-05,
      "loss": 0.0469,
      "step": 3103
    },
    {
      "epoch": 1.204,
      "grad_norm": 3.109325885772705,
      "learning_rate": 1.9346938775510203e-05,
      "loss": 0.2601,
      "step": 3104
    },
    {
      "epoch": 1.2042,
      "grad_norm": 9.423465728759766,
      "learning_rate": 1.933673469387755e-05,
      "loss": 2.4327,
      "step": 3105
    },
    {
      "epoch": 1.2044,
      "grad_norm": 15.970878601074219,
      "learning_rate": 1.9326530612244902e-05,
      "loss": 3.0273,
      "step": 3106
    },
    {
      "epoch": 1.2046000000000001,
      "grad_norm": 2.112394332885742,
      "learning_rate": 1.9316326530612248e-05,
      "loss": 0.1451,
      "step": 3107
    },
    {
      "epoch": 1.2048,
      "grad_norm": 2.8906023502349854,
      "learning_rate": 1.9306122448979593e-05,
      "loss": 0.2964,
      "step": 3108
    },
    {
      "epoch": 1.205,
      "grad_norm": 5.92968225479126,
      "learning_rate": 1.929591836734694e-05,
      "loss": 1.036,
      "step": 3109
    },
    {
      "epoch": 1.2052,
      "grad_norm": 8.283905982971191,
      "learning_rate": 1.928571428571429e-05,
      "loss": 3.0009,
      "step": 3110
    },
    {
      "epoch": 1.2054,
      "grad_norm": 2.000481605529785,
      "learning_rate": 1.9275510204081634e-05,
      "loss": 0.139,
      "step": 3111
    },
    {
      "epoch": 1.2056,
      "grad_norm": 7.078057289123535,
      "learning_rate": 1.926530612244898e-05,
      "loss": 1.4782,
      "step": 3112
    },
    {
      "epoch": 1.2058,
      "grad_norm": 9.67080307006836,
      "learning_rate": 1.925510204081633e-05,
      "loss": 2.2358,
      "step": 3113
    },
    {
      "epoch": 1.206,
      "grad_norm": 4.307836532592773,
      "learning_rate": 1.9244897959183675e-05,
      "loss": 0.1114,
      "step": 3114
    },
    {
      "epoch": 1.2062,
      "grad_norm": 5.063274383544922,
      "learning_rate": 1.923469387755102e-05,
      "loss": 0.3036,
      "step": 3115
    },
    {
      "epoch": 1.2064,
      "grad_norm": 1.4876151084899902,
      "learning_rate": 1.922448979591837e-05,
      "loss": 0.1199,
      "step": 3116
    },
    {
      "epoch": 1.2066,
      "grad_norm": 1.2019920349121094,
      "learning_rate": 1.9214285714285716e-05,
      "loss": 0.0794,
      "step": 3117
    },
    {
      "epoch": 1.2068,
      "grad_norm": 1.368234395980835,
      "learning_rate": 1.920408163265306e-05,
      "loss": 0.049,
      "step": 3118
    },
    {
      "epoch": 1.207,
      "grad_norm": 1.189566731452942,
      "learning_rate": 1.9193877551020407e-05,
      "loss": 0.061,
      "step": 3119
    },
    {
      "epoch": 1.2072,
      "grad_norm": 3.7581393718719482,
      "learning_rate": 1.9183673469387756e-05,
      "loss": 0.3694,
      "step": 3120
    },
    {
      "epoch": 1.2074,
      "grad_norm": 5.4688520431518555,
      "learning_rate": 1.9173469387755102e-05,
      "loss": 1.4537,
      "step": 3121
    },
    {
      "epoch": 1.2076,
      "grad_norm": 1.7890475988388062,
      "learning_rate": 1.9163265306122448e-05,
      "loss": 0.1057,
      "step": 3122
    },
    {
      "epoch": 1.2078,
      "grad_norm": 1.502280354499817,
      "learning_rate": 1.9153061224489797e-05,
      "loss": 0.1204,
      "step": 3123
    },
    {
      "epoch": 1.208,
      "grad_norm": 3.024693012237549,
      "learning_rate": 1.9142857142857143e-05,
      "loss": 0.0935,
      "step": 3124
    },
    {
      "epoch": 1.2082,
      "grad_norm": 10.54287052154541,
      "learning_rate": 1.913265306122449e-05,
      "loss": 0.7667,
      "step": 3125
    },
    {
      "epoch": 1.2084,
      "grad_norm": 14.563943862915039,
      "learning_rate": 1.9122448979591838e-05,
      "loss": 2.016,
      "step": 3126
    },
    {
      "epoch": 1.2086000000000001,
      "grad_norm": 1.2480556964874268,
      "learning_rate": 1.9112244897959184e-05,
      "loss": 0.0672,
      "step": 3127
    },
    {
      "epoch": 1.2088,
      "grad_norm": 6.187782287597656,
      "learning_rate": 1.910204081632653e-05,
      "loss": 0.9447,
      "step": 3128
    },
    {
      "epoch": 1.209,
      "grad_norm": 11.717743873596191,
      "learning_rate": 1.909183673469388e-05,
      "loss": 3.6272,
      "step": 3129
    },
    {
      "epoch": 1.2092,
      "grad_norm": 3.94441294670105,
      "learning_rate": 1.9081632653061228e-05,
      "loss": 0.4909,
      "step": 3130
    },
    {
      "epoch": 1.2094,
      "grad_norm": 5.881809711456299,
      "learning_rate": 1.9071428571428574e-05,
      "loss": 0.7599,
      "step": 3131
    },
    {
      "epoch": 1.2096,
      "grad_norm": 1.982026219367981,
      "learning_rate": 1.906122448979592e-05,
      "loss": 0.1455,
      "step": 3132
    },
    {
      "epoch": 1.2098,
      "grad_norm": 2.9072039127349854,
      "learning_rate": 1.905102040816327e-05,
      "loss": 0.5168,
      "step": 3133
    },
    {
      "epoch": 1.21,
      "grad_norm": 1.7545095682144165,
      "learning_rate": 1.9040816326530614e-05,
      "loss": 0.0821,
      "step": 3134
    },
    {
      "epoch": 1.2102,
      "grad_norm": 1.8028314113616943,
      "learning_rate": 1.903061224489796e-05,
      "loss": 0.1375,
      "step": 3135
    },
    {
      "epoch": 1.2104,
      "grad_norm": 4.452255725860596,
      "learning_rate": 1.9020408163265306e-05,
      "loss": 0.6812,
      "step": 3136
    },
    {
      "epoch": 1.2106,
      "grad_norm": 10.38892936706543,
      "learning_rate": 1.9010204081632655e-05,
      "loss": 1.4032,
      "step": 3137
    },
    {
      "epoch": 1.2107999999999999,
      "grad_norm": 9.387345314025879,
      "learning_rate": 1.9e-05,
      "loss": 3.0628,
      "step": 3138
    },
    {
      "epoch": 1.211,
      "grad_norm": 5.306800365447998,
      "learning_rate": 1.8989795918367347e-05,
      "loss": 0.6412,
      "step": 3139
    },
    {
      "epoch": 1.2112,
      "grad_norm": 3.5087547302246094,
      "learning_rate": 1.8979591836734696e-05,
      "loss": 0.2057,
      "step": 3140
    },
    {
      "epoch": 1.2114,
      "grad_norm": 5.041624546051025,
      "learning_rate": 1.8969387755102042e-05,
      "loss": 0.6413,
      "step": 3141
    },
    {
      "epoch": 1.2116,
      "grad_norm": 1.5640407800674438,
      "learning_rate": 1.8959183673469388e-05,
      "loss": 0.1225,
      "step": 3142
    },
    {
      "epoch": 1.2118,
      "grad_norm": 1.5499358177185059,
      "learning_rate": 1.8948979591836737e-05,
      "loss": 0.1756,
      "step": 3143
    },
    {
      "epoch": 1.212,
      "grad_norm": 3.4542722702026367,
      "learning_rate": 1.8938775510204083e-05,
      "loss": 0.0812,
      "step": 3144
    },
    {
      "epoch": 1.2122,
      "grad_norm": 6.98813009262085,
      "learning_rate": 1.892857142857143e-05,
      "loss": 0.3454,
      "step": 3145
    },
    {
      "epoch": 1.2124,
      "grad_norm": 2.5921764373779297,
      "learning_rate": 1.8918367346938774e-05,
      "loss": 0.463,
      "step": 3146
    },
    {
      "epoch": 1.2126000000000001,
      "grad_norm": 0.8930935859680176,
      "learning_rate": 1.8908163265306123e-05,
      "loss": 0.0271,
      "step": 3147
    },
    {
      "epoch": 1.2128,
      "grad_norm": 2.3683059215545654,
      "learning_rate": 1.889795918367347e-05,
      "loss": 0.2736,
      "step": 3148
    },
    {
      "epoch": 1.213,
      "grad_norm": 5.169450283050537,
      "learning_rate": 1.8887755102040815e-05,
      "loss": 0.8165,
      "step": 3149
    },
    {
      "epoch": 1.2132,
      "grad_norm": 2.7456088066101074,
      "learning_rate": 1.8877551020408164e-05,
      "loss": 0.4483,
      "step": 3150
    },
    {
      "epoch": 1.2134,
      "grad_norm": 2.067122220993042,
      "learning_rate": 1.886734693877551e-05,
      "loss": 0.2132,
      "step": 3151
    },
    {
      "epoch": 1.2136,
      "grad_norm": 3.7471401691436768,
      "learning_rate": 1.885714285714286e-05,
      "loss": 0.6301,
      "step": 3152
    },
    {
      "epoch": 1.2138,
      "grad_norm": 2.1313626766204834,
      "learning_rate": 1.8846938775510205e-05,
      "loss": 0.1247,
      "step": 3153
    },
    {
      "epoch": 1.214,
      "grad_norm": 8.123455047607422,
      "learning_rate": 1.8836734693877554e-05,
      "loss": 0.9875,
      "step": 3154
    },
    {
      "epoch": 1.2142,
      "grad_norm": 11.68111801147461,
      "learning_rate": 1.88265306122449e-05,
      "loss": 1.5304,
      "step": 3155
    },
    {
      "epoch": 1.2144,
      "grad_norm": 7.343622207641602,
      "learning_rate": 1.8816326530612246e-05,
      "loss": 1.0192,
      "step": 3156
    },
    {
      "epoch": 1.2146,
      "grad_norm": 6.457760334014893,
      "learning_rate": 1.8806122448979595e-05,
      "loss": 3.1121,
      "step": 3157
    },
    {
      "epoch": 1.2147999999999999,
      "grad_norm": 2.7293622493743896,
      "learning_rate": 1.879591836734694e-05,
      "loss": 0.5017,
      "step": 3158
    },
    {
      "epoch": 1.215,
      "grad_norm": 1.0072730779647827,
      "learning_rate": 1.8785714285714286e-05,
      "loss": 0.0282,
      "step": 3159
    },
    {
      "epoch": 1.2152,
      "grad_norm": 3.2834362983703613,
      "learning_rate": 1.8775510204081632e-05,
      "loss": 1.0516,
      "step": 3160
    },
    {
      "epoch": 1.2154,
      "grad_norm": 11.370521545410156,
      "learning_rate": 1.876530612244898e-05,
      "loss": 3.0999,
      "step": 3161
    },
    {
      "epoch": 1.2156,
      "grad_norm": 1.2034019231796265,
      "learning_rate": 1.8755102040816327e-05,
      "loss": 0.0787,
      "step": 3162
    },
    {
      "epoch": 1.2158,
      "grad_norm": 1.3203438520431519,
      "learning_rate": 1.8744897959183673e-05,
      "loss": 0.0865,
      "step": 3163
    },
    {
      "epoch": 1.216,
      "grad_norm": 1.1307048797607422,
      "learning_rate": 1.8734693877551022e-05,
      "loss": 0.071,
      "step": 3164
    },
    {
      "epoch": 1.2162,
      "grad_norm": 0.9662017822265625,
      "learning_rate": 1.8724489795918368e-05,
      "loss": 0.0276,
      "step": 3165
    },
    {
      "epoch": 1.2164,
      "grad_norm": 1.1824790239334106,
      "learning_rate": 1.8714285714285714e-05,
      "loss": 0.0763,
      "step": 3166
    },
    {
      "epoch": 1.2166,
      "grad_norm": 1.7388250827789307,
      "learning_rate": 1.8704081632653063e-05,
      "loss": 0.1711,
      "step": 3167
    },
    {
      "epoch": 1.2168,
      "grad_norm": 2.5671846866607666,
      "learning_rate": 1.869387755102041e-05,
      "loss": 0.4834,
      "step": 3168
    },
    {
      "epoch": 1.217,
      "grad_norm": 2.9493188858032227,
      "learning_rate": 1.8683673469387754e-05,
      "loss": 0.4221,
      "step": 3169
    },
    {
      "epoch": 1.2172,
      "grad_norm": 6.83276891708374,
      "learning_rate": 1.8673469387755104e-05,
      "loss": 1.5277,
      "step": 3170
    },
    {
      "epoch": 1.2174,
      "grad_norm": 5.947171688079834,
      "learning_rate": 1.866326530612245e-05,
      "loss": 0.7498,
      "step": 3171
    },
    {
      "epoch": 1.2176,
      "grad_norm": 1.5263842344284058,
      "learning_rate": 1.8653061224489795e-05,
      "loss": 0.1066,
      "step": 3172
    },
    {
      "epoch": 1.2178,
      "grad_norm": 7.348236083984375,
      "learning_rate": 1.864285714285714e-05,
      "loss": 1.0489,
      "step": 3173
    },
    {
      "epoch": 1.218,
      "grad_norm": 12.764039039611816,
      "learning_rate": 1.8632653061224494e-05,
      "loss": 2.8689,
      "step": 3174
    },
    {
      "epoch": 1.2182,
      "grad_norm": 9.676426887512207,
      "learning_rate": 1.862244897959184e-05,
      "loss": 1.3229,
      "step": 3175
    },
    {
      "epoch": 1.2184,
      "grad_norm": 1.6813863515853882,
      "learning_rate": 1.8612244897959185e-05,
      "loss": 0.0723,
      "step": 3176
    },
    {
      "epoch": 1.2186,
      "grad_norm": 3.3537490367889404,
      "learning_rate": 1.860204081632653e-05,
      "loss": 0.3726,
      "step": 3177
    },
    {
      "epoch": 1.2187999999999999,
      "grad_norm": 5.313870429992676,
      "learning_rate": 1.859183673469388e-05,
      "loss": 0.7126,
      "step": 3178
    },
    {
      "epoch": 1.219,
      "grad_norm": 1.637223720550537,
      "learning_rate": 1.8581632653061226e-05,
      "loss": 0.0945,
      "step": 3179
    },
    {
      "epoch": 1.2192,
      "grad_norm": 2.391230821609497,
      "learning_rate": 1.8571428571428572e-05,
      "loss": 0.1888,
      "step": 3180
    },
    {
      "epoch": 1.2194,
      "grad_norm": 4.959512233734131,
      "learning_rate": 1.856122448979592e-05,
      "loss": 0.6768,
      "step": 3181
    },
    {
      "epoch": 1.2196,
      "grad_norm": 4.827409267425537,
      "learning_rate": 1.8551020408163267e-05,
      "loss": 1.2034,
      "step": 3182
    },
    {
      "epoch": 1.2198,
      "grad_norm": 9.361382484436035,
      "learning_rate": 1.8540816326530613e-05,
      "loss": 2.1611,
      "step": 3183
    },
    {
      "epoch": 1.22,
      "grad_norm": 10.543557167053223,
      "learning_rate": 1.853061224489796e-05,
      "loss": 2.0855,
      "step": 3184
    },
    {
      "epoch": 1.2202,
      "grad_norm": 17.42910385131836,
      "learning_rate": 1.8520408163265307e-05,
      "loss": 1.6268,
      "step": 3185
    },
    {
      "epoch": 1.2204,
      "grad_norm": 1.5924044847488403,
      "learning_rate": 1.8510204081632653e-05,
      "loss": 0.1284,
      "step": 3186
    },
    {
      "epoch": 1.2206,
      "grad_norm": 1.471555471420288,
      "learning_rate": 1.85e-05,
      "loss": 0.1076,
      "step": 3187
    },
    {
      "epoch": 1.2208,
      "grad_norm": 1.4289698600769043,
      "learning_rate": 1.8489795918367348e-05,
      "loss": 0.105,
      "step": 3188
    },
    {
      "epoch": 1.221,
      "grad_norm": 1.8644956350326538,
      "learning_rate": 1.8479591836734694e-05,
      "loss": 0.1873,
      "step": 3189
    },
    {
      "epoch": 1.2212,
      "grad_norm": 3.7627358436584473,
      "learning_rate": 1.846938775510204e-05,
      "loss": 0.7693,
      "step": 3190
    },
    {
      "epoch": 1.2214,
      "grad_norm": 5.294459819793701,
      "learning_rate": 1.845918367346939e-05,
      "loss": 0.7467,
      "step": 3191
    },
    {
      "epoch": 1.2216,
      "grad_norm": 3.0752527713775635,
      "learning_rate": 1.8448979591836735e-05,
      "loss": 0.5994,
      "step": 3192
    },
    {
      "epoch": 1.2218,
      "grad_norm": 4.9536967277526855,
      "learning_rate": 1.843877551020408e-05,
      "loss": 0.9815,
      "step": 3193
    },
    {
      "epoch": 1.222,
      "grad_norm": 9.496366500854492,
      "learning_rate": 1.842857142857143e-05,
      "loss": 2.0773,
      "step": 3194
    },
    {
      "epoch": 1.2222,
      "grad_norm": 5.776390552520752,
      "learning_rate": 1.8418367346938776e-05,
      "loss": 1.5251,
      "step": 3195
    },
    {
      "epoch": 1.2224,
      "grad_norm": 3.0072059631347656,
      "learning_rate": 1.840816326530612e-05,
      "loss": 0.5866,
      "step": 3196
    },
    {
      "epoch": 1.2226,
      "grad_norm": 1.4602209329605103,
      "learning_rate": 1.839795918367347e-05,
      "loss": 0.1017,
      "step": 3197
    },
    {
      "epoch": 1.2227999999999999,
      "grad_norm": 1.5261403322219849,
      "learning_rate": 1.838775510204082e-05,
      "loss": 0.0664,
      "step": 3198
    },
    {
      "epoch": 1.223,
      "grad_norm": 1.498910665512085,
      "learning_rate": 1.8377551020408165e-05,
      "loss": 0.1027,
      "step": 3199
    },
    {
      "epoch": 1.2232,
      "grad_norm": 1.4415857791900635,
      "learning_rate": 1.836734693877551e-05,
      "loss": 0.0466,
      "step": 3200
    },
    {
      "epoch": 1.2234,
      "grad_norm": 1.0863139629364014,
      "learning_rate": 1.835714285714286e-05,
      "loss": 0.0707,
      "step": 3201
    },
    {
      "epoch": 1.2236,
      "grad_norm": 1.3481663465499878,
      "learning_rate": 1.8346938775510206e-05,
      "loss": 0.053,
      "step": 3202
    },
    {
      "epoch": 1.2238,
      "grad_norm": 9.405648231506348,
      "learning_rate": 1.8336734693877552e-05,
      "loss": 0.6962,
      "step": 3203
    },
    {
      "epoch": 1.224,
      "grad_norm": 17.991626739501953,
      "learning_rate": 1.8326530612244898e-05,
      "loss": 2.3324,
      "step": 3204
    },
    {
      "epoch": 1.2242,
      "grad_norm": 5.935586452484131,
      "learning_rate": 1.8316326530612247e-05,
      "loss": 0.7266,
      "step": 3205
    },
    {
      "epoch": 1.2244,
      "grad_norm": 6.7307610511779785,
      "learning_rate": 1.8306122448979593e-05,
      "loss": 0.4605,
      "step": 3206
    },
    {
      "epoch": 1.2246,
      "grad_norm": 10.014396667480469,
      "learning_rate": 1.829591836734694e-05,
      "loss": 1.3713,
      "step": 3207
    },
    {
      "epoch": 1.2248,
      "grad_norm": 2.9402997493743896,
      "learning_rate": 1.8285714285714288e-05,
      "loss": 0.5045,
      "step": 3208
    },
    {
      "epoch": 1.225,
      "grad_norm": 1.3361576795578003,
      "learning_rate": 1.8275510204081634e-05,
      "loss": 0.102,
      "step": 3209
    },
    {
      "epoch": 1.2252,
      "grad_norm": 1.4597071409225464,
      "learning_rate": 1.826530612244898e-05,
      "loss": 0.0999,
      "step": 3210
    },
    {
      "epoch": 1.2254,
      "grad_norm": 1.4014943838119507,
      "learning_rate": 1.825510204081633e-05,
      "loss": 0.0607,
      "step": 3211
    },
    {
      "epoch": 1.2256,
      "grad_norm": 1.4707921743392944,
      "learning_rate": 1.8244897959183674e-05,
      "loss": 0.085,
      "step": 3212
    },
    {
      "epoch": 1.2258,
      "grad_norm": 1.6909431219100952,
      "learning_rate": 1.823469387755102e-05,
      "loss": 0.1145,
      "step": 3213
    },
    {
      "epoch": 1.226,
      "grad_norm": 2.412825584411621,
      "learning_rate": 1.8224489795918366e-05,
      "loss": 0.1601,
      "step": 3214
    },
    {
      "epoch": 1.2262,
      "grad_norm": 2.01613187789917,
      "learning_rate": 1.8214285714285715e-05,
      "loss": 0.1601,
      "step": 3215
    },
    {
      "epoch": 1.2264,
      "grad_norm": 4.6927900314331055,
      "learning_rate": 1.820408163265306e-05,
      "loss": 0.8213,
      "step": 3216
    },
    {
      "epoch": 1.2266,
      "grad_norm": 7.865774154663086,
      "learning_rate": 1.8193877551020407e-05,
      "loss": 1.802,
      "step": 3217
    },
    {
      "epoch": 1.2268,
      "grad_norm": 11.120959281921387,
      "learning_rate": 1.8183673469387756e-05,
      "loss": 3.1244,
      "step": 3218
    },
    {
      "epoch": 1.227,
      "grad_norm": 1.3915950059890747,
      "learning_rate": 1.81734693877551e-05,
      "loss": 0.0592,
      "step": 3219
    },
    {
      "epoch": 1.2272,
      "grad_norm": 2.2663471698760986,
      "learning_rate": 1.816326530612245e-05,
      "loss": 0.1695,
      "step": 3220
    },
    {
      "epoch": 1.2274,
      "grad_norm": 5.7890801429748535,
      "learning_rate": 1.8153061224489797e-05,
      "loss": 0.9364,
      "step": 3221
    },
    {
      "epoch": 1.2276,
      "grad_norm": 4.851029872894287,
      "learning_rate": 1.8142857142857146e-05,
      "loss": 0.7313,
      "step": 3222
    },
    {
      "epoch": 1.2278,
      "grad_norm": 1.829401969909668,
      "learning_rate": 1.813265306122449e-05,
      "loss": 0.1682,
      "step": 3223
    },
    {
      "epoch": 1.228,
      "grad_norm": 2.4404027462005615,
      "learning_rate": 1.8122448979591837e-05,
      "loss": 0.1551,
      "step": 3224
    },
    {
      "epoch": 1.2282,
      "grad_norm": 3.285078763961792,
      "learning_rate": 1.8112244897959187e-05,
      "loss": 0.4083,
      "step": 3225
    },
    {
      "epoch": 1.2284,
      "grad_norm": 5.276399612426758,
      "learning_rate": 1.8102040816326532e-05,
      "loss": 1.4821,
      "step": 3226
    },
    {
      "epoch": 1.2286,
      "grad_norm": 2.0326952934265137,
      "learning_rate": 1.8091836734693878e-05,
      "loss": 0.2251,
      "step": 3227
    },
    {
      "epoch": 1.2288000000000001,
      "grad_norm": 5.430430889129639,
      "learning_rate": 1.8081632653061227e-05,
      "loss": 1.5465,
      "step": 3228
    },
    {
      "epoch": 1.229,
      "grad_norm": 10.54371166229248,
      "learning_rate": 1.8071428571428573e-05,
      "loss": 1.6414,
      "step": 3229
    },
    {
      "epoch": 1.2292,
      "grad_norm": 1.2699222564697266,
      "learning_rate": 1.806122448979592e-05,
      "loss": 0.0412,
      "step": 3230
    },
    {
      "epoch": 1.2294,
      "grad_norm": 2.6544322967529297,
      "learning_rate": 1.8051020408163265e-05,
      "loss": 0.2643,
      "step": 3231
    },
    {
      "epoch": 1.2296,
      "grad_norm": 2.2963390350341797,
      "learning_rate": 1.8040816326530614e-05,
      "loss": 0.5428,
      "step": 3232
    },
    {
      "epoch": 1.2298,
      "grad_norm": 1.345909595489502,
      "learning_rate": 1.803061224489796e-05,
      "loss": 0.0937,
      "step": 3233
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.867246150970459,
      "learning_rate": 1.8020408163265305e-05,
      "loss": 0.1299,
      "step": 3234
    },
    {
      "epoch": 1.2302,
      "grad_norm": 1.998002529144287,
      "learning_rate": 1.8010204081632655e-05,
      "loss": 0.2057,
      "step": 3235
    },
    {
      "epoch": 1.2304,
      "grad_norm": 3.9584403038024902,
      "learning_rate": 1.8e-05,
      "loss": 0.6425,
      "step": 3236
    },
    {
      "epoch": 1.2306,
      "grad_norm": 6.284635543823242,
      "learning_rate": 1.7989795918367346e-05,
      "loss": 0.7627,
      "step": 3237
    },
    {
      "epoch": 1.2308,
      "grad_norm": 2.366175651550293,
      "learning_rate": 1.7979591836734695e-05,
      "loss": 0.1369,
      "step": 3238
    },
    {
      "epoch": 1.231,
      "grad_norm": 5.042405128479004,
      "learning_rate": 1.796938775510204e-05,
      "loss": 0.6354,
      "step": 3239
    },
    {
      "epoch": 1.2312,
      "grad_norm": 2.543888568878174,
      "learning_rate": 1.7959183673469387e-05,
      "loss": 0.1593,
      "step": 3240
    },
    {
      "epoch": 1.2314,
      "grad_norm": 4.086974620819092,
      "learning_rate": 1.7948979591836733e-05,
      "loss": 0.2952,
      "step": 3241
    },
    {
      "epoch": 1.2316,
      "grad_norm": 2.588975429534912,
      "learning_rate": 1.7938775510204085e-05,
      "loss": 0.1849,
      "step": 3242
    },
    {
      "epoch": 1.2318,
      "grad_norm": 8.656416893005371,
      "learning_rate": 1.792857142857143e-05,
      "loss": 0.7699,
      "step": 3243
    },
    {
      "epoch": 1.232,
      "grad_norm": 4.926800727844238,
      "learning_rate": 1.7918367346938777e-05,
      "loss": 0.6368,
      "step": 3244
    },
    {
      "epoch": 1.2322,
      "grad_norm": 1.3040525913238525,
      "learning_rate": 1.7908163265306123e-05,
      "loss": 0.0759,
      "step": 3245
    },
    {
      "epoch": 1.2324,
      "grad_norm": 1.3231332302093506,
      "learning_rate": 1.7897959183673472e-05,
      "loss": 0.0512,
      "step": 3246
    },
    {
      "epoch": 1.2326,
      "grad_norm": 1.4585577249526978,
      "learning_rate": 1.7887755102040818e-05,
      "loss": 0.0938,
      "step": 3247
    },
    {
      "epoch": 1.2328000000000001,
      "grad_norm": 2.9919850826263428,
      "learning_rate": 1.7877551020408164e-05,
      "loss": 0.5489,
      "step": 3248
    },
    {
      "epoch": 1.233,
      "grad_norm": 1.8937764167785645,
      "learning_rate": 1.7867346938775513e-05,
      "loss": 0.1754,
      "step": 3249
    },
    {
      "epoch": 1.2332,
      "grad_norm": 2.412397861480713,
      "learning_rate": 1.785714285714286e-05,
      "loss": 0.2007,
      "step": 3250
    },
    {
      "epoch": 1.2334,
      "grad_norm": 4.665653228759766,
      "learning_rate": 1.7846938775510204e-05,
      "loss": 0.4089,
      "step": 3251
    },
    {
      "epoch": 1.2336,
      "grad_norm": 6.121510028839111,
      "learning_rate": 1.7836734693877553e-05,
      "loss": 0.786,
      "step": 3252
    },
    {
      "epoch": 1.2338,
      "grad_norm": 4.21718168258667,
      "learning_rate": 1.78265306122449e-05,
      "loss": 0.5882,
      "step": 3253
    },
    {
      "epoch": 1.234,
      "grad_norm": 1.6234577894210815,
      "learning_rate": 1.7816326530612245e-05,
      "loss": 0.1117,
      "step": 3254
    },
    {
      "epoch": 1.2342,
      "grad_norm": 1.4547427892684937,
      "learning_rate": 1.780612244897959e-05,
      "loss": 0.0564,
      "step": 3255
    },
    {
      "epoch": 1.2344,
      "grad_norm": 1.4127044677734375,
      "learning_rate": 1.779591836734694e-05,
      "loss": 0.1022,
      "step": 3256
    },
    {
      "epoch": 1.2346,
      "grad_norm": 5.04308557510376,
      "learning_rate": 1.7785714285714286e-05,
      "loss": 0.8675,
      "step": 3257
    },
    {
      "epoch": 1.2348,
      "grad_norm": 8.409255981445312,
      "learning_rate": 1.777551020408163e-05,
      "loss": 1.7431,
      "step": 3258
    },
    {
      "epoch": 1.2349999999999999,
      "grad_norm": 3.6114909648895264,
      "learning_rate": 1.776530612244898e-05,
      "loss": 0.8014,
      "step": 3259
    },
    {
      "epoch": 1.2352,
      "grad_norm": 4.705353736877441,
      "learning_rate": 1.7755102040816327e-05,
      "loss": 1.4045,
      "step": 3260
    },
    {
      "epoch": 1.2354,
      "grad_norm": 1.5038385391235352,
      "learning_rate": 1.7744897959183672e-05,
      "loss": 0.1013,
      "step": 3261
    },
    {
      "epoch": 1.2356,
      "grad_norm": 2.7289583683013916,
      "learning_rate": 1.773469387755102e-05,
      "loss": 0.479,
      "step": 3262
    },
    {
      "epoch": 1.2358,
      "grad_norm": 6.702466011047363,
      "learning_rate": 1.7724489795918367e-05,
      "loss": 0.9266,
      "step": 3263
    },
    {
      "epoch": 1.236,
      "grad_norm": 12.450255393981934,
      "learning_rate": 1.7714285714285713e-05,
      "loss": 3.5627,
      "step": 3264
    },
    {
      "epoch": 1.2362,
      "grad_norm": 3.067748785018921,
      "learning_rate": 1.7704081632653062e-05,
      "loss": 0.3582,
      "step": 3265
    },
    {
      "epoch": 1.2364,
      "grad_norm": 2.5850586891174316,
      "learning_rate": 1.769387755102041e-05,
      "loss": 0.1721,
      "step": 3266
    },
    {
      "epoch": 1.2366,
      "grad_norm": 5.197973728179932,
      "learning_rate": 1.7683673469387757e-05,
      "loss": 0.6389,
      "step": 3267
    },
    {
      "epoch": 1.2368000000000001,
      "grad_norm": 1.110595464706421,
      "learning_rate": 1.7673469387755103e-05,
      "loss": 0.033,
      "step": 3268
    },
    {
      "epoch": 1.237,
      "grad_norm": 1.408860445022583,
      "learning_rate": 1.7663265306122452e-05,
      "loss": 0.0923,
      "step": 3269
    },
    {
      "epoch": 1.2372,
      "grad_norm": 1.7110408544540405,
      "learning_rate": 1.7653061224489798e-05,
      "loss": 0.1991,
      "step": 3270
    },
    {
      "epoch": 1.2374,
      "grad_norm": 2.680989980697632,
      "learning_rate": 1.7642857142857144e-05,
      "loss": 0.4824,
      "step": 3271
    },
    {
      "epoch": 1.2376,
      "grad_norm": 1.338800072669983,
      "learning_rate": 1.763265306122449e-05,
      "loss": 0.0762,
      "step": 3272
    },
    {
      "epoch": 1.2378,
      "grad_norm": 3.547950506210327,
      "learning_rate": 1.762244897959184e-05,
      "loss": 0.4761,
      "step": 3273
    },
    {
      "epoch": 1.238,
      "grad_norm": 5.276917934417725,
      "learning_rate": 1.7612244897959185e-05,
      "loss": 1.4496,
      "step": 3274
    },
    {
      "epoch": 1.2382,
      "grad_norm": 1.638073444366455,
      "learning_rate": 1.760204081632653e-05,
      "loss": 0.1042,
      "step": 3275
    },
    {
      "epoch": 1.2384,
      "grad_norm": 0.9599982500076294,
      "learning_rate": 1.759183673469388e-05,
      "loss": 0.0289,
      "step": 3276
    },
    {
      "epoch": 1.2386,
      "grad_norm": 1.8225157260894775,
      "learning_rate": 1.7581632653061225e-05,
      "loss": 0.133,
      "step": 3277
    },
    {
      "epoch": 1.2388,
      "grad_norm": 1.4022389650344849,
      "learning_rate": 1.757142857142857e-05,
      "loss": 0.0832,
      "step": 3278
    },
    {
      "epoch": 1.2389999999999999,
      "grad_norm": 1.6842702627182007,
      "learning_rate": 1.756122448979592e-05,
      "loss": 0.1013,
      "step": 3279
    },
    {
      "epoch": 1.2392,
      "grad_norm": 1.0950801372528076,
      "learning_rate": 1.7551020408163266e-05,
      "loss": 0.0562,
      "step": 3280
    },
    {
      "epoch": 1.2394,
      "grad_norm": 5.457022666931152,
      "learning_rate": 1.7540816326530612e-05,
      "loss": 0.8113,
      "step": 3281
    },
    {
      "epoch": 1.2396,
      "grad_norm": 11.909248352050781,
      "learning_rate": 1.7530612244897958e-05,
      "loss": 2.2316,
      "step": 3282
    },
    {
      "epoch": 1.2398,
      "grad_norm": 7.8928422927856445,
      "learning_rate": 1.7520408163265307e-05,
      "loss": 1.4548,
      "step": 3283
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.9538159370422363,
      "learning_rate": 1.7510204081632653e-05,
      "loss": 0.6393,
      "step": 3284
    },
    {
      "epoch": 1.2402,
      "grad_norm": 1.4128315448760986,
      "learning_rate": 1.75e-05,
      "loss": 0.0856,
      "step": 3285
    },
    {
      "epoch": 1.2404,
      "grad_norm": 1.7415316104888916,
      "learning_rate": 1.7489795918367348e-05,
      "loss": 0.1573,
      "step": 3286
    },
    {
      "epoch": 1.2406,
      "grad_norm": 3.041987419128418,
      "learning_rate": 1.7479591836734693e-05,
      "loss": 0.5616,
      "step": 3287
    },
    {
      "epoch": 1.2408,
      "grad_norm": 2.1510403156280518,
      "learning_rate": 1.7469387755102043e-05,
      "loss": 0.1567,
      "step": 3288
    },
    {
      "epoch": 1.241,
      "grad_norm": 4.927199363708496,
      "learning_rate": 1.745918367346939e-05,
      "loss": 0.6416,
      "step": 3289
    },
    {
      "epoch": 1.2412,
      "grad_norm": 1.535884141921997,
      "learning_rate": 1.7448979591836738e-05,
      "loss": 0.1165,
      "step": 3290
    },
    {
      "epoch": 1.2414,
      "grad_norm": 1.228538155555725,
      "learning_rate": 1.7438775510204083e-05,
      "loss": 0.0394,
      "step": 3291
    },
    {
      "epoch": 1.2416,
      "grad_norm": 1.5323854684829712,
      "learning_rate": 1.742857142857143e-05,
      "loss": 0.1044,
      "step": 3292
    },
    {
      "epoch": 1.2418,
      "grad_norm": 1.7505450248718262,
      "learning_rate": 1.741836734693878e-05,
      "loss": 0.0901,
      "step": 3293
    },
    {
      "epoch": 1.242,
      "grad_norm": 1.385218858718872,
      "learning_rate": 1.7408163265306124e-05,
      "loss": 0.0434,
      "step": 3294
    },
    {
      "epoch": 1.2422,
      "grad_norm": 1.3750401735305786,
      "learning_rate": 1.739795918367347e-05,
      "loss": 0.0524,
      "step": 3295
    },
    {
      "epoch": 1.2424,
      "grad_norm": 4.895515441894531,
      "learning_rate": 1.738775510204082e-05,
      "loss": 0.9659,
      "step": 3296
    },
    {
      "epoch": 1.2426,
      "grad_norm": 7.652602672576904,
      "learning_rate": 1.7377551020408165e-05,
      "loss": 1.8379,
      "step": 3297
    },
    {
      "epoch": 1.2428,
      "grad_norm": 1.8131300210952759,
      "learning_rate": 1.736734693877551e-05,
      "loss": 0.1161,
      "step": 3298
    },
    {
      "epoch": 1.2429999999999999,
      "grad_norm": 1.3964576721191406,
      "learning_rate": 1.7357142857142856e-05,
      "loss": 0.0571,
      "step": 3299
    },
    {
      "epoch": 1.2432,
      "grad_norm": 1.4084566831588745,
      "learning_rate": 1.7346938775510206e-05,
      "loss": 0.1014,
      "step": 3300
    },
    {
      "epoch": 1.2434,
      "grad_norm": 1.425021767616272,
      "learning_rate": 1.733673469387755e-05,
      "loss": 0.0705,
      "step": 3301
    },
    {
      "epoch": 1.2436,
      "grad_norm": 4.891021251678467,
      "learning_rate": 1.7326530612244897e-05,
      "loss": 0.9931,
      "step": 3302
    },
    {
      "epoch": 1.2438,
      "grad_norm": 9.658141136169434,
      "learning_rate": 1.7316326530612246e-05,
      "loss": 3.5513,
      "step": 3303
    },
    {
      "epoch": 1.244,
      "grad_norm": 2.6448793411254883,
      "learning_rate": 1.7306122448979592e-05,
      "loss": 0.4798,
      "step": 3304
    },
    {
      "epoch": 1.2442,
      "grad_norm": 1.2641750574111938,
      "learning_rate": 1.7295918367346938e-05,
      "loss": 0.0766,
      "step": 3305
    },
    {
      "epoch": 1.2444,
      "grad_norm": 5.329100608825684,
      "learning_rate": 1.7285714285714287e-05,
      "loss": 0.9358,
      "step": 3306
    },
    {
      "epoch": 1.2446,
      "grad_norm": 8.251531600952148,
      "learning_rate": 1.7275510204081633e-05,
      "loss": 3.4661,
      "step": 3307
    },
    {
      "epoch": 1.2448,
      "grad_norm": 1.6422741413116455,
      "learning_rate": 1.726530612244898e-05,
      "loss": 0.0595,
      "step": 3308
    },
    {
      "epoch": 1.245,
      "grad_norm": 1.5212047100067139,
      "learning_rate": 1.7255102040816325e-05,
      "loss": 0.1585,
      "step": 3309
    },
    {
      "epoch": 1.2452,
      "grad_norm": 2.3259589672088623,
      "learning_rate": 1.7244897959183677e-05,
      "loss": 0.4884,
      "step": 3310
    },
    {
      "epoch": 1.2454,
      "grad_norm": 1.1222413778305054,
      "learning_rate": 1.7234693877551023e-05,
      "loss": 0.0563,
      "step": 3311
    },
    {
      "epoch": 1.2456,
      "grad_norm": 1.3256467580795288,
      "learning_rate": 1.722448979591837e-05,
      "loss": 0.086,
      "step": 3312
    },
    {
      "epoch": 1.2458,
      "grad_norm": 1.7415765523910522,
      "learning_rate": 1.7214285714285715e-05,
      "loss": 0.1463,
      "step": 3313
    },
    {
      "epoch": 1.246,
      "grad_norm": 2.6794984340667725,
      "learning_rate": 1.7204081632653064e-05,
      "loss": 0.4776,
      "step": 3314
    },
    {
      "epoch": 1.2462,
      "grad_norm": 1.9408143758773804,
      "learning_rate": 1.719387755102041e-05,
      "loss": 0.1916,
      "step": 3315
    },
    {
      "epoch": 1.2464,
      "grad_norm": 9.202816009521484,
      "learning_rate": 1.7183673469387755e-05,
      "loss": 0.7523,
      "step": 3316
    },
    {
      "epoch": 1.2466,
      "grad_norm": 7.594089984893799,
      "learning_rate": 1.7173469387755104e-05,
      "loss": 1.3364,
      "step": 3317
    },
    {
      "epoch": 1.2468,
      "grad_norm": 12.51789379119873,
      "learning_rate": 1.716326530612245e-05,
      "loss": 3.4165,
      "step": 3318
    },
    {
      "epoch": 1.2469999999999999,
      "grad_norm": 1.0752918720245361,
      "learning_rate": 1.7153061224489796e-05,
      "loss": 0.0322,
      "step": 3319
    },
    {
      "epoch": 1.2472,
      "grad_norm": 2.296510934829712,
      "learning_rate": 1.7142857142857145e-05,
      "loss": 0.1935,
      "step": 3320
    },
    {
      "epoch": 1.2474,
      "grad_norm": 5.305103778839111,
      "learning_rate": 1.713265306122449e-05,
      "loss": 0.6499,
      "step": 3321
    },
    {
      "epoch": 1.2476,
      "grad_norm": 1.4535934925079346,
      "learning_rate": 1.7122448979591837e-05,
      "loss": 0.0932,
      "step": 3322
    },
    {
      "epoch": 1.2478,
      "grad_norm": 3.06929612159729,
      "learning_rate": 1.7112244897959183e-05,
      "loss": 0.3918,
      "step": 3323
    },
    {
      "epoch": 1.248,
      "grad_norm": 5.280209541320801,
      "learning_rate": 1.7102040816326532e-05,
      "loss": 1.4359,
      "step": 3324
    },
    {
      "epoch": 1.2482,
      "grad_norm": 1.0554393529891968,
      "learning_rate": 1.7091836734693878e-05,
      "loss": 0.0297,
      "step": 3325
    },
    {
      "epoch": 1.2484,
      "grad_norm": 1.6997267007827759,
      "learning_rate": 1.7081632653061223e-05,
      "loss": 0.0987,
      "step": 3326
    },
    {
      "epoch": 1.2486,
      "grad_norm": 2.0310628414154053,
      "learning_rate": 1.7071428571428573e-05,
      "loss": 0.1564,
      "step": 3327
    },
    {
      "epoch": 1.2488,
      "grad_norm": 4.670139312744141,
      "learning_rate": 1.706122448979592e-05,
      "loss": 0.8549,
      "step": 3328
    },
    {
      "epoch": 1.249,
      "grad_norm": 4.195288181304932,
      "learning_rate": 1.7051020408163264e-05,
      "loss": 0.7412,
      "step": 3329
    },
    {
      "epoch": 1.2492,
      "grad_norm": 11.199814796447754,
      "learning_rate": 1.7040816326530613e-05,
      "loss": 1.9341,
      "step": 3330
    },
    {
      "epoch": 1.2494,
      "grad_norm": 15.108905792236328,
      "learning_rate": 1.703061224489796e-05,
      "loss": 1.6621,
      "step": 3331
    },
    {
      "epoch": 1.2496,
      "grad_norm": 9.078606605529785,
      "learning_rate": 1.7020408163265305e-05,
      "loss": 0.401,
      "step": 3332
    },
    {
      "epoch": 1.2498,
      "grad_norm": 17.929407119750977,
      "learning_rate": 1.7010204081632654e-05,
      "loss": 1.9839,
      "step": 3333
    },
    {
      "epoch": 1.25,
      "grad_norm": 5.080728530883789,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 1.3457,
      "step": 3334
    },
    {
      "epoch": 1.2502,
      "grad_norm": 1.418836236000061,
      "learning_rate": 1.698979591836735e-05,
      "loss": 0.0805,
      "step": 3335
    },
    {
      "epoch": 1.2504,
      "grad_norm": 1.9480489492416382,
      "learning_rate": 1.6979591836734695e-05,
      "loss": 0.1381,
      "step": 3336
    },
    {
      "epoch": 1.2506,
      "grad_norm": 1.7020474672317505,
      "learning_rate": 1.6969387755102044e-05,
      "loss": 0.1188,
      "step": 3337
    },
    {
      "epoch": 1.2508,
      "grad_norm": 2.2416229248046875,
      "learning_rate": 1.695918367346939e-05,
      "loss": 0.2585,
      "step": 3338
    },
    {
      "epoch": 1.251,
      "grad_norm": 3.647240400314331,
      "learning_rate": 1.6948979591836736e-05,
      "loss": 0.7583,
      "step": 3339
    },
    {
      "epoch": 1.2511999999999999,
      "grad_norm": 1.6109038591384888,
      "learning_rate": 1.693877551020408e-05,
      "loss": 0.1145,
      "step": 3340
    },
    {
      "epoch": 1.2514,
      "grad_norm": 2.7248711585998535,
      "learning_rate": 1.692857142857143e-05,
      "loss": 0.2185,
      "step": 3341
    },
    {
      "epoch": 1.2516,
      "grad_norm": 4.435530185699463,
      "learning_rate": 1.6918367346938776e-05,
      "loss": 0.4876,
      "step": 3342
    },
    {
      "epoch": 1.2518,
      "grad_norm": 3.164839744567871,
      "learning_rate": 1.6908163265306122e-05,
      "loss": 0.602,
      "step": 3343
    },
    {
      "epoch": 1.252,
      "grad_norm": 1.6481194496154785,
      "learning_rate": 1.689795918367347e-05,
      "loss": 0.1389,
      "step": 3344
    },
    {
      "epoch": 1.2522,
      "grad_norm": 8.841111183166504,
      "learning_rate": 1.6887755102040817e-05,
      "loss": 0.6078,
      "step": 3345
    },
    {
      "epoch": 1.2524,
      "grad_norm": 13.993014335632324,
      "learning_rate": 1.6877551020408163e-05,
      "loss": 1.9227,
      "step": 3346
    },
    {
      "epoch": 1.2526,
      "grad_norm": 1.9829548597335815,
      "learning_rate": 1.6867346938775512e-05,
      "loss": 0.2012,
      "step": 3347
    },
    {
      "epoch": 1.2528000000000001,
      "grad_norm": 2.797745704650879,
      "learning_rate": 1.6857142857142858e-05,
      "loss": 0.4567,
      "step": 3348
    },
    {
      "epoch": 1.2530000000000001,
      "grad_norm": 3.1786699295043945,
      "learning_rate": 1.6846938775510204e-05,
      "loss": 0.3397,
      "step": 3349
    },
    {
      "epoch": 1.2532,
      "grad_norm": 12.750046730041504,
      "learning_rate": 1.683673469387755e-05,
      "loss": 1.9939,
      "step": 3350
    },
    {
      "epoch": 1.2534,
      "grad_norm": 12.33714485168457,
      "learning_rate": 1.68265306122449e-05,
      "loss": 1.5595,
      "step": 3351
    },
    {
      "epoch": 1.2536,
      "grad_norm": 2.612457275390625,
      "learning_rate": 1.6816326530612244e-05,
      "loss": 0.5026,
      "step": 3352
    },
    {
      "epoch": 1.2538,
      "grad_norm": 2.011706590652466,
      "learning_rate": 1.680612244897959e-05,
      "loss": 0.1798,
      "step": 3353
    },
    {
      "epoch": 1.254,
      "grad_norm": 3.1007089614868164,
      "learning_rate": 1.679591836734694e-05,
      "loss": 0.5509,
      "step": 3354
    },
    {
      "epoch": 1.2542,
      "grad_norm": 1.5206526517868042,
      "learning_rate": 1.6785714285714285e-05,
      "loss": 0.1098,
      "step": 3355
    },
    {
      "epoch": 1.2544,
      "grad_norm": 1.6399301290512085,
      "learning_rate": 1.6775510204081634e-05,
      "loss": 0.1127,
      "step": 3356
    },
    {
      "epoch": 1.2546,
      "grad_norm": 1.327517032623291,
      "learning_rate": 1.676530612244898e-05,
      "loss": 0.0481,
      "step": 3357
    },
    {
      "epoch": 1.2548,
      "grad_norm": 1.5172820091247559,
      "learning_rate": 1.675510204081633e-05,
      "loss": 0.1195,
      "step": 3358
    },
    {
      "epoch": 1.255,
      "grad_norm": 1.360618233680725,
      "learning_rate": 1.6744897959183675e-05,
      "loss": 0.0576,
      "step": 3359
    },
    {
      "epoch": 1.2551999999999999,
      "grad_norm": 2.326371431350708,
      "learning_rate": 1.673469387755102e-05,
      "loss": 0.1188,
      "step": 3360
    },
    {
      "epoch": 1.2554,
      "grad_norm": 5.61547327041626,
      "learning_rate": 1.672448979591837e-05,
      "loss": 0.7798,
      "step": 3361
    },
    {
      "epoch": 1.2556,
      "grad_norm": 5.132718563079834,
      "learning_rate": 1.6714285714285716e-05,
      "loss": 0.6264,
      "step": 3362
    },
    {
      "epoch": 1.2558,
      "grad_norm": 8.089487075805664,
      "learning_rate": 1.6704081632653062e-05,
      "loss": 0.7934,
      "step": 3363
    },
    {
      "epoch": 1.256,
      "grad_norm": 1.711835265159607,
      "learning_rate": 1.669387755102041e-05,
      "loss": 0.1151,
      "step": 3364
    },
    {
      "epoch": 1.2562,
      "grad_norm": 4.24153995513916,
      "learning_rate": 1.6683673469387757e-05,
      "loss": 0.9138,
      "step": 3365
    },
    {
      "epoch": 1.2564,
      "grad_norm": 9.217023849487305,
      "learning_rate": 1.6673469387755102e-05,
      "loss": 3.4371,
      "step": 3366
    },
    {
      "epoch": 1.2566,
      "grad_norm": 3.100184679031372,
      "learning_rate": 1.6663265306122448e-05,
      "loss": 0.3243,
      "step": 3367
    },
    {
      "epoch": 1.2568,
      "grad_norm": 2.0953240394592285,
      "learning_rate": 1.6653061224489797e-05,
      "loss": 0.1945,
      "step": 3368
    },
    {
      "epoch": 1.2570000000000001,
      "grad_norm": 3.7271270751953125,
      "learning_rate": 1.6642857142857143e-05,
      "loss": 0.7156,
      "step": 3369
    },
    {
      "epoch": 1.2572,
      "grad_norm": 2.181183099746704,
      "learning_rate": 1.663265306122449e-05,
      "loss": 0.1978,
      "step": 3370
    },
    {
      "epoch": 1.2574,
      "grad_norm": 2.6793060302734375,
      "learning_rate": 1.6622448979591838e-05,
      "loss": 0.4631,
      "step": 3371
    },
    {
      "epoch": 1.2576,
      "grad_norm": 1.3397785425186157,
      "learning_rate": 1.6612244897959184e-05,
      "loss": 0.0393,
      "step": 3372
    },
    {
      "epoch": 1.2578,
      "grad_norm": 1.6295970678329468,
      "learning_rate": 1.660204081632653e-05,
      "loss": 0.1165,
      "step": 3373
    },
    {
      "epoch": 1.258,
      "grad_norm": 1.4244837760925293,
      "learning_rate": 1.659183673469388e-05,
      "loss": 0.0769,
      "step": 3374
    },
    {
      "epoch": 1.2582,
      "grad_norm": 1.6356065273284912,
      "learning_rate": 1.6581632653061225e-05,
      "loss": 0.1071,
      "step": 3375
    },
    {
      "epoch": 1.2584,
      "grad_norm": 5.48160457611084,
      "learning_rate": 1.657142857142857e-05,
      "loss": 1.4024,
      "step": 3376
    },
    {
      "epoch": 1.2586,
      "grad_norm": 8.133343696594238,
      "learning_rate": 1.6561224489795916e-05,
      "loss": 1.9838,
      "step": 3377
    },
    {
      "epoch": 1.2588,
      "grad_norm": 2.3891677856445312,
      "learning_rate": 1.6551020408163266e-05,
      "loss": 0.1329,
      "step": 3378
    },
    {
      "epoch": 1.259,
      "grad_norm": 1.8075857162475586,
      "learning_rate": 1.6540816326530615e-05,
      "loss": 0.069,
      "step": 3379
    },
    {
      "epoch": 1.2591999999999999,
      "grad_norm": 1.2503466606140137,
      "learning_rate": 1.653061224489796e-05,
      "loss": 0.0409,
      "step": 3380
    },
    {
      "epoch": 1.2594,
      "grad_norm": 1.4084055423736572,
      "learning_rate": 1.6520408163265306e-05,
      "loss": 0.082,
      "step": 3381
    },
    {
      "epoch": 1.2596,
      "grad_norm": 2.6745080947875977,
      "learning_rate": 1.6510204081632655e-05,
      "loss": 0.1938,
      "step": 3382
    },
    {
      "epoch": 1.2598,
      "grad_norm": 5.423195838928223,
      "learning_rate": 1.65e-05,
      "loss": 0.7431,
      "step": 3383
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.263220191001892,
      "learning_rate": 1.6489795918367347e-05,
      "loss": 0.0898,
      "step": 3384
    },
    {
      "epoch": 1.2602,
      "grad_norm": 1.5418835878372192,
      "learning_rate": 1.6479591836734696e-05,
      "loss": 0.0626,
      "step": 3385
    },
    {
      "epoch": 1.2604,
      "grad_norm": 1.559310793876648,
      "learning_rate": 1.6469387755102042e-05,
      "loss": 0.0583,
      "step": 3386
    },
    {
      "epoch": 1.2606,
      "grad_norm": 1.265760898590088,
      "learning_rate": 1.6459183673469388e-05,
      "loss": 0.042,
      "step": 3387
    },
    {
      "epoch": 1.2608,
      "grad_norm": 1.7681432962417603,
      "learning_rate": 1.6448979591836737e-05,
      "loss": 0.1167,
      "step": 3388
    },
    {
      "epoch": 1.2610000000000001,
      "grad_norm": 2.5975985527038574,
      "learning_rate": 1.6438775510204083e-05,
      "loss": 0.1504,
      "step": 3389
    },
    {
      "epoch": 1.2612,
      "grad_norm": 2.9997403621673584,
      "learning_rate": 1.642857142857143e-05,
      "loss": 0.4652,
      "step": 3390
    },
    {
      "epoch": 1.2614,
      "grad_norm": 5.229567050933838,
      "learning_rate": 1.6418367346938778e-05,
      "loss": 1.5219,
      "step": 3391
    },
    {
      "epoch": 1.2616,
      "grad_norm": 4.952725410461426,
      "learning_rate": 1.6408163265306124e-05,
      "loss": 0.6158,
      "step": 3392
    },
    {
      "epoch": 1.2618,
      "grad_norm": 1.8712340593338013,
      "learning_rate": 1.639795918367347e-05,
      "loss": 0.1638,
      "step": 3393
    },
    {
      "epoch": 1.262,
      "grad_norm": 2.623368501663208,
      "learning_rate": 1.6387755102040815e-05,
      "loss": 0.4956,
      "step": 3394
    },
    {
      "epoch": 1.2622,
      "grad_norm": 6.142118453979492,
      "learning_rate": 1.6377551020408164e-05,
      "loss": 1.0003,
      "step": 3395
    },
    {
      "epoch": 1.2624,
      "grad_norm": 16.167905807495117,
      "learning_rate": 1.636734693877551e-05,
      "loss": 3.6286,
      "step": 3396
    },
    {
      "epoch": 1.2626,
      "grad_norm": 6.149655818939209,
      "learning_rate": 1.6357142857142856e-05,
      "loss": 1.6503,
      "step": 3397
    },
    {
      "epoch": 1.2628,
      "grad_norm": 10.63801097869873,
      "learning_rate": 1.6346938775510205e-05,
      "loss": 3.1041,
      "step": 3398
    },
    {
      "epoch": 1.263,
      "grad_norm": 10.49075984954834,
      "learning_rate": 1.633673469387755e-05,
      "loss": 3.8939,
      "step": 3399
    },
    {
      "epoch": 1.2631999999999999,
      "grad_norm": 9.431962966918945,
      "learning_rate": 1.6326530612244897e-05,
      "loss": 3.4103,
      "step": 3400
    },
    {
      "epoch": 1.2634,
      "grad_norm": 1.8127503395080566,
      "learning_rate": 1.6316326530612246e-05,
      "loss": 0.144,
      "step": 3401
    },
    {
      "epoch": 1.2636,
      "grad_norm": 3.003711223602295,
      "learning_rate": 1.6306122448979595e-05,
      "loss": 0.354,
      "step": 3402
    },
    {
      "epoch": 1.2638,
      "grad_norm": 1.5330774784088135,
      "learning_rate": 1.629591836734694e-05,
      "loss": 0.1217,
      "step": 3403
    },
    {
      "epoch": 1.264,
      "grad_norm": 2.6452956199645996,
      "learning_rate": 1.6285714285714287e-05,
      "loss": 0.18,
      "step": 3404
    },
    {
      "epoch": 1.2642,
      "grad_norm": 6.225276947021484,
      "learning_rate": 1.6275510204081636e-05,
      "loss": 0.8655,
      "step": 3405
    },
    {
      "epoch": 1.2644,
      "grad_norm": 4.707012176513672,
      "learning_rate": 1.626530612244898e-05,
      "loss": 0.7126,
      "step": 3406
    },
    {
      "epoch": 1.2646,
      "grad_norm": 1.9787338972091675,
      "learning_rate": 1.6255102040816327e-05,
      "loss": 0.1043,
      "step": 3407
    },
    {
      "epoch": 1.2648,
      "grad_norm": 3.948756694793701,
      "learning_rate": 1.6244897959183673e-05,
      "loss": 0.4824,
      "step": 3408
    },
    {
      "epoch": 1.2650000000000001,
      "grad_norm": 4.800436019897461,
      "learning_rate": 1.6234693877551022e-05,
      "loss": 0.6866,
      "step": 3409
    },
    {
      "epoch": 1.2652,
      "grad_norm": 1.5941983461380005,
      "learning_rate": 1.6224489795918368e-05,
      "loss": 0.1029,
      "step": 3410
    },
    {
      "epoch": 1.2654,
      "grad_norm": 1.7154693603515625,
      "learning_rate": 1.6214285714285714e-05,
      "loss": 0.115,
      "step": 3411
    },
    {
      "epoch": 1.2656,
      "grad_norm": 3.7221055030822754,
      "learning_rate": 1.6204081632653063e-05,
      "loss": 0.6513,
      "step": 3412
    },
    {
      "epoch": 1.2658,
      "grad_norm": 8.22921085357666,
      "learning_rate": 1.619387755102041e-05,
      "loss": 0.7431,
      "step": 3413
    },
    {
      "epoch": 1.266,
      "grad_norm": 1.383772611618042,
      "learning_rate": 1.6183673469387755e-05,
      "loss": 0.086,
      "step": 3414
    },
    {
      "epoch": 1.2662,
      "grad_norm": 1.8097155094146729,
      "learning_rate": 1.6173469387755104e-05,
      "loss": 0.1126,
      "step": 3415
    },
    {
      "epoch": 1.2664,
      "grad_norm": 2.1031274795532227,
      "learning_rate": 1.616326530612245e-05,
      "loss": 0.1949,
      "step": 3416
    },
    {
      "epoch": 1.2666,
      "grad_norm": 7.550622463226318,
      "learning_rate": 1.6153061224489795e-05,
      "loss": 0.7638,
      "step": 3417
    },
    {
      "epoch": 1.2668,
      "grad_norm": 3.7834267616271973,
      "learning_rate": 1.614285714285714e-05,
      "loss": 0.6121,
      "step": 3418
    },
    {
      "epoch": 1.267,
      "grad_norm": 4.540528297424316,
      "learning_rate": 1.613265306122449e-05,
      "loss": 0.5646,
      "step": 3419
    },
    {
      "epoch": 1.2671999999999999,
      "grad_norm": 1.4493329524993896,
      "learning_rate": 1.6122448979591836e-05,
      "loss": 0.0882,
      "step": 3420
    },
    {
      "epoch": 1.2674,
      "grad_norm": 2.163473129272461,
      "learning_rate": 1.6112244897959182e-05,
      "loss": 0.1259,
      "step": 3421
    },
    {
      "epoch": 1.2676,
      "grad_norm": 4.591796398162842,
      "learning_rate": 1.610204081632653e-05,
      "loss": 0.5881,
      "step": 3422
    },
    {
      "epoch": 1.2678,
      "grad_norm": 3.2775096893310547,
      "learning_rate": 1.6091836734693877e-05,
      "loss": 0.1809,
      "step": 3423
    },
    {
      "epoch": 1.268,
      "grad_norm": 6.489138603210449,
      "learning_rate": 1.6081632653061226e-05,
      "loss": 0.7169,
      "step": 3424
    },
    {
      "epoch": 1.2682,
      "grad_norm": 1.3613406419754028,
      "learning_rate": 1.6071428571428572e-05,
      "loss": 0.0461,
      "step": 3425
    },
    {
      "epoch": 1.2684,
      "grad_norm": 2.6900811195373535,
      "learning_rate": 1.606122448979592e-05,
      "loss": 0.1681,
      "step": 3426
    },
    {
      "epoch": 1.2686,
      "grad_norm": 6.287497520446777,
      "learning_rate": 1.6051020408163267e-05,
      "loss": 0.7438,
      "step": 3427
    },
    {
      "epoch": 1.2688,
      "grad_norm": 1.5202996730804443,
      "learning_rate": 1.6040816326530613e-05,
      "loss": 0.0881,
      "step": 3428
    },
    {
      "epoch": 1.2690000000000001,
      "grad_norm": 1.743475079536438,
      "learning_rate": 1.6030612244897962e-05,
      "loss": 0.0626,
      "step": 3429
    },
    {
      "epoch": 1.2692,
      "grad_norm": 1.582897663116455,
      "learning_rate": 1.6020408163265308e-05,
      "loss": 0.0645,
      "step": 3430
    },
    {
      "epoch": 1.2694,
      "grad_norm": 7.358126640319824,
      "learning_rate": 1.6010204081632653e-05,
      "loss": 0.5717,
      "step": 3431
    },
    {
      "epoch": 1.2696,
      "grad_norm": 13.37587833404541,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.0154,
      "step": 3432
    },
    {
      "epoch": 1.2698,
      "grad_norm": 2.172822952270508,
      "learning_rate": 1.598979591836735e-05,
      "loss": 0.1618,
      "step": 3433
    },
    {
      "epoch": 1.27,
      "grad_norm": 2.9426889419555664,
      "learning_rate": 1.5979591836734694e-05,
      "loss": 0.2733,
      "step": 3434
    },
    {
      "epoch": 1.2702,
      "grad_norm": 1.3792476654052734,
      "learning_rate": 1.596938775510204e-05,
      "loss": 0.0719,
      "step": 3435
    },
    {
      "epoch": 1.2704,
      "grad_norm": 1.2770227193832397,
      "learning_rate": 1.595918367346939e-05,
      "loss": 0.0714,
      "step": 3436
    },
    {
      "epoch": 1.2706,
      "grad_norm": 1.3073512315750122,
      "learning_rate": 1.5948979591836735e-05,
      "loss": 0.1088,
      "step": 3437
    },
    {
      "epoch": 1.2708,
      "grad_norm": 1.2940077781677246,
      "learning_rate": 1.593877551020408e-05,
      "loss": 0.0461,
      "step": 3438
    },
    {
      "epoch": 1.271,
      "grad_norm": 2.482409715652466,
      "learning_rate": 1.592857142857143e-05,
      "loss": 0.1819,
      "step": 3439
    },
    {
      "epoch": 1.2711999999999999,
      "grad_norm": 4.856838703155518,
      "learning_rate": 1.5918367346938776e-05,
      "loss": 0.689,
      "step": 3440
    },
    {
      "epoch": 1.2713999999999999,
      "grad_norm": 2.8870060443878174,
      "learning_rate": 1.590816326530612e-05,
      "loss": 0.4883,
      "step": 3441
    },
    {
      "epoch": 1.2716,
      "grad_norm": 1.156765103340149,
      "learning_rate": 1.589795918367347e-05,
      "loss": 0.0341,
      "step": 3442
    },
    {
      "epoch": 1.2718,
      "grad_norm": 2.10788893699646,
      "learning_rate": 1.5887755102040817e-05,
      "loss": 0.2436,
      "step": 3443
    },
    {
      "epoch": 1.272,
      "grad_norm": 2.6830666065216064,
      "learning_rate": 1.5877551020408162e-05,
      "loss": 0.479,
      "step": 3444
    },
    {
      "epoch": 1.2722,
      "grad_norm": 1.2923004627227783,
      "learning_rate": 1.5867346938775508e-05,
      "loss": 0.1194,
      "step": 3445
    },
    {
      "epoch": 1.2724,
      "grad_norm": 1.9385242462158203,
      "learning_rate": 1.5857142857142857e-05,
      "loss": 0.1861,
      "step": 3446
    },
    {
      "epoch": 1.2726,
      "grad_norm": 3.1385715007781982,
      "learning_rate": 1.5846938775510206e-05,
      "loss": 0.6352,
      "step": 3447
    },
    {
      "epoch": 1.2728,
      "grad_norm": 1.4715768098831177,
      "learning_rate": 1.5836734693877552e-05,
      "loss": 0.0476,
      "step": 3448
    },
    {
      "epoch": 1.2730000000000001,
      "grad_norm": 1.7465702295303345,
      "learning_rate": 1.5826530612244898e-05,
      "loss": 0.0899,
      "step": 3449
    },
    {
      "epoch": 1.2732,
      "grad_norm": 3.3258235454559326,
      "learning_rate": 1.5816326530612247e-05,
      "loss": 0.3732,
      "step": 3450
    },
    {
      "epoch": 1.2734,
      "grad_norm": 4.332332134246826,
      "learning_rate": 1.5806122448979593e-05,
      "loss": 1.1368,
      "step": 3451
    },
    {
      "epoch": 1.2736,
      "grad_norm": 8.17160701751709,
      "learning_rate": 1.579591836734694e-05,
      "loss": 2.0268,
      "step": 3452
    },
    {
      "epoch": 1.2738,
      "grad_norm": 1.359741449356079,
      "learning_rate": 1.5785714285714288e-05,
      "loss": 0.0842,
      "step": 3453
    },
    {
      "epoch": 1.274,
      "grad_norm": 1.439969778060913,
      "learning_rate": 1.5775510204081634e-05,
      "loss": 0.0787,
      "step": 3454
    },
    {
      "epoch": 1.2742,
      "grad_norm": 3.2931435108184814,
      "learning_rate": 1.576530612244898e-05,
      "loss": 0.2335,
      "step": 3455
    },
    {
      "epoch": 1.2744,
      "grad_norm": 7.263006210327148,
      "learning_rate": 1.575510204081633e-05,
      "loss": 1.3831,
      "step": 3456
    },
    {
      "epoch": 1.2746,
      "grad_norm": 2.6793131828308105,
      "learning_rate": 1.5744897959183675e-05,
      "loss": 0.3533,
      "step": 3457
    },
    {
      "epoch": 1.2748,
      "grad_norm": 6.873786926269531,
      "learning_rate": 1.573469387755102e-05,
      "loss": 1.512,
      "step": 3458
    },
    {
      "epoch": 1.275,
      "grad_norm": 5.337397575378418,
      "learning_rate": 1.572448979591837e-05,
      "loss": 1.2574,
      "step": 3459
    },
    {
      "epoch": 1.2752,
      "grad_norm": 9.182343482971191,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 2.6326,
      "step": 3460
    },
    {
      "epoch": 1.2753999999999999,
      "grad_norm": 6.159662246704102,
      "learning_rate": 1.570408163265306e-05,
      "loss": 2.4802,
      "step": 3461
    },
    {
      "epoch": 1.2756,
      "grad_norm": 17.988075256347656,
      "learning_rate": 1.5693877551020407e-05,
      "loss": 3.4773,
      "step": 3462
    },
    {
      "epoch": 1.2758,
      "grad_norm": 33.38669204711914,
      "learning_rate": 1.5683673469387756e-05,
      "loss": 3.9685,
      "step": 3463
    },
    {
      "epoch": 1.276,
      "grad_norm": 10.34990406036377,
      "learning_rate": 1.5673469387755102e-05,
      "loss": 1.2912,
      "step": 3464
    },
    {
      "epoch": 1.2762,
      "grad_norm": 7.714395046234131,
      "learning_rate": 1.5663265306122448e-05,
      "loss": 0.9114,
      "step": 3465
    },
    {
      "epoch": 1.2764,
      "grad_norm": 3.0588672161102295,
      "learning_rate": 1.5653061224489797e-05,
      "loss": 0.5662,
      "step": 3466
    },
    {
      "epoch": 1.2766,
      "grad_norm": 6.391806125640869,
      "learning_rate": 1.5642857142857143e-05,
      "loss": 0.9289,
      "step": 3467
    },
    {
      "epoch": 1.2768,
      "grad_norm": 16.139328002929688,
      "learning_rate": 1.563265306122449e-05,
      "loss": 3.4601,
      "step": 3468
    },
    {
      "epoch": 1.2770000000000001,
      "grad_norm": 1.3934966325759888,
      "learning_rate": 1.5622448979591838e-05,
      "loss": 0.0688,
      "step": 3469
    },
    {
      "epoch": 1.2772000000000001,
      "grad_norm": 5.1799821853637695,
      "learning_rate": 1.5612244897959187e-05,
      "loss": 0.8917,
      "step": 3470
    },
    {
      "epoch": 1.2774,
      "grad_norm": 8.424983978271484,
      "learning_rate": 1.5602040816326533e-05,
      "loss": 3.3538,
      "step": 3471
    },
    {
      "epoch": 1.2776,
      "grad_norm": 6.700727462768555,
      "learning_rate": 1.559183673469388e-05,
      "loss": 1.2781,
      "step": 3472
    },
    {
      "epoch": 1.2778,
      "grad_norm": 8.653555870056152,
      "learning_rate": 1.5581632653061228e-05,
      "loss": 2.5732,
      "step": 3473
    },
    {
      "epoch": 1.278,
      "grad_norm": 1.482365608215332,
      "learning_rate": 1.5571428571428573e-05,
      "loss": 0.0817,
      "step": 3474
    },
    {
      "epoch": 1.2782,
      "grad_norm": 1.7122341394424438,
      "learning_rate": 1.556122448979592e-05,
      "loss": 0.0784,
      "step": 3475
    },
    {
      "epoch": 1.2784,
      "grad_norm": 2.120478630065918,
      "learning_rate": 1.5551020408163265e-05,
      "loss": 0.124,
      "step": 3476
    },
    {
      "epoch": 1.2786,
      "grad_norm": 2.957033157348633,
      "learning_rate": 1.5540816326530614e-05,
      "loss": 0.3313,
      "step": 3477
    },
    {
      "epoch": 1.2788,
      "grad_norm": 3.401562452316284,
      "learning_rate": 1.553061224489796e-05,
      "loss": 0.1388,
      "step": 3478
    },
    {
      "epoch": 1.279,
      "grad_norm": 7.845393657684326,
      "learning_rate": 1.5520408163265306e-05,
      "loss": 0.7818,
      "step": 3479
    },
    {
      "epoch": 1.2792,
      "grad_norm": 1.0962510108947754,
      "learning_rate": 1.5510204081632655e-05,
      "loss": 0.0656,
      "step": 3480
    },
    {
      "epoch": 1.2793999999999999,
      "grad_norm": 2.6956839561462402,
      "learning_rate": 1.55e-05,
      "loss": 0.3444,
      "step": 3481
    },
    {
      "epoch": 1.2796,
      "grad_norm": 7.444957256317139,
      "learning_rate": 1.5489795918367346e-05,
      "loss": 1.4494,
      "step": 3482
    },
    {
      "epoch": 1.2798,
      "grad_norm": 1.4879704713821411,
      "learning_rate": 1.5479591836734696e-05,
      "loss": 0.0571,
      "step": 3483
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.6962056159973145,
      "learning_rate": 1.546938775510204e-05,
      "loss": 0.172,
      "step": 3484
    },
    {
      "epoch": 1.2802,
      "grad_norm": 1.4376516342163086,
      "learning_rate": 1.5459183673469387e-05,
      "loss": 0.0517,
      "step": 3485
    },
    {
      "epoch": 1.2804,
      "grad_norm": 1.297108769416809,
      "learning_rate": 1.5448979591836733e-05,
      "loss": 0.0451,
      "step": 3486
    },
    {
      "epoch": 1.2806,
      "grad_norm": 4.2883830070495605,
      "learning_rate": 1.5438775510204082e-05,
      "loss": 1.1647,
      "step": 3487
    },
    {
      "epoch": 1.2808,
      "grad_norm": 11.137147903442383,
      "learning_rate": 1.5428571428571428e-05,
      "loss": 2.7671,
      "step": 3488
    },
    {
      "epoch": 1.2810000000000001,
      "grad_norm": 1.6357594728469849,
      "learning_rate": 1.5418367346938774e-05,
      "loss": 0.119,
      "step": 3489
    },
    {
      "epoch": 1.2812000000000001,
      "grad_norm": 2.295539379119873,
      "learning_rate": 1.5408163265306123e-05,
      "loss": 0.1267,
      "step": 3490
    },
    {
      "epoch": 1.2814,
      "grad_norm": 8.366477966308594,
      "learning_rate": 1.539795918367347e-05,
      "loss": 1.4248,
      "step": 3491
    },
    {
      "epoch": 1.2816,
      "grad_norm": 21.025161743164062,
      "learning_rate": 1.5387755102040818e-05,
      "loss": 3.7029,
      "step": 3492
    },
    {
      "epoch": 1.2818,
      "grad_norm": 7.5603532791137695,
      "learning_rate": 1.5377551020408164e-05,
      "loss": 0.7515,
      "step": 3493
    },
    {
      "epoch": 1.282,
      "grad_norm": 4.353247165679932,
      "learning_rate": 1.5367346938775513e-05,
      "loss": 0.5647,
      "step": 3494
    },
    {
      "epoch": 1.2822,
      "grad_norm": 1.7207931280136108,
      "learning_rate": 1.535714285714286e-05,
      "loss": 0.1705,
      "step": 3495
    },
    {
      "epoch": 1.2824,
      "grad_norm": 1.7064158916473389,
      "learning_rate": 1.5346938775510204e-05,
      "loss": 0.1937,
      "step": 3496
    },
    {
      "epoch": 1.2826,
      "grad_norm": 2.7570483684539795,
      "learning_rate": 1.5336734693877554e-05,
      "loss": 0.4847,
      "step": 3497
    },
    {
      "epoch": 1.2828,
      "grad_norm": 3.3982269763946533,
      "learning_rate": 1.53265306122449e-05,
      "loss": 0.2327,
      "step": 3498
    },
    {
      "epoch": 1.283,
      "grad_norm": 7.127781867980957,
      "learning_rate": 1.5316326530612245e-05,
      "loss": 1.3224,
      "step": 3499
    },
    {
      "epoch": 1.2832,
      "grad_norm": 3.0783791542053223,
      "learning_rate": 1.5306122448979594e-05,
      "loss": 0.6105,
      "step": 3500
    },
    {
      "epoch": 1.2833999999999999,
      "grad_norm": 3.3459722995758057,
      "learning_rate": 1.529591836734694e-05,
      "loss": 0.4059,
      "step": 3501
    },
    {
      "epoch": 1.2836,
      "grad_norm": 4.628581523895264,
      "learning_rate": 1.5285714285714286e-05,
      "loss": 1.4011,
      "step": 3502
    },
    {
      "epoch": 1.2838,
      "grad_norm": 1.6847960948944092,
      "learning_rate": 1.5275510204081632e-05,
      "loss": 0.1031,
      "step": 3503
    },
    {
      "epoch": 1.284,
      "grad_norm": 1.219497799873352,
      "learning_rate": 1.526530612244898e-05,
      "loss": 0.0378,
      "step": 3504
    },
    {
      "epoch": 1.2842,
      "grad_norm": 1.7342090606689453,
      "learning_rate": 1.5255102040816327e-05,
      "loss": 0.1441,
      "step": 3505
    },
    {
      "epoch": 1.2844,
      "grad_norm": 1.6367411613464355,
      "learning_rate": 1.5244897959183674e-05,
      "loss": 0.0833,
      "step": 3506
    },
    {
      "epoch": 1.2846,
      "grad_norm": 3.5634076595306396,
      "learning_rate": 1.523469387755102e-05,
      "loss": 0.4575,
      "step": 3507
    },
    {
      "epoch": 1.2848,
      "grad_norm": 5.396846294403076,
      "learning_rate": 1.5224489795918368e-05,
      "loss": 0.7511,
      "step": 3508
    },
    {
      "epoch": 1.285,
      "grad_norm": 5.117428302764893,
      "learning_rate": 1.5214285714285715e-05,
      "loss": 0.7507,
      "step": 3509
    },
    {
      "epoch": 1.2852000000000001,
      "grad_norm": 4.2475666999816895,
      "learning_rate": 1.520408163265306e-05,
      "loss": 0.4041,
      "step": 3510
    },
    {
      "epoch": 1.2854,
      "grad_norm": 10.006048202514648,
      "learning_rate": 1.5193877551020408e-05,
      "loss": 1.283,
      "step": 3511
    },
    {
      "epoch": 1.2856,
      "grad_norm": 6.676290512084961,
      "learning_rate": 1.5183673469387754e-05,
      "loss": 1.648,
      "step": 3512
    },
    {
      "epoch": 1.2858,
      "grad_norm": 10.92365550994873,
      "learning_rate": 1.5173469387755102e-05,
      "loss": 3.6243,
      "step": 3513
    },
    {
      "epoch": 1.286,
      "grad_norm": 11.564417839050293,
      "learning_rate": 1.5163265306122449e-05,
      "loss": 4.3346,
      "step": 3514
    },
    {
      "epoch": 1.2862,
      "grad_norm": 9.778779029846191,
      "learning_rate": 1.5153061224489798e-05,
      "loss": 2.6538,
      "step": 3515
    },
    {
      "epoch": 1.2864,
      "grad_norm": 8.761896133422852,
      "learning_rate": 1.5142857142857144e-05,
      "loss": 1.4395,
      "step": 3516
    },
    {
      "epoch": 1.2866,
      "grad_norm": 11.730488777160645,
      "learning_rate": 1.5132653061224492e-05,
      "loss": 3.4833,
      "step": 3517
    },
    {
      "epoch": 1.2868,
      "grad_norm": 4.998980522155762,
      "learning_rate": 1.5122448979591839e-05,
      "loss": 0.5691,
      "step": 3518
    },
    {
      "epoch": 1.287,
      "grad_norm": 1.5715025663375854,
      "learning_rate": 1.5112244897959185e-05,
      "loss": 0.077,
      "step": 3519
    },
    {
      "epoch": 1.2872,
      "grad_norm": 1.9681670665740967,
      "learning_rate": 1.5102040816326532e-05,
      "loss": 0.0896,
      "step": 3520
    },
    {
      "epoch": 1.2873999999999999,
      "grad_norm": 1.2198389768600464,
      "learning_rate": 1.5091836734693878e-05,
      "loss": 0.0338,
      "step": 3521
    },
    {
      "epoch": 1.2876,
      "grad_norm": 1.6890063285827637,
      "learning_rate": 1.5081632653061226e-05,
      "loss": 0.1405,
      "step": 3522
    },
    {
      "epoch": 1.2878,
      "grad_norm": 1.5838737487792969,
      "learning_rate": 1.5071428571428573e-05,
      "loss": 0.1021,
      "step": 3523
    },
    {
      "epoch": 1.288,
      "grad_norm": 1.7186545133590698,
      "learning_rate": 1.5061224489795919e-05,
      "loss": 0.1365,
      "step": 3524
    },
    {
      "epoch": 1.2882,
      "grad_norm": 1.1837531328201294,
      "learning_rate": 1.5051020408163266e-05,
      "loss": 0.0649,
      "step": 3525
    },
    {
      "epoch": 1.2884,
      "grad_norm": 1.6602630615234375,
      "learning_rate": 1.5040816326530612e-05,
      "loss": 0.1081,
      "step": 3526
    },
    {
      "epoch": 1.2886,
      "grad_norm": 1.467596173286438,
      "learning_rate": 1.503061224489796e-05,
      "loss": 0.0973,
      "step": 3527
    },
    {
      "epoch": 1.2888,
      "grad_norm": 1.367651104927063,
      "learning_rate": 1.5020408163265307e-05,
      "loss": 0.0538,
      "step": 3528
    },
    {
      "epoch": 1.289,
      "grad_norm": 1.1187045574188232,
      "learning_rate": 1.5010204081632653e-05,
      "loss": 0.0379,
      "step": 3529
    },
    {
      "epoch": 1.2892000000000001,
      "grad_norm": 7.7804718017578125,
      "learning_rate": 1.5e-05,
      "loss": 0.9649,
      "step": 3530
    },
    {
      "epoch": 1.2894,
      "grad_norm": 17.334747314453125,
      "learning_rate": 1.4989795918367346e-05,
      "loss": 3.3545,
      "step": 3531
    },
    {
      "epoch": 1.2896,
      "grad_norm": 1.814028263092041,
      "learning_rate": 1.4979591836734694e-05,
      "loss": 0.1417,
      "step": 3532
    },
    {
      "epoch": 1.2898,
      "grad_norm": 1.3732644319534302,
      "learning_rate": 1.4969387755102041e-05,
      "loss": 0.0833,
      "step": 3533
    },
    {
      "epoch": 1.29,
      "grad_norm": 7.133530616760254,
      "learning_rate": 1.4959183673469387e-05,
      "loss": 1.2164,
      "step": 3534
    },
    {
      "epoch": 1.2902,
      "grad_norm": 10.225381851196289,
      "learning_rate": 1.4948979591836734e-05,
      "loss": 2.5016,
      "step": 3535
    },
    {
      "epoch": 1.2904,
      "grad_norm": 4.777650833129883,
      "learning_rate": 1.493877551020408e-05,
      "loss": 0.8102,
      "step": 3536
    },
    {
      "epoch": 1.2906,
      "grad_norm": 11.105450630187988,
      "learning_rate": 1.4928571428571431e-05,
      "loss": 2.552,
      "step": 3537
    },
    {
      "epoch": 1.2908,
      "grad_norm": 1.9035511016845703,
      "learning_rate": 1.4918367346938777e-05,
      "loss": 0.132,
      "step": 3538
    },
    {
      "epoch": 1.291,
      "grad_norm": 3.101081371307373,
      "learning_rate": 1.4908163265306124e-05,
      "loss": 0.325,
      "step": 3539
    },
    {
      "epoch": 1.2912,
      "grad_norm": 1.6706990003585815,
      "learning_rate": 1.4897959183673472e-05,
      "loss": 0.1237,
      "step": 3540
    },
    {
      "epoch": 1.2913999999999999,
      "grad_norm": 0.9322052001953125,
      "learning_rate": 1.4887755102040818e-05,
      "loss": 0.0218,
      "step": 3541
    },
    {
      "epoch": 1.2916,
      "grad_norm": 1.633112907409668,
      "learning_rate": 1.4877551020408165e-05,
      "loss": 0.1139,
      "step": 3542
    },
    {
      "epoch": 1.2918,
      "grad_norm": 1.4652941226959229,
      "learning_rate": 1.4867346938775511e-05,
      "loss": 0.0537,
      "step": 3543
    },
    {
      "epoch": 1.292,
      "grad_norm": 2.806101083755493,
      "learning_rate": 1.4857142857142858e-05,
      "loss": 0.375,
      "step": 3544
    },
    {
      "epoch": 1.2922,
      "grad_norm": 5.178884983062744,
      "learning_rate": 1.4846938775510206e-05,
      "loss": 1.5283,
      "step": 3545
    },
    {
      "epoch": 1.2924,
      "grad_norm": 3.7781739234924316,
      "learning_rate": 1.4836734693877552e-05,
      "loss": 0.6923,
      "step": 3546
    },
    {
      "epoch": 1.2926,
      "grad_norm": 1.3200960159301758,
      "learning_rate": 1.48265306122449e-05,
      "loss": 0.1126,
      "step": 3547
    },
    {
      "epoch": 1.2928,
      "grad_norm": 4.988454341888428,
      "learning_rate": 1.4816326530612245e-05,
      "loss": 0.3449,
      "step": 3548
    },
    {
      "epoch": 1.293,
      "grad_norm": 10.687281608581543,
      "learning_rate": 1.4806122448979592e-05,
      "loss": 1.2852,
      "step": 3549
    },
    {
      "epoch": 1.2932000000000001,
      "grad_norm": 6.6570940017700195,
      "learning_rate": 1.479591836734694e-05,
      "loss": 1.1017,
      "step": 3550
    },
    {
      "epoch": 1.2934,
      "grad_norm": 19.316333770751953,
      "learning_rate": 1.4785714285714286e-05,
      "loss": 3.1181,
      "step": 3551
    },
    {
      "epoch": 1.2936,
      "grad_norm": 3.2980775833129883,
      "learning_rate": 1.4775510204081633e-05,
      "loss": 0.2851,
      "step": 3552
    },
    {
      "epoch": 1.2938,
      "grad_norm": 7.51362943649292,
      "learning_rate": 1.4765306122448979e-05,
      "loss": 1.3359,
      "step": 3553
    },
    {
      "epoch": 1.294,
      "grad_norm": 1.1324206590652466,
      "learning_rate": 1.4755102040816326e-05,
      "loss": 0.0294,
      "step": 3554
    },
    {
      "epoch": 1.2942,
      "grad_norm": 2.5744659900665283,
      "learning_rate": 1.4744897959183674e-05,
      "loss": 0.2059,
      "step": 3555
    },
    {
      "epoch": 1.2944,
      "grad_norm": 4.4462504386901855,
      "learning_rate": 1.473469387755102e-05,
      "loss": 0.593,
      "step": 3556
    },
    {
      "epoch": 1.2946,
      "grad_norm": 2.342153549194336,
      "learning_rate": 1.4724489795918367e-05,
      "loss": 0.2257,
      "step": 3557
    },
    {
      "epoch": 1.2948,
      "grad_norm": 3.1876916885375977,
      "learning_rate": 1.4714285714285713e-05,
      "loss": 0.3238,
      "step": 3558
    },
    {
      "epoch": 1.295,
      "grad_norm": 1.0605589151382446,
      "learning_rate": 1.470408163265306e-05,
      "loss": 0.0264,
      "step": 3559
    },
    {
      "epoch": 1.2952,
      "grad_norm": 1.5809507369995117,
      "learning_rate": 1.469387755102041e-05,
      "loss": 0.0874,
      "step": 3560
    },
    {
      "epoch": 1.2953999999999999,
      "grad_norm": 5.060691833496094,
      "learning_rate": 1.4683673469387757e-05,
      "loss": 0.3005,
      "step": 3561
    },
    {
      "epoch": 1.2955999999999999,
      "grad_norm": 10.105025291442871,
      "learning_rate": 1.4673469387755103e-05,
      "loss": 1.2824,
      "step": 3562
    },
    {
      "epoch": 1.2958,
      "grad_norm": 1.5753602981567383,
      "learning_rate": 1.466326530612245e-05,
      "loss": 0.0595,
      "step": 3563
    },
    {
      "epoch": 1.296,
      "grad_norm": 2.5145034790039062,
      "learning_rate": 1.4653061224489798e-05,
      "loss": 0.1782,
      "step": 3564
    },
    {
      "epoch": 1.2962,
      "grad_norm": 4.732728481292725,
      "learning_rate": 1.4642857142857144e-05,
      "loss": 0.6166,
      "step": 3565
    },
    {
      "epoch": 1.2964,
      "grad_norm": 3.8714406490325928,
      "learning_rate": 1.4632653061224491e-05,
      "loss": 1.0394,
      "step": 3566
    },
    {
      "epoch": 1.2966,
      "grad_norm": 12.83499813079834,
      "learning_rate": 1.4622448979591839e-05,
      "loss": 3.0973,
      "step": 3567
    },
    {
      "epoch": 1.2968,
      "grad_norm": 1.1112841367721558,
      "learning_rate": 1.4612244897959185e-05,
      "loss": 0.0258,
      "step": 3568
    },
    {
      "epoch": 1.297,
      "grad_norm": 1.4227652549743652,
      "learning_rate": 1.4602040816326532e-05,
      "loss": 0.1001,
      "step": 3569
    },
    {
      "epoch": 1.2972000000000001,
      "grad_norm": 2.9268789291381836,
      "learning_rate": 1.4591836734693878e-05,
      "loss": 0.4727,
      "step": 3570
    },
    {
      "epoch": 1.2974,
      "grad_norm": 5.912209987640381,
      "learning_rate": 1.4581632653061225e-05,
      "loss": 1.6332,
      "step": 3571
    },
    {
      "epoch": 1.2976,
      "grad_norm": 5.406517028808594,
      "learning_rate": 1.4571428571428573e-05,
      "loss": 0.658,
      "step": 3572
    },
    {
      "epoch": 1.2978,
      "grad_norm": 1.526465892791748,
      "learning_rate": 1.4561224489795919e-05,
      "loss": 0.1091,
      "step": 3573
    },
    {
      "epoch": 1.298,
      "grad_norm": 1.7140930891036987,
      "learning_rate": 1.4551020408163266e-05,
      "loss": 0.0577,
      "step": 3574
    },
    {
      "epoch": 1.2982,
      "grad_norm": 1.4619375467300415,
      "learning_rate": 1.4540816326530612e-05,
      "loss": 0.0507,
      "step": 3575
    },
    {
      "epoch": 1.2984,
      "grad_norm": 1.4371378421783447,
      "learning_rate": 1.453061224489796e-05,
      "loss": 0.0845,
      "step": 3576
    },
    {
      "epoch": 1.2986,
      "grad_norm": 2.455461025238037,
      "learning_rate": 1.4520408163265307e-05,
      "loss": 0.2963,
      "step": 3577
    },
    {
      "epoch": 1.2988,
      "grad_norm": 5.196033477783203,
      "learning_rate": 1.4510204081632653e-05,
      "loss": 1.4591,
      "step": 3578
    },
    {
      "epoch": 1.299,
      "grad_norm": 1.3846713304519653,
      "learning_rate": 1.45e-05,
      "loss": 0.0728,
      "step": 3579
    },
    {
      "epoch": 1.2992,
      "grad_norm": 1.5920789241790771,
      "learning_rate": 1.4489795918367346e-05,
      "loss": 0.1345,
      "step": 3580
    },
    {
      "epoch": 1.2993999999999999,
      "grad_norm": 1.436944842338562,
      "learning_rate": 1.4479591836734693e-05,
      "loss": 0.083,
      "step": 3581
    },
    {
      "epoch": 1.2995999999999999,
      "grad_norm": 1.6569437980651855,
      "learning_rate": 1.4469387755102041e-05,
      "loss": 0.1281,
      "step": 3582
    },
    {
      "epoch": 1.2998,
      "grad_norm": 1.71175217628479,
      "learning_rate": 1.445918367346939e-05,
      "loss": 0.1145,
      "step": 3583
    },
    {
      "epoch": 1.3,
      "grad_norm": 5.769881248474121,
      "learning_rate": 1.4448979591836736e-05,
      "loss": 0.9269,
      "step": 3584
    },
    {
      "epoch": 1.3002,
      "grad_norm": 11.544594764709473,
      "learning_rate": 1.4438775510204083e-05,
      "loss": 3.452,
      "step": 3585
    },
    {
      "epoch": 1.3004,
      "grad_norm": 3.0036733150482178,
      "learning_rate": 1.442857142857143e-05,
      "loss": 0.5211,
      "step": 3586
    },
    {
      "epoch": 1.3006,
      "grad_norm": 2.4071826934814453,
      "learning_rate": 1.4418367346938777e-05,
      "loss": 0.2461,
      "step": 3587
    },
    {
      "epoch": 1.3008,
      "grad_norm": 6.731344223022461,
      "learning_rate": 1.4408163265306124e-05,
      "loss": 0.8418,
      "step": 3588
    },
    {
      "epoch": 1.301,
      "grad_norm": 4.861983299255371,
      "learning_rate": 1.439795918367347e-05,
      "loss": 0.5717,
      "step": 3589
    },
    {
      "epoch": 1.3012000000000001,
      "grad_norm": 1.4551787376403809,
      "learning_rate": 1.4387755102040817e-05,
      "loss": 0.0891,
      "step": 3590
    },
    {
      "epoch": 1.3014000000000001,
      "grad_norm": 2.8933231830596924,
      "learning_rate": 1.4377551020408165e-05,
      "loss": 0.2431,
      "step": 3591
    },
    {
      "epoch": 1.3016,
      "grad_norm": 4.9407453536987305,
      "learning_rate": 1.436734693877551e-05,
      "loss": 0.65,
      "step": 3592
    },
    {
      "epoch": 1.3018,
      "grad_norm": 1.4356982707977295,
      "learning_rate": 1.4357142857142858e-05,
      "loss": 0.0407,
      "step": 3593
    },
    {
      "epoch": 1.302,
      "grad_norm": 3.079118490219116,
      "learning_rate": 1.4346938775510204e-05,
      "loss": 0.2953,
      "step": 3594
    },
    {
      "epoch": 1.3022,
      "grad_norm": 5.729078769683838,
      "learning_rate": 1.4336734693877551e-05,
      "loss": 0.7755,
      "step": 3595
    },
    {
      "epoch": 1.3024,
      "grad_norm": 1.4708061218261719,
      "learning_rate": 1.4326530612244899e-05,
      "loss": 0.0989,
      "step": 3596
    },
    {
      "epoch": 1.3026,
      "grad_norm": 1.650211215019226,
      "learning_rate": 1.4316326530612245e-05,
      "loss": 0.0479,
      "step": 3597
    },
    {
      "epoch": 1.3028,
      "grad_norm": 1.3778326511383057,
      "learning_rate": 1.4306122448979592e-05,
      "loss": 0.0676,
      "step": 3598
    },
    {
      "epoch": 1.303,
      "grad_norm": 1.4415026903152466,
      "learning_rate": 1.4295918367346938e-05,
      "loss": 0.0994,
      "step": 3599
    },
    {
      "epoch": 1.3032,
      "grad_norm": 9.081709861755371,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.5877,
      "step": 3600
    },
    {
      "epoch": 1.3034,
      "grad_norm": 15.51332950592041,
      "learning_rate": 1.4275510204081633e-05,
      "loss": 2.0339,
      "step": 3601
    },
    {
      "epoch": 1.3035999999999999,
      "grad_norm": 9.164117813110352,
      "learning_rate": 1.4265306122448979e-05,
      "loss": 1.4634,
      "step": 3602
    },
    {
      "epoch": 1.3038,
      "grad_norm": 20.999588012695312,
      "learning_rate": 1.4255102040816326e-05,
      "loss": 2.6444,
      "step": 3603
    },
    {
      "epoch": 1.304,
      "grad_norm": 4.09861421585083,
      "learning_rate": 1.4244897959183674e-05,
      "loss": 0.599,
      "step": 3604
    },
    {
      "epoch": 1.3042,
      "grad_norm": 5.565969467163086,
      "learning_rate": 1.4234693877551023e-05,
      "loss": 0.9323,
      "step": 3605
    },
    {
      "epoch": 1.3044,
      "grad_norm": 15.917082786560059,
      "learning_rate": 1.4224489795918369e-05,
      "loss": 3.9851,
      "step": 3606
    },
    {
      "epoch": 1.3046,
      "grad_norm": 16.127853393554688,
      "learning_rate": 1.4214285714285716e-05,
      "loss": 2.1158,
      "step": 3607
    },
    {
      "epoch": 1.3048,
      "grad_norm": 6.162598133087158,
      "learning_rate": 1.4204081632653064e-05,
      "loss": 1.0179,
      "step": 3608
    },
    {
      "epoch": 1.305,
      "grad_norm": 5.599311828613281,
      "learning_rate": 1.419387755102041e-05,
      "loss": 0.6646,
      "step": 3609
    },
    {
      "epoch": 1.3052000000000001,
      "grad_norm": 2.427150011062622,
      "learning_rate": 1.4183673469387757e-05,
      "loss": 0.3712,
      "step": 3610
    },
    {
      "epoch": 1.3054000000000001,
      "grad_norm": 4.940189838409424,
      "learning_rate": 1.4173469387755103e-05,
      "loss": 1.4755,
      "step": 3611
    },
    {
      "epoch": 1.3056,
      "grad_norm": 2.0217015743255615,
      "learning_rate": 1.416326530612245e-05,
      "loss": 0.1821,
      "step": 3612
    },
    {
      "epoch": 1.3058,
      "grad_norm": 3.79567813873291,
      "learning_rate": 1.4153061224489798e-05,
      "loss": 0.665,
      "step": 3613
    },
    {
      "epoch": 1.306,
      "grad_norm": 5.068130970001221,
      "learning_rate": 1.4142857142857143e-05,
      "loss": 0.7768,
      "step": 3614
    },
    {
      "epoch": 1.3062,
      "grad_norm": 7.242030620574951,
      "learning_rate": 1.4132653061224491e-05,
      "loss": 2.3842,
      "step": 3615
    },
    {
      "epoch": 1.3064,
      "grad_norm": 1.7902650833129883,
      "learning_rate": 1.4122448979591837e-05,
      "loss": 0.1467,
      "step": 3616
    },
    {
      "epoch": 1.3066,
      "grad_norm": 1.5323320627212524,
      "learning_rate": 1.4112244897959184e-05,
      "loss": 0.1101,
      "step": 3617
    },
    {
      "epoch": 1.3068,
      "grad_norm": 6.208742141723633,
      "learning_rate": 1.4102040816326532e-05,
      "loss": 1.0017,
      "step": 3618
    },
    {
      "epoch": 1.307,
      "grad_norm": 10.6547212600708,
      "learning_rate": 1.4091836734693877e-05,
      "loss": 3.3961,
      "step": 3619
    },
    {
      "epoch": 1.3072,
      "grad_norm": 1.841823697090149,
      "learning_rate": 1.4081632653061225e-05,
      "loss": 0.187,
      "step": 3620
    },
    {
      "epoch": 1.3074,
      "grad_norm": 3.2964346408843994,
      "learning_rate": 1.407142857142857e-05,
      "loss": 0.6089,
      "step": 3621
    },
    {
      "epoch": 1.3075999999999999,
      "grad_norm": 2.113030433654785,
      "learning_rate": 1.4061224489795918e-05,
      "loss": 0.139,
      "step": 3622
    },
    {
      "epoch": 1.3078,
      "grad_norm": 3.6273980140686035,
      "learning_rate": 1.4051020408163266e-05,
      "loss": 0.4046,
      "step": 3623
    },
    {
      "epoch": 1.308,
      "grad_norm": 1.473108172416687,
      "learning_rate": 1.4040816326530612e-05,
      "loss": 0.1369,
      "step": 3624
    },
    {
      "epoch": 1.3082,
      "grad_norm": 2.0529394149780273,
      "learning_rate": 1.4030612244897959e-05,
      "loss": 0.1391,
      "step": 3625
    },
    {
      "epoch": 1.3084,
      "grad_norm": 3.2491400241851807,
      "learning_rate": 1.4020408163265305e-05,
      "loss": 0.3464,
      "step": 3626
    },
    {
      "epoch": 1.3086,
      "grad_norm": 2.3641955852508545,
      "learning_rate": 1.4010204081632652e-05,
      "loss": 0.3468,
      "step": 3627
    },
    {
      "epoch": 1.3088,
      "grad_norm": 5.182273864746094,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 1.41,
      "step": 3628
    },
    {
      "epoch": 1.309,
      "grad_norm": 1.8585495948791504,
      "learning_rate": 1.3989795918367349e-05,
      "loss": 0.1579,
      "step": 3629
    },
    {
      "epoch": 1.3092,
      "grad_norm": 4.79997444152832,
      "learning_rate": 1.3979591836734696e-05,
      "loss": 0.78,
      "step": 3630
    },
    {
      "epoch": 1.3094000000000001,
      "grad_norm": 7.107030868530273,
      "learning_rate": 1.3969387755102042e-05,
      "loss": 0.4945,
      "step": 3631
    },
    {
      "epoch": 1.3096,
      "grad_norm": 6.470427989959717,
      "learning_rate": 1.395918367346939e-05,
      "loss": 0.7194,
      "step": 3632
    },
    {
      "epoch": 1.3098,
      "grad_norm": 1.762215495109558,
      "learning_rate": 1.3948979591836736e-05,
      "loss": 0.1192,
      "step": 3633
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.6998673677444458,
      "learning_rate": 1.3938775510204083e-05,
      "loss": 0.107,
      "step": 3634
    },
    {
      "epoch": 1.3102,
      "grad_norm": 3.6796700954437256,
      "learning_rate": 1.392857142857143e-05,
      "loss": 0.2478,
      "step": 3635
    },
    {
      "epoch": 1.3104,
      "grad_norm": 5.984239101409912,
      "learning_rate": 1.3918367346938776e-05,
      "loss": 0.8121,
      "step": 3636
    },
    {
      "epoch": 1.3106,
      "grad_norm": 5.5593695640563965,
      "learning_rate": 1.3908163265306124e-05,
      "loss": 0.8158,
      "step": 3637
    },
    {
      "epoch": 1.3108,
      "grad_norm": 8.170197486877441,
      "learning_rate": 1.389795918367347e-05,
      "loss": 1.161,
      "step": 3638
    },
    {
      "epoch": 1.311,
      "grad_norm": 1.3757100105285645,
      "learning_rate": 1.3887755102040817e-05,
      "loss": 0.0896,
      "step": 3639
    },
    {
      "epoch": 1.3112,
      "grad_norm": 1.858641266822815,
      "learning_rate": 1.3877551020408165e-05,
      "loss": 0.1156,
      "step": 3640
    },
    {
      "epoch": 1.3114,
      "grad_norm": 1.5672694444656372,
      "learning_rate": 1.386734693877551e-05,
      "loss": 0.0827,
      "step": 3641
    },
    {
      "epoch": 1.3115999999999999,
      "grad_norm": 1.9095299243927002,
      "learning_rate": 1.3857142857142858e-05,
      "loss": 0.1906,
      "step": 3642
    },
    {
      "epoch": 1.3118,
      "grad_norm": 3.8802802562713623,
      "learning_rate": 1.3846938775510204e-05,
      "loss": 0.9082,
      "step": 3643
    },
    {
      "epoch": 1.312,
      "grad_norm": 3.765718936920166,
      "learning_rate": 1.3836734693877551e-05,
      "loss": 0.6387,
      "step": 3644
    },
    {
      "epoch": 1.3122,
      "grad_norm": 4.606767177581787,
      "learning_rate": 1.3826530612244899e-05,
      "loss": 0.5988,
      "step": 3645
    },
    {
      "epoch": 1.3124,
      "grad_norm": 1.8502494096755981,
      "learning_rate": 1.3816326530612244e-05,
      "loss": 0.1423,
      "step": 3646
    },
    {
      "epoch": 1.3126,
      "grad_norm": 1.8141132593154907,
      "learning_rate": 1.3806122448979592e-05,
      "loss": 0.1143,
      "step": 3647
    },
    {
      "epoch": 1.3128,
      "grad_norm": 1.5949076414108276,
      "learning_rate": 1.3795918367346938e-05,
      "loss": 0.111,
      "step": 3648
    },
    {
      "epoch": 1.313,
      "grad_norm": 1.3847777843475342,
      "learning_rate": 1.3785714285714285e-05,
      "loss": 0.0915,
      "step": 3649
    },
    {
      "epoch": 1.3132,
      "grad_norm": 1.3259258270263672,
      "learning_rate": 1.3775510204081633e-05,
      "loss": 0.0673,
      "step": 3650
    },
    {
      "epoch": 1.3134000000000001,
      "grad_norm": 1.9255751371383667,
      "learning_rate": 1.3765306122448982e-05,
      "loss": 0.1472,
      "step": 3651
    },
    {
      "epoch": 1.3136,
      "grad_norm": 1.9907561540603638,
      "learning_rate": 1.3755102040816328e-05,
      "loss": 0.1404,
      "step": 3652
    },
    {
      "epoch": 1.3138,
      "grad_norm": 3.772474765777588,
      "learning_rate": 1.3744897959183675e-05,
      "loss": 0.4566,
      "step": 3653
    },
    {
      "epoch": 1.314,
      "grad_norm": 4.542640686035156,
      "learning_rate": 1.3734693877551023e-05,
      "loss": 0.5977,
      "step": 3654
    },
    {
      "epoch": 1.3142,
      "grad_norm": 1.2687718868255615,
      "learning_rate": 1.3724489795918368e-05,
      "loss": 0.0769,
      "step": 3655
    },
    {
      "epoch": 1.3144,
      "grad_norm": 7.83428430557251,
      "learning_rate": 1.3714285714285716e-05,
      "loss": 1.029,
      "step": 3656
    },
    {
      "epoch": 1.3146,
      "grad_norm": 10.699220657348633,
      "learning_rate": 1.3704081632653062e-05,
      "loss": 2.4053,
      "step": 3657
    },
    {
      "epoch": 1.3148,
      "grad_norm": 1.6182563304901123,
      "learning_rate": 1.3693877551020409e-05,
      "loss": 0.1276,
      "step": 3658
    },
    {
      "epoch": 1.315,
      "grad_norm": 2.7827768325805664,
      "learning_rate": 1.3683673469387757e-05,
      "loss": 0.4677,
      "step": 3659
    },
    {
      "epoch": 1.3152,
      "grad_norm": 1.7074637413024902,
      "learning_rate": 1.3673469387755102e-05,
      "loss": 0.1219,
      "step": 3660
    },
    {
      "epoch": 1.3154,
      "grad_norm": 1.3121364116668701,
      "learning_rate": 1.366326530612245e-05,
      "loss": 0.0634,
      "step": 3661
    },
    {
      "epoch": 1.3155999999999999,
      "grad_norm": 1.4462305307388306,
      "learning_rate": 1.3653061224489796e-05,
      "loss": 0.1197,
      "step": 3662
    },
    {
      "epoch": 1.3158,
      "grad_norm": 2.5625054836273193,
      "learning_rate": 1.3642857142857143e-05,
      "loss": 0.3954,
      "step": 3663
    },
    {
      "epoch": 1.316,
      "grad_norm": 6.2406816482543945,
      "learning_rate": 1.363265306122449e-05,
      "loss": 1.5542,
      "step": 3664
    },
    {
      "epoch": 1.3162,
      "grad_norm": 5.714224338531494,
      "learning_rate": 1.3622448979591836e-05,
      "loss": 0.7596,
      "step": 3665
    },
    {
      "epoch": 1.3164,
      "grad_norm": 3.18538498878479,
      "learning_rate": 1.3612244897959184e-05,
      "loss": 0.6324,
      "step": 3666
    },
    {
      "epoch": 1.3166,
      "grad_norm": 1.0018253326416016,
      "learning_rate": 1.3602040816326531e-05,
      "loss": 0.0642,
      "step": 3667
    },
    {
      "epoch": 1.3168,
      "grad_norm": 1.6076992750167847,
      "learning_rate": 1.3591836734693877e-05,
      "loss": 0.0985,
      "step": 3668
    },
    {
      "epoch": 1.317,
      "grad_norm": 1.3837776184082031,
      "learning_rate": 1.3581632653061225e-05,
      "loss": 0.0813,
      "step": 3669
    },
    {
      "epoch": 1.3172,
      "grad_norm": 1.8163518905639648,
      "learning_rate": 1.357142857142857e-05,
      "loss": 0.147,
      "step": 3670
    },
    {
      "epoch": 1.3174000000000001,
      "grad_norm": 2.740192413330078,
      "learning_rate": 1.3561224489795918e-05,
      "loss": 0.5002,
      "step": 3671
    },
    {
      "epoch": 1.3176,
      "grad_norm": 7.440036773681641,
      "learning_rate": 1.3551020408163265e-05,
      "loss": 0.9479,
      "step": 3672
    },
    {
      "epoch": 1.3178,
      "grad_norm": 11.16032600402832,
      "learning_rate": 1.3540816326530615e-05,
      "loss": 3.4297,
      "step": 3673
    },
    {
      "epoch": 1.318,
      "grad_norm": 4.578827381134033,
      "learning_rate": 1.353061224489796e-05,
      "loss": 0.8416,
      "step": 3674
    },
    {
      "epoch": 1.3182,
      "grad_norm": 7.189239501953125,
      "learning_rate": 1.3520408163265308e-05,
      "loss": 2.1488,
      "step": 3675
    },
    {
      "epoch": 1.3184,
      "grad_norm": 3.0757930278778076,
      "learning_rate": 1.3510204081632655e-05,
      "loss": 0.5967,
      "step": 3676
    },
    {
      "epoch": 1.3186,
      "grad_norm": 2.4431116580963135,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.1462,
      "step": 3677
    },
    {
      "epoch": 1.3188,
      "grad_norm": 5.975914001464844,
      "learning_rate": 1.3489795918367349e-05,
      "loss": 0.4564,
      "step": 3678
    },
    {
      "epoch": 1.319,
      "grad_norm": 3.4433000087738037,
      "learning_rate": 1.3479591836734694e-05,
      "loss": 0.8065,
      "step": 3679
    },
    {
      "epoch": 1.3192,
      "grad_norm": 4.407238960266113,
      "learning_rate": 1.3469387755102042e-05,
      "loss": 0.6133,
      "step": 3680
    },
    {
      "epoch": 1.3194,
      "grad_norm": 4.648280143737793,
      "learning_rate": 1.345918367346939e-05,
      "loss": 0.5786,
      "step": 3681
    },
    {
      "epoch": 1.3195999999999999,
      "grad_norm": 4.969205856323242,
      "learning_rate": 1.3448979591836735e-05,
      "loss": 0.3892,
      "step": 3682
    },
    {
      "epoch": 1.3197999999999999,
      "grad_norm": 1.3952200412750244,
      "learning_rate": 1.3438775510204083e-05,
      "loss": 0.0894,
      "step": 3683
    },
    {
      "epoch": 1.32,
      "grad_norm": 3.4232287406921387,
      "learning_rate": 1.3428571428571429e-05,
      "loss": 0.2368,
      "step": 3684
    },
    {
      "epoch": 1.3202,
      "grad_norm": 5.8791728019714355,
      "learning_rate": 1.3418367346938776e-05,
      "loss": 0.851,
      "step": 3685
    },
    {
      "epoch": 1.3204,
      "grad_norm": 2.5943305492401123,
      "learning_rate": 1.3408163265306123e-05,
      "loss": 0.2203,
      "step": 3686
    },
    {
      "epoch": 1.3206,
      "grad_norm": 5.222429275512695,
      "learning_rate": 1.339795918367347e-05,
      "loss": 0.8633,
      "step": 3687
    },
    {
      "epoch": 1.3208,
      "grad_norm": 3.6457314491271973,
      "learning_rate": 1.3387755102040817e-05,
      "loss": 0.7084,
      "step": 3688
    },
    {
      "epoch": 1.321,
      "grad_norm": 6.079892635345459,
      "learning_rate": 1.3377551020408163e-05,
      "loss": 1.2813,
      "step": 3689
    },
    {
      "epoch": 1.3212,
      "grad_norm": 9.036714553833008,
      "learning_rate": 1.336734693877551e-05,
      "loss": 3.1631,
      "step": 3690
    },
    {
      "epoch": 1.3214000000000001,
      "grad_norm": 19.296865463256836,
      "learning_rate": 1.3357142857142858e-05,
      "loss": 4.8567,
      "step": 3691
    },
    {
      "epoch": 1.3216,
      "grad_norm": 12.835603713989258,
      "learning_rate": 1.3346938775510203e-05,
      "loss": 3.3974,
      "step": 3692
    },
    {
      "epoch": 1.3218,
      "grad_norm": 3.113570213317871,
      "learning_rate": 1.333673469387755e-05,
      "loss": 0.1844,
      "step": 3693
    },
    {
      "epoch": 1.322,
      "grad_norm": 7.056849956512451,
      "learning_rate": 1.3326530612244897e-05,
      "loss": 0.7761,
      "step": 3694
    },
    {
      "epoch": 1.3222,
      "grad_norm": 1.145198106765747,
      "learning_rate": 1.3316326530612244e-05,
      "loss": 0.0626,
      "step": 3695
    },
    {
      "epoch": 1.3224,
      "grad_norm": 1.8467090129852295,
      "learning_rate": 1.3306122448979593e-05,
      "loss": 0.0861,
      "step": 3696
    },
    {
      "epoch": 1.3226,
      "grad_norm": 5.691300392150879,
      "learning_rate": 1.329591836734694e-05,
      "loss": 0.8919,
      "step": 3697
    },
    {
      "epoch": 1.3228,
      "grad_norm": 10.928914070129395,
      "learning_rate": 1.3285714285714288e-05,
      "loss": 3.3144,
      "step": 3698
    },
    {
      "epoch": 1.323,
      "grad_norm": 2.7220211029052734,
      "learning_rate": 1.3275510204081634e-05,
      "loss": 0.2012,
      "step": 3699
    },
    {
      "epoch": 1.3232,
      "grad_norm": 6.318994045257568,
      "learning_rate": 1.3265306122448982e-05,
      "loss": 0.797,
      "step": 3700
    },
    {
      "epoch": 1.3234,
      "grad_norm": 7.228105068206787,
      "learning_rate": 1.3255102040816327e-05,
      "loss": 0.7282,
      "step": 3701
    },
    {
      "epoch": 1.3235999999999999,
      "grad_norm": 1.5779112577438354,
      "learning_rate": 1.3244897959183675e-05,
      "loss": 0.0974,
      "step": 3702
    },
    {
      "epoch": 1.3237999999999999,
      "grad_norm": 1.5987306833267212,
      "learning_rate": 1.3234693877551022e-05,
      "loss": 0.0883,
      "step": 3703
    },
    {
      "epoch": 1.324,
      "grad_norm": 2.9678375720977783,
      "learning_rate": 1.3224489795918368e-05,
      "loss": 0.4179,
      "step": 3704
    },
    {
      "epoch": 1.3242,
      "grad_norm": 5.287540912628174,
      "learning_rate": 1.3214285714285716e-05,
      "loss": 1.488,
      "step": 3705
    },
    {
      "epoch": 1.3244,
      "grad_norm": 2.976473093032837,
      "learning_rate": 1.3204081632653061e-05,
      "loss": 0.599,
      "step": 3706
    },
    {
      "epoch": 1.3246,
      "grad_norm": 1.6741358041763306,
      "learning_rate": 1.3193877551020409e-05,
      "loss": 0.1117,
      "step": 3707
    },
    {
      "epoch": 1.3248,
      "grad_norm": 1.5690118074417114,
      "learning_rate": 1.3183673469387756e-05,
      "loss": 0.0883,
      "step": 3708
    },
    {
      "epoch": 1.325,
      "grad_norm": 4.477802276611328,
      "learning_rate": 1.3173469387755102e-05,
      "loss": 0.6296,
      "step": 3709
    },
    {
      "epoch": 1.3252,
      "grad_norm": 6.388333797454834,
      "learning_rate": 1.316326530612245e-05,
      "loss": 2.0726,
      "step": 3710
    },
    {
      "epoch": 1.3254000000000001,
      "grad_norm": 5.454859733581543,
      "learning_rate": 1.3153061224489795e-05,
      "loss": 0.7201,
      "step": 3711
    },
    {
      "epoch": 1.3256000000000001,
      "grad_norm": 1.328463077545166,
      "learning_rate": 1.3142857142857143e-05,
      "loss": 0.0711,
      "step": 3712
    },
    {
      "epoch": 1.3258,
      "grad_norm": 1.4585798978805542,
      "learning_rate": 1.313265306122449e-05,
      "loss": 0.0998,
      "step": 3713
    },
    {
      "epoch": 1.326,
      "grad_norm": 1.403083086013794,
      "learning_rate": 1.3122448979591836e-05,
      "loss": 0.0689,
      "step": 3714
    },
    {
      "epoch": 1.3262,
      "grad_norm": 1.627443552017212,
      "learning_rate": 1.3112244897959184e-05,
      "loss": 0.1369,
      "step": 3715
    },
    {
      "epoch": 1.3264,
      "grad_norm": 5.76826810836792,
      "learning_rate": 1.310204081632653e-05,
      "loss": 0.5985,
      "step": 3716
    },
    {
      "epoch": 1.3266,
      "grad_norm": 3.087222099304199,
      "learning_rate": 1.3091836734693877e-05,
      "loss": 0.5963,
      "step": 3717
    },
    {
      "epoch": 1.3268,
      "grad_norm": 2.640472888946533,
      "learning_rate": 1.3081632653061224e-05,
      "loss": 0.1815,
      "step": 3718
    },
    {
      "epoch": 1.327,
      "grad_norm": 4.838239669799805,
      "learning_rate": 1.3071428571428574e-05,
      "loss": 0.6387,
      "step": 3719
    },
    {
      "epoch": 1.3272,
      "grad_norm": 2.5957911014556885,
      "learning_rate": 1.306122448979592e-05,
      "loss": 0.2261,
      "step": 3720
    },
    {
      "epoch": 1.3274,
      "grad_norm": 4.602501392364502,
      "learning_rate": 1.3051020408163267e-05,
      "loss": 0.6065,
      "step": 3721
    },
    {
      "epoch": 1.3276,
      "grad_norm": 1.1260366439819336,
      "learning_rate": 1.3040816326530614e-05,
      "loss": 0.0524,
      "step": 3722
    },
    {
      "epoch": 1.3277999999999999,
      "grad_norm": 1.6132937669754028,
      "learning_rate": 1.303061224489796e-05,
      "loss": 0.096,
      "step": 3723
    },
    {
      "epoch": 1.328,
      "grad_norm": 1.4178515672683716,
      "learning_rate": 1.3020408163265308e-05,
      "loss": 0.0875,
      "step": 3724
    },
    {
      "epoch": 1.3282,
      "grad_norm": 0.9025347828865051,
      "learning_rate": 1.3010204081632653e-05,
      "loss": 0.0325,
      "step": 3725
    },
    {
      "epoch": 1.3284,
      "grad_norm": 1.5743216276168823,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.0891,
      "step": 3726
    },
    {
      "epoch": 1.3286,
      "grad_norm": 1.3139675855636597,
      "learning_rate": 1.2989795918367348e-05,
      "loss": 0.1146,
      "step": 3727
    },
    {
      "epoch": 1.3288,
      "grad_norm": 1.460642695426941,
      "learning_rate": 1.2979591836734694e-05,
      "loss": 0.0889,
      "step": 3728
    },
    {
      "epoch": 1.329,
      "grad_norm": 1.1795800924301147,
      "learning_rate": 1.2969387755102042e-05,
      "loss": 0.0618,
      "step": 3729
    },
    {
      "epoch": 1.3292,
      "grad_norm": 1.6362366676330566,
      "learning_rate": 1.2959183673469389e-05,
      "loss": 0.1454,
      "step": 3730
    },
    {
      "epoch": 1.3294000000000001,
      "grad_norm": 2.8314414024353027,
      "learning_rate": 1.2948979591836735e-05,
      "loss": 0.4987,
      "step": 3731
    },
    {
      "epoch": 1.3296000000000001,
      "grad_norm": 1.8919548988342285,
      "learning_rate": 1.2938775510204082e-05,
      "loss": 0.1928,
      "step": 3732
    },
    {
      "epoch": 1.3298,
      "grad_norm": 2.7952616214752197,
      "learning_rate": 1.2928571428571428e-05,
      "loss": 0.5315,
      "step": 3733
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.3814488649368286,
      "learning_rate": 1.2918367346938776e-05,
      "loss": 0.0935,
      "step": 3734
    },
    {
      "epoch": 1.3302,
      "grad_norm": 1.2890468835830688,
      "learning_rate": 1.2908163265306123e-05,
      "loss": 0.0548,
      "step": 3735
    },
    {
      "epoch": 1.3304,
      "grad_norm": 1.5378800630569458,
      "learning_rate": 1.2897959183673469e-05,
      "loss": 0.0899,
      "step": 3736
    },
    {
      "epoch": 1.3306,
      "grad_norm": 1.6947510242462158,
      "learning_rate": 1.2887755102040816e-05,
      "loss": 0.1284,
      "step": 3737
    },
    {
      "epoch": 1.3308,
      "grad_norm": 1.4605859518051147,
      "learning_rate": 1.2877551020408162e-05,
      "loss": 0.0965,
      "step": 3738
    },
    {
      "epoch": 1.331,
      "grad_norm": 1.8479279279708862,
      "learning_rate": 1.286734693877551e-05,
      "loss": 0.1376,
      "step": 3739
    },
    {
      "epoch": 1.3312,
      "grad_norm": 6.6838459968566895,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 1.0591,
      "step": 3740
    },
    {
      "epoch": 1.3314,
      "grad_norm": 7.753203868865967,
      "learning_rate": 1.2846938775510206e-05,
      "loss": 1.8856,
      "step": 3741
    },
    {
      "epoch": 1.3316,
      "grad_norm": 1.8943339586257935,
      "learning_rate": 1.2836734693877552e-05,
      "loss": 0.1905,
      "step": 3742
    },
    {
      "epoch": 1.3317999999999999,
      "grad_norm": 3.1973650455474854,
      "learning_rate": 1.28265306122449e-05,
      "loss": 0.6096,
      "step": 3743
    },
    {
      "epoch": 1.332,
      "grad_norm": 1.7720997333526611,
      "learning_rate": 1.2816326530612247e-05,
      "loss": 0.106,
      "step": 3744
    },
    {
      "epoch": 1.3322,
      "grad_norm": 1.6186349391937256,
      "learning_rate": 1.2806122448979593e-05,
      "loss": 0.1065,
      "step": 3745
    },
    {
      "epoch": 1.3324,
      "grad_norm": 1.695291519165039,
      "learning_rate": 1.279591836734694e-05,
      "loss": 0.112,
      "step": 3746
    },
    {
      "epoch": 1.3326,
      "grad_norm": 1.3820916414260864,
      "learning_rate": 1.2785714285714286e-05,
      "loss": 0.0755,
      "step": 3747
    },
    {
      "epoch": 1.3328,
      "grad_norm": 1.4397777318954468,
      "learning_rate": 1.2775510204081634e-05,
      "loss": 0.0657,
      "step": 3748
    },
    {
      "epoch": 1.333,
      "grad_norm": 1.9753934144973755,
      "learning_rate": 1.2765306122448981e-05,
      "loss": 0.1784,
      "step": 3749
    },
    {
      "epoch": 1.3332,
      "grad_norm": 3.92526912689209,
      "learning_rate": 1.2755102040816327e-05,
      "loss": 0.6451,
      "step": 3750
    },
    {
      "epoch": 1.3334,
      "grad_norm": 3.340161085128784,
      "learning_rate": 1.2744897959183674e-05,
      "loss": 0.8984,
      "step": 3751
    },
    {
      "epoch": 1.3336000000000001,
      "grad_norm": 4.063291072845459,
      "learning_rate": 1.273469387755102e-05,
      "loss": 0.9231,
      "step": 3752
    },
    {
      "epoch": 1.3338,
      "grad_norm": 4.840859889984131,
      "learning_rate": 1.2724489795918368e-05,
      "loss": 1.4109,
      "step": 3753
    },
    {
      "epoch": 1.334,
      "grad_norm": 1.761492133140564,
      "learning_rate": 1.2714285714285715e-05,
      "loss": 0.1427,
      "step": 3754
    },
    {
      "epoch": 1.3342,
      "grad_norm": 4.310940742492676,
      "learning_rate": 1.2704081632653061e-05,
      "loss": 0.64,
      "step": 3755
    },
    {
      "epoch": 1.3344,
      "grad_norm": 5.979731559753418,
      "learning_rate": 1.2693877551020409e-05,
      "loss": 0.7816,
      "step": 3756
    },
    {
      "epoch": 1.3346,
      "grad_norm": 1.581800937652588,
      "learning_rate": 1.2683673469387754e-05,
      "loss": 0.1127,
      "step": 3757
    },
    {
      "epoch": 1.3348,
      "grad_norm": 1.3641674518585205,
      "learning_rate": 1.2673469387755102e-05,
      "loss": 0.077,
      "step": 3758
    },
    {
      "epoch": 1.335,
      "grad_norm": 2.465879201889038,
      "learning_rate": 1.266326530612245e-05,
      "loss": 0.3563,
      "step": 3759
    },
    {
      "epoch": 1.3352,
      "grad_norm": 4.839778423309326,
      "learning_rate": 1.2653061224489795e-05,
      "loss": 1.443,
      "step": 3760
    },
    {
      "epoch": 1.3354,
      "grad_norm": 2.8524973392486572,
      "learning_rate": 1.2642857142857143e-05,
      "loss": 0.2132,
      "step": 3761
    },
    {
      "epoch": 1.3356,
      "grad_norm": 7.233521938323975,
      "learning_rate": 1.263265306122449e-05,
      "loss": 1.3656,
      "step": 3762
    },
    {
      "epoch": 1.3357999999999999,
      "grad_norm": 9.550909042358398,
      "learning_rate": 1.2622448979591836e-05,
      "loss": 2.0322,
      "step": 3763
    },
    {
      "epoch": 1.336,
      "grad_norm": 5.07677698135376,
      "learning_rate": 1.2612244897959185e-05,
      "loss": 0.665,
      "step": 3764
    },
    {
      "epoch": 1.3362,
      "grad_norm": 2.5103707313537598,
      "learning_rate": 1.2602040816326533e-05,
      "loss": 0.2274,
      "step": 3765
    },
    {
      "epoch": 1.3364,
      "grad_norm": 4.901525974273682,
      "learning_rate": 1.259183673469388e-05,
      "loss": 0.6456,
      "step": 3766
    },
    {
      "epoch": 1.3366,
      "grad_norm": 1.6755197048187256,
      "learning_rate": 1.2581632653061226e-05,
      "loss": 0.1005,
      "step": 3767
    },
    {
      "epoch": 1.3368,
      "grad_norm": 2.485529899597168,
      "learning_rate": 1.2571428571428573e-05,
      "loss": 0.1637,
      "step": 3768
    },
    {
      "epoch": 1.337,
      "grad_norm": 7.253251552581787,
      "learning_rate": 1.2561224489795919e-05,
      "loss": 0.8481,
      "step": 3769
    },
    {
      "epoch": 1.3372,
      "grad_norm": 5.495018482208252,
      "learning_rate": 1.2551020408163267e-05,
      "loss": 0.8316,
      "step": 3770
    },
    {
      "epoch": 1.3374,
      "grad_norm": 2.9167444705963135,
      "learning_rate": 1.2540816326530614e-05,
      "loss": 0.5102,
      "step": 3771
    },
    {
      "epoch": 1.3376000000000001,
      "grad_norm": 1.8908495903015137,
      "learning_rate": 1.253061224489796e-05,
      "loss": 0.1286,
      "step": 3772
    },
    {
      "epoch": 1.3378,
      "grad_norm": 1.4749886989593506,
      "learning_rate": 1.2520408163265307e-05,
      "loss": 0.0961,
      "step": 3773
    },
    {
      "epoch": 1.338,
      "grad_norm": 1.9437893629074097,
      "learning_rate": 1.2510204081632653e-05,
      "loss": 0.1222,
      "step": 3774
    },
    {
      "epoch": 1.3382,
      "grad_norm": 1.6354379653930664,
      "learning_rate": 1.25e-05,
      "loss": 0.107,
      "step": 3775
    },
    {
      "epoch": 1.3384,
      "grad_norm": 1.515820026397705,
      "learning_rate": 1.2489795918367348e-05,
      "loss": 0.1014,
      "step": 3776
    },
    {
      "epoch": 1.3386,
      "grad_norm": 2.15248441696167,
      "learning_rate": 1.2479591836734694e-05,
      "loss": 0.2705,
      "step": 3777
    },
    {
      "epoch": 1.3388,
      "grad_norm": 3.161968469619751,
      "learning_rate": 1.2469387755102041e-05,
      "loss": 0.3206,
      "step": 3778
    },
    {
      "epoch": 1.339,
      "grad_norm": 1.6812952756881714,
      "learning_rate": 1.2459183673469387e-05,
      "loss": 0.1136,
      "step": 3779
    },
    {
      "epoch": 1.3392,
      "grad_norm": 1.4099876880645752,
      "learning_rate": 1.2448979591836735e-05,
      "loss": 0.1028,
      "step": 3780
    },
    {
      "epoch": 1.3394,
      "grad_norm": 3.105964183807373,
      "learning_rate": 1.2438775510204082e-05,
      "loss": 0.4409,
      "step": 3781
    },
    {
      "epoch": 1.3396,
      "grad_norm": 4.7388596534729,
      "learning_rate": 1.242857142857143e-05,
      "loss": 1.4694,
      "step": 3782
    },
    {
      "epoch": 1.3397999999999999,
      "grad_norm": 3.004696846008301,
      "learning_rate": 1.2418367346938777e-05,
      "loss": 0.5646,
      "step": 3783
    },
    {
      "epoch": 1.34,
      "grad_norm": 1.8444324731826782,
      "learning_rate": 1.2408163265306123e-05,
      "loss": 0.1194,
      "step": 3784
    },
    {
      "epoch": 1.3402,
      "grad_norm": 1.3238780498504639,
      "learning_rate": 1.239795918367347e-05,
      "loss": 0.0807,
      "step": 3785
    },
    {
      "epoch": 1.3404,
      "grad_norm": 1.6439921855926514,
      "learning_rate": 1.2387755102040816e-05,
      "loss": 0.1438,
      "step": 3786
    },
    {
      "epoch": 1.3406,
      "grad_norm": 2.3479502201080322,
      "learning_rate": 1.2377551020408164e-05,
      "loss": 0.3354,
      "step": 3787
    },
    {
      "epoch": 1.3408,
      "grad_norm": 3.0557663440704346,
      "learning_rate": 1.2367346938775511e-05,
      "loss": 0.587,
      "step": 3788
    },
    {
      "epoch": 1.341,
      "grad_norm": 2.2186567783355713,
      "learning_rate": 1.2357142857142857e-05,
      "loss": 0.1944,
      "step": 3789
    },
    {
      "epoch": 1.3412,
      "grad_norm": 4.796391487121582,
      "learning_rate": 1.2346938775510204e-05,
      "loss": 0.612,
      "step": 3790
    },
    {
      "epoch": 1.3414,
      "grad_norm": 4.777772426605225,
      "learning_rate": 1.233673469387755e-05,
      "loss": 0.8965,
      "step": 3791
    },
    {
      "epoch": 1.3416000000000001,
      "grad_norm": 11.811017036437988,
      "learning_rate": 1.2326530612244898e-05,
      "loss": 3.2756,
      "step": 3792
    },
    {
      "epoch": 1.3418,
      "grad_norm": 8.805556297302246,
      "learning_rate": 1.2316326530612247e-05,
      "loss": 2.5276,
      "step": 3793
    },
    {
      "epoch": 1.342,
      "grad_norm": 1.7621744871139526,
      "learning_rate": 1.2306122448979593e-05,
      "loss": 0.1785,
      "step": 3794
    },
    {
      "epoch": 1.3422,
      "grad_norm": 2.9811997413635254,
      "learning_rate": 1.229591836734694e-05,
      "loss": 0.4932,
      "step": 3795
    },
    {
      "epoch": 1.3424,
      "grad_norm": 1.205106258392334,
      "learning_rate": 1.2285714285714286e-05,
      "loss": 0.06,
      "step": 3796
    },
    {
      "epoch": 1.3426,
      "grad_norm": 2.116752862930298,
      "learning_rate": 1.2275510204081633e-05,
      "loss": 0.2296,
      "step": 3797
    },
    {
      "epoch": 1.3428,
      "grad_norm": 2.8943114280700684,
      "learning_rate": 1.2265306122448981e-05,
      "loss": 0.5197,
      "step": 3798
    },
    {
      "epoch": 1.343,
      "grad_norm": 4.272083282470703,
      "learning_rate": 1.2255102040816327e-05,
      "loss": 0.6344,
      "step": 3799
    },
    {
      "epoch": 1.3432,
      "grad_norm": 8.491045951843262,
      "learning_rate": 1.2244897959183674e-05,
      "loss": 2.5191,
      "step": 3800
    },
    {
      "epoch": 1.3434,
      "grad_norm": 1.980488657951355,
      "learning_rate": 1.223469387755102e-05,
      "loss": 0.2081,
      "step": 3801
    },
    {
      "epoch": 1.3436,
      "grad_norm": 3.655423164367676,
      "learning_rate": 1.2224489795918367e-05,
      "loss": 0.6737,
      "step": 3802
    },
    {
      "epoch": 1.3437999999999999,
      "grad_norm": 5.269712448120117,
      "learning_rate": 1.2214285714285715e-05,
      "loss": 0.7562,
      "step": 3803
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 2.799565553665161,
      "learning_rate": 1.2204081632653062e-05,
      "loss": 0.5657,
      "step": 3804
    },
    {
      "epoch": 1.3442,
      "grad_norm": 1.1319365501403809,
      "learning_rate": 1.219387755102041e-05,
      "loss": 0.0589,
      "step": 3805
    },
    {
      "epoch": 1.3444,
      "grad_norm": 1.70322585105896,
      "learning_rate": 1.2183673469387756e-05,
      "loss": 0.0815,
      "step": 3806
    },
    {
      "epoch": 1.3446,
      "grad_norm": 1.3940032720565796,
      "learning_rate": 1.2173469387755103e-05,
      "loss": 0.0749,
      "step": 3807
    },
    {
      "epoch": 1.3448,
      "grad_norm": 1.9430924654006958,
      "learning_rate": 1.2163265306122449e-05,
      "loss": 0.1559,
      "step": 3808
    },
    {
      "epoch": 1.345,
      "grad_norm": 3.467571973800659,
      "learning_rate": 1.2153061224489796e-05,
      "loss": 0.6982,
      "step": 3809
    },
    {
      "epoch": 1.3452,
      "grad_norm": 5.4435224533081055,
      "learning_rate": 1.2142857142857144e-05,
      "loss": 1.0138,
      "step": 3810
    },
    {
      "epoch": 1.3454,
      "grad_norm": 9.767927169799805,
      "learning_rate": 1.213265306122449e-05,
      "loss": 1.7398,
      "step": 3811
    },
    {
      "epoch": 1.3456000000000001,
      "grad_norm": 5.552506923675537,
      "learning_rate": 1.2122448979591837e-05,
      "loss": 0.8101,
      "step": 3812
    },
    {
      "epoch": 1.3458,
      "grad_norm": 3.9132087230682373,
      "learning_rate": 1.2112244897959183e-05,
      "loss": 0.478,
      "step": 3813
    },
    {
      "epoch": 1.346,
      "grad_norm": 3.659240961074829,
      "learning_rate": 1.210204081632653e-05,
      "loss": 0.7716,
      "step": 3814
    },
    {
      "epoch": 1.3462,
      "grad_norm": 5.815901756286621,
      "learning_rate": 1.2091836734693878e-05,
      "loss": 1.4835,
      "step": 3815
    },
    {
      "epoch": 1.3464,
      "grad_norm": 3.8814995288848877,
      "learning_rate": 1.2081632653061225e-05,
      "loss": 0.4621,
      "step": 3816
    },
    {
      "epoch": 1.3466,
      "grad_norm": 5.397446632385254,
      "learning_rate": 1.2071428571428573e-05,
      "loss": 0.7511,
      "step": 3817
    },
    {
      "epoch": 1.3468,
      "grad_norm": 4.171359539031982,
      "learning_rate": 1.2061224489795919e-05,
      "loss": 0.5696,
      "step": 3818
    },
    {
      "epoch": 1.347,
      "grad_norm": 3.8873350620269775,
      "learning_rate": 1.2051020408163266e-05,
      "loss": 0.5423,
      "step": 3819
    },
    {
      "epoch": 1.3472,
      "grad_norm": 3.4660348892211914,
      "learning_rate": 1.2040816326530612e-05,
      "loss": 0.4298,
      "step": 3820
    },
    {
      "epoch": 1.3474,
      "grad_norm": 2.9098331928253174,
      "learning_rate": 1.203061224489796e-05,
      "loss": 0.5387,
      "step": 3821
    },
    {
      "epoch": 1.3476,
      "grad_norm": 2.7750661373138428,
      "learning_rate": 1.2020408163265307e-05,
      "loss": 0.2058,
      "step": 3822
    },
    {
      "epoch": 1.3477999999999999,
      "grad_norm": 5.425154209136963,
      "learning_rate": 1.2010204081632653e-05,
      "loss": 0.7798,
      "step": 3823
    },
    {
      "epoch": 1.3479999999999999,
      "grad_norm": 3.839717149734497,
      "learning_rate": 1.2e-05,
      "loss": 0.4634,
      "step": 3824
    },
    {
      "epoch": 1.3482,
      "grad_norm": 4.481932640075684,
      "learning_rate": 1.1989795918367348e-05,
      "loss": 0.5418,
      "step": 3825
    },
    {
      "epoch": 1.3484,
      "grad_norm": 1.583490014076233,
      "learning_rate": 1.1979591836734694e-05,
      "loss": 0.1359,
      "step": 3826
    },
    {
      "epoch": 1.3486,
      "grad_norm": 3.9904890060424805,
      "learning_rate": 1.1969387755102043e-05,
      "loss": 0.7038,
      "step": 3827
    },
    {
      "epoch": 1.3488,
      "grad_norm": 6.746601104736328,
      "learning_rate": 1.1959183673469389e-05,
      "loss": 0.8479,
      "step": 3828
    },
    {
      "epoch": 1.349,
      "grad_norm": 2.903327703475952,
      "learning_rate": 1.1948979591836736e-05,
      "loss": 0.5251,
      "step": 3829
    },
    {
      "epoch": 1.3492,
      "grad_norm": 3.328550338745117,
      "learning_rate": 1.1938775510204082e-05,
      "loss": 0.3319,
      "step": 3830
    },
    {
      "epoch": 1.3494,
      "grad_norm": 6.920244216918945,
      "learning_rate": 1.192857142857143e-05,
      "loss": 1.7583,
      "step": 3831
    },
    {
      "epoch": 1.3496000000000001,
      "grad_norm": 1.6275689601898193,
      "learning_rate": 1.1918367346938777e-05,
      "loss": 0.1088,
      "step": 3832
    },
    {
      "epoch": 1.3498,
      "grad_norm": 1.926766037940979,
      "learning_rate": 1.1908163265306123e-05,
      "loss": 0.1258,
      "step": 3833
    },
    {
      "epoch": 1.35,
      "grad_norm": 3.261030673980713,
      "learning_rate": 1.189795918367347e-05,
      "loss": 0.3952,
      "step": 3834
    },
    {
      "epoch": 1.3502,
      "grad_norm": 1.8843517303466797,
      "learning_rate": 1.1887755102040816e-05,
      "loss": 0.1201,
      "step": 3835
    },
    {
      "epoch": 1.3504,
      "grad_norm": 4.032954692840576,
      "learning_rate": 1.1877551020408163e-05,
      "loss": 0.5584,
      "step": 3836
    },
    {
      "epoch": 1.3506,
      "grad_norm": 4.915853500366211,
      "learning_rate": 1.186734693877551e-05,
      "loss": 0.5074,
      "step": 3837
    },
    {
      "epoch": 1.3508,
      "grad_norm": 5.292396068572998,
      "learning_rate": 1.1857142857142858e-05,
      "loss": 0.8171,
      "step": 3838
    },
    {
      "epoch": 1.351,
      "grad_norm": 2.990004301071167,
      "learning_rate": 1.1846938775510206e-05,
      "loss": 0.5732,
      "step": 3839
    },
    {
      "epoch": 1.3512,
      "grad_norm": 4.736379146575928,
      "learning_rate": 1.1836734693877552e-05,
      "loss": 0.5969,
      "step": 3840
    },
    {
      "epoch": 1.3514,
      "grad_norm": 8.278603553771973,
      "learning_rate": 1.1826530612244899e-05,
      "loss": 2.2096,
      "step": 3841
    },
    {
      "epoch": 1.3516,
      "grad_norm": 5.408187389373779,
      "learning_rate": 1.1816326530612245e-05,
      "loss": 0.9051,
      "step": 3842
    },
    {
      "epoch": 1.3518,
      "grad_norm": 4.891638278961182,
      "learning_rate": 1.1806122448979592e-05,
      "loss": 1.3975,
      "step": 3843
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 1.6680200099945068,
      "learning_rate": 1.179591836734694e-05,
      "loss": 0.1287,
      "step": 3844
    },
    {
      "epoch": 1.3522,
      "grad_norm": 1.410937786102295,
      "learning_rate": 1.1785714285714286e-05,
      "loss": 0.078,
      "step": 3845
    },
    {
      "epoch": 1.3524,
      "grad_norm": 3.5020029544830322,
      "learning_rate": 1.1775510204081633e-05,
      "loss": 0.5996,
      "step": 3846
    },
    {
      "epoch": 1.3526,
      "grad_norm": 7.307976722717285,
      "learning_rate": 1.1765306122448979e-05,
      "loss": 2.084,
      "step": 3847
    },
    {
      "epoch": 1.3528,
      "grad_norm": 3.692384719848633,
      "learning_rate": 1.1755102040816326e-05,
      "loss": 0.6842,
      "step": 3848
    },
    {
      "epoch": 1.353,
      "grad_norm": 3.2924370765686035,
      "learning_rate": 1.1744897959183674e-05,
      "loss": 0.3576,
      "step": 3849
    },
    {
      "epoch": 1.3532,
      "grad_norm": 1.646892786026001,
      "learning_rate": 1.1734693877551021e-05,
      "loss": 0.1366,
      "step": 3850
    },
    {
      "epoch": 1.3534,
      "grad_norm": 2.786102533340454,
      "learning_rate": 1.1724489795918369e-05,
      "loss": 0.2471,
      "step": 3851
    },
    {
      "epoch": 1.3536000000000001,
      "grad_norm": 5.256365776062012,
      "learning_rate": 1.1714285714285715e-05,
      "loss": 0.6327,
      "step": 3852
    },
    {
      "epoch": 1.3538000000000001,
      "grad_norm": 2.0491883754730225,
      "learning_rate": 1.1704081632653062e-05,
      "loss": 0.214,
      "step": 3853
    },
    {
      "epoch": 1.354,
      "grad_norm": 3.036485433578491,
      "learning_rate": 1.1693877551020408e-05,
      "loss": 0.5884,
      "step": 3854
    },
    {
      "epoch": 1.3542,
      "grad_norm": 1.5038535594940186,
      "learning_rate": 1.1683673469387755e-05,
      "loss": 0.1011,
      "step": 3855
    },
    {
      "epoch": 1.3544,
      "grad_norm": 1.5168235301971436,
      "learning_rate": 1.1673469387755103e-05,
      "loss": 0.09,
      "step": 3856
    },
    {
      "epoch": 1.3546,
      "grad_norm": 1.7048945426940918,
      "learning_rate": 1.1663265306122449e-05,
      "loss": 0.143,
      "step": 3857
    },
    {
      "epoch": 1.3548,
      "grad_norm": 1.708801031112671,
      "learning_rate": 1.1653061224489796e-05,
      "loss": 0.1426,
      "step": 3858
    },
    {
      "epoch": 1.355,
      "grad_norm": 3.5101585388183594,
      "learning_rate": 1.1642857142857144e-05,
      "loss": 0.6942,
      "step": 3859
    },
    {
      "epoch": 1.3552,
      "grad_norm": 7.236852169036865,
      "learning_rate": 1.163265306122449e-05,
      "loss": 2.1071,
      "step": 3860
    },
    {
      "epoch": 1.3554,
      "grad_norm": 2.947705030441284,
      "learning_rate": 1.1622448979591839e-05,
      "loss": 0.5327,
      "step": 3861
    },
    {
      "epoch": 1.3556,
      "grad_norm": 2.4596071243286133,
      "learning_rate": 1.1612244897959184e-05,
      "loss": 0.3587,
      "step": 3862
    },
    {
      "epoch": 1.3558,
      "grad_norm": 4.5340256690979,
      "learning_rate": 1.1602040816326532e-05,
      "loss": 1.5091,
      "step": 3863
    },
    {
      "epoch": 1.3559999999999999,
      "grad_norm": 1.7653383016586304,
      "learning_rate": 1.1591836734693878e-05,
      "loss": 0.1534,
      "step": 3864
    },
    {
      "epoch": 1.3562,
      "grad_norm": 3.754122734069824,
      "learning_rate": 1.1581632653061225e-05,
      "loss": 0.6394,
      "step": 3865
    },
    {
      "epoch": 1.3564,
      "grad_norm": 6.031116962432861,
      "learning_rate": 1.1571428571428573e-05,
      "loss": 0.8656,
      "step": 3866
    },
    {
      "epoch": 1.3566,
      "grad_norm": 4.633213043212891,
      "learning_rate": 1.1561224489795918e-05,
      "loss": 0.7653,
      "step": 3867
    },
    {
      "epoch": 1.3568,
      "grad_norm": 3.4678215980529785,
      "learning_rate": 1.1551020408163266e-05,
      "loss": 0.6047,
      "step": 3868
    },
    {
      "epoch": 1.357,
      "grad_norm": 3.8937063217163086,
      "learning_rate": 1.1540816326530612e-05,
      "loss": 0.445,
      "step": 3869
    },
    {
      "epoch": 1.3572,
      "grad_norm": 4.51537561416626,
      "learning_rate": 1.153061224489796e-05,
      "loss": 1.0653,
      "step": 3870
    },
    {
      "epoch": 1.3574,
      "grad_norm": 8.521768569946289,
      "learning_rate": 1.1520408163265307e-05,
      "loss": 2.3745,
      "step": 3871
    },
    {
      "epoch": 1.3576,
      "grad_norm": 3.4433677196502686,
      "learning_rate": 1.1510204081632654e-05,
      "loss": 0.6414,
      "step": 3872
    },
    {
      "epoch": 1.3578000000000001,
      "grad_norm": 6.296442031860352,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.7838,
      "step": 3873
    },
    {
      "epoch": 1.358,
      "grad_norm": 5.0307207107543945,
      "learning_rate": 1.1489795918367347e-05,
      "loss": 0.7592,
      "step": 3874
    },
    {
      "epoch": 1.3582,
      "grad_norm": 3.6523263454437256,
      "learning_rate": 1.1479591836734695e-05,
      "loss": 0.63,
      "step": 3875
    },
    {
      "epoch": 1.3584,
      "grad_norm": 5.82245397567749,
      "learning_rate": 1.146938775510204e-05,
      "loss": 0.8227,
      "step": 3876
    },
    {
      "epoch": 1.3586,
      "grad_norm": 5.670694351196289,
      "learning_rate": 1.1459183673469388e-05,
      "loss": 0.7424,
      "step": 3877
    },
    {
      "epoch": 1.3588,
      "grad_norm": 5.125217437744141,
      "learning_rate": 1.1448979591836736e-05,
      "loss": 0.7306,
      "step": 3878
    },
    {
      "epoch": 1.359,
      "grad_norm": 4.8379435539245605,
      "learning_rate": 1.1438775510204082e-05,
      "loss": 0.8655,
      "step": 3879
    },
    {
      "epoch": 1.3592,
      "grad_norm": 9.271891593933105,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 3.0184,
      "step": 3880
    },
    {
      "epoch": 1.3594,
      "grad_norm": 8.492823600769043,
      "learning_rate": 1.1418367346938775e-05,
      "loss": 2.1837,
      "step": 3881
    },
    {
      "epoch": 1.3596,
      "grad_norm": 3.692840099334717,
      "learning_rate": 1.1408163265306122e-05,
      "loss": 0.4024,
      "step": 3882
    },
    {
      "epoch": 1.3598,
      "grad_norm": 4.732965469360352,
      "learning_rate": 1.139795918367347e-05,
      "loss": 0.5713,
      "step": 3883
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 5.9341583251953125,
      "learning_rate": 1.1387755102040817e-05,
      "loss": 0.9768,
      "step": 3884
    },
    {
      "epoch": 1.3602,
      "grad_norm": 4.89736270904541,
      "learning_rate": 1.1377551020408165e-05,
      "loss": 1.623,
      "step": 3885
    },
    {
      "epoch": 1.3604,
      "grad_norm": 4.786985874176025,
      "learning_rate": 1.136734693877551e-05,
      "loss": 0.6604,
      "step": 3886
    },
    {
      "epoch": 1.3606,
      "grad_norm": 5.9557623863220215,
      "learning_rate": 1.1357142857142858e-05,
      "loss": 0.8251,
      "step": 3887
    },
    {
      "epoch": 1.3608,
      "grad_norm": 4.59980583190918,
      "learning_rate": 1.1346938775510206e-05,
      "loss": 0.8647,
      "step": 3888
    },
    {
      "epoch": 1.361,
      "grad_norm": 4.638906002044678,
      "learning_rate": 1.1336734693877551e-05,
      "loss": 1.4674,
      "step": 3889
    },
    {
      "epoch": 1.3612,
      "grad_norm": 5.079880237579346,
      "learning_rate": 1.1326530612244899e-05,
      "loss": 0.7242,
      "step": 3890
    },
    {
      "epoch": 1.3614,
      "grad_norm": 5.692809104919434,
      "learning_rate": 1.1316326530612245e-05,
      "loss": 0.7839,
      "step": 3891
    },
    {
      "epoch": 1.3616,
      "grad_norm": 6.237496852874756,
      "learning_rate": 1.1306122448979592e-05,
      "loss": 1.2235,
      "step": 3892
    },
    {
      "epoch": 1.3618000000000001,
      "grad_norm": 8.457006454467773,
      "learning_rate": 1.129591836734694e-05,
      "loss": 2.2326,
      "step": 3893
    },
    {
      "epoch": 1.362,
      "grad_norm": 9.149642944335938,
      "learning_rate": 1.1285714285714285e-05,
      "loss": 0.6763,
      "step": 3894
    },
    {
      "epoch": 1.3622,
      "grad_norm": 3.8505139350891113,
      "learning_rate": 1.1275510204081635e-05,
      "loss": 0.4128,
      "step": 3895
    },
    {
      "epoch": 1.3624,
      "grad_norm": 6.831646919250488,
      "learning_rate": 1.126530612244898e-05,
      "loss": 0.8069,
      "step": 3896
    },
    {
      "epoch": 1.3626,
      "grad_norm": 4.0196146965026855,
      "learning_rate": 1.1255102040816328e-05,
      "loss": 0.4681,
      "step": 3897
    },
    {
      "epoch": 1.3628,
      "grad_norm": 7.241461277008057,
      "learning_rate": 1.1244897959183674e-05,
      "loss": 1.3956,
      "step": 3898
    },
    {
      "epoch": 1.363,
      "grad_norm": 9.091708183288574,
      "learning_rate": 1.1234693877551021e-05,
      "loss": 2.7268,
      "step": 3899
    },
    {
      "epoch": 1.3632,
      "grad_norm": 9.10069751739502,
      "learning_rate": 1.1224489795918369e-05,
      "loss": 2.2056,
      "step": 3900
    },
    {
      "epoch": 1.3634,
      "grad_norm": 6.977329730987549,
      "learning_rate": 1.1214285714285714e-05,
      "loss": 1.2128,
      "step": 3901
    },
    {
      "epoch": 1.3636,
      "grad_norm": 10.115520477294922,
      "learning_rate": 1.1204081632653062e-05,
      "loss": 2.788,
      "step": 3902
    },
    {
      "epoch": 1.3638,
      "grad_norm": 8.028214454650879,
      "learning_rate": 1.1193877551020408e-05,
      "loss": 2.0324,
      "step": 3903
    },
    {
      "epoch": 1.3639999999999999,
      "grad_norm": 4.721216678619385,
      "learning_rate": 1.1183673469387755e-05,
      "loss": 0.6433,
      "step": 3904
    },
    {
      "epoch": 1.3642,
      "grad_norm": 7.163980484008789,
      "learning_rate": 1.1173469387755103e-05,
      "loss": 1.4074,
      "step": 3905
    },
    {
      "epoch": 1.3644,
      "grad_norm": 8.084197998046875,
      "learning_rate": 1.116326530612245e-05,
      "loss": 2.0196,
      "step": 3906
    },
    {
      "epoch": 1.3646,
      "grad_norm": 4.783535480499268,
      "learning_rate": 1.1153061224489798e-05,
      "loss": 0.7092,
      "step": 3907
    },
    {
      "epoch": 1.3648,
      "grad_norm": 4.9050092697143555,
      "learning_rate": 1.1142857142857143e-05,
      "loss": 0.7534,
      "step": 3908
    },
    {
      "epoch": 1.365,
      "grad_norm": 7.531850337982178,
      "learning_rate": 1.1132653061224491e-05,
      "loss": 1.2272,
      "step": 3909
    },
    {
      "epoch": 1.3652,
      "grad_norm": 10.261857986450195,
      "learning_rate": 1.1122448979591837e-05,
      "loss": 1.9668,
      "step": 3910
    },
    {
      "epoch": 1.3654,
      "grad_norm": 3.57900333404541,
      "learning_rate": 1.1112244897959184e-05,
      "loss": 0.6528,
      "step": 3911
    },
    {
      "epoch": 1.3656,
      "grad_norm": 3.6782445907592773,
      "learning_rate": 1.1102040816326532e-05,
      "loss": 0.4855,
      "step": 3912
    },
    {
      "epoch": 1.3658000000000001,
      "grad_norm": 4.000214576721191,
      "learning_rate": 1.1091836734693877e-05,
      "loss": 0.4317,
      "step": 3913
    },
    {
      "epoch": 1.366,
      "grad_norm": 7.233711242675781,
      "learning_rate": 1.1081632653061225e-05,
      "loss": 1.1248,
      "step": 3914
    },
    {
      "epoch": 1.3662,
      "grad_norm": 9.311083793640137,
      "learning_rate": 1.107142857142857e-05,
      "loss": 2.0811,
      "step": 3915
    },
    {
      "epoch": 1.3664,
      "grad_norm": 5.855093955993652,
      "learning_rate": 1.1061224489795918e-05,
      "loss": 0.7559,
      "step": 3916
    },
    {
      "epoch": 1.3666,
      "grad_norm": 3.242572546005249,
      "learning_rate": 1.1051020408163266e-05,
      "loss": 0.5673,
      "step": 3917
    },
    {
      "epoch": 1.3668,
      "grad_norm": 3.610198974609375,
      "learning_rate": 1.1040816326530613e-05,
      "loss": 0.4493,
      "step": 3918
    },
    {
      "epoch": 1.367,
      "grad_norm": 3.7920875549316406,
      "learning_rate": 1.103061224489796e-05,
      "loss": 0.7715,
      "step": 3919
    },
    {
      "epoch": 1.3672,
      "grad_norm": 5.547231197357178,
      "learning_rate": 1.1020408163265306e-05,
      "loss": 0.8798,
      "step": 3920
    },
    {
      "epoch": 1.3674,
      "grad_norm": 8.020981788635254,
      "learning_rate": 1.1010204081632654e-05,
      "loss": 0.7225,
      "step": 3921
    },
    {
      "epoch": 1.3676,
      "grad_norm": 10.814255714416504,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 1.1109,
      "step": 3922
    },
    {
      "epoch": 1.3678,
      "grad_norm": 6.38167667388916,
      "learning_rate": 1.0989795918367347e-05,
      "loss": 0.945,
      "step": 3923
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 9.843852996826172,
      "learning_rate": 1.0979591836734695e-05,
      "loss": 2.1945,
      "step": 3924
    },
    {
      "epoch": 1.3682,
      "grad_norm": 6.032860279083252,
      "learning_rate": 1.096938775510204e-05,
      "loss": 0.7325,
      "step": 3925
    },
    {
      "epoch": 1.3684,
      "grad_norm": 8.657896041870117,
      "learning_rate": 1.0959183673469388e-05,
      "loss": 1.1041,
      "step": 3926
    },
    {
      "epoch": 1.3686,
      "grad_norm": 5.5394062995910645,
      "learning_rate": 1.0948979591836735e-05,
      "loss": 1.0176,
      "step": 3927
    },
    {
      "epoch": 1.3688,
      "grad_norm": 9.15536117553711,
      "learning_rate": 1.0938775510204081e-05,
      "loss": 1.9866,
      "step": 3928
    },
    {
      "epoch": 1.369,
      "grad_norm": 3.647921562194824,
      "learning_rate": 1.092857142857143e-05,
      "loss": 0.4283,
      "step": 3929
    },
    {
      "epoch": 1.3692,
      "grad_norm": 5.62546968460083,
      "learning_rate": 1.0918367346938776e-05,
      "loss": 0.7953,
      "step": 3930
    },
    {
      "epoch": 1.3694,
      "grad_norm": 3.3547279834747314,
      "learning_rate": 1.0908163265306124e-05,
      "loss": 0.5298,
      "step": 3931
    },
    {
      "epoch": 1.3696,
      "grad_norm": 3.8890135288238525,
      "learning_rate": 1.089795918367347e-05,
      "loss": 0.482,
      "step": 3932
    },
    {
      "epoch": 1.3698000000000001,
      "grad_norm": 3.9916648864746094,
      "learning_rate": 1.0887755102040817e-05,
      "loss": 0.6226,
      "step": 3933
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.355892181396484,
      "learning_rate": 1.0877551020408164e-05,
      "loss": 0.8334,
      "step": 3934
    },
    {
      "epoch": 1.3702,
      "grad_norm": 7.691184997558594,
      "learning_rate": 1.086734693877551e-05,
      "loss": 2.0341,
      "step": 3935
    },
    {
      "epoch": 1.3704,
      "grad_norm": 3.757251501083374,
      "learning_rate": 1.0857142857142858e-05,
      "loss": 0.6091,
      "step": 3936
    },
    {
      "epoch": 1.3706,
      "grad_norm": 5.115704536437988,
      "learning_rate": 1.0846938775510204e-05,
      "loss": 0.7629,
      "step": 3937
    },
    {
      "epoch": 1.3708,
      "grad_norm": 3.836716413497925,
      "learning_rate": 1.0836734693877551e-05,
      "loss": 0.7387,
      "step": 3938
    },
    {
      "epoch": 1.371,
      "grad_norm": 5.407971382141113,
      "learning_rate": 1.0826530612244899e-05,
      "loss": 0.7584,
      "step": 3939
    },
    {
      "epoch": 1.3712,
      "grad_norm": 3.6939923763275146,
      "learning_rate": 1.0816326530612246e-05,
      "loss": 0.6111,
      "step": 3940
    },
    {
      "epoch": 1.3714,
      "grad_norm": 5.296103477478027,
      "learning_rate": 1.0806122448979593e-05,
      "loss": 0.7072,
      "step": 3941
    },
    {
      "epoch": 1.3716,
      "grad_norm": 9.211636543273926,
      "learning_rate": 1.079591836734694e-05,
      "loss": 2.5747,
      "step": 3942
    },
    {
      "epoch": 1.3718,
      "grad_norm": 5.751867771148682,
      "learning_rate": 1.0785714285714287e-05,
      "loss": 0.8185,
      "step": 3943
    },
    {
      "epoch": 1.3719999999999999,
      "grad_norm": 4.8876237869262695,
      "learning_rate": 1.0775510204081633e-05,
      "loss": 0.9352,
      "step": 3944
    },
    {
      "epoch": 1.3721999999999999,
      "grad_norm": 9.685640335083008,
      "learning_rate": 1.076530612244898e-05,
      "loss": 2.645,
      "step": 3945
    },
    {
      "epoch": 1.3724,
      "grad_norm": 4.994866371154785,
      "learning_rate": 1.0755102040816328e-05,
      "loss": 0.6653,
      "step": 3946
    },
    {
      "epoch": 1.3726,
      "grad_norm": 4.2619805335998535,
      "learning_rate": 1.0744897959183673e-05,
      "loss": 0.7159,
      "step": 3947
    },
    {
      "epoch": 1.3728,
      "grad_norm": 5.770644664764404,
      "learning_rate": 1.073469387755102e-05,
      "loss": 0.8857,
      "step": 3948
    },
    {
      "epoch": 1.373,
      "grad_norm": 5.596279621124268,
      "learning_rate": 1.0724489795918367e-05,
      "loss": 0.7091,
      "step": 3949
    },
    {
      "epoch": 1.3732,
      "grad_norm": 3.7536487579345703,
      "learning_rate": 1.0714285714285714e-05,
      "loss": 0.4113,
      "step": 3950
    },
    {
      "epoch": 1.3734,
      "grad_norm": 4.525453090667725,
      "learning_rate": 1.0704081632653063e-05,
      "loss": 0.6371,
      "step": 3951
    },
    {
      "epoch": 1.3736,
      "grad_norm": 3.5729410648345947,
      "learning_rate": 1.0693877551020409e-05,
      "loss": 0.6469,
      "step": 3952
    },
    {
      "epoch": 1.3738000000000001,
      "grad_norm": 6.834310531616211,
      "learning_rate": 1.0683673469387757e-05,
      "loss": 1.0561,
      "step": 3953
    },
    {
      "epoch": 1.374,
      "grad_norm": 9.00754451751709,
      "learning_rate": 1.0673469387755102e-05,
      "loss": 2.6829,
      "step": 3954
    },
    {
      "epoch": 1.3742,
      "grad_norm": 4.654486179351807,
      "learning_rate": 1.066326530612245e-05,
      "loss": 0.7045,
      "step": 3955
    },
    {
      "epoch": 1.3744,
      "grad_norm": 3.5815670490264893,
      "learning_rate": 1.0653061224489797e-05,
      "loss": 0.6919,
      "step": 3956
    },
    {
      "epoch": 1.3746,
      "grad_norm": 3.902749538421631,
      "learning_rate": 1.0642857142857143e-05,
      "loss": 0.4963,
      "step": 3957
    },
    {
      "epoch": 1.3748,
      "grad_norm": 3.7020041942596436,
      "learning_rate": 1.063265306122449e-05,
      "loss": 0.6277,
      "step": 3958
    },
    {
      "epoch": 1.375,
      "grad_norm": 5.73671293258667,
      "learning_rate": 1.0622448979591836e-05,
      "loss": 0.6776,
      "step": 3959
    },
    {
      "epoch": 1.3752,
      "grad_norm": 5.688353538513184,
      "learning_rate": 1.0612244897959184e-05,
      "loss": 0.7026,
      "step": 3960
    },
    {
      "epoch": 1.3754,
      "grad_norm": 7.400761127471924,
      "learning_rate": 1.0602040816326531e-05,
      "loss": 1.2377,
      "step": 3961
    },
    {
      "epoch": 1.3756,
      "grad_norm": 11.227447509765625,
      "learning_rate": 1.0591836734693877e-05,
      "loss": 2.7975,
      "step": 3962
    },
    {
      "epoch": 1.3758,
      "grad_norm": 3.6369383335113525,
      "learning_rate": 1.0581632653061226e-05,
      "loss": 0.6545,
      "step": 3963
    },
    {
      "epoch": 1.376,
      "grad_norm": 3.8340251445770264,
      "learning_rate": 1.0571428571428572e-05,
      "loss": 0.7249,
      "step": 3964
    },
    {
      "epoch": 1.3761999999999999,
      "grad_norm": 3.273732900619507,
      "learning_rate": 1.056122448979592e-05,
      "loss": 0.6021,
      "step": 3965
    },
    {
      "epoch": 1.3764,
      "grad_norm": 5.729831695556641,
      "learning_rate": 1.0551020408163265e-05,
      "loss": 0.7489,
      "step": 3966
    },
    {
      "epoch": 1.3766,
      "grad_norm": 9.157611846923828,
      "learning_rate": 1.0540816326530613e-05,
      "loss": 2.1601,
      "step": 3967
    },
    {
      "epoch": 1.3768,
      "grad_norm": 5.697342395782471,
      "learning_rate": 1.053061224489796e-05,
      "loss": 0.8413,
      "step": 3968
    },
    {
      "epoch": 1.377,
      "grad_norm": 5.372077941894531,
      "learning_rate": 1.0520408163265306e-05,
      "loss": 0.6816,
      "step": 3969
    },
    {
      "epoch": 1.3772,
      "grad_norm": 4.570939064025879,
      "learning_rate": 1.0510204081632654e-05,
      "loss": 0.5851,
      "step": 3970
    },
    {
      "epoch": 1.3774,
      "grad_norm": 5.907211780548096,
      "learning_rate": 1.05e-05,
      "loss": 0.8755,
      "step": 3971
    },
    {
      "epoch": 1.3776,
      "grad_norm": 11.042017936706543,
      "learning_rate": 1.0489795918367347e-05,
      "loss": 2.9378,
      "step": 3972
    },
    {
      "epoch": 1.3778000000000001,
      "grad_norm": 10.968971252441406,
      "learning_rate": 1.0479591836734694e-05,
      "loss": 2.9056,
      "step": 3973
    },
    {
      "epoch": 1.3780000000000001,
      "grad_norm": 10.674814224243164,
      "learning_rate": 1.0469387755102042e-05,
      "loss": 2.6566,
      "step": 3974
    },
    {
      "epoch": 1.3782,
      "grad_norm": 6.546092510223389,
      "learning_rate": 1.045918367346939e-05,
      "loss": 0.9165,
      "step": 3975
    },
    {
      "epoch": 1.3784,
      "grad_norm": 12.73819351196289,
      "learning_rate": 1.0448979591836735e-05,
      "loss": 3.0223,
      "step": 3976
    },
    {
      "epoch": 1.3786,
      "grad_norm": 10.385857582092285,
      "learning_rate": 1.0438775510204083e-05,
      "loss": 2.549,
      "step": 3977
    },
    {
      "epoch": 1.3788,
      "grad_norm": 3.7381415367126465,
      "learning_rate": 1.0428571428571428e-05,
      "loss": 0.7069,
      "step": 3978
    },
    {
      "epoch": 1.379,
      "grad_norm": 5.126061916351318,
      "learning_rate": 1.0418367346938776e-05,
      "loss": 0.6788,
      "step": 3979
    },
    {
      "epoch": 1.3792,
      "grad_norm": 5.441112995147705,
      "learning_rate": 1.0408163265306123e-05,
      "loss": 0.7679,
      "step": 3980
    },
    {
      "epoch": 1.3794,
      "grad_norm": 5.262506008148193,
      "learning_rate": 1.039795918367347e-05,
      "loss": 0.7664,
      "step": 3981
    },
    {
      "epoch": 1.3796,
      "grad_norm": 3.900252103805542,
      "learning_rate": 1.0387755102040817e-05,
      "loss": 0.6374,
      "step": 3982
    },
    {
      "epoch": 1.3798,
      "grad_norm": 5.284982681274414,
      "learning_rate": 1.0377551020408162e-05,
      "loss": 0.7284,
      "step": 3983
    },
    {
      "epoch": 1.38,
      "grad_norm": 6.346417427062988,
      "learning_rate": 1.036734693877551e-05,
      "loss": 0.9852,
      "step": 3984
    },
    {
      "epoch": 1.3801999999999999,
      "grad_norm": 5.45538854598999,
      "learning_rate": 1.0357142857142859e-05,
      "loss": 0.8302,
      "step": 3985
    },
    {
      "epoch": 1.3804,
      "grad_norm": 5.084419250488281,
      "learning_rate": 1.0346938775510205e-05,
      "loss": 0.647,
      "step": 3986
    },
    {
      "epoch": 1.3806,
      "grad_norm": 6.146941661834717,
      "learning_rate": 1.0336734693877552e-05,
      "loss": 0.9438,
      "step": 3987
    },
    {
      "epoch": 1.3808,
      "grad_norm": 12.002094268798828,
      "learning_rate": 1.0326530612244898e-05,
      "loss": 2.7541,
      "step": 3988
    },
    {
      "epoch": 1.381,
      "grad_norm": 3.95780611038208,
      "learning_rate": 1.0316326530612246e-05,
      "loss": 0.5239,
      "step": 3989
    },
    {
      "epoch": 1.3812,
      "grad_norm": 3.9181201457977295,
      "learning_rate": 1.0306122448979593e-05,
      "loss": 0.4687,
      "step": 3990
    },
    {
      "epoch": 1.3814,
      "grad_norm": 5.811712265014648,
      "learning_rate": 1.0295918367346939e-05,
      "loss": 0.9536,
      "step": 3991
    },
    {
      "epoch": 1.3816,
      "grad_norm": 4.9396562576293945,
      "learning_rate": 1.0285714285714286e-05,
      "loss": 0.6593,
      "step": 3992
    },
    {
      "epoch": 1.3818,
      "grad_norm": 3.514286756515503,
      "learning_rate": 1.0275510204081632e-05,
      "loss": 0.7083,
      "step": 3993
    },
    {
      "epoch": 1.3820000000000001,
      "grad_norm": 3.364076614379883,
      "learning_rate": 1.026530612244898e-05,
      "loss": 0.5974,
      "step": 3994
    },
    {
      "epoch": 1.3822,
      "grad_norm": 3.648085832595825,
      "learning_rate": 1.0255102040816327e-05,
      "loss": 0.6451,
      "step": 3995
    },
    {
      "epoch": 1.3824,
      "grad_norm": 4.9181694984436035,
      "learning_rate": 1.0244897959183673e-05,
      "loss": 0.8782,
      "step": 3996
    },
    {
      "epoch": 1.3826,
      "grad_norm": 4.802428722381592,
      "learning_rate": 1.0234693877551022e-05,
      "loss": 0.734,
      "step": 3997
    },
    {
      "epoch": 1.3828,
      "grad_norm": 3.6626720428466797,
      "learning_rate": 1.0224489795918368e-05,
      "loss": 0.7175,
      "step": 3998
    },
    {
      "epoch": 1.383,
      "grad_norm": 4.834207534790039,
      "learning_rate": 1.0214285714285715e-05,
      "loss": 0.6871,
      "step": 3999
    },
    {
      "epoch": 1.3832,
      "grad_norm": 3.8979711532592773,
      "learning_rate": 1.0204081632653061e-05,
      "loss": 0.5119,
      "step": 4000
    },
    {
      "epoch": 1.3834,
      "grad_norm": 3.582329511642456,
      "learning_rate": 1.0193877551020409e-05,
      "loss": 0.3988,
      "step": 4001
    },
    {
      "epoch": 1.3836,
      "grad_norm": 3.8930623531341553,
      "learning_rate": 1.0183673469387756e-05,
      "loss": 0.6107,
      "step": 4002
    },
    {
      "epoch": 1.3838,
      "grad_norm": 5.010544300079346,
      "learning_rate": 1.0173469387755102e-05,
      "loss": 0.732,
      "step": 4003
    },
    {
      "epoch": 1.384,
      "grad_norm": 3.3792223930358887,
      "learning_rate": 1.016326530612245e-05,
      "loss": 0.5825,
      "step": 4004
    },
    {
      "epoch": 1.3841999999999999,
      "grad_norm": 6.155489444732666,
      "learning_rate": 1.0153061224489795e-05,
      "loss": 0.8847,
      "step": 4005
    },
    {
      "epoch": 1.3844,
      "grad_norm": 9.239078521728516,
      "learning_rate": 1.0142857142857143e-05,
      "loss": 2.9732,
      "step": 4006
    },
    {
      "epoch": 1.3846,
      "grad_norm": 5.7950358390808105,
      "learning_rate": 1.013265306122449e-05,
      "loss": 0.8344,
      "step": 4007
    },
    {
      "epoch": 1.3848,
      "grad_norm": 4.781053066253662,
      "learning_rate": 1.0122448979591838e-05,
      "loss": 0.6476,
      "step": 4008
    },
    {
      "epoch": 1.385,
      "grad_norm": 6.565587997436523,
      "learning_rate": 1.0112244897959185e-05,
      "loss": 1.0583,
      "step": 4009
    },
    {
      "epoch": 1.3852,
      "grad_norm": 9.15428638458252,
      "learning_rate": 1.0102040816326531e-05,
      "loss": 2.5636,
      "step": 4010
    },
    {
      "epoch": 1.3854,
      "grad_norm": 5.543002128601074,
      "learning_rate": 1.0091836734693879e-05,
      "loss": 0.8628,
      "step": 4011
    },
    {
      "epoch": 1.3856,
      "grad_norm": 5.14238977432251,
      "learning_rate": 1.0081632653061224e-05,
      "loss": 0.6357,
      "step": 4012
    },
    {
      "epoch": 1.3858,
      "grad_norm": 3.532233715057373,
      "learning_rate": 1.0071428571428572e-05,
      "loss": 0.569,
      "step": 4013
    },
    {
      "epoch": 1.3860000000000001,
      "grad_norm": 3.352722644805908,
      "learning_rate": 1.006122448979592e-05,
      "loss": 0.3895,
      "step": 4014
    },
    {
      "epoch": 1.3862,
      "grad_norm": 3.551661729812622,
      "learning_rate": 1.0051020408163265e-05,
      "loss": 0.6503,
      "step": 4015
    },
    {
      "epoch": 1.3864,
      "grad_norm": 5.485884666442871,
      "learning_rate": 1.0040816326530613e-05,
      "loss": 0.8738,
      "step": 4016
    },
    {
      "epoch": 1.3866,
      "grad_norm": 7.892735004425049,
      "learning_rate": 1.0030612244897958e-05,
      "loss": 1.035,
      "step": 4017
    },
    {
      "epoch": 1.3868,
      "grad_norm": 9.256954193115234,
      "learning_rate": 1.0020408163265306e-05,
      "loss": 2.4708,
      "step": 4018
    },
    {
      "epoch": 1.387,
      "grad_norm": 4.731393337249756,
      "learning_rate": 1.0010204081632655e-05,
      "loss": 0.8429,
      "step": 4019
    },
    {
      "epoch": 1.3872,
      "grad_norm": 9.638980865478516,
      "learning_rate": 1e-05,
      "loss": 2.3211,
      "step": 4020
    },
    {
      "epoch": 1.3874,
      "grad_norm": 5.631479740142822,
      "learning_rate": 9.989795918367348e-06,
      "loss": 0.699,
      "step": 4021
    },
    {
      "epoch": 1.3876,
      "grad_norm": 6.148746490478516,
      "learning_rate": 9.979591836734694e-06,
      "loss": 0.7217,
      "step": 4022
    },
    {
      "epoch": 1.3878,
      "grad_norm": 6.269064903259277,
      "learning_rate": 9.969387755102042e-06,
      "loss": 0.7055,
      "step": 4023
    },
    {
      "epoch": 1.388,
      "grad_norm": 5.382347106933594,
      "learning_rate": 9.959183673469389e-06,
      "loss": 0.7104,
      "step": 4024
    },
    {
      "epoch": 1.3881999999999999,
      "grad_norm": 5.656306743621826,
      "learning_rate": 9.948979591836735e-06,
      "loss": 0.836,
      "step": 4025
    },
    {
      "epoch": 1.3884,
      "grad_norm": 5.1849870681762695,
      "learning_rate": 9.938775510204082e-06,
      "loss": 0.7665,
      "step": 4026
    },
    {
      "epoch": 1.3886,
      "grad_norm": 5.041769981384277,
      "learning_rate": 9.928571428571428e-06,
      "loss": 0.7329,
      "step": 4027
    },
    {
      "epoch": 1.3888,
      "grad_norm": 3.5364441871643066,
      "learning_rate": 9.918367346938776e-06,
      "loss": 0.4226,
      "step": 4028
    },
    {
      "epoch": 1.389,
      "grad_norm": 3.902554750442505,
      "learning_rate": 9.908163265306123e-06,
      "loss": 0.6374,
      "step": 4029
    },
    {
      "epoch": 1.3892,
      "grad_norm": 4.9168829917907715,
      "learning_rate": 9.897959183673469e-06,
      "loss": 0.7221,
      "step": 4030
    },
    {
      "epoch": 1.3894,
      "grad_norm": 5.48930025100708,
      "learning_rate": 9.887755102040818e-06,
      "loss": 0.6971,
      "step": 4031
    },
    {
      "epoch": 1.3896,
      "grad_norm": 3.7762134075164795,
      "learning_rate": 9.877551020408164e-06,
      "loss": 0.4239,
      "step": 4032
    },
    {
      "epoch": 1.3898,
      "grad_norm": 6.708253383636475,
      "learning_rate": 9.867346938775511e-06,
      "loss": 0.9419,
      "step": 4033
    },
    {
      "epoch": 1.3900000000000001,
      "grad_norm": 11.21085262298584,
      "learning_rate": 9.857142857142857e-06,
      "loss": 2.8889,
      "step": 4034
    },
    {
      "epoch": 1.3902,
      "grad_norm": 9.258325576782227,
      "learning_rate": 9.846938775510205e-06,
      "loss": 2.289,
      "step": 4035
    },
    {
      "epoch": 1.3904,
      "grad_norm": 3.815279722213745,
      "learning_rate": 9.836734693877552e-06,
      "loss": 0.5744,
      "step": 4036
    },
    {
      "epoch": 1.3906,
      "grad_norm": 5.737252712249756,
      "learning_rate": 9.826530612244898e-06,
      "loss": 0.7616,
      "step": 4037
    },
    {
      "epoch": 1.3908,
      "grad_norm": 3.4769980907440186,
      "learning_rate": 9.816326530612245e-06,
      "loss": 0.4377,
      "step": 4038
    },
    {
      "epoch": 1.391,
      "grad_norm": 3.816934823989868,
      "learning_rate": 9.806122448979591e-06,
      "loss": 0.6121,
      "step": 4039
    },
    {
      "epoch": 1.3912,
      "grad_norm": 6.097662448883057,
      "learning_rate": 9.795918367346939e-06,
      "loss": 1.1407,
      "step": 4040
    },
    {
      "epoch": 1.3914,
      "grad_norm": 5.40137243270874,
      "learning_rate": 9.785714285714286e-06,
      "loss": 0.8819,
      "step": 4041
    },
    {
      "epoch": 1.3916,
      "grad_norm": 5.388091564178467,
      "learning_rate": 9.775510204081634e-06,
      "loss": 0.8149,
      "step": 4042
    },
    {
      "epoch": 1.3918,
      "grad_norm": 4.445162296295166,
      "learning_rate": 9.765306122448981e-06,
      "loss": 0.8164,
      "step": 4043
    },
    {
      "epoch": 1.392,
      "grad_norm": 3.53629469871521,
      "learning_rate": 9.755102040816327e-06,
      "loss": 0.6779,
      "step": 4044
    },
    {
      "epoch": 1.3921999999999999,
      "grad_norm": 5.086369514465332,
      "learning_rate": 9.744897959183674e-06,
      "loss": 0.7485,
      "step": 4045
    },
    {
      "epoch": 1.3924,
      "grad_norm": 3.230525493621826,
      "learning_rate": 9.73469387755102e-06,
      "loss": 0.6028,
      "step": 4046
    },
    {
      "epoch": 1.3926,
      "grad_norm": 6.732891082763672,
      "learning_rate": 9.724489795918368e-06,
      "loss": 1.0818,
      "step": 4047
    },
    {
      "epoch": 1.3928,
      "grad_norm": 9.322139739990234,
      "learning_rate": 9.714285714285715e-06,
      "loss": 2.5275,
      "step": 4048
    },
    {
      "epoch": 1.393,
      "grad_norm": 5.124191761016846,
      "learning_rate": 9.704081632653061e-06,
      "loss": 0.6462,
      "step": 4049
    },
    {
      "epoch": 1.3932,
      "grad_norm": 3.3168301582336426,
      "learning_rate": 9.693877551020408e-06,
      "loss": 0.5227,
      "step": 4050
    },
    {
      "epoch": 1.3934,
      "grad_norm": 4.323263168334961,
      "learning_rate": 9.683673469387756e-06,
      "loss": 0.4867,
      "step": 4051
    },
    {
      "epoch": 1.3936,
      "grad_norm": 5.344505786895752,
      "learning_rate": 9.673469387755102e-06,
      "loss": 0.797,
      "step": 4052
    },
    {
      "epoch": 1.3938,
      "grad_norm": 4.83172607421875,
      "learning_rate": 9.663265306122451e-06,
      "loss": 0.6209,
      "step": 4053
    },
    {
      "epoch": 1.3940000000000001,
      "grad_norm": 5.8424391746521,
      "learning_rate": 9.653061224489797e-06,
      "loss": 0.8825,
      "step": 4054
    },
    {
      "epoch": 1.3942,
      "grad_norm": 11.91540813446045,
      "learning_rate": 9.642857142857144e-06,
      "loss": 2.654,
      "step": 4055
    },
    {
      "epoch": 1.3944,
      "grad_norm": 3.9795920848846436,
      "learning_rate": 9.63265306122449e-06,
      "loss": 0.538,
      "step": 4056
    },
    {
      "epoch": 1.3946,
      "grad_norm": 4.208872318267822,
      "learning_rate": 9.622448979591837e-06,
      "loss": 0.4365,
      "step": 4057
    },
    {
      "epoch": 1.3948,
      "grad_norm": 6.616360664367676,
      "learning_rate": 9.612244897959185e-06,
      "loss": 1.0681,
      "step": 4058
    },
    {
      "epoch": 1.395,
      "grad_norm": 6.30941915512085,
      "learning_rate": 9.60204081632653e-06,
      "loss": 0.8877,
      "step": 4059
    },
    {
      "epoch": 1.3952,
      "grad_norm": 5.092617511749268,
      "learning_rate": 9.591836734693878e-06,
      "loss": 0.8003,
      "step": 4060
    },
    {
      "epoch": 1.3954,
      "grad_norm": 3.4397690296173096,
      "learning_rate": 9.581632653061224e-06,
      "loss": 0.6257,
      "step": 4061
    },
    {
      "epoch": 1.3956,
      "grad_norm": 4.654313564300537,
      "learning_rate": 9.571428571428572e-06,
      "loss": 0.7335,
      "step": 4062
    },
    {
      "epoch": 1.3958,
      "grad_norm": 4.091364860534668,
      "learning_rate": 9.561224489795919e-06,
      "loss": 0.9497,
      "step": 4063
    },
    {
      "epoch": 1.396,
      "grad_norm": 3.311901092529297,
      "learning_rate": 9.551020408163265e-06,
      "loss": 0.6365,
      "step": 4064
    },
    {
      "epoch": 1.3961999999999999,
      "grad_norm": 3.5903725624084473,
      "learning_rate": 9.540816326530614e-06,
      "loss": 0.4147,
      "step": 4065
    },
    {
      "epoch": 1.3963999999999999,
      "grad_norm": 4.654419898986816,
      "learning_rate": 9.53061224489796e-06,
      "loss": 0.6337,
      "step": 4066
    },
    {
      "epoch": 1.3966,
      "grad_norm": 3.7038626670837402,
      "learning_rate": 9.520408163265307e-06,
      "loss": 0.4614,
      "step": 4067
    },
    {
      "epoch": 1.3968,
      "grad_norm": 6.082508087158203,
      "learning_rate": 9.510204081632653e-06,
      "loss": 1.1482,
      "step": 4068
    },
    {
      "epoch": 1.397,
      "grad_norm": 9.412351608276367,
      "learning_rate": 9.5e-06,
      "loss": 2.536,
      "step": 4069
    },
    {
      "epoch": 1.3972,
      "grad_norm": 3.590003728866577,
      "learning_rate": 9.489795918367348e-06,
      "loss": 0.8897,
      "step": 4070
    },
    {
      "epoch": 1.3974,
      "grad_norm": 5.914702892303467,
      "learning_rate": 9.479591836734694e-06,
      "loss": 1.4219,
      "step": 4071
    },
    {
      "epoch": 1.3976,
      "grad_norm": 13.116612434387207,
      "learning_rate": 9.469387755102041e-06,
      "loss": 3.6732,
      "step": 4072
    },
    {
      "epoch": 1.3978,
      "grad_norm": 11.515417098999023,
      "learning_rate": 9.459183673469387e-06,
      "loss": 2.5414,
      "step": 4073
    },
    {
      "epoch": 1.3980000000000001,
      "grad_norm": 4.8623223304748535,
      "learning_rate": 9.448979591836735e-06,
      "loss": 0.7811,
      "step": 4074
    },
    {
      "epoch": 1.3982,
      "grad_norm": 5.086510181427002,
      "learning_rate": 9.438775510204082e-06,
      "loss": 0.7708,
      "step": 4075
    },
    {
      "epoch": 1.3984,
      "grad_norm": 3.5381362438201904,
      "learning_rate": 9.42857142857143e-06,
      "loss": 0.7936,
      "step": 4076
    },
    {
      "epoch": 1.3986,
      "grad_norm": 4.599300861358643,
      "learning_rate": 9.418367346938777e-06,
      "loss": 0.7034,
      "step": 4077
    },
    {
      "epoch": 1.3988,
      "grad_norm": 5.645698070526123,
      "learning_rate": 9.408163265306123e-06,
      "loss": 0.8146,
      "step": 4078
    },
    {
      "epoch": 1.399,
      "grad_norm": 3.565887212753296,
      "learning_rate": 9.39795918367347e-06,
      "loss": 0.6053,
      "step": 4079
    },
    {
      "epoch": 1.3992,
      "grad_norm": 4.043582916259766,
      "learning_rate": 9.387755102040816e-06,
      "loss": 0.5256,
      "step": 4080
    },
    {
      "epoch": 1.3994,
      "grad_norm": 3.9750559329986572,
      "learning_rate": 9.377551020408164e-06,
      "loss": 0.4568,
      "step": 4081
    },
    {
      "epoch": 1.3996,
      "grad_norm": 4.978681564331055,
      "learning_rate": 9.367346938775511e-06,
      "loss": 0.6702,
      "step": 4082
    },
    {
      "epoch": 1.3998,
      "grad_norm": 3.766176223754883,
      "learning_rate": 9.357142857142857e-06,
      "loss": 0.6297,
      "step": 4083
    },
    {
      "epoch": 1.4,
      "grad_norm": 6.139642238616943,
      "learning_rate": 9.346938775510204e-06,
      "loss": 0.8615,
      "step": 4084
    },
    {
      "epoch": 1.4002,
      "grad_norm": 5.712095737457275,
      "learning_rate": 9.336734693877552e-06,
      "loss": 0.7664,
      "step": 4085
    },
    {
      "epoch": 1.4003999999999999,
      "grad_norm": 4.974853038787842,
      "learning_rate": 9.326530612244898e-06,
      "loss": 0.7525,
      "step": 4086
    },
    {
      "epoch": 1.4006,
      "grad_norm": 3.8871517181396484,
      "learning_rate": 9.316326530612247e-06,
      "loss": 0.7542,
      "step": 4087
    },
    {
      "epoch": 1.4008,
      "grad_norm": 3.846158027648926,
      "learning_rate": 9.306122448979593e-06,
      "loss": 0.6825,
      "step": 4088
    },
    {
      "epoch": 1.401,
      "grad_norm": 5.497731685638428,
      "learning_rate": 9.29591836734694e-06,
      "loss": 0.7653,
      "step": 4089
    },
    {
      "epoch": 1.4012,
      "grad_norm": 4.662613391876221,
      "learning_rate": 9.285714285714286e-06,
      "loss": 0.7129,
      "step": 4090
    },
    {
      "epoch": 1.4014,
      "grad_norm": 3.373971462249756,
      "learning_rate": 9.275510204081633e-06,
      "loss": 0.5762,
      "step": 4091
    },
    {
      "epoch": 1.4016,
      "grad_norm": 4.859783172607422,
      "learning_rate": 9.26530612244898e-06,
      "loss": 0.7057,
      "step": 4092
    },
    {
      "epoch": 1.4018,
      "grad_norm": 6.266166687011719,
      "learning_rate": 9.255102040816327e-06,
      "loss": 0.8399,
      "step": 4093
    },
    {
      "epoch": 1.4020000000000001,
      "grad_norm": 3.591654062271118,
      "learning_rate": 9.244897959183674e-06,
      "loss": 0.6136,
      "step": 4094
    },
    {
      "epoch": 1.4022000000000001,
      "grad_norm": 3.9025590419769287,
      "learning_rate": 9.23469387755102e-06,
      "loss": 0.5225,
      "step": 4095
    },
    {
      "epoch": 1.4024,
      "grad_norm": 3.5609548091888428,
      "learning_rate": 9.224489795918367e-06,
      "loss": 0.4294,
      "step": 4096
    },
    {
      "epoch": 1.4026,
      "grad_norm": 4.027820587158203,
      "learning_rate": 9.214285714285715e-06,
      "loss": 0.8391,
      "step": 4097
    },
    {
      "epoch": 1.4028,
      "grad_norm": 3.616084337234497,
      "learning_rate": 9.20408163265306e-06,
      "loss": 0.7161,
      "step": 4098
    },
    {
      "epoch": 1.403,
      "grad_norm": 5.0737385749816895,
      "learning_rate": 9.19387755102041e-06,
      "loss": 0.7033,
      "step": 4099
    },
    {
      "epoch": 1.4032,
      "grad_norm": 3.8645384311676025,
      "learning_rate": 9.183673469387756e-06,
      "loss": 0.4397,
      "step": 4100
    },
    {
      "epoch": 1.4034,
      "grad_norm": 4.167079925537109,
      "learning_rate": 9.173469387755103e-06,
      "loss": 0.7173,
      "step": 4101
    },
    {
      "epoch": 1.4036,
      "grad_norm": 5.854258060455322,
      "learning_rate": 9.163265306122449e-06,
      "loss": 0.7312,
      "step": 4102
    },
    {
      "epoch": 1.4038,
      "grad_norm": 3.604153871536255,
      "learning_rate": 9.153061224489796e-06,
      "loss": 0.617,
      "step": 4103
    },
    {
      "epoch": 1.404,
      "grad_norm": 5.196364879608154,
      "learning_rate": 9.142857142857144e-06,
      "loss": 0.7194,
      "step": 4104
    },
    {
      "epoch": 1.4042,
      "grad_norm": 4.474340915679932,
      "learning_rate": 9.13265306122449e-06,
      "loss": 0.6409,
      "step": 4105
    },
    {
      "epoch": 1.4043999999999999,
      "grad_norm": 8.195387840270996,
      "learning_rate": 9.122448979591837e-06,
      "loss": 1.1349,
      "step": 4106
    },
    {
      "epoch": 1.4046,
      "grad_norm": 8.900938034057617,
      "learning_rate": 9.112244897959183e-06,
      "loss": 1.1237,
      "step": 4107
    },
    {
      "epoch": 1.4048,
      "grad_norm": 8.176107406616211,
      "learning_rate": 9.10204081632653e-06,
      "loss": 0.8196,
      "step": 4108
    },
    {
      "epoch": 1.405,
      "grad_norm": 6.472485542297363,
      "learning_rate": 9.091836734693878e-06,
      "loss": 0.7335,
      "step": 4109
    },
    {
      "epoch": 1.4052,
      "grad_norm": 5.660496234893799,
      "learning_rate": 9.081632653061225e-06,
      "loss": 0.7534,
      "step": 4110
    },
    {
      "epoch": 1.4054,
      "grad_norm": 5.223132133483887,
      "learning_rate": 9.071428571428573e-06,
      "loss": 0.8232,
      "step": 4111
    },
    {
      "epoch": 1.4056,
      "grad_norm": 4.634860515594482,
      "learning_rate": 9.061224489795919e-06,
      "loss": 0.6467,
      "step": 4112
    },
    {
      "epoch": 1.4058,
      "grad_norm": 7.191314220428467,
      "learning_rate": 9.051020408163266e-06,
      "loss": 0.7423,
      "step": 4113
    },
    {
      "epoch": 1.4060000000000001,
      "grad_norm": 3.7934906482696533,
      "learning_rate": 9.040816326530614e-06,
      "loss": 0.441,
      "step": 4114
    },
    {
      "epoch": 1.4062000000000001,
      "grad_norm": 5.6176910400390625,
      "learning_rate": 9.03061224489796e-06,
      "loss": 0.7425,
      "step": 4115
    },
    {
      "epoch": 1.4064,
      "grad_norm": 3.6926732063293457,
      "learning_rate": 9.020408163265307e-06,
      "loss": 0.8754,
      "step": 4116
    },
    {
      "epoch": 1.4066,
      "grad_norm": 4.297953128814697,
      "learning_rate": 9.010204081632653e-06,
      "loss": 0.6508,
      "step": 4117
    },
    {
      "epoch": 1.4068,
      "grad_norm": 2.9790539741516113,
      "learning_rate": 9e-06,
      "loss": 0.5164,
      "step": 4118
    },
    {
      "epoch": 1.407,
      "grad_norm": 3.659420967102051,
      "learning_rate": 8.989795918367348e-06,
      "loss": 0.4337,
      "step": 4119
    },
    {
      "epoch": 1.4072,
      "grad_norm": 3.8091375827789307,
      "learning_rate": 8.979591836734694e-06,
      "loss": 0.7421,
      "step": 4120
    },
    {
      "epoch": 1.4074,
      "grad_norm": 3.657268524169922,
      "learning_rate": 8.969387755102043e-06,
      "loss": 0.6517,
      "step": 4121
    },
    {
      "epoch": 1.4076,
      "grad_norm": 3.3698313236236572,
      "learning_rate": 8.959183673469388e-06,
      "loss": 0.619,
      "step": 4122
    },
    {
      "epoch": 1.4078,
      "grad_norm": 3.6699438095092773,
      "learning_rate": 8.948979591836736e-06,
      "loss": 0.4399,
      "step": 4123
    },
    {
      "epoch": 1.408,
      "grad_norm": 3.251619338989258,
      "learning_rate": 8.938775510204082e-06,
      "loss": 0.5502,
      "step": 4124
    },
    {
      "epoch": 1.4082,
      "grad_norm": 3.619903326034546,
      "learning_rate": 8.92857142857143e-06,
      "loss": 0.3735,
      "step": 4125
    },
    {
      "epoch": 1.4083999999999999,
      "grad_norm": 3.4549977779388428,
      "learning_rate": 8.918367346938777e-06,
      "loss": 0.5709,
      "step": 4126
    },
    {
      "epoch": 1.4086,
      "grad_norm": 3.7849767208099365,
      "learning_rate": 8.908163265306123e-06,
      "loss": 0.6561,
      "step": 4127
    },
    {
      "epoch": 1.4088,
      "grad_norm": 3.783862352371216,
      "learning_rate": 8.89795918367347e-06,
      "loss": 0.5563,
      "step": 4128
    },
    {
      "epoch": 1.409,
      "grad_norm": 3.595287799835205,
      "learning_rate": 8.887755102040816e-06,
      "loss": 0.9047,
      "step": 4129
    },
    {
      "epoch": 1.4092,
      "grad_norm": 3.724360466003418,
      "learning_rate": 8.877551020408163e-06,
      "loss": 0.6573,
      "step": 4130
    },
    {
      "epoch": 1.4094,
      "grad_norm": 4.294783115386963,
      "learning_rate": 8.86734693877551e-06,
      "loss": 0.6083,
      "step": 4131
    },
    {
      "epoch": 1.4096,
      "grad_norm": 3.358783006668091,
      "learning_rate": 8.857142857142857e-06,
      "loss": 0.3971,
      "step": 4132
    },
    {
      "epoch": 1.4098,
      "grad_norm": 3.7183899879455566,
      "learning_rate": 8.846938775510206e-06,
      "loss": 0.8904,
      "step": 4133
    },
    {
      "epoch": 1.41,
      "grad_norm": 3.684494972229004,
      "learning_rate": 8.836734693877552e-06,
      "loss": 0.7093,
      "step": 4134
    },
    {
      "epoch": 1.4102000000000001,
      "grad_norm": 5.470197677612305,
      "learning_rate": 8.826530612244899e-06,
      "loss": 0.7394,
      "step": 4135
    },
    {
      "epoch": 1.4104,
      "grad_norm": 3.2948615550994873,
      "learning_rate": 8.816326530612245e-06,
      "loss": 0.5397,
      "step": 4136
    },
    {
      "epoch": 1.4106,
      "grad_norm": 4.346402168273926,
      "learning_rate": 8.806122448979592e-06,
      "loss": 0.4657,
      "step": 4137
    },
    {
      "epoch": 1.4108,
      "grad_norm": 12.534156799316406,
      "learning_rate": 8.79591836734694e-06,
      "loss": 0.7853,
      "step": 4138
    },
    {
      "epoch": 1.411,
      "grad_norm": 3.222975015640259,
      "learning_rate": 8.785714285714286e-06,
      "loss": 0.5772,
      "step": 4139
    },
    {
      "epoch": 1.4112,
      "grad_norm": 3.6416501998901367,
      "learning_rate": 8.775510204081633e-06,
      "loss": 0.6124,
      "step": 4140
    },
    {
      "epoch": 1.4114,
      "grad_norm": 5.341848373413086,
      "learning_rate": 8.765306122448979e-06,
      "loss": 0.8143,
      "step": 4141
    },
    {
      "epoch": 1.4116,
      "grad_norm": 3.1770570278167725,
      "learning_rate": 8.755102040816326e-06,
      "loss": 0.5841,
      "step": 4142
    },
    {
      "epoch": 1.4118,
      "grad_norm": 3.6214945316314697,
      "learning_rate": 8.744897959183674e-06,
      "loss": 0.435,
      "step": 4143
    },
    {
      "epoch": 1.412,
      "grad_norm": 3.3380117416381836,
      "learning_rate": 8.734693877551021e-06,
      "loss": 0.5665,
      "step": 4144
    },
    {
      "epoch": 1.4122,
      "grad_norm": 4.026908874511719,
      "learning_rate": 8.724489795918369e-06,
      "loss": 0.5302,
      "step": 4145
    },
    {
      "epoch": 1.4123999999999999,
      "grad_norm": 4.105876445770264,
      "learning_rate": 8.714285714285715e-06,
      "loss": 0.4974,
      "step": 4146
    },
    {
      "epoch": 1.4126,
      "grad_norm": 4.341339111328125,
      "learning_rate": 8.704081632653062e-06,
      "loss": 0.5447,
      "step": 4147
    },
    {
      "epoch": 1.4128,
      "grad_norm": 6.510222911834717,
      "learning_rate": 8.69387755102041e-06,
      "loss": 0.4419,
      "step": 4148
    },
    {
      "epoch": 1.413,
      "grad_norm": 6.3148579597473145,
      "learning_rate": 8.683673469387755e-06,
      "loss": 1.0734,
      "step": 4149
    },
    {
      "epoch": 1.4132,
      "grad_norm": 6.665963649749756,
      "learning_rate": 8.673469387755103e-06,
      "loss": 1.1138,
      "step": 4150
    },
    {
      "epoch": 1.4134,
      "grad_norm": 5.742279052734375,
      "learning_rate": 8.663265306122449e-06,
      "loss": 0.725,
      "step": 4151
    },
    {
      "epoch": 1.4136,
      "grad_norm": 3.820310354232788,
      "learning_rate": 8.653061224489796e-06,
      "loss": 0.6811,
      "step": 4152
    },
    {
      "epoch": 1.4138,
      "grad_norm": 5.027363300323486,
      "learning_rate": 8.642857142857144e-06,
      "loss": 0.7551,
      "step": 4153
    },
    {
      "epoch": 1.414,
      "grad_norm": 4.043216228485107,
      "learning_rate": 8.63265306122449e-06,
      "loss": 0.5471,
      "step": 4154
    },
    {
      "epoch": 1.4142000000000001,
      "grad_norm": 6.13849401473999,
      "learning_rate": 8.622448979591839e-06,
      "loss": 0.7599,
      "step": 4155
    },
    {
      "epoch": 1.4144,
      "grad_norm": 4.4201812744140625,
      "learning_rate": 8.612244897959184e-06,
      "loss": 0.5476,
      "step": 4156
    },
    {
      "epoch": 1.4146,
      "grad_norm": 4.416066646575928,
      "learning_rate": 8.602040816326532e-06,
      "loss": 0.5428,
      "step": 4157
    },
    {
      "epoch": 1.4148,
      "grad_norm": 4.235730171203613,
      "learning_rate": 8.591836734693878e-06,
      "loss": 0.5485,
      "step": 4158
    },
    {
      "epoch": 1.415,
      "grad_norm": 4.274699687957764,
      "learning_rate": 8.581632653061225e-06,
      "loss": 0.5564,
      "step": 4159
    },
    {
      "epoch": 1.4152,
      "grad_norm": 4.5172224044799805,
      "learning_rate": 8.571428571428573e-06,
      "loss": 0.6145,
      "step": 4160
    },
    {
      "epoch": 1.4154,
      "grad_norm": 4.030407428741455,
      "learning_rate": 8.561224489795918e-06,
      "loss": 0.4312,
      "step": 4161
    },
    {
      "epoch": 1.4156,
      "grad_norm": 8.166952133178711,
      "learning_rate": 8.551020408163266e-06,
      "loss": 1.1238,
      "step": 4162
    },
    {
      "epoch": 1.4158,
      "grad_norm": 5.0052337646484375,
      "learning_rate": 8.540816326530612e-06,
      "loss": 1.0528,
      "step": 4163
    },
    {
      "epoch": 1.416,
      "grad_norm": 11.661174774169922,
      "learning_rate": 8.53061224489796e-06,
      "loss": 1.1472,
      "step": 4164
    },
    {
      "epoch": 1.4162,
      "grad_norm": 5.528961181640625,
      "learning_rate": 8.520408163265307e-06,
      "loss": 1.1316,
      "step": 4165
    },
    {
      "epoch": 1.4163999999999999,
      "grad_norm": 5.657470703125,
      "learning_rate": 8.510204081632652e-06,
      "loss": 1.0495,
      "step": 4166
    },
    {
      "epoch": 1.4166,
      "grad_norm": 4.580628395080566,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.6827,
      "step": 4167
    },
    {
      "epoch": 1.4168,
      "grad_norm": 7.498333930969238,
      "learning_rate": 8.489795918367347e-06,
      "loss": 1.3054,
      "step": 4168
    },
    {
      "epoch": 2.0002,
      "grad_norm": 2.02801513671875,
      "learning_rate": 8.479591836734695e-06,
      "loss": 0.1952,
      "step": 4169
    },
    {
      "epoch": 2.0004,
      "grad_norm": 11.26914119720459,
      "learning_rate": 8.46938775510204e-06,
      "loss": 2.2445,
      "step": 4170
    },
    {
      "epoch": 2.0006,
      "grad_norm": 2.6974380016326904,
      "learning_rate": 8.459183673469388e-06,
      "loss": 0.2363,
      "step": 4171
    },
    {
      "epoch": 2.0008,
      "grad_norm": 3.35848069190979,
      "learning_rate": 8.448979591836736e-06,
      "loss": 0.3612,
      "step": 4172
    },
    {
      "epoch": 2.001,
      "grad_norm": 4.7232279777526855,
      "learning_rate": 8.438775510204081e-06,
      "loss": 0.307,
      "step": 4173
    },
    {
      "epoch": 2.0012,
      "grad_norm": 0.9839540719985962,
      "learning_rate": 8.428571428571429e-06,
      "loss": 0.0254,
      "step": 4174
    },
    {
      "epoch": 2.0014,
      "grad_norm": 2.924433708190918,
      "learning_rate": 8.418367346938775e-06,
      "loss": 0.4624,
      "step": 4175
    },
    {
      "epoch": 2.0016,
      "grad_norm": 2.3263041973114014,
      "learning_rate": 8.408163265306122e-06,
      "loss": 0.2279,
      "step": 4176
    },
    {
      "epoch": 2.0018,
      "grad_norm": 4.812216281890869,
      "learning_rate": 8.39795918367347e-06,
      "loss": 1.4139,
      "step": 4177
    },
    {
      "epoch": 2.002,
      "grad_norm": 2.2195217609405518,
      "learning_rate": 8.387755102040817e-06,
      "loss": 0.1295,
      "step": 4178
    },
    {
      "epoch": 2.0022,
      "grad_norm": 2.8188059329986572,
      "learning_rate": 8.377551020408165e-06,
      "loss": 0.4121,
      "step": 4179
    },
    {
      "epoch": 2.0024,
      "grad_norm": 1.218191146850586,
      "learning_rate": 8.36734693877551e-06,
      "loss": 0.0442,
      "step": 4180
    },
    {
      "epoch": 2.0026,
      "grad_norm": 3.5840089321136475,
      "learning_rate": 8.357142857142858e-06,
      "loss": 1.4749,
      "step": 4181
    },
    {
      "epoch": 2.0028,
      "grad_norm": 2.198009967803955,
      "learning_rate": 8.346938775510205e-06,
      "loss": 0.2235,
      "step": 4182
    },
    {
      "epoch": 2.003,
      "grad_norm": 4.087214946746826,
      "learning_rate": 8.336734693877551e-06,
      "loss": 0.5768,
      "step": 4183
    },
    {
      "epoch": 2.0032,
      "grad_norm": 6.406209945678711,
      "learning_rate": 8.326530612244899e-06,
      "loss": 1.4211,
      "step": 4184
    },
    {
      "epoch": 2.0034,
      "grad_norm": 7.409274578094482,
      "learning_rate": 8.316326530612245e-06,
      "loss": 1.9132,
      "step": 4185
    },
    {
      "epoch": 2.0036,
      "grad_norm": 9.410537719726562,
      "learning_rate": 8.306122448979592e-06,
      "loss": 3.1996,
      "step": 4186
    },
    {
      "epoch": 2.0038,
      "grad_norm": 1.7160249948501587,
      "learning_rate": 8.29591836734694e-06,
      "loss": 0.1781,
      "step": 4187
    },
    {
      "epoch": 2.004,
      "grad_norm": 8.489348411560059,
      "learning_rate": 8.285714285714285e-06,
      "loss": 2.0185,
      "step": 4188
    },
    {
      "epoch": 2.0042,
      "grad_norm": 2.1543900966644287,
      "learning_rate": 8.275510204081633e-06,
      "loss": 0.1903,
      "step": 4189
    },
    {
      "epoch": 2.0044,
      "grad_norm": 18.228240966796875,
      "learning_rate": 8.26530612244898e-06,
      "loss": 4.8913,
      "step": 4190
    },
    {
      "epoch": 2.0046,
      "grad_norm": 1.31488037109375,
      "learning_rate": 8.255102040816328e-06,
      "loss": 0.0419,
      "step": 4191
    },
    {
      "epoch": 2.0048,
      "grad_norm": 2.7342636585235596,
      "learning_rate": 8.244897959183674e-06,
      "loss": 0.2603,
      "step": 4192
    },
    {
      "epoch": 2.005,
      "grad_norm": 3.703019142150879,
      "learning_rate": 8.234693877551021e-06,
      "loss": 0.2315,
      "step": 4193
    },
    {
      "epoch": 2.0052,
      "grad_norm": 11.856758117675781,
      "learning_rate": 8.224489795918369e-06,
      "loss": 2.0342,
      "step": 4194
    },
    {
      "epoch": 2.0054,
      "grad_norm": 3.2413318157196045,
      "learning_rate": 8.214285714285714e-06,
      "loss": 0.2025,
      "step": 4195
    },
    {
      "epoch": 2.0056,
      "grad_norm": 21.850154876708984,
      "learning_rate": 8.204081632653062e-06,
      "loss": 5.154,
      "step": 4196
    },
    {
      "epoch": 2.0058,
      "grad_norm": 3.440854549407959,
      "learning_rate": 8.193877551020408e-06,
      "loss": 0.3227,
      "step": 4197
    },
    {
      "epoch": 2.006,
      "grad_norm": 3.0768935680389404,
      "learning_rate": 8.183673469387755e-06,
      "loss": 0.3272,
      "step": 4198
    },
    {
      "epoch": 2.0062,
      "grad_norm": 3.454883575439453,
      "learning_rate": 8.173469387755103e-06,
      "loss": 0.4007,
      "step": 4199
    },
    {
      "epoch": 2.0064,
      "grad_norm": 1.5237864255905151,
      "learning_rate": 8.163265306122448e-06,
      "loss": 0.0421,
      "step": 4200
    },
    {
      "epoch": 2.0066,
      "grad_norm": 7.132639408111572,
      "learning_rate": 8.153061224489798e-06,
      "loss": 1.1057,
      "step": 4201
    },
    {
      "epoch": 2.0068,
      "grad_norm": 3.31197452545166,
      "learning_rate": 8.142857142857143e-06,
      "loss": 0.2151,
      "step": 4202
    },
    {
      "epoch": 2.007,
      "grad_norm": 3.2910704612731934,
      "learning_rate": 8.13265306122449e-06,
      "loss": 0.4873,
      "step": 4203
    },
    {
      "epoch": 2.0072,
      "grad_norm": 5.533116340637207,
      "learning_rate": 8.122448979591837e-06,
      "loss": 0.6616,
      "step": 4204
    },
    {
      "epoch": 2.0074,
      "grad_norm": 5.0104265213012695,
      "learning_rate": 8.112244897959184e-06,
      "loss": 1.5563,
      "step": 4205
    },
    {
      "epoch": 2.0076,
      "grad_norm": 2.2984113693237305,
      "learning_rate": 8.102040816326532e-06,
      "loss": 0.565,
      "step": 4206
    },
    {
      "epoch": 2.0078,
      "grad_norm": 5.124585151672363,
      "learning_rate": 8.091836734693877e-06,
      "loss": 0.9707,
      "step": 4207
    },
    {
      "epoch": 2.008,
      "grad_norm": 3.2739717960357666,
      "learning_rate": 8.081632653061225e-06,
      "loss": 1.4951,
      "step": 4208
    },
    {
      "epoch": 2.0082,
      "grad_norm": 9.073833465576172,
      "learning_rate": 8.07142857142857e-06,
      "loss": 2.1831,
      "step": 4209
    },
    {
      "epoch": 2.0084,
      "grad_norm": 3.1194498538970947,
      "learning_rate": 8.061224489795918e-06,
      "loss": 0.3266,
      "step": 4210
    },
    {
      "epoch": 2.0086,
      "grad_norm": 3.2492129802703857,
      "learning_rate": 8.051020408163266e-06,
      "loss": 0.2966,
      "step": 4211
    },
    {
      "epoch": 2.0088,
      "grad_norm": 9.486139297485352,
      "learning_rate": 8.040816326530613e-06,
      "loss": 1.0458,
      "step": 4212
    },
    {
      "epoch": 2.009,
      "grad_norm": 1.5922856330871582,
      "learning_rate": 8.03061224489796e-06,
      "loss": 0.2079,
      "step": 4213
    },
    {
      "epoch": 2.0092,
      "grad_norm": 5.670705795288086,
      "learning_rate": 8.020408163265306e-06,
      "loss": 1.0761,
      "step": 4214
    },
    {
      "epoch": 2.0094,
      "grad_norm": 3.3283936977386475,
      "learning_rate": 8.010204081632654e-06,
      "loss": 0.1212,
      "step": 4215
    },
    {
      "epoch": 2.0096,
      "grad_norm": 11.98408317565918,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.9528,
      "step": 4216
    },
    {
      "epoch": 2.0098,
      "grad_norm": 3.9587674140930176,
      "learning_rate": 7.989795918367347e-06,
      "loss": 0.4188,
      "step": 4217
    },
    {
      "epoch": 2.01,
      "grad_norm": 5.890332221984863,
      "learning_rate": 7.979591836734695e-06,
      "loss": 1.4773,
      "step": 4218
    },
    {
      "epoch": 2.0102,
      "grad_norm": 4.72938346862793,
      "learning_rate": 7.96938775510204e-06,
      "loss": 0.6635,
      "step": 4219
    },
    {
      "epoch": 2.0104,
      "grad_norm": 6.216917514801025,
      "learning_rate": 7.959183673469388e-06,
      "loss": 1.1733,
      "step": 4220
    },
    {
      "epoch": 2.0106,
      "grad_norm": 2.5533437728881836,
      "learning_rate": 7.948979591836735e-06,
      "loss": 0.3137,
      "step": 4221
    },
    {
      "epoch": 2.0108,
      "grad_norm": 3.2857799530029297,
      "learning_rate": 7.938775510204081e-06,
      "loss": 1.569,
      "step": 4222
    },
    {
      "epoch": 2.011,
      "grad_norm": 2.7982468605041504,
      "learning_rate": 7.928571428571429e-06,
      "loss": 0.3094,
      "step": 4223
    },
    {
      "epoch": 2.0112,
      "grad_norm": 2.6701512336730957,
      "learning_rate": 7.918367346938776e-06,
      "loss": 0.6731,
      "step": 4224
    },
    {
      "epoch": 2.0114,
      "grad_norm": 3.838090181350708,
      "learning_rate": 7.908163265306124e-06,
      "loss": 0.5143,
      "step": 4225
    },
    {
      "epoch": 2.0116,
      "grad_norm": 7.445224285125732,
      "learning_rate": 7.89795918367347e-06,
      "loss": 1.9191,
      "step": 4226
    },
    {
      "epoch": 2.0118,
      "grad_norm": 5.421460151672363,
      "learning_rate": 7.887755102040817e-06,
      "loss": 1.8454,
      "step": 4227
    },
    {
      "epoch": 2.012,
      "grad_norm": 11.494588851928711,
      "learning_rate": 7.877551020408164e-06,
      "loss": 1.2474,
      "step": 4228
    },
    {
      "epoch": 2.0122,
      "grad_norm": 4.028449058532715,
      "learning_rate": 7.86734693877551e-06,
      "loss": 2.0761,
      "step": 4229
    },
    {
      "epoch": 2.0124,
      "grad_norm": 2.2099153995513916,
      "learning_rate": 7.857142857142858e-06,
      "loss": 0.1245,
      "step": 4230
    },
    {
      "epoch": 2.0126,
      "grad_norm": 3.8839001655578613,
      "learning_rate": 7.846938775510203e-06,
      "loss": 1.7844,
      "step": 4231
    },
    {
      "epoch": 2.0128,
      "grad_norm": 3.8286972045898438,
      "learning_rate": 7.836734693877551e-06,
      "loss": 0.6683,
      "step": 4232
    },
    {
      "epoch": 2.013,
      "grad_norm": 3.683375835418701,
      "learning_rate": 7.826530612244898e-06,
      "loss": 2.0083,
      "step": 4233
    },
    {
      "epoch": 2.0132,
      "grad_norm": 6.994391918182373,
      "learning_rate": 7.816326530612244e-06,
      "loss": 2.0595,
      "step": 4234
    },
    {
      "epoch": 2.0134,
      "grad_norm": 8.173694610595703,
      "learning_rate": 7.806122448979593e-06,
      "loss": 2.3681,
      "step": 4235
    },
    {
      "epoch": 2.0136,
      "grad_norm": 9.8168363571167,
      "learning_rate": 7.79591836734694e-06,
      "loss": 1.6911,
      "step": 4236
    },
    {
      "epoch": 2.0138,
      "grad_norm": 9.103837013244629,
      "learning_rate": 7.785714285714287e-06,
      "loss": 3.4232,
      "step": 4237
    },
    {
      "epoch": 2.014,
      "grad_norm": 9.685304641723633,
      "learning_rate": 7.775510204081632e-06,
      "loss": 3.0462,
      "step": 4238
    },
    {
      "epoch": 2.0142,
      "grad_norm": 2.858280897140503,
      "learning_rate": 7.76530612244898e-06,
      "loss": 0.4622,
      "step": 4239
    },
    {
      "epoch": 2.0144,
      "grad_norm": 7.876575946807861,
      "learning_rate": 7.755102040816327e-06,
      "loss": 2.1807,
      "step": 4240
    },
    {
      "epoch": 2.0146,
      "grad_norm": 3.4903438091278076,
      "learning_rate": 7.744897959183673e-06,
      "loss": 1.124,
      "step": 4241
    },
    {
      "epoch": 2.0148,
      "grad_norm": 6.04346227645874,
      "learning_rate": 7.73469387755102e-06,
      "loss": 1.4367,
      "step": 4242
    },
    {
      "epoch": 2.015,
      "grad_norm": 25.000646591186523,
      "learning_rate": 7.724489795918367e-06,
      "loss": 3.2368,
      "step": 4243
    },
    {
      "epoch": 2.0152,
      "grad_norm": 9.479928016662598,
      "learning_rate": 7.714285714285714e-06,
      "loss": 2.1588,
      "step": 4244
    },
    {
      "epoch": 2.0154,
      "grad_norm": 8.344285011291504,
      "learning_rate": 7.704081632653061e-06,
      "loss": 1.5787,
      "step": 4245
    },
    {
      "epoch": 2.0156,
      "grad_norm": 3.25274920463562,
      "learning_rate": 7.693877551020409e-06,
      "loss": 1.4998,
      "step": 4246
    },
    {
      "epoch": 2.0158,
      "grad_norm": 8.191744804382324,
      "learning_rate": 7.683673469387756e-06,
      "loss": 2.2567,
      "step": 4247
    },
    {
      "epoch": 2.016,
      "grad_norm": 4.326716899871826,
      "learning_rate": 7.673469387755102e-06,
      "loss": 0.8182,
      "step": 4248
    },
    {
      "epoch": 2.0162,
      "grad_norm": 3.1718268394470215,
      "learning_rate": 7.66326530612245e-06,
      "loss": 1.5348,
      "step": 4249
    },
    {
      "epoch": 2.0164,
      "grad_norm": 6.938101291656494,
      "learning_rate": 7.653061224489797e-06,
      "loss": 2.21,
      "step": 4250
    },
    {
      "epoch": 2.0166,
      "grad_norm": 5.614826679229736,
      "learning_rate": 7.642857142857143e-06,
      "loss": 1.0724,
      "step": 4251
    },
    {
      "epoch": 2.0168,
      "grad_norm": 4.613611221313477,
      "learning_rate": 7.63265306122449e-06,
      "loss": 0.8283,
      "step": 4252
    },
    {
      "epoch": 2.017,
      "grad_norm": 8.282724380493164,
      "learning_rate": 7.622448979591837e-06,
      "loss": 3.3605,
      "step": 4253
    },
    {
      "epoch": 2.0172,
      "grad_norm": 5.70613431930542,
      "learning_rate": 7.612244897959184e-06,
      "loss": 0.3506,
      "step": 4254
    },
    {
      "epoch": 2.0174,
      "grad_norm": 2.392794609069824,
      "learning_rate": 7.60204081632653e-06,
      "loss": 0.2306,
      "step": 4255
    },
    {
      "epoch": 2.0176,
      "grad_norm": 3.4948360919952393,
      "learning_rate": 7.591836734693877e-06,
      "loss": 0.2047,
      "step": 4256
    },
    {
      "epoch": 2.0178,
      "grad_norm": 4.542838096618652,
      "learning_rate": 7.5816326530612245e-06,
      "loss": 0.5617,
      "step": 4257
    },
    {
      "epoch": 2.018,
      "grad_norm": 4.445930004119873,
      "learning_rate": 7.571428571428572e-06,
      "loss": 0.6029,
      "step": 4258
    },
    {
      "epoch": 2.0182,
      "grad_norm": 3.355067491531372,
      "learning_rate": 7.5612244897959195e-06,
      "loss": 0.3332,
      "step": 4259
    },
    {
      "epoch": 2.0184,
      "grad_norm": 2.9681215286254883,
      "learning_rate": 7.551020408163266e-06,
      "loss": 0.2963,
      "step": 4260
    },
    {
      "epoch": 2.0186,
      "grad_norm": 4.1483988761901855,
      "learning_rate": 7.540816326530613e-06,
      "loss": 0.4423,
      "step": 4261
    },
    {
      "epoch": 2.0188,
      "grad_norm": 3.583317756652832,
      "learning_rate": 7.5306122448979594e-06,
      "loss": 0.211,
      "step": 4262
    },
    {
      "epoch": 2.019,
      "grad_norm": 8.856846809387207,
      "learning_rate": 7.520408163265306e-06,
      "loss": 2.3365,
      "step": 4263
    },
    {
      "epoch": 2.0192,
      "grad_norm": 4.985814571380615,
      "learning_rate": 7.5102040816326536e-06,
      "loss": 0.8046,
      "step": 4264
    },
    {
      "epoch": 2.0194,
      "grad_norm": 10.49881649017334,
      "learning_rate": 7.5e-06,
      "loss": 5.0147,
      "step": 4265
    },
    {
      "epoch": 2.0196,
      "grad_norm": 8.625734329223633,
      "learning_rate": 7.489795918367347e-06,
      "loss": 0.2431,
      "step": 4266
    },
    {
      "epoch": 2.0198,
      "grad_norm": 8.84630298614502,
      "learning_rate": 7.4795918367346935e-06,
      "loss": 1.6197,
      "step": 4267
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.5825251340866089,
      "learning_rate": 7.46938775510204e-06,
      "loss": 0.0774,
      "step": 4268
    },
    {
      "epoch": 2.0202,
      "grad_norm": 1.0772866010665894,
      "learning_rate": 7.4591836734693884e-06,
      "loss": 0.0352,
      "step": 4269
    },
    {
      "epoch": 2.0204,
      "grad_norm": 3.064786911010742,
      "learning_rate": 7.448979591836736e-06,
      "loss": 0.3147,
      "step": 4270
    },
    {
      "epoch": 2.0206,
      "grad_norm": 4.729042053222656,
      "learning_rate": 7.4387755102040826e-06,
      "loss": 0.1488,
      "step": 4271
    },
    {
      "epoch": 2.0208,
      "grad_norm": 0.9412669539451599,
      "learning_rate": 7.428571428571429e-06,
      "loss": 0.0267,
      "step": 4272
    },
    {
      "epoch": 2.021,
      "grad_norm": 2.5195536613464355,
      "learning_rate": 7.418367346938776e-06,
      "loss": 0.2004,
      "step": 4273
    },
    {
      "epoch": 2.0212,
      "grad_norm": 4.93627405166626,
      "learning_rate": 7.4081632653061225e-06,
      "loss": 1.0138,
      "step": 4274
    },
    {
      "epoch": 2.0214,
      "grad_norm": 3.3213465213775635,
      "learning_rate": 7.39795918367347e-06,
      "loss": 0.2463,
      "step": 4275
    },
    {
      "epoch": 2.0216,
      "grad_norm": 9.367770195007324,
      "learning_rate": 7.387755102040817e-06,
      "loss": 1.5291,
      "step": 4276
    },
    {
      "epoch": 2.0218,
      "grad_norm": 5.484495639801025,
      "learning_rate": 7.377551020408163e-06,
      "loss": 0.3543,
      "step": 4277
    },
    {
      "epoch": 2.022,
      "grad_norm": 2.817056894302368,
      "learning_rate": 7.36734693877551e-06,
      "loss": 0.2793,
      "step": 4278
    },
    {
      "epoch": 2.0222,
      "grad_norm": 4.89556884765625,
      "learning_rate": 7.3571428571428565e-06,
      "loss": 0.4653,
      "step": 4279
    },
    {
      "epoch": 2.0224,
      "grad_norm": 5.620337009429932,
      "learning_rate": 7.346938775510205e-06,
      "loss": 1.0702,
      "step": 4280
    },
    {
      "epoch": 2.0226,
      "grad_norm": 8.641432762145996,
      "learning_rate": 7.3367346938775515e-06,
      "loss": 1.0402,
      "step": 4281
    },
    {
      "epoch": 2.0228,
      "grad_norm": 3.3844666481018066,
      "learning_rate": 7.326530612244899e-06,
      "loss": 1.4472,
      "step": 4282
    },
    {
      "epoch": 2.023,
      "grad_norm": 4.433985710144043,
      "learning_rate": 7.316326530612246e-06,
      "loss": 0.6459,
      "step": 4283
    },
    {
      "epoch": 2.0232,
      "grad_norm": 1.9350392818450928,
      "learning_rate": 7.306122448979592e-06,
      "loss": 0.1746,
      "step": 4284
    },
    {
      "epoch": 2.0234,
      "grad_norm": 6.165185451507568,
      "learning_rate": 7.295918367346939e-06,
      "loss": 0.6697,
      "step": 4285
    },
    {
      "epoch": 2.0236,
      "grad_norm": 2.852959156036377,
      "learning_rate": 7.285714285714286e-06,
      "loss": 0.3198,
      "step": 4286
    },
    {
      "epoch": 2.0238,
      "grad_norm": 8.078595161437988,
      "learning_rate": 7.275510204081633e-06,
      "loss": 2.5873,
      "step": 4287
    },
    {
      "epoch": 2.024,
      "grad_norm": 4.600830554962158,
      "learning_rate": 7.26530612244898e-06,
      "loss": 1.077,
      "step": 4288
    },
    {
      "epoch": 2.0242,
      "grad_norm": 10.893990516662598,
      "learning_rate": 7.255102040816326e-06,
      "loss": 1.9659,
      "step": 4289
    },
    {
      "epoch": 2.0244,
      "grad_norm": 5.1831207275390625,
      "learning_rate": 7.244897959183673e-06,
      "loss": 2.1732,
      "step": 4290
    },
    {
      "epoch": 2.0246,
      "grad_norm": 25.182043075561523,
      "learning_rate": 7.2346938775510204e-06,
      "loss": 3.6428,
      "step": 4291
    },
    {
      "epoch": 2.0248,
      "grad_norm": 10.35595703125,
      "learning_rate": 7.224489795918368e-06,
      "loss": 2.1839,
      "step": 4292
    },
    {
      "epoch": 2.025,
      "grad_norm": 32.797996520996094,
      "learning_rate": 7.214285714285715e-06,
      "loss": 6.636,
      "step": 4293
    },
    {
      "epoch": 2.0252,
      "grad_norm": 4.553895473480225,
      "learning_rate": 7.204081632653062e-06,
      "loss": 0.5681,
      "step": 4294
    },
    {
      "epoch": 2.0254,
      "grad_norm": 9.21674633026123,
      "learning_rate": 7.193877551020409e-06,
      "loss": 4.7964,
      "step": 4295
    },
    {
      "epoch": 2.0256,
      "grad_norm": 4.0077714920043945,
      "learning_rate": 7.183673469387755e-06,
      "loss": 0.841,
      "step": 4296
    },
    {
      "epoch": 2.0258,
      "grad_norm": 4.008864402770996,
      "learning_rate": 7.173469387755102e-06,
      "loss": 0.5305,
      "step": 4297
    },
    {
      "epoch": 2.026,
      "grad_norm": 3.588186740875244,
      "learning_rate": 7.1632653061224494e-06,
      "loss": 1.5291,
      "step": 4298
    },
    {
      "epoch": 2.0262000000000002,
      "grad_norm": 4.0893330574035645,
      "learning_rate": 7.153061224489796e-06,
      "loss": 1.1,
      "step": 4299
    },
    {
      "epoch": 2.0264,
      "grad_norm": 3.504427194595337,
      "learning_rate": 7.142857142857143e-06,
      "loss": 1.3353,
      "step": 4300
    },
    {
      "epoch": 2.0266,
      "grad_norm": 3.810716390609741,
      "learning_rate": 7.132653061224489e-06,
      "loss": 2.1167,
      "step": 4301
    },
    {
      "epoch": 2.0268,
      "grad_norm": 12.104032516479492,
      "learning_rate": 7.122448979591837e-06,
      "loss": 5.6261,
      "step": 4302
    },
    {
      "epoch": 2.027,
      "grad_norm": 3.746196985244751,
      "learning_rate": 7.112244897959184e-06,
      "loss": 1.511,
      "step": 4303
    },
    {
      "epoch": 2.0272,
      "grad_norm": 12.053168296813965,
      "learning_rate": 7.102040816326532e-06,
      "loss": 3.4344,
      "step": 4304
    },
    {
      "epoch": 2.0274,
      "grad_norm": 1.4310916662216187,
      "learning_rate": 7.0918367346938785e-06,
      "loss": 0.0547,
      "step": 4305
    },
    {
      "epoch": 2.0276,
      "grad_norm": 7.563154697418213,
      "learning_rate": 7.081632653061225e-06,
      "loss": 0.9291,
      "step": 4306
    },
    {
      "epoch": 2.0278,
      "grad_norm": 2.7030062675476074,
      "learning_rate": 7.071428571428572e-06,
      "loss": 0.3102,
      "step": 4307
    },
    {
      "epoch": 2.028,
      "grad_norm": 16.48634910583496,
      "learning_rate": 7.061224489795918e-06,
      "loss": 2.2507,
      "step": 4308
    },
    {
      "epoch": 2.0282,
      "grad_norm": 3.9196617603302,
      "learning_rate": 7.051020408163266e-06,
      "loss": 0.5105,
      "step": 4309
    },
    {
      "epoch": 2.0284,
      "grad_norm": 4.114338397979736,
      "learning_rate": 7.0408163265306125e-06,
      "loss": 0.9735,
      "step": 4310
    },
    {
      "epoch": 2.0286,
      "grad_norm": 1.3699504137039185,
      "learning_rate": 7.030612244897959e-06,
      "loss": 0.0524,
      "step": 4311
    },
    {
      "epoch": 2.0288,
      "grad_norm": 3.6880979537963867,
      "learning_rate": 7.020408163265306e-06,
      "loss": 1.5327,
      "step": 4312
    },
    {
      "epoch": 2.029,
      "grad_norm": 5.881893157958984,
      "learning_rate": 7.010204081632652e-06,
      "loss": 1.4355,
      "step": 4313
    },
    {
      "epoch": 2.0292,
      "grad_norm": 4.015309810638428,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.6515,
      "step": 4314
    },
    {
      "epoch": 2.0294,
      "grad_norm": 9.127671241760254,
      "learning_rate": 6.989795918367348e-06,
      "loss": 5.0853,
      "step": 4315
    },
    {
      "epoch": 2.0296,
      "grad_norm": 9.522642135620117,
      "learning_rate": 6.979591836734695e-06,
      "loss": 2.508,
      "step": 4316
    },
    {
      "epoch": 2.0298,
      "grad_norm": 8.305075645446777,
      "learning_rate": 6.9693877551020415e-06,
      "loss": 2.0008,
      "step": 4317
    },
    {
      "epoch": 2.03,
      "grad_norm": 10.065943717956543,
      "learning_rate": 6.959183673469388e-06,
      "loss": 3.3854,
      "step": 4318
    },
    {
      "epoch": 2.0302,
      "grad_norm": 3.0334036350250244,
      "learning_rate": 6.948979591836735e-06,
      "loss": 0.51,
      "step": 4319
    },
    {
      "epoch": 2.0304,
      "grad_norm": 5.859947681427002,
      "learning_rate": 6.938775510204082e-06,
      "loss": 0.4441,
      "step": 4320
    },
    {
      "epoch": 2.0306,
      "grad_norm": 5.833090305328369,
      "learning_rate": 6.928571428571429e-06,
      "loss": 1.9727,
      "step": 4321
    },
    {
      "epoch": 2.0308,
      "grad_norm": 14.231902122497559,
      "learning_rate": 6.9183673469387755e-06,
      "loss": 2.0807,
      "step": 4322
    },
    {
      "epoch": 2.031,
      "grad_norm": 3.9816040992736816,
      "learning_rate": 6.908163265306122e-06,
      "loss": 2.2134,
      "step": 4323
    },
    {
      "epoch": 2.0312,
      "grad_norm": 10.230045318603516,
      "learning_rate": 6.897959183673469e-06,
      "loss": 1.8335,
      "step": 4324
    },
    {
      "epoch": 2.0314,
      "grad_norm": 3.276740789413452,
      "learning_rate": 6.887755102040816e-06,
      "loss": 1.4815,
      "step": 4325
    },
    {
      "epoch": 2.0316,
      "grad_norm": 6.0765790939331055,
      "learning_rate": 6.877551020408164e-06,
      "loss": 0.6174,
      "step": 4326
    },
    {
      "epoch": 2.0318,
      "grad_norm": 1.807294249534607,
      "learning_rate": 6.867346938775511e-06,
      "loss": 0.1718,
      "step": 4327
    },
    {
      "epoch": 2.032,
      "grad_norm": 12.442965507507324,
      "learning_rate": 6.857142857142858e-06,
      "loss": 1.442,
      "step": 4328
    },
    {
      "epoch": 2.0322,
      "grad_norm": 14.195085525512695,
      "learning_rate": 6.8469387755102046e-06,
      "loss": 0.8755,
      "step": 4329
    },
    {
      "epoch": 2.0324,
      "grad_norm": 2.847365617752075,
      "learning_rate": 6.836734693877551e-06,
      "loss": 0.4877,
      "step": 4330
    },
    {
      "epoch": 2.0326,
      "grad_norm": 11.989447593688965,
      "learning_rate": 6.826530612244898e-06,
      "loss": 3.3478,
      "step": 4331
    },
    {
      "epoch": 2.0328,
      "grad_norm": 1.4652307033538818,
      "learning_rate": 6.816326530612245e-06,
      "loss": 0.0239,
      "step": 4332
    },
    {
      "epoch": 2.033,
      "grad_norm": 3.8287265300750732,
      "learning_rate": 6.806122448979592e-06,
      "loss": 0.3573,
      "step": 4333
    },
    {
      "epoch": 2.0332,
      "grad_norm": 2.04620099067688,
      "learning_rate": 6.795918367346939e-06,
      "loss": 0.1761,
      "step": 4334
    },
    {
      "epoch": 2.0334,
      "grad_norm": 10.700292587280273,
      "learning_rate": 6.785714285714285e-06,
      "loss": 1.4991,
      "step": 4335
    },
    {
      "epoch": 2.0336,
      "grad_norm": 2.6859498023986816,
      "learning_rate": 6.775510204081633e-06,
      "loss": 0.1894,
      "step": 4336
    },
    {
      "epoch": 2.0338,
      "grad_norm": 7.146368503570557,
      "learning_rate": 6.76530612244898e-06,
      "loss": 1.0313,
      "step": 4337
    },
    {
      "epoch": 2.034,
      "grad_norm": 4.037665367126465,
      "learning_rate": 6.755102040816328e-06,
      "loss": 0.556,
      "step": 4338
    },
    {
      "epoch": 2.0342,
      "grad_norm": 3.5893335342407227,
      "learning_rate": 6.744897959183674e-06,
      "loss": 0.6636,
      "step": 4339
    },
    {
      "epoch": 2.0344,
      "grad_norm": 3.3621785640716553,
      "learning_rate": 6.734693877551021e-06,
      "loss": 1.4062,
      "step": 4340
    },
    {
      "epoch": 2.0346,
      "grad_norm": 5.736871242523193,
      "learning_rate": 6.724489795918368e-06,
      "loss": 1.1887,
      "step": 4341
    },
    {
      "epoch": 2.0348,
      "grad_norm": 8.566832542419434,
      "learning_rate": 6.714285714285714e-06,
      "loss": 5.0112,
      "step": 4342
    },
    {
      "epoch": 2.035,
      "grad_norm": 7.10464334487915,
      "learning_rate": 6.704081632653062e-06,
      "loss": 1.9142,
      "step": 4343
    },
    {
      "epoch": 2.0352,
      "grad_norm": 1.1951695680618286,
      "learning_rate": 6.693877551020408e-06,
      "loss": 0.0413,
      "step": 4344
    },
    {
      "epoch": 2.0354,
      "grad_norm": 1.8170075416564941,
      "learning_rate": 6.683673469387755e-06,
      "loss": 0.1659,
      "step": 4345
    },
    {
      "epoch": 2.0356,
      "grad_norm": 1.655696988105774,
      "learning_rate": 6.673469387755102e-06,
      "loss": 0.1428,
      "step": 4346
    },
    {
      "epoch": 2.0358,
      "grad_norm": 2.352597236633301,
      "learning_rate": 6.663265306122448e-06,
      "loss": 0.2514,
      "step": 4347
    },
    {
      "epoch": 2.036,
      "grad_norm": 3.6672985553741455,
      "learning_rate": 6.653061224489797e-06,
      "loss": 0.3943,
      "step": 4348
    },
    {
      "epoch": 2.0362,
      "grad_norm": 4.368422985076904,
      "learning_rate": 6.642857142857144e-06,
      "loss": 0.6111,
      "step": 4349
    },
    {
      "epoch": 2.0364,
      "grad_norm": 6.8434624671936035,
      "learning_rate": 6.632653061224491e-06,
      "loss": 1.3778,
      "step": 4350
    },
    {
      "epoch": 2.0366,
      "grad_norm": 1.9027135372161865,
      "learning_rate": 6.622448979591837e-06,
      "loss": 0.1582,
      "step": 4351
    },
    {
      "epoch": 2.0368,
      "grad_norm": 3.6676418781280518,
      "learning_rate": 6.612244897959184e-06,
      "loss": 1.6667,
      "step": 4352
    },
    {
      "epoch": 2.037,
      "grad_norm": 1.3977506160736084,
      "learning_rate": 6.602040816326531e-06,
      "loss": 0.0251,
      "step": 4353
    },
    {
      "epoch": 2.0372,
      "grad_norm": 4.393439769744873,
      "learning_rate": 6.591836734693878e-06,
      "loss": 0.2797,
      "step": 4354
    },
    {
      "epoch": 2.0374,
      "grad_norm": 1.9742120504379272,
      "learning_rate": 6.581632653061225e-06,
      "loss": 0.1793,
      "step": 4355
    },
    {
      "epoch": 2.0376,
      "grad_norm": 3.285804510116577,
      "learning_rate": 6.5714285714285714e-06,
      "loss": 0.2586,
      "step": 4356
    },
    {
      "epoch": 2.0378,
      "grad_norm": 3.293999195098877,
      "learning_rate": 6.561224489795918e-06,
      "loss": 0.1964,
      "step": 4357
    },
    {
      "epoch": 2.038,
      "grad_norm": 3.4198780059814453,
      "learning_rate": 6.551020408163265e-06,
      "loss": 0.5805,
      "step": 4358
    },
    {
      "epoch": 2.0382,
      "grad_norm": 10.4772367477417,
      "learning_rate": 6.540816326530612e-06,
      "loss": 1.8666,
      "step": 4359
    },
    {
      "epoch": 2.0384,
      "grad_norm": 2.4163289070129395,
      "learning_rate": 6.53061224489796e-06,
      "loss": 0.2647,
      "step": 4360
    },
    {
      "epoch": 2.0386,
      "grad_norm": 3.692894220352173,
      "learning_rate": 6.520408163265307e-06,
      "loss": 0.4535,
      "step": 4361
    },
    {
      "epoch": 2.0388,
      "grad_norm": 4.505298614501953,
      "learning_rate": 6.510204081632654e-06,
      "loss": 0.5907,
      "step": 4362
    },
    {
      "epoch": 2.039,
      "grad_norm": 4.578971862792969,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.7895,
      "step": 4363
    },
    {
      "epoch": 2.0392,
      "grad_norm": 1.5840128660202026,
      "learning_rate": 6.489795918367347e-06,
      "loss": 0.1479,
      "step": 4364
    },
    {
      "epoch": 2.0394,
      "grad_norm": 4.764891624450684,
      "learning_rate": 6.4795918367346946e-06,
      "loss": 0.6096,
      "step": 4365
    },
    {
      "epoch": 2.0396,
      "grad_norm": 2.5816330909729004,
      "learning_rate": 6.469387755102041e-06,
      "loss": 0.3104,
      "step": 4366
    },
    {
      "epoch": 2.0398,
      "grad_norm": 2.8380391597747803,
      "learning_rate": 6.459183673469388e-06,
      "loss": 0.3112,
      "step": 4367
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.092507839202881,
      "learning_rate": 6.4489795918367345e-06,
      "loss": 0.6007,
      "step": 4368
    },
    {
      "epoch": 2.0402,
      "grad_norm": 5.286302089691162,
      "learning_rate": 6.438775510204081e-06,
      "loss": 1.4574,
      "step": 4369
    },
    {
      "epoch": 2.0404,
      "grad_norm": 11.976883888244629,
      "learning_rate": 6.428571428571429e-06,
      "loss": 0.9541,
      "step": 4370
    },
    {
      "epoch": 2.0406,
      "grad_norm": 3.1366777420043945,
      "learning_rate": 6.418367346938776e-06,
      "loss": 0.5663,
      "step": 4371
    },
    {
      "epoch": 2.0408,
      "grad_norm": 5.570676326751709,
      "learning_rate": 6.408163265306124e-06,
      "loss": 1.6063,
      "step": 4372
    },
    {
      "epoch": 2.041,
      "grad_norm": 3.534306764602661,
      "learning_rate": 6.39795918367347e-06,
      "loss": 0.4283,
      "step": 4373
    },
    {
      "epoch": 2.0412,
      "grad_norm": 8.754057884216309,
      "learning_rate": 6.387755102040817e-06,
      "loss": 2.1288,
      "step": 4374
    },
    {
      "epoch": 2.0414,
      "grad_norm": 4.957884788513184,
      "learning_rate": 6.3775510204081635e-06,
      "loss": 0.5723,
      "step": 4375
    },
    {
      "epoch": 2.0416,
      "grad_norm": 3.3927721977233887,
      "learning_rate": 6.36734693877551e-06,
      "loss": 1.4173,
      "step": 4376
    },
    {
      "epoch": 2.0418,
      "grad_norm": 4.619457244873047,
      "learning_rate": 6.357142857142858e-06,
      "loss": 0.5361,
      "step": 4377
    },
    {
      "epoch": 2.042,
      "grad_norm": 2.0513007640838623,
      "learning_rate": 6.346938775510204e-06,
      "loss": 0.2143,
      "step": 4378
    },
    {
      "epoch": 2.0422,
      "grad_norm": 2.817387580871582,
      "learning_rate": 6.336734693877551e-06,
      "loss": 0.3838,
      "step": 4379
    },
    {
      "epoch": 2.0424,
      "grad_norm": 3.58909273147583,
      "learning_rate": 6.3265306122448975e-06,
      "loss": 0.3807,
      "step": 4380
    },
    {
      "epoch": 2.0426,
      "grad_norm": 4.798856735229492,
      "learning_rate": 6.316326530612245e-06,
      "loss": 1.434,
      "step": 4381
    },
    {
      "epoch": 2.0428,
      "grad_norm": 8.82713794708252,
      "learning_rate": 6.3061224489795925e-06,
      "loss": 1.4132,
      "step": 4382
    },
    {
      "epoch": 2.043,
      "grad_norm": 2.2979350090026855,
      "learning_rate": 6.29591836734694e-06,
      "loss": 0.1361,
      "step": 4383
    },
    {
      "epoch": 2.0432,
      "grad_norm": 2.487318754196167,
      "learning_rate": 6.285714285714287e-06,
      "loss": 0.5233,
      "step": 4384
    },
    {
      "epoch": 2.0434,
      "grad_norm": 4.03679084777832,
      "learning_rate": 6.275510204081633e-06,
      "loss": 0.4972,
      "step": 4385
    },
    {
      "epoch": 2.0436,
      "grad_norm": 5.528019905090332,
      "learning_rate": 6.26530612244898e-06,
      "loss": 2.2495,
      "step": 4386
    },
    {
      "epoch": 2.0438,
      "grad_norm": 1.5046391487121582,
      "learning_rate": 6.2551020408163266e-06,
      "loss": 0.0472,
      "step": 4387
    },
    {
      "epoch": 2.044,
      "grad_norm": 9.946911811828613,
      "learning_rate": 6.244897959183674e-06,
      "loss": 2.3159,
      "step": 4388
    },
    {
      "epoch": 2.0442,
      "grad_norm": 2.3810222148895264,
      "learning_rate": 6.234693877551021e-06,
      "loss": 0.183,
      "step": 4389
    },
    {
      "epoch": 2.0444,
      "grad_norm": 12.703330039978027,
      "learning_rate": 6.224489795918367e-06,
      "loss": 3.7456,
      "step": 4390
    },
    {
      "epoch": 2.0446,
      "grad_norm": 4.766584873199463,
      "learning_rate": 6.214285714285715e-06,
      "loss": 0.9412,
      "step": 4391
    },
    {
      "epoch": 2.0448,
      "grad_norm": 13.101330757141113,
      "learning_rate": 6.2040816326530614e-06,
      "loss": 1.9205,
      "step": 4392
    },
    {
      "epoch": 2.045,
      "grad_norm": 4.829402446746826,
      "learning_rate": 6.193877551020408e-06,
      "loss": 1.543,
      "step": 4393
    },
    {
      "epoch": 2.0452,
      "grad_norm": 7.299891948699951,
      "learning_rate": 6.1836734693877556e-06,
      "loss": 2.6336,
      "step": 4394
    },
    {
      "epoch": 2.0454,
      "grad_norm": 1.5839056968688965,
      "learning_rate": 6.173469387755102e-06,
      "loss": 0.0655,
      "step": 4395
    },
    {
      "epoch": 2.0456,
      "grad_norm": 2.8372104167938232,
      "learning_rate": 6.163265306122449e-06,
      "loss": 1.3619,
      "step": 4396
    },
    {
      "epoch": 2.0458,
      "grad_norm": 1.482968807220459,
      "learning_rate": 6.153061224489796e-06,
      "loss": 0.0501,
      "step": 4397
    },
    {
      "epoch": 2.046,
      "grad_norm": 2.1378843784332275,
      "learning_rate": 6.142857142857143e-06,
      "loss": 0.2139,
      "step": 4398
    },
    {
      "epoch": 2.0462,
      "grad_norm": 4.614658832550049,
      "learning_rate": 6.1326530612244905e-06,
      "loss": 0.2673,
      "step": 4399
    },
    {
      "epoch": 2.0464,
      "grad_norm": 6.131068229675293,
      "learning_rate": 6.122448979591837e-06,
      "loss": 0.5613,
      "step": 4400
    },
    {
      "epoch": 2.0466,
      "grad_norm": 7.097366809844971,
      "learning_rate": 6.112244897959184e-06,
      "loss": 0.7178,
      "step": 4401
    },
    {
      "epoch": 2.0468,
      "grad_norm": 13.37352180480957,
      "learning_rate": 6.102040816326531e-06,
      "loss": 1.3431,
      "step": 4402
    },
    {
      "epoch": 2.047,
      "grad_norm": 2.1020140647888184,
      "learning_rate": 6.091836734693878e-06,
      "loss": 0.2304,
      "step": 4403
    },
    {
      "epoch": 2.0472,
      "grad_norm": 2.519576072692871,
      "learning_rate": 6.0816326530612245e-06,
      "loss": 0.3093,
      "step": 4404
    },
    {
      "epoch": 2.0474,
      "grad_norm": 3.063638925552368,
      "learning_rate": 6.071428571428572e-06,
      "loss": 0.5728,
      "step": 4405
    },
    {
      "epoch": 2.0476,
      "grad_norm": 5.231879234313965,
      "learning_rate": 6.061224489795919e-06,
      "loss": 0.9442,
      "step": 4406
    },
    {
      "epoch": 2.0478,
      "grad_norm": 4.134194850921631,
      "learning_rate": 6.051020408163265e-06,
      "loss": 0.5767,
      "step": 4407
    },
    {
      "epoch": 2.048,
      "grad_norm": 4.078321933746338,
      "learning_rate": 6.040816326530613e-06,
      "loss": 0.6297,
      "step": 4408
    },
    {
      "epoch": 2.0482,
      "grad_norm": 8.251264572143555,
      "learning_rate": 6.030612244897959e-06,
      "loss": 2.754,
      "step": 4409
    },
    {
      "epoch": 2.0484,
      "grad_norm": 2.3369216918945312,
      "learning_rate": 6.020408163265306e-06,
      "loss": 0.4342,
      "step": 4410
    },
    {
      "epoch": 2.0486,
      "grad_norm": 6.78234338760376,
      "learning_rate": 6.0102040816326535e-06,
      "loss": 1.9212,
      "step": 4411
    },
    {
      "epoch": 2.0488,
      "grad_norm": 3.5200674533843994,
      "learning_rate": 6e-06,
      "loss": 1.628,
      "step": 4412
    },
    {
      "epoch": 2.049,
      "grad_norm": 2.2730045318603516,
      "learning_rate": 5.989795918367347e-06,
      "loss": 0.2568,
      "step": 4413
    },
    {
      "epoch": 2.0492,
      "grad_norm": 2.6582682132720947,
      "learning_rate": 5.979591836734694e-06,
      "loss": 0.2148,
      "step": 4414
    },
    {
      "epoch": 2.0494,
      "grad_norm": 5.059128761291504,
      "learning_rate": 5.969387755102041e-06,
      "loss": 0.5851,
      "step": 4415
    },
    {
      "epoch": 2.0496,
      "grad_norm": 2.932605266571045,
      "learning_rate": 5.959183673469388e-06,
      "loss": 0.5677,
      "step": 4416
    },
    {
      "epoch": 2.0498,
      "grad_norm": 2.9726316928863525,
      "learning_rate": 5.948979591836735e-06,
      "loss": 0.2871,
      "step": 4417
    },
    {
      "epoch": 2.05,
      "grad_norm": 3.800670623779297,
      "learning_rate": 5.938775510204082e-06,
      "loss": 1.1672,
      "step": 4418
    },
    {
      "epoch": 2.0502,
      "grad_norm": 7.1238579750061035,
      "learning_rate": 5.928571428571429e-06,
      "loss": 0.815,
      "step": 4419
    },
    {
      "epoch": 2.0504,
      "grad_norm": 15.937649726867676,
      "learning_rate": 5.918367346938776e-06,
      "loss": 3.0679,
      "step": 4420
    },
    {
      "epoch": 2.0506,
      "grad_norm": 3.0840086936950684,
      "learning_rate": 5.9081632653061224e-06,
      "loss": 1.4612,
      "step": 4421
    },
    {
      "epoch": 2.0508,
      "grad_norm": 4.5798187255859375,
      "learning_rate": 5.89795918367347e-06,
      "loss": 0.648,
      "step": 4422
    },
    {
      "epoch": 2.051,
      "grad_norm": 1.7499949932098389,
      "learning_rate": 5.8877551020408166e-06,
      "loss": 0.1632,
      "step": 4423
    },
    {
      "epoch": 2.0512,
      "grad_norm": 6.960896015167236,
      "learning_rate": 5.877551020408163e-06,
      "loss": 1.4873,
      "step": 4424
    },
    {
      "epoch": 2.0514,
      "grad_norm": 4.876622676849365,
      "learning_rate": 5.867346938775511e-06,
      "loss": 0.8274,
      "step": 4425
    },
    {
      "epoch": 2.0516,
      "grad_norm": 4.382575988769531,
      "learning_rate": 5.857142857142857e-06,
      "loss": 0.7828,
      "step": 4426
    },
    {
      "epoch": 2.0518,
      "grad_norm": 8.884764671325684,
      "learning_rate": 5.846938775510204e-06,
      "loss": 1.8684,
      "step": 4427
    },
    {
      "epoch": 2.052,
      "grad_norm": 6.382287502288818,
      "learning_rate": 5.8367346938775515e-06,
      "loss": 1.1565,
      "step": 4428
    },
    {
      "epoch": 2.0522,
      "grad_norm": 10.393208503723145,
      "learning_rate": 5.826530612244898e-06,
      "loss": 2.3059,
      "step": 4429
    },
    {
      "epoch": 2.0524,
      "grad_norm": 9.367992401123047,
      "learning_rate": 5.816326530612245e-06,
      "loss": 3.4319,
      "step": 4430
    },
    {
      "epoch": 2.0526,
      "grad_norm": 24.33891487121582,
      "learning_rate": 5.806122448979592e-06,
      "loss": 5.3695,
      "step": 4431
    },
    {
      "epoch": 2.0528,
      "grad_norm": 4.95985746383667,
      "learning_rate": 5.795918367346939e-06,
      "loss": 1.0752,
      "step": 4432
    },
    {
      "epoch": 2.053,
      "grad_norm": 20.802255630493164,
      "learning_rate": 5.785714285714286e-06,
      "loss": 3.8515,
      "step": 4433
    },
    {
      "epoch": 2.0532,
      "grad_norm": 3.7617321014404297,
      "learning_rate": 5.775510204081633e-06,
      "loss": 1.4798,
      "step": 4434
    },
    {
      "epoch": 2.0534,
      "grad_norm": 1.6784859895706177,
      "learning_rate": 5.76530612244898e-06,
      "loss": 0.0366,
      "step": 4435
    },
    {
      "epoch": 2.0536,
      "grad_norm": 4.394393444061279,
      "learning_rate": 5.755102040816327e-06,
      "loss": 0.5598,
      "step": 4436
    },
    {
      "epoch": 2.0538,
      "grad_norm": 6.262113094329834,
      "learning_rate": 5.744897959183674e-06,
      "loss": 1.5221,
      "step": 4437
    },
    {
      "epoch": 2.054,
      "grad_norm": 2.5081212520599365,
      "learning_rate": 5.73469387755102e-06,
      "loss": 0.2736,
      "step": 4438
    },
    {
      "epoch": 2.0542,
      "grad_norm": 15.906336784362793,
      "learning_rate": 5.724489795918368e-06,
      "loss": 5.0865,
      "step": 4439
    },
    {
      "epoch": 2.0544,
      "grad_norm": 4.739946365356445,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.5971,
      "step": 4440
    },
    {
      "epoch": 2.0546,
      "grad_norm": 3.405334234237671,
      "learning_rate": 5.704081632653061e-06,
      "loss": 1.0074,
      "step": 4441
    },
    {
      "epoch": 2.0548,
      "grad_norm": 20.478851318359375,
      "learning_rate": 5.693877551020409e-06,
      "loss": 1.5863,
      "step": 4442
    },
    {
      "epoch": 2.055,
      "grad_norm": 3.57926344871521,
      "learning_rate": 5.683673469387755e-06,
      "loss": 1.4315,
      "step": 4443
    },
    {
      "epoch": 2.0552,
      "grad_norm": 18.06096839904785,
      "learning_rate": 5.673469387755103e-06,
      "loss": 3.774,
      "step": 4444
    },
    {
      "epoch": 2.0554,
      "grad_norm": 1.1555712223052979,
      "learning_rate": 5.663265306122449e-06,
      "loss": 0.0234,
      "step": 4445
    },
    {
      "epoch": 2.0556,
      "grad_norm": 12.743829727172852,
      "learning_rate": 5.653061224489796e-06,
      "loss": 1.3324,
      "step": 4446
    },
    {
      "epoch": 2.0558,
      "grad_norm": 1.9675103425979614,
      "learning_rate": 5.642857142857143e-06,
      "loss": 0.1721,
      "step": 4447
    },
    {
      "epoch": 2.056,
      "grad_norm": 11.829596519470215,
      "learning_rate": 5.63265306122449e-06,
      "loss": 3.8803,
      "step": 4448
    },
    {
      "epoch": 2.0562,
      "grad_norm": 1.3106498718261719,
      "learning_rate": 5.622448979591837e-06,
      "loss": 0.0457,
      "step": 4449
    },
    {
      "epoch": 2.0564,
      "grad_norm": 4.3609938621521,
      "learning_rate": 5.612244897959184e-06,
      "loss": 0.2479,
      "step": 4450
    },
    {
      "epoch": 2.0566,
      "grad_norm": 1.6185758113861084,
      "learning_rate": 5.602040816326531e-06,
      "loss": 0.1516,
      "step": 4451
    },
    {
      "epoch": 2.0568,
      "grad_norm": 2.02146577835083,
      "learning_rate": 5.5918367346938776e-06,
      "loss": 0.1894,
      "step": 4452
    },
    {
      "epoch": 2.057,
      "grad_norm": 1.1370601654052734,
      "learning_rate": 5.581632653061225e-06,
      "loss": 0.0356,
      "step": 4453
    },
    {
      "epoch": 2.0572,
      "grad_norm": 2.578493118286133,
      "learning_rate": 5.571428571428572e-06,
      "loss": 0.6012,
      "step": 4454
    },
    {
      "epoch": 2.0574,
      "grad_norm": 1.9479355812072754,
      "learning_rate": 5.561224489795918e-06,
      "loss": 0.2416,
      "step": 4455
    },
    {
      "epoch": 2.0576,
      "grad_norm": 4.031513690948486,
      "learning_rate": 5.551020408163266e-06,
      "loss": 1.8204,
      "step": 4456
    },
    {
      "epoch": 2.0578,
      "grad_norm": 2.8992726802825928,
      "learning_rate": 5.5408163265306125e-06,
      "loss": 0.3126,
      "step": 4457
    },
    {
      "epoch": 2.058,
      "grad_norm": 5.120978355407715,
      "learning_rate": 5.530612244897959e-06,
      "loss": 1.4985,
      "step": 4458
    },
    {
      "epoch": 2.0582,
      "grad_norm": 4.285439491271973,
      "learning_rate": 5.520408163265307e-06,
      "loss": 0.553,
      "step": 4459
    },
    {
      "epoch": 2.0584,
      "grad_norm": 1.931534767150879,
      "learning_rate": 5.510204081632653e-06,
      "loss": 0.1953,
      "step": 4460
    },
    {
      "epoch": 2.0586,
      "grad_norm": 3.271833658218384,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.5028,
      "step": 4461
    },
    {
      "epoch": 2.0588,
      "grad_norm": 1.9770781993865967,
      "learning_rate": 5.489795918367347e-06,
      "loss": 0.0458,
      "step": 4462
    },
    {
      "epoch": 2.059,
      "grad_norm": 4.2079997062683105,
      "learning_rate": 5.479591836734694e-06,
      "loss": 1.8211,
      "step": 4463
    },
    {
      "epoch": 2.0592,
      "grad_norm": 2.775291919708252,
      "learning_rate": 5.469387755102041e-06,
      "loss": 0.2137,
      "step": 4464
    },
    {
      "epoch": 2.0594,
      "grad_norm": 6.623566627502441,
      "learning_rate": 5.459183673469388e-06,
      "loss": 1.5583,
      "step": 4465
    },
    {
      "epoch": 2.0596,
      "grad_norm": 3.7688889503479004,
      "learning_rate": 5.448979591836735e-06,
      "loss": 0.1495,
      "step": 4466
    },
    {
      "epoch": 2.0598,
      "grad_norm": 3.026883363723755,
      "learning_rate": 5.438775510204082e-06,
      "loss": 0.593,
      "step": 4467
    },
    {
      "epoch": 2.06,
      "grad_norm": 5.533717155456543,
      "learning_rate": 5.428571428571429e-06,
      "loss": 1.3528,
      "step": 4468
    },
    {
      "epoch": 2.0602,
      "grad_norm": 2.3275177478790283,
      "learning_rate": 5.4183673469387755e-06,
      "loss": 0.0615,
      "step": 4469
    },
    {
      "epoch": 2.0604,
      "grad_norm": 8.552267074584961,
      "learning_rate": 5.408163265306123e-06,
      "loss": 4.6218,
      "step": 4470
    },
    {
      "epoch": 2.0606,
      "grad_norm": 2.0251433849334717,
      "learning_rate": 5.39795918367347e-06,
      "loss": 0.0536,
      "step": 4471
    },
    {
      "epoch": 2.0608,
      "grad_norm": 2.1941115856170654,
      "learning_rate": 5.387755102040816e-06,
      "loss": 0.1264,
      "step": 4472
    },
    {
      "epoch": 2.061,
      "grad_norm": 4.157203197479248,
      "learning_rate": 5.377551020408164e-06,
      "loss": 1.3165,
      "step": 4473
    },
    {
      "epoch": 2.0612,
      "grad_norm": 4.577666759490967,
      "learning_rate": 5.36734693877551e-06,
      "loss": 0.568,
      "step": 4474
    },
    {
      "epoch": 2.0614,
      "grad_norm": 8.491950035095215,
      "learning_rate": 5.357142857142857e-06,
      "loss": 4.998,
      "step": 4475
    },
    {
      "epoch": 2.0616,
      "grad_norm": 1.6711488962173462,
      "learning_rate": 5.3469387755102045e-06,
      "loss": 0.1915,
      "step": 4476
    },
    {
      "epoch": 2.0618,
      "grad_norm": 3.7359209060668945,
      "learning_rate": 5.336734693877551e-06,
      "loss": 1.6961,
      "step": 4477
    },
    {
      "epoch": 2.062,
      "grad_norm": 5.737436294555664,
      "learning_rate": 5.326530612244899e-06,
      "loss": 0.8686,
      "step": 4478
    },
    {
      "epoch": 2.0622,
      "grad_norm": 4.680469512939453,
      "learning_rate": 5.316326530612245e-06,
      "loss": 0.7464,
      "step": 4479
    },
    {
      "epoch": 2.0624,
      "grad_norm": 8.432143211364746,
      "learning_rate": 5.306122448979592e-06,
      "loss": 1.9435,
      "step": 4480
    },
    {
      "epoch": 2.0626,
      "grad_norm": 1.610304594039917,
      "learning_rate": 5.2959183673469386e-06,
      "loss": 0.111,
      "step": 4481
    },
    {
      "epoch": 2.0628,
      "grad_norm": 6.408649921417236,
      "learning_rate": 5.285714285714286e-06,
      "loss": 0.8789,
      "step": 4482
    },
    {
      "epoch": 2.063,
      "grad_norm": 2.3763010501861572,
      "learning_rate": 5.275510204081633e-06,
      "loss": 0.3241,
      "step": 4483
    },
    {
      "epoch": 2.0632,
      "grad_norm": 5.558871746063232,
      "learning_rate": 5.26530612244898e-06,
      "loss": 1.4693,
      "step": 4484
    },
    {
      "epoch": 2.0634,
      "grad_norm": 4.030745506286621,
      "learning_rate": 5.255102040816327e-06,
      "loss": 0.6144,
      "step": 4485
    },
    {
      "epoch": 2.0636,
      "grad_norm": 14.399699211120605,
      "learning_rate": 5.2448979591836735e-06,
      "loss": 0.8409,
      "step": 4486
    },
    {
      "epoch": 2.0638,
      "grad_norm": 2.289337635040283,
      "learning_rate": 5.234693877551021e-06,
      "loss": 0.192,
      "step": 4487
    },
    {
      "epoch": 2.064,
      "grad_norm": 26.556169509887695,
      "learning_rate": 5.224489795918368e-06,
      "loss": 2.6163,
      "step": 4488
    },
    {
      "epoch": 2.0642,
      "grad_norm": 1.5711888074874878,
      "learning_rate": 5.214285714285714e-06,
      "loss": 0.0689,
      "step": 4489
    },
    {
      "epoch": 2.0644,
      "grad_norm": 4.504660606384277,
      "learning_rate": 5.204081632653062e-06,
      "loss": 1.2876,
      "step": 4490
    },
    {
      "epoch": 2.0646,
      "grad_norm": 1.9299784898757935,
      "learning_rate": 5.193877551020408e-06,
      "loss": 0.491,
      "step": 4491
    },
    {
      "epoch": 2.0648,
      "grad_norm": 2.0123250484466553,
      "learning_rate": 5.183673469387755e-06,
      "loss": 0.2154,
      "step": 4492
    },
    {
      "epoch": 2.065,
      "grad_norm": 3.2566661834716797,
      "learning_rate": 5.1734693877551025e-06,
      "loss": 1.411,
      "step": 4493
    },
    {
      "epoch": 2.0652,
      "grad_norm": 1.9258058071136475,
      "learning_rate": 5.163265306122449e-06,
      "loss": 0.2116,
      "step": 4494
    },
    {
      "epoch": 2.0654,
      "grad_norm": 2.1975531578063965,
      "learning_rate": 5.153061224489797e-06,
      "loss": 0.2916,
      "step": 4495
    },
    {
      "epoch": 2.0656,
      "grad_norm": 1.9354606866836548,
      "learning_rate": 5.142857142857143e-06,
      "loss": 0.4395,
      "step": 4496
    },
    {
      "epoch": 2.0658,
      "grad_norm": 3.242257833480835,
      "learning_rate": 5.13265306122449e-06,
      "loss": 0.5785,
      "step": 4497
    },
    {
      "epoch": 2.066,
      "grad_norm": 3.7497684955596924,
      "learning_rate": 5.1224489795918365e-06,
      "loss": 1.5256,
      "step": 4498
    },
    {
      "epoch": 2.0662,
      "grad_norm": 6.796024322509766,
      "learning_rate": 5.112244897959184e-06,
      "loss": 0.4456,
      "step": 4499
    },
    {
      "epoch": 2.0664,
      "grad_norm": 4.2627716064453125,
      "learning_rate": 5.102040816326531e-06,
      "loss": 1.1087,
      "step": 4500
    },
    {
      "epoch": 2.0666,
      "grad_norm": 3.804938316345215,
      "learning_rate": 5.091836734693878e-06,
      "loss": 0.6744,
      "step": 4501
    },
    {
      "epoch": 2.0668,
      "grad_norm": 2.9033427238464355,
      "learning_rate": 5.081632653061225e-06,
      "loss": 1.4458,
      "step": 4502
    },
    {
      "epoch": 2.067,
      "grad_norm": 1.893961787223816,
      "learning_rate": 5.071428571428571e-06,
      "loss": 0.1196,
      "step": 4503
    },
    {
      "epoch": 2.0672,
      "grad_norm": 2.7537388801574707,
      "learning_rate": 5.061224489795919e-06,
      "loss": 0.2561,
      "step": 4504
    },
    {
      "epoch": 2.0674,
      "grad_norm": 2.0070817470550537,
      "learning_rate": 5.0510204081632655e-06,
      "loss": 0.1651,
      "step": 4505
    },
    {
      "epoch": 2.0676,
      "grad_norm": 7.835712909698486,
      "learning_rate": 5.040816326530612e-06,
      "loss": 0.67,
      "step": 4506
    },
    {
      "epoch": 2.0678,
      "grad_norm": 3.468743085861206,
      "learning_rate": 5.03061224489796e-06,
      "loss": 0.2437,
      "step": 4507
    },
    {
      "epoch": 2.068,
      "grad_norm": 8.640397071838379,
      "learning_rate": 5.020408163265306e-06,
      "loss": 1.3277,
      "step": 4508
    },
    {
      "epoch": 2.0682,
      "grad_norm": 9.059866905212402,
      "learning_rate": 5.010204081632653e-06,
      "loss": 1.7499,
      "step": 4509
    },
    {
      "epoch": 2.0684,
      "grad_norm": 4.851001739501953,
      "learning_rate": 5e-06,
      "loss": 0.273,
      "step": 4510
    },
    {
      "epoch": 2.0686,
      "grad_norm": 4.267689228057861,
      "learning_rate": 4.989795918367347e-06,
      "loss": 2.3196,
      "step": 4511
    },
    {
      "epoch": 2.0688,
      "grad_norm": 4.466909885406494,
      "learning_rate": 4.9795918367346945e-06,
      "loss": 1.4522,
      "step": 4512
    },
    {
      "epoch": 2.069,
      "grad_norm": 3.5190610885620117,
      "learning_rate": 4.969387755102041e-06,
      "loss": 1.5381,
      "step": 4513
    },
    {
      "epoch": 2.0692,
      "grad_norm": 7.272828102111816,
      "learning_rate": 4.959183673469388e-06,
      "loss": 4.6036,
      "step": 4514
    },
    {
      "epoch": 2.0694,
      "grad_norm": 4.943753242492676,
      "learning_rate": 4.9489795918367345e-06,
      "loss": 0.8271,
      "step": 4515
    },
    {
      "epoch": 2.0696,
      "grad_norm": 4.36967134475708,
      "learning_rate": 4.938775510204082e-06,
      "loss": 1.4336,
      "step": 4516
    },
    {
      "epoch": 2.0698,
      "grad_norm": 8.297110557556152,
      "learning_rate": 4.9285714285714286e-06,
      "loss": 2.122,
      "step": 4517
    },
    {
      "epoch": 2.07,
      "grad_norm": 7.8115668296813965,
      "learning_rate": 4.918367346938776e-06,
      "loss": 4.7454,
      "step": 4518
    },
    {
      "epoch": 2.0702,
      "grad_norm": 3.2708168029785156,
      "learning_rate": 4.908163265306123e-06,
      "loss": 0.311,
      "step": 4519
    },
    {
      "epoch": 2.0704,
      "grad_norm": 3.0260918140411377,
      "learning_rate": 4.897959183673469e-06,
      "loss": 0.2899,
      "step": 4520
    },
    {
      "epoch": 2.0705999999999998,
      "grad_norm": 1.0097763538360596,
      "learning_rate": 4.887755102040817e-06,
      "loss": 0.0318,
      "step": 4521
    },
    {
      "epoch": 2.0708,
      "grad_norm": 1.5027214288711548,
      "learning_rate": 4.8775510204081635e-06,
      "loss": 0.0214,
      "step": 4522
    },
    {
      "epoch": 2.071,
      "grad_norm": 1.852122187614441,
      "learning_rate": 4.86734693877551e-06,
      "loss": 0.1686,
      "step": 4523
    },
    {
      "epoch": 2.0712,
      "grad_norm": 1.601088047027588,
      "learning_rate": 4.857142857142858e-06,
      "loss": 0.0679,
      "step": 4524
    },
    {
      "epoch": 2.0714,
      "grad_norm": 2.005523681640625,
      "learning_rate": 4.846938775510204e-06,
      "loss": 0.1974,
      "step": 4525
    },
    {
      "epoch": 2.0716,
      "grad_norm": 1.9652719497680664,
      "learning_rate": 4.836734693877551e-06,
      "loss": 0.1106,
      "step": 4526
    },
    {
      "epoch": 2.0718,
      "grad_norm": 0.9379339218139648,
      "learning_rate": 4.826530612244898e-06,
      "loss": 0.0182,
      "step": 4527
    },
    {
      "epoch": 2.072,
      "grad_norm": 1.3320714235305786,
      "learning_rate": 4.816326530612245e-06,
      "loss": 0.0359,
      "step": 4528
    },
    {
      "epoch": 2.0722,
      "grad_norm": 3.7485291957855225,
      "learning_rate": 4.8061224489795925e-06,
      "loss": 0.4026,
      "step": 4529
    },
    {
      "epoch": 2.0724,
      "grad_norm": 0.9966763854026794,
      "learning_rate": 4.795918367346939e-06,
      "loss": 0.0292,
      "step": 4530
    },
    {
      "epoch": 2.0726,
      "grad_norm": 9.170167922973633,
      "learning_rate": 4.785714285714286e-06,
      "loss": 1.6943,
      "step": 4531
    },
    {
      "epoch": 2.0728,
      "grad_norm": 3.7105157375335693,
      "learning_rate": 4.775510204081632e-06,
      "loss": 0.3304,
      "step": 4532
    },
    {
      "epoch": 2.073,
      "grad_norm": 4.878239154815674,
      "learning_rate": 4.76530612244898e-06,
      "loss": 1.7712,
      "step": 4533
    },
    {
      "epoch": 2.0732,
      "grad_norm": 7.267502307891846,
      "learning_rate": 4.7551020408163265e-06,
      "loss": 1.1406,
      "step": 4534
    },
    {
      "epoch": 2.0734,
      "grad_norm": 7.83075475692749,
      "learning_rate": 4.744897959183674e-06,
      "loss": 0.7267,
      "step": 4535
    },
    {
      "epoch": 2.0736,
      "grad_norm": 2.710083484649658,
      "learning_rate": 4.734693877551021e-06,
      "loss": 0.2014,
      "step": 4536
    },
    {
      "epoch": 2.0738,
      "grad_norm": 2.002182960510254,
      "learning_rate": 4.724489795918367e-06,
      "loss": 0.1058,
      "step": 4537
    },
    {
      "epoch": 2.074,
      "grad_norm": 2.4177191257476807,
      "learning_rate": 4.714285714285715e-06,
      "loss": 0.131,
      "step": 4538
    },
    {
      "epoch": 2.0742,
      "grad_norm": 2.5525918006896973,
      "learning_rate": 4.704081632653061e-06,
      "loss": 0.115,
      "step": 4539
    },
    {
      "epoch": 2.0744,
      "grad_norm": 5.234180450439453,
      "learning_rate": 4.693877551020408e-06,
      "loss": 0.3899,
      "step": 4540
    },
    {
      "epoch": 2.0746,
      "grad_norm": 2.046069383621216,
      "learning_rate": 4.6836734693877555e-06,
      "loss": 0.1686,
      "step": 4541
    },
    {
      "epoch": 2.0748,
      "grad_norm": 11.975163459777832,
      "learning_rate": 4.673469387755102e-06,
      "loss": 1.2345,
      "step": 4542
    },
    {
      "epoch": 2.075,
      "grad_norm": 5.11647891998291,
      "learning_rate": 4.663265306122449e-06,
      "loss": 0.7455,
      "step": 4543
    },
    {
      "epoch": 2.0752,
      "grad_norm": 1.9567712545394897,
      "learning_rate": 4.653061224489796e-06,
      "loss": 0.1626,
      "step": 4544
    },
    {
      "epoch": 2.0754,
      "grad_norm": 10.514437675476074,
      "learning_rate": 4.642857142857143e-06,
      "loss": 2.4514,
      "step": 4545
    },
    {
      "epoch": 2.0756,
      "grad_norm": 1.8375921249389648,
      "learning_rate": 4.63265306122449e-06,
      "loss": 0.1867,
      "step": 4546
    },
    {
      "epoch": 2.0758,
      "grad_norm": 10.785018920898438,
      "learning_rate": 4.622448979591837e-06,
      "loss": 1.6582,
      "step": 4547
    },
    {
      "epoch": 2.076,
      "grad_norm": 2.0468430519104004,
      "learning_rate": 4.612244897959184e-06,
      "loss": 0.4524,
      "step": 4548
    },
    {
      "epoch": 2.0762,
      "grad_norm": 4.889520645141602,
      "learning_rate": 4.60204081632653e-06,
      "loss": 1.7329,
      "step": 4549
    },
    {
      "epoch": 2.0764,
      "grad_norm": 3.7147347927093506,
      "learning_rate": 4.591836734693878e-06,
      "loss": 1.6102,
      "step": 4550
    },
    {
      "epoch": 2.0766,
      "grad_norm": 9.234541893005371,
      "learning_rate": 4.5816326530612245e-06,
      "loss": 1.1917,
      "step": 4551
    },
    {
      "epoch": 2.0768,
      "grad_norm": 2.258267641067505,
      "learning_rate": 4.571428571428572e-06,
      "loss": 0.2806,
      "step": 4552
    },
    {
      "epoch": 2.077,
      "grad_norm": 11.382367134094238,
      "learning_rate": 4.561224489795919e-06,
      "loss": 3.6602,
      "step": 4553
    },
    {
      "epoch": 2.0772,
      "grad_norm": 3.3093225955963135,
      "learning_rate": 4.551020408163265e-06,
      "loss": 0.5698,
      "step": 4554
    },
    {
      "epoch": 2.0774,
      "grad_norm": 5.779459476470947,
      "learning_rate": 4.540816326530613e-06,
      "loss": 1.5482,
      "step": 4555
    },
    {
      "epoch": 2.0776,
      "grad_norm": 1.640430212020874,
      "learning_rate": 4.530612244897959e-06,
      "loss": 0.1241,
      "step": 4556
    },
    {
      "epoch": 2.0778,
      "grad_norm": 6.966116428375244,
      "learning_rate": 4.520408163265307e-06,
      "loss": 1.0226,
      "step": 4557
    },
    {
      "epoch": 2.078,
      "grad_norm": 5.381195068359375,
      "learning_rate": 4.5102040816326535e-06,
      "loss": 0.8309,
      "step": 4558
    },
    {
      "epoch": 2.0782,
      "grad_norm": 4.727212429046631,
      "learning_rate": 4.5e-06,
      "loss": 1.6594,
      "step": 4559
    },
    {
      "epoch": 2.0784,
      "grad_norm": 21.576250076293945,
      "learning_rate": 4.489795918367347e-06,
      "loss": 4.3874,
      "step": 4560
    },
    {
      "epoch": 2.0786,
      "grad_norm": 7.968533992767334,
      "learning_rate": 4.479591836734694e-06,
      "loss": 0.782,
      "step": 4561
    },
    {
      "epoch": 2.0788,
      "grad_norm": 10.656570434570312,
      "learning_rate": 4.469387755102041e-06,
      "loss": 3.7518,
      "step": 4562
    },
    {
      "epoch": 2.079,
      "grad_norm": 1.8644603490829468,
      "learning_rate": 4.459183673469388e-06,
      "loss": 0.1694,
      "step": 4563
    },
    {
      "epoch": 2.0792,
      "grad_norm": 11.678762435913086,
      "learning_rate": 4.448979591836735e-06,
      "loss": 1.506,
      "step": 4564
    },
    {
      "epoch": 2.0794,
      "grad_norm": 1.744140386581421,
      "learning_rate": 4.438775510204082e-06,
      "loss": 0.1903,
      "step": 4565
    },
    {
      "epoch": 2.0796,
      "grad_norm": 19.20170021057129,
      "learning_rate": 4.428571428571428e-06,
      "loss": 3.6634,
      "step": 4566
    },
    {
      "epoch": 2.0798,
      "grad_norm": 2.1338698863983154,
      "learning_rate": 4.418367346938776e-06,
      "loss": 0.2085,
      "step": 4567
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.9909071922302246,
      "learning_rate": 4.408163265306122e-06,
      "loss": 0.0245,
      "step": 4568
    },
    {
      "epoch": 2.0802,
      "grad_norm": 5.628470420837402,
      "learning_rate": 4.39795918367347e-06,
      "loss": 1.6193,
      "step": 4569
    },
    {
      "epoch": 2.0804,
      "grad_norm": 3.133054494857788,
      "learning_rate": 4.3877551020408165e-06,
      "loss": 0.2746,
      "step": 4570
    },
    {
      "epoch": 2.0806,
      "grad_norm": 12.851790428161621,
      "learning_rate": 4.377551020408163e-06,
      "loss": 5.1795,
      "step": 4571
    },
    {
      "epoch": 2.0808,
      "grad_norm": 4.939739227294922,
      "learning_rate": 4.367346938775511e-06,
      "loss": 0.5878,
      "step": 4572
    },
    {
      "epoch": 2.081,
      "grad_norm": 8.724700927734375,
      "learning_rate": 4.357142857142857e-06,
      "loss": 1.3007,
      "step": 4573
    },
    {
      "epoch": 2.0812,
      "grad_norm": 3.8831076622009277,
      "learning_rate": 4.346938775510205e-06,
      "loss": 0.443,
      "step": 4574
    },
    {
      "epoch": 2.0814,
      "grad_norm": 2.3261559009552,
      "learning_rate": 4.336734693877551e-06,
      "loss": 0.244,
      "step": 4575
    },
    {
      "epoch": 2.0816,
      "grad_norm": 1.6815943717956543,
      "learning_rate": 4.326530612244898e-06,
      "loss": 0.1314,
      "step": 4576
    },
    {
      "epoch": 2.0818,
      "grad_norm": 3.9276928901672363,
      "learning_rate": 4.316326530612245e-06,
      "loss": 0.4005,
      "step": 4577
    },
    {
      "epoch": 2.082,
      "grad_norm": 5.823033809661865,
      "learning_rate": 4.306122448979592e-06,
      "loss": 1.0409,
      "step": 4578
    },
    {
      "epoch": 2.0822,
      "grad_norm": 5.226304054260254,
      "learning_rate": 4.295918367346939e-06,
      "loss": 0.6593,
      "step": 4579
    },
    {
      "epoch": 2.0824,
      "grad_norm": 53.483036041259766,
      "learning_rate": 4.285714285714286e-06,
      "loss": 4.1761,
      "step": 4580
    },
    {
      "epoch": 2.0826000000000002,
      "grad_norm": 3.3650310039520264,
      "learning_rate": 4.275510204081633e-06,
      "loss": 1.4406,
      "step": 4581
    },
    {
      "epoch": 2.0828,
      "grad_norm": 16.876192092895508,
      "learning_rate": 4.26530612244898e-06,
      "loss": 3.6484,
      "step": 4582
    },
    {
      "epoch": 2.083,
      "grad_norm": 7.014874458312988,
      "learning_rate": 4.255102040816326e-06,
      "loss": 5.0853,
      "step": 4583
    },
    {
      "epoch": 2.0832,
      "grad_norm": 3.952516555786133,
      "learning_rate": 4.244897959183674e-06,
      "loss": 0.2555,
      "step": 4584
    },
    {
      "epoch": 2.0834,
      "grad_norm": 5.884845733642578,
      "learning_rate": 4.23469387755102e-06,
      "loss": 0.8335,
      "step": 4585
    },
    {
      "epoch": 2.0836,
      "grad_norm": 5.378721714019775,
      "learning_rate": 4.224489795918368e-06,
      "loss": 0.2679,
      "step": 4586
    },
    {
      "epoch": 2.0838,
      "grad_norm": 4.5112385749816895,
      "learning_rate": 4.2142857142857145e-06,
      "loss": 0.5756,
      "step": 4587
    },
    {
      "epoch": 2.084,
      "grad_norm": 1.023730754852295,
      "learning_rate": 4.204081632653061e-06,
      "loss": 0.0301,
      "step": 4588
    },
    {
      "epoch": 2.0842,
      "grad_norm": 4.677689552307129,
      "learning_rate": 4.193877551020409e-06,
      "loss": 0.5565,
      "step": 4589
    },
    {
      "epoch": 2.0844,
      "grad_norm": 2.3065829277038574,
      "learning_rate": 4.183673469387755e-06,
      "loss": 0.2316,
      "step": 4590
    },
    {
      "epoch": 2.0846,
      "grad_norm": 3.4042725563049316,
      "learning_rate": 4.173469387755103e-06,
      "loss": 1.5457,
      "step": 4591
    },
    {
      "epoch": 2.0848,
      "grad_norm": 3.8823788166046143,
      "learning_rate": 4.163265306122449e-06,
      "loss": 0.7111,
      "step": 4592
    },
    {
      "epoch": 2.085,
      "grad_norm": 1.4060784578323364,
      "learning_rate": 4.153061224489796e-06,
      "loss": 0.0451,
      "step": 4593
    },
    {
      "epoch": 2.0852,
      "grad_norm": 3.823343515396118,
      "learning_rate": 4.142857142857143e-06,
      "loss": 1.6003,
      "step": 4594
    },
    {
      "epoch": 2.0854,
      "grad_norm": 2.5086982250213623,
      "learning_rate": 4.13265306122449e-06,
      "loss": 0.1459,
      "step": 4595
    },
    {
      "epoch": 2.0856,
      "grad_norm": 3.642223834991455,
      "learning_rate": 4.122448979591837e-06,
      "loss": 0.3265,
      "step": 4596
    },
    {
      "epoch": 2.0858,
      "grad_norm": 6.024643421173096,
      "learning_rate": 4.112244897959184e-06,
      "loss": 0.7844,
      "step": 4597
    },
    {
      "epoch": 2.086,
      "grad_norm": 3.7316811084747314,
      "learning_rate": 4.102040816326531e-06,
      "loss": 0.4592,
      "step": 4598
    },
    {
      "epoch": 2.0862,
      "grad_norm": 10.370810508728027,
      "learning_rate": 4.0918367346938775e-06,
      "loss": 2.3321,
      "step": 4599
    },
    {
      "epoch": 2.0864,
      "grad_norm": 10.592586517333984,
      "learning_rate": 4.081632653061224e-06,
      "loss": 1.8294,
      "step": 4600
    },
    {
      "epoch": 2.0866,
      "grad_norm": 10.37260627746582,
      "learning_rate": 4.071428571428572e-06,
      "loss": 2.6426,
      "step": 4601
    },
    {
      "epoch": 2.0868,
      "grad_norm": 7.045802593231201,
      "learning_rate": 4.061224489795918e-06,
      "loss": 2.4168,
      "step": 4602
    },
    {
      "epoch": 2.087,
      "grad_norm": 2.7917001247406006,
      "learning_rate": 4.051020408163266e-06,
      "loss": 0.4655,
      "step": 4603
    },
    {
      "epoch": 2.0872,
      "grad_norm": 8.666906356811523,
      "learning_rate": 4.040816326530612e-06,
      "loss": 3.4238,
      "step": 4604
    },
    {
      "epoch": 2.0874,
      "grad_norm": 2.353450298309326,
      "learning_rate": 4.030612244897959e-06,
      "loss": 0.4478,
      "step": 4605
    },
    {
      "epoch": 2.0876,
      "grad_norm": 3.345656156539917,
      "learning_rate": 4.0204081632653065e-06,
      "loss": 0.801,
      "step": 4606
    },
    {
      "epoch": 2.0878,
      "grad_norm": 3.443507432937622,
      "learning_rate": 4.010204081632653e-06,
      "loss": 1.5316,
      "step": 4607
    },
    {
      "epoch": 2.088,
      "grad_norm": 4.042265892028809,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.6096,
      "step": 4608
    },
    {
      "epoch": 2.0882,
      "grad_norm": 2.2342569828033447,
      "learning_rate": 3.989795918367347e-06,
      "loss": 0.3755,
      "step": 4609
    },
    {
      "epoch": 2.0884,
      "grad_norm": 4.226164817810059,
      "learning_rate": 3.979591836734694e-06,
      "loss": 0.5237,
      "step": 4610
    },
    {
      "epoch": 2.0886,
      "grad_norm": 2.139458417892456,
      "learning_rate": 3.969387755102041e-06,
      "loss": 0.2093,
      "step": 4611
    },
    {
      "epoch": 2.0888,
      "grad_norm": 2.6331608295440674,
      "learning_rate": 3.959183673469388e-06,
      "loss": 0.2545,
      "step": 4612
    },
    {
      "epoch": 2.089,
      "grad_norm": 1.0588756799697876,
      "learning_rate": 3.948979591836735e-06,
      "loss": 0.0267,
      "step": 4613
    },
    {
      "epoch": 2.0892,
      "grad_norm": 3.786996603012085,
      "learning_rate": 3.938775510204082e-06,
      "loss": 0.4619,
      "step": 4614
    },
    {
      "epoch": 2.0894,
      "grad_norm": 2.729401111602783,
      "learning_rate": 3.928571428571429e-06,
      "loss": 0.2752,
      "step": 4615
    },
    {
      "epoch": 2.0896,
      "grad_norm": 4.36923885345459,
      "learning_rate": 3.9183673469387755e-06,
      "loss": 0.3585,
      "step": 4616
    },
    {
      "epoch": 2.0898,
      "grad_norm": 8.847342491149902,
      "learning_rate": 3.908163265306122e-06,
      "loss": 0.8233,
      "step": 4617
    },
    {
      "epoch": 2.09,
      "grad_norm": 7.369173526763916,
      "learning_rate": 3.89795918367347e-06,
      "loss": 1.1151,
      "step": 4618
    },
    {
      "epoch": 2.0902,
      "grad_norm": 20.38437843322754,
      "learning_rate": 3.887755102040816e-06,
      "loss": 2.2948,
      "step": 4619
    },
    {
      "epoch": 2.0904,
      "grad_norm": 2.462642192840576,
      "learning_rate": 3.877551020408164e-06,
      "loss": 0.2833,
      "step": 4620
    },
    {
      "epoch": 2.0906,
      "grad_norm": 9.721959114074707,
      "learning_rate": 3.86734693877551e-06,
      "loss": 2.6022,
      "step": 4621
    },
    {
      "epoch": 2.0908,
      "grad_norm": 5.129932880401611,
      "learning_rate": 3.857142857142857e-06,
      "loss": 0.8727,
      "step": 4622
    },
    {
      "epoch": 2.091,
      "grad_norm": 2.118366003036499,
      "learning_rate": 3.8469387755102045e-06,
      "loss": 0.1915,
      "step": 4623
    },
    {
      "epoch": 2.0912,
      "grad_norm": 8.516611099243164,
      "learning_rate": 3.836734693877551e-06,
      "loss": 0.7175,
      "step": 4624
    },
    {
      "epoch": 2.0914,
      "grad_norm": 1.7975201606750488,
      "learning_rate": 3.826530612244899e-06,
      "loss": 0.2119,
      "step": 4625
    },
    {
      "epoch": 2.0916,
      "grad_norm": 2.179056167602539,
      "learning_rate": 3.816326530612245e-06,
      "loss": 0.1473,
      "step": 4626
    },
    {
      "epoch": 2.0918,
      "grad_norm": 3.825058698654175,
      "learning_rate": 3.806122448979592e-06,
      "loss": 0.2241,
      "step": 4627
    },
    {
      "epoch": 2.092,
      "grad_norm": 3.3491287231445312,
      "learning_rate": 3.7959183673469385e-06,
      "loss": 0.3319,
      "step": 4628
    },
    {
      "epoch": 2.0922,
      "grad_norm": 8.091402053833008,
      "learning_rate": 3.785714285714286e-06,
      "loss": 1.0905,
      "step": 4629
    },
    {
      "epoch": 2.0924,
      "grad_norm": 0.9689975380897522,
      "learning_rate": 3.775510204081633e-06,
      "loss": 0.0246,
      "step": 4630
    },
    {
      "epoch": 2.0926,
      "grad_norm": 1.3830915689468384,
      "learning_rate": 3.7653061224489797e-06,
      "loss": 0.0428,
      "step": 4631
    },
    {
      "epoch": 2.0928,
      "grad_norm": 2.716122627258301,
      "learning_rate": 3.7551020408163268e-06,
      "loss": 0.5996,
      "step": 4632
    },
    {
      "epoch": 2.093,
      "grad_norm": 3.4472105503082275,
      "learning_rate": 3.7448979591836734e-06,
      "loss": 0.1459,
      "step": 4633
    },
    {
      "epoch": 2.0932,
      "grad_norm": 4.0314788818359375,
      "learning_rate": 3.73469387755102e-06,
      "loss": 1.5885,
      "step": 4634
    },
    {
      "epoch": 2.0934,
      "grad_norm": 8.527509689331055,
      "learning_rate": 3.724489795918368e-06,
      "loss": 0.8418,
      "step": 4635
    },
    {
      "epoch": 2.0936,
      "grad_norm": 5.430739402770996,
      "learning_rate": 3.7142857142857146e-06,
      "loss": 0.7711,
      "step": 4636
    },
    {
      "epoch": 2.0938,
      "grad_norm": 4.720272064208984,
      "learning_rate": 3.7040816326530612e-06,
      "loss": 0.7726,
      "step": 4637
    },
    {
      "epoch": 2.094,
      "grad_norm": 3.743134021759033,
      "learning_rate": 3.6938775510204083e-06,
      "loss": 0.626,
      "step": 4638
    },
    {
      "epoch": 2.0942,
      "grad_norm": 1.433454155921936,
      "learning_rate": 3.683673469387755e-06,
      "loss": 0.105,
      "step": 4639
    },
    {
      "epoch": 2.0944,
      "grad_norm": 1.5286427736282349,
      "learning_rate": 3.6734693877551024e-06,
      "loss": 0.0555,
      "step": 4640
    },
    {
      "epoch": 2.0946,
      "grad_norm": 2.7598135471343994,
      "learning_rate": 3.6632653061224495e-06,
      "loss": 0.4154,
      "step": 4641
    },
    {
      "epoch": 2.0948,
      "grad_norm": 1.687098741531372,
      "learning_rate": 3.653061224489796e-06,
      "loss": 0.0738,
      "step": 4642
    },
    {
      "epoch": 2.095,
      "grad_norm": 5.211580276489258,
      "learning_rate": 3.642857142857143e-06,
      "loss": 1.4158,
      "step": 4643
    },
    {
      "epoch": 2.0952,
      "grad_norm": 7.764120578765869,
      "learning_rate": 3.63265306122449e-06,
      "loss": 0.5652,
      "step": 4644
    },
    {
      "epoch": 2.0954,
      "grad_norm": 0.9991572499275208,
      "learning_rate": 3.6224489795918365e-06,
      "loss": 0.0194,
      "step": 4645
    },
    {
      "epoch": 2.0956,
      "grad_norm": 15.356642723083496,
      "learning_rate": 3.612244897959184e-06,
      "loss": 1.9036,
      "step": 4646
    },
    {
      "epoch": 2.0958,
      "grad_norm": 2.4500508308410645,
      "learning_rate": 3.602040816326531e-06,
      "loss": 0.1955,
      "step": 4647
    },
    {
      "epoch": 2.096,
      "grad_norm": 3.684885025024414,
      "learning_rate": 3.5918367346938777e-06,
      "loss": 0.3032,
      "step": 4648
    },
    {
      "epoch": 2.0962,
      "grad_norm": 3.018481492996216,
      "learning_rate": 3.5816326530612247e-06,
      "loss": 0.3123,
      "step": 4649
    },
    {
      "epoch": 2.0964,
      "grad_norm": 5.848809719085693,
      "learning_rate": 3.5714285714285714e-06,
      "loss": 1.9129,
      "step": 4650
    },
    {
      "epoch": 2.0966,
      "grad_norm": 1.8316609859466553,
      "learning_rate": 3.5612244897959184e-06,
      "loss": 0.1484,
      "step": 4651
    },
    {
      "epoch": 2.0968,
      "grad_norm": 6.228904724121094,
      "learning_rate": 3.551020408163266e-06,
      "loss": 5.0758,
      "step": 4652
    },
    {
      "epoch": 2.097,
      "grad_norm": 3.67339825630188,
      "learning_rate": 3.5408163265306125e-06,
      "loss": 0.6337,
      "step": 4653
    },
    {
      "epoch": 2.0972,
      "grad_norm": 8.086236953735352,
      "learning_rate": 3.530612244897959e-06,
      "loss": 0.8362,
      "step": 4654
    },
    {
      "epoch": 2.0974,
      "grad_norm": 4.554009437561035,
      "learning_rate": 3.5204081632653062e-06,
      "loss": 0.6495,
      "step": 4655
    },
    {
      "epoch": 2.0976,
      "grad_norm": 18.859203338623047,
      "learning_rate": 3.510204081632653e-06,
      "loss": 2.0373,
      "step": 4656
    },
    {
      "epoch": 2.0978,
      "grad_norm": 3.6999504566192627,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 1.0081,
      "step": 4657
    },
    {
      "epoch": 2.098,
      "grad_norm": 2.6290180683135986,
      "learning_rate": 3.4897959183673474e-06,
      "loss": 0.594,
      "step": 4658
    },
    {
      "epoch": 2.0982,
      "grad_norm": 4.123224258422852,
      "learning_rate": 3.479591836734694e-06,
      "loss": 2.1693,
      "step": 4659
    },
    {
      "epoch": 2.0984,
      "grad_norm": 3.309077024459839,
      "learning_rate": 3.469387755102041e-06,
      "loss": 1.3687,
      "step": 4660
    },
    {
      "epoch": 2.0986,
      "grad_norm": 8.282438278198242,
      "learning_rate": 3.4591836734693878e-06,
      "loss": 2.6114,
      "step": 4661
    },
    {
      "epoch": 2.0987999999999998,
      "grad_norm": 2.382996082305908,
      "learning_rate": 3.4489795918367344e-06,
      "loss": 0.2528,
      "step": 4662
    },
    {
      "epoch": 2.099,
      "grad_norm": 11.994235038757324,
      "learning_rate": 3.438775510204082e-06,
      "loss": 3.0576,
      "step": 4663
    },
    {
      "epoch": 2.0992,
      "grad_norm": 2.2513766288757324,
      "learning_rate": 3.428571428571429e-06,
      "loss": 0.0989,
      "step": 4664
    },
    {
      "epoch": 2.0994,
      "grad_norm": 1.784716248512268,
      "learning_rate": 3.4183673469387756e-06,
      "loss": 0.2105,
      "step": 4665
    },
    {
      "epoch": 2.0996,
      "grad_norm": 3.966416120529175,
      "learning_rate": 3.4081632653061227e-06,
      "loss": 0.4555,
      "step": 4666
    },
    {
      "epoch": 2.0998,
      "grad_norm": 2.5667736530303955,
      "learning_rate": 3.3979591836734693e-06,
      "loss": 0.2572,
      "step": 4667
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.7077491283416748,
      "learning_rate": 3.3877551020408164e-06,
      "loss": 0.1606,
      "step": 4668
    },
    {
      "epoch": 2.1002,
      "grad_norm": 3.992199659347534,
      "learning_rate": 3.377551020408164e-06,
      "loss": 0.479,
      "step": 4669
    },
    {
      "epoch": 2.1004,
      "grad_norm": 2.081674575805664,
      "learning_rate": 3.3673469387755105e-06,
      "loss": 0.1013,
      "step": 4670
    },
    {
      "epoch": 2.1006,
      "grad_norm": 2.2576310634613037,
      "learning_rate": 3.357142857142857e-06,
      "loss": 0.2264,
      "step": 4671
    },
    {
      "epoch": 2.1008,
      "grad_norm": 4.365758895874023,
      "learning_rate": 3.346938775510204e-06,
      "loss": 0.5402,
      "step": 4672
    },
    {
      "epoch": 2.101,
      "grad_norm": 4.182484149932861,
      "learning_rate": 3.336734693877551e-06,
      "loss": 0.236,
      "step": 4673
    },
    {
      "epoch": 2.1012,
      "grad_norm": 1.898068904876709,
      "learning_rate": 3.3265306122448983e-06,
      "loss": 0.0834,
      "step": 4674
    },
    {
      "epoch": 2.1014,
      "grad_norm": 2.2708635330200195,
      "learning_rate": 3.3163265306122454e-06,
      "loss": 0.1356,
      "step": 4675
    },
    {
      "epoch": 2.1016,
      "grad_norm": 6.384464263916016,
      "learning_rate": 3.306122448979592e-06,
      "loss": 1.4563,
      "step": 4676
    },
    {
      "epoch": 2.1018,
      "grad_norm": 4.844964981079102,
      "learning_rate": 3.295918367346939e-06,
      "loss": 0.678,
      "step": 4677
    },
    {
      "epoch": 2.102,
      "grad_norm": 9.251545906066895,
      "learning_rate": 3.2857142857142857e-06,
      "loss": 3.3184,
      "step": 4678
    },
    {
      "epoch": 2.1022,
      "grad_norm": 3.1075327396392822,
      "learning_rate": 3.2755102040816324e-06,
      "loss": 0.1363,
      "step": 4679
    },
    {
      "epoch": 2.1024,
      "grad_norm": 5.791533946990967,
      "learning_rate": 3.26530612244898e-06,
      "loss": 1.0356,
      "step": 4680
    },
    {
      "epoch": 2.1026,
      "grad_norm": 7.99643611907959,
      "learning_rate": 3.255102040816327e-06,
      "loss": 0.6833,
      "step": 4681
    },
    {
      "epoch": 2.1028000000000002,
      "grad_norm": 9.739289283752441,
      "learning_rate": 3.2448979591836735e-06,
      "loss": 2.835,
      "step": 4682
    },
    {
      "epoch": 2.103,
      "grad_norm": 1.0964337587356567,
      "learning_rate": 3.2346938775510206e-06,
      "loss": 0.0257,
      "step": 4683
    },
    {
      "epoch": 2.1032,
      "grad_norm": 1.8878360986709595,
      "learning_rate": 3.2244897959183672e-06,
      "loss": 0.1182,
      "step": 4684
    },
    {
      "epoch": 2.1034,
      "grad_norm": 2.076249837875366,
      "learning_rate": 3.2142857142857143e-06,
      "loss": 0.1854,
      "step": 4685
    },
    {
      "epoch": 2.1036,
      "grad_norm": 1.6759300231933594,
      "learning_rate": 3.204081632653062e-06,
      "loss": 0.1459,
      "step": 4686
    },
    {
      "epoch": 2.1038,
      "grad_norm": 1.6163229942321777,
      "learning_rate": 3.1938775510204084e-06,
      "loss": 0.1178,
      "step": 4687
    },
    {
      "epoch": 2.104,
      "grad_norm": 0.738377571105957,
      "learning_rate": 3.183673469387755e-06,
      "loss": 0.0223,
      "step": 4688
    },
    {
      "epoch": 2.1042,
      "grad_norm": 6.086853981018066,
      "learning_rate": 3.173469387755102e-06,
      "loss": 0.5203,
      "step": 4689
    },
    {
      "epoch": 2.1044,
      "grad_norm": 1.979962706565857,
      "learning_rate": 3.1632653061224488e-06,
      "loss": 0.3415,
      "step": 4690
    },
    {
      "epoch": 2.1046,
      "grad_norm": 9.70959186553955,
      "learning_rate": 3.1530612244897963e-06,
      "loss": 1.558,
      "step": 4691
    },
    {
      "epoch": 2.1048,
      "grad_norm": 2.2602646350860596,
      "learning_rate": 3.1428571428571433e-06,
      "loss": 0.2151,
      "step": 4692
    },
    {
      "epoch": 2.105,
      "grad_norm": 3.5727615356445312,
      "learning_rate": 3.13265306122449e-06,
      "loss": 1.6043,
      "step": 4693
    },
    {
      "epoch": 2.1052,
      "grad_norm": 5.286078929901123,
      "learning_rate": 3.122448979591837e-06,
      "loss": 0.828,
      "step": 4694
    },
    {
      "epoch": 2.1054,
      "grad_norm": 1.0418708324432373,
      "learning_rate": 3.1122448979591837e-06,
      "loss": 0.0245,
      "step": 4695
    },
    {
      "epoch": 2.1056,
      "grad_norm": 10.79141902923584,
      "learning_rate": 3.1020408163265307e-06,
      "loss": 2.8219,
      "step": 4696
    },
    {
      "epoch": 2.1058,
      "grad_norm": 5.624372959136963,
      "learning_rate": 3.0918367346938778e-06,
      "loss": 0.3448,
      "step": 4697
    },
    {
      "epoch": 2.106,
      "grad_norm": 3.7493929862976074,
      "learning_rate": 3.0816326530612244e-06,
      "loss": 1.5757,
      "step": 4698
    },
    {
      "epoch": 2.1062,
      "grad_norm": 7.292983055114746,
      "learning_rate": 3.0714285714285715e-06,
      "loss": 1.0822,
      "step": 4699
    },
    {
      "epoch": 2.1064,
      "grad_norm": 6.196479320526123,
      "learning_rate": 3.0612244897959185e-06,
      "loss": 5.0199,
      "step": 4700
    },
    {
      "epoch": 2.1066,
      "grad_norm": 1.9192054271697998,
      "learning_rate": 3.0510204081632656e-06,
      "loss": 0.1834,
      "step": 4701
    },
    {
      "epoch": 2.1068,
      "grad_norm": 4.074975490570068,
      "learning_rate": 3.0408163265306122e-06,
      "loss": 0.5157,
      "step": 4702
    },
    {
      "epoch": 2.107,
      "grad_norm": 1.673568606376648,
      "learning_rate": 3.0306122448979593e-06,
      "loss": 0.0709,
      "step": 4703
    },
    {
      "epoch": 2.1072,
      "grad_norm": 3.941925048828125,
      "learning_rate": 3.0204081632653064e-06,
      "loss": 0.125,
      "step": 4704
    },
    {
      "epoch": 2.1074,
      "grad_norm": 9.043163299560547,
      "learning_rate": 3.010204081632653e-06,
      "loss": 0.4157,
      "step": 4705
    },
    {
      "epoch": 2.1076,
      "grad_norm": 1.8842334747314453,
      "learning_rate": 3e-06,
      "loss": 0.134,
      "step": 4706
    },
    {
      "epoch": 2.1078,
      "grad_norm": 14.688392639160156,
      "learning_rate": 2.989795918367347e-06,
      "loss": 1.8615,
      "step": 4707
    },
    {
      "epoch": 2.108,
      "grad_norm": 1.6997333765029907,
      "learning_rate": 2.979591836734694e-06,
      "loss": 0.115,
      "step": 4708
    },
    {
      "epoch": 2.1082,
      "grad_norm": 5.09561014175415,
      "learning_rate": 2.969387755102041e-06,
      "loss": 0.8584,
      "step": 4709
    },
    {
      "epoch": 2.1084,
      "grad_norm": 1.824609637260437,
      "learning_rate": 2.959183673469388e-06,
      "loss": 0.1815,
      "step": 4710
    },
    {
      "epoch": 2.1086,
      "grad_norm": 12.072440147399902,
      "learning_rate": 2.948979591836735e-06,
      "loss": 3.6057,
      "step": 4711
    },
    {
      "epoch": 2.1088,
      "grad_norm": 1.943498969078064,
      "learning_rate": 2.9387755102040816e-06,
      "loss": 0.0558,
      "step": 4712
    },
    {
      "epoch": 2.109,
      "grad_norm": 10.319755554199219,
      "learning_rate": 2.9285714285714287e-06,
      "loss": 1.573,
      "step": 4713
    },
    {
      "epoch": 2.1092,
      "grad_norm": 1.8012566566467285,
      "learning_rate": 2.9183673469387757e-06,
      "loss": 0.1308,
      "step": 4714
    },
    {
      "epoch": 2.1094,
      "grad_norm": 1.4766422510147095,
      "learning_rate": 2.9081632653061224e-06,
      "loss": 0.1336,
      "step": 4715
    },
    {
      "epoch": 2.1096,
      "grad_norm": 1.7192660570144653,
      "learning_rate": 2.8979591836734694e-06,
      "loss": 0.1554,
      "step": 4716
    },
    {
      "epoch": 2.1098,
      "grad_norm": 2.284416913986206,
      "learning_rate": 2.8877551020408165e-06,
      "loss": 0.2452,
      "step": 4717
    },
    {
      "epoch": 2.11,
      "grad_norm": 2.098954200744629,
      "learning_rate": 2.8775510204081636e-06,
      "loss": 0.1814,
      "step": 4718
    },
    {
      "epoch": 2.1102,
      "grad_norm": 4.4768967628479,
      "learning_rate": 2.86734693877551e-06,
      "loss": 0.6873,
      "step": 4719
    },
    {
      "epoch": 2.1104,
      "grad_norm": 4.6320109367370605,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.9844,
      "step": 4720
    },
    {
      "epoch": 2.1106,
      "grad_norm": 7.1224365234375,
      "learning_rate": 2.8469387755102043e-06,
      "loss": 0.7156,
      "step": 4721
    },
    {
      "epoch": 2.1108,
      "grad_norm": 9.049817085266113,
      "learning_rate": 2.8367346938775514e-06,
      "loss": 3.3146,
      "step": 4722
    },
    {
      "epoch": 2.111,
      "grad_norm": 2.0975863933563232,
      "learning_rate": 2.826530612244898e-06,
      "loss": 0.4132,
      "step": 4723
    },
    {
      "epoch": 2.1112,
      "grad_norm": 2.4252097606658936,
      "learning_rate": 2.816326530612245e-06,
      "loss": 0.2678,
      "step": 4724
    },
    {
      "epoch": 2.1114,
      "grad_norm": 3.6085386276245117,
      "learning_rate": 2.806122448979592e-06,
      "loss": 1.6607,
      "step": 4725
    },
    {
      "epoch": 2.1116,
      "grad_norm": 3.3134396076202393,
      "learning_rate": 2.7959183673469388e-06,
      "loss": 0.5871,
      "step": 4726
    },
    {
      "epoch": 2.1118,
      "grad_norm": 1.4712363481521606,
      "learning_rate": 2.785714285714286e-06,
      "loss": 0.0912,
      "step": 4727
    },
    {
      "epoch": 2.112,
      "grad_norm": 6.200575351715088,
      "learning_rate": 2.775510204081633e-06,
      "loss": 0.9258,
      "step": 4728
    },
    {
      "epoch": 2.1122,
      "grad_norm": 1.3659968376159668,
      "learning_rate": 2.7653061224489795e-06,
      "loss": 0.0313,
      "step": 4729
    },
    {
      "epoch": 2.1124,
      "grad_norm": 10.170721054077148,
      "learning_rate": 2.7551020408163266e-06,
      "loss": 2.3293,
      "step": 4730
    },
    {
      "epoch": 2.1126,
      "grad_norm": 7.839609622955322,
      "learning_rate": 2.7448979591836737e-06,
      "loss": 0.6248,
      "step": 4731
    },
    {
      "epoch": 2.1128,
      "grad_norm": 4.498930931091309,
      "learning_rate": 2.7346938775510203e-06,
      "loss": 0.562,
      "step": 4732
    },
    {
      "epoch": 2.113,
      "grad_norm": 18.936641693115234,
      "learning_rate": 2.7244897959183674e-06,
      "loss": 1.9553,
      "step": 4733
    },
    {
      "epoch": 2.1132,
      "grad_norm": 2.0487070083618164,
      "learning_rate": 2.7142857142857144e-06,
      "loss": 0.1837,
      "step": 4734
    },
    {
      "epoch": 2.1134,
      "grad_norm": 3.12445330619812,
      "learning_rate": 2.7040816326530615e-06,
      "loss": 0.4411,
      "step": 4735
    },
    {
      "epoch": 2.1136,
      "grad_norm": 1.5472902059555054,
      "learning_rate": 2.693877551020408e-06,
      "loss": 0.0534,
      "step": 4736
    },
    {
      "epoch": 2.1138,
      "grad_norm": 4.787519931793213,
      "learning_rate": 2.683673469387755e-06,
      "loss": 1.3217,
      "step": 4737
    },
    {
      "epoch": 2.114,
      "grad_norm": 1.6154824495315552,
      "learning_rate": 2.6734693877551023e-06,
      "loss": 0.0597,
      "step": 4738
    },
    {
      "epoch": 2.1142,
      "grad_norm": 2.8686463832855225,
      "learning_rate": 2.6632653061224493e-06,
      "loss": 0.304,
      "step": 4739
    },
    {
      "epoch": 2.1144,
      "grad_norm": 5.769339084625244,
      "learning_rate": 2.653061224489796e-06,
      "loss": 0.9847,
      "step": 4740
    },
    {
      "epoch": 2.1146,
      "grad_norm": 6.174430847167969,
      "learning_rate": 2.642857142857143e-06,
      "loss": 1.4963,
      "step": 4741
    },
    {
      "epoch": 2.1148,
      "grad_norm": 11.479657173156738,
      "learning_rate": 2.63265306122449e-06,
      "loss": 3.2436,
      "step": 4742
    },
    {
      "epoch": 2.115,
      "grad_norm": 5.433708190917969,
      "learning_rate": 2.6224489795918367e-06,
      "loss": 0.6161,
      "step": 4743
    },
    {
      "epoch": 2.1152,
      "grad_norm": 6.422898292541504,
      "learning_rate": 2.612244897959184e-06,
      "loss": 0.6999,
      "step": 4744
    },
    {
      "epoch": 2.1154,
      "grad_norm": 1.6894084215164185,
      "learning_rate": 2.602040816326531e-06,
      "loss": 0.0894,
      "step": 4745
    },
    {
      "epoch": 2.1156,
      "grad_norm": 5.903936862945557,
      "learning_rate": 2.5918367346938775e-06,
      "loss": 0.7027,
      "step": 4746
    },
    {
      "epoch": 2.1158,
      "grad_norm": 2.2160937786102295,
      "learning_rate": 2.5816326530612246e-06,
      "loss": 0.1945,
      "step": 4747
    },
    {
      "epoch": 2.116,
      "grad_norm": 8.383194923400879,
      "learning_rate": 2.5714285714285716e-06,
      "loss": 2.4187,
      "step": 4748
    },
    {
      "epoch": 2.1162,
      "grad_norm": 1.5930655002593994,
      "learning_rate": 2.5612244897959183e-06,
      "loss": 0.1469,
      "step": 4749
    },
    {
      "epoch": 2.1164,
      "grad_norm": 3.298297643661499,
      "learning_rate": 2.5510204081632653e-06,
      "loss": 0.5827,
      "step": 4750
    },
    {
      "epoch": 2.1166,
      "grad_norm": 2.2217905521392822,
      "learning_rate": 2.5408163265306124e-06,
      "loss": 0.4792,
      "step": 4751
    },
    {
      "epoch": 2.1168,
      "grad_norm": 2.5096020698547363,
      "learning_rate": 2.5306122448979594e-06,
      "loss": 0.267,
      "step": 4752
    },
    {
      "epoch": 2.117,
      "grad_norm": 3.6636083126068115,
      "learning_rate": 2.520408163265306e-06,
      "loss": 1.53,
      "step": 4753
    },
    {
      "epoch": 2.1172,
      "grad_norm": 3.336641788482666,
      "learning_rate": 2.510204081632653e-06,
      "loss": 0.5622,
      "step": 4754
    },
    {
      "epoch": 2.1174,
      "grad_norm": 8.178468704223633,
      "learning_rate": 2.5e-06,
      "loss": 1.1144,
      "step": 4755
    },
    {
      "epoch": 2.1176,
      "grad_norm": 1.8707865476608276,
      "learning_rate": 2.4897959183673473e-06,
      "loss": 0.1159,
      "step": 4756
    },
    {
      "epoch": 2.1178,
      "grad_norm": 10.267629623413086,
      "learning_rate": 2.479591836734694e-06,
      "loss": 2.3957,
      "step": 4757
    },
    {
      "epoch": 2.118,
      "grad_norm": 2.2508950233459473,
      "learning_rate": 2.469387755102041e-06,
      "loss": 0.1804,
      "step": 4758
    },
    {
      "epoch": 2.1182,
      "grad_norm": 2.4401466846466064,
      "learning_rate": 2.459183673469388e-06,
      "loss": 0.2665,
      "step": 4759
    },
    {
      "epoch": 2.1184,
      "grad_norm": 3.918395757675171,
      "learning_rate": 2.4489795918367347e-06,
      "loss": 0.1593,
      "step": 4760
    },
    {
      "epoch": 2.1186,
      "grad_norm": 0.7459062933921814,
      "learning_rate": 2.4387755102040817e-06,
      "loss": 0.0159,
      "step": 4761
    },
    {
      "epoch": 2.1188,
      "grad_norm": 1.8292782306671143,
      "learning_rate": 2.428571428571429e-06,
      "loss": 0.1906,
      "step": 4762
    },
    {
      "epoch": 2.1189999999999998,
      "grad_norm": 2.560241460800171,
      "learning_rate": 2.4183673469387754e-06,
      "loss": 0.2444,
      "step": 4763
    },
    {
      "epoch": 2.1192,
      "grad_norm": 3.5728046894073486,
      "learning_rate": 2.4081632653061225e-06,
      "loss": 0.2309,
      "step": 4764
    },
    {
      "epoch": 2.1194,
      "grad_norm": 5.053526401519775,
      "learning_rate": 2.3979591836734696e-06,
      "loss": 0.9814,
      "step": 4765
    },
    {
      "epoch": 2.1196,
      "grad_norm": 5.086238384246826,
      "learning_rate": 2.387755102040816e-06,
      "loss": 0.3073,
      "step": 4766
    },
    {
      "epoch": 2.1198,
      "grad_norm": 5.67470121383667,
      "learning_rate": 2.3775510204081633e-06,
      "loss": 2.2228,
      "step": 4767
    },
    {
      "epoch": 2.12,
      "grad_norm": 1.1746867895126343,
      "learning_rate": 2.3673469387755103e-06,
      "loss": 0.0368,
      "step": 4768
    },
    {
      "epoch": 2.1202,
      "grad_norm": 9.472219467163086,
      "learning_rate": 2.3571428571428574e-06,
      "loss": 3.4311,
      "step": 4769
    },
    {
      "epoch": 2.1204,
      "grad_norm": 1.9404643774032593,
      "learning_rate": 2.346938775510204e-06,
      "loss": 0.1586,
      "step": 4770
    },
    {
      "epoch": 2.1206,
      "grad_norm": 6.1719255447387695,
      "learning_rate": 2.336734693877551e-06,
      "loss": 1.236,
      "step": 4771
    },
    {
      "epoch": 2.1208,
      "grad_norm": 3.216226816177368,
      "learning_rate": 2.326530612244898e-06,
      "loss": 0.577,
      "step": 4772
    },
    {
      "epoch": 2.121,
      "grad_norm": 12.40564250946045,
      "learning_rate": 2.316326530612245e-06,
      "loss": 3.3144,
      "step": 4773
    },
    {
      "epoch": 2.1212,
      "grad_norm": 2.8573033809661865,
      "learning_rate": 2.306122448979592e-06,
      "loss": 0.1612,
      "step": 4774
    },
    {
      "epoch": 2.1214,
      "grad_norm": 12.39941120147705,
      "learning_rate": 2.295918367346939e-06,
      "loss": 2.1376,
      "step": 4775
    },
    {
      "epoch": 2.1216,
      "grad_norm": 2.2197749614715576,
      "learning_rate": 2.285714285714286e-06,
      "loss": 0.1219,
      "step": 4776
    },
    {
      "epoch": 2.1218,
      "grad_norm": 6.356390953063965,
      "learning_rate": 2.2755102040816326e-06,
      "loss": 2.936,
      "step": 4777
    },
    {
      "epoch": 2.122,
      "grad_norm": 6.300841331481934,
      "learning_rate": 2.2653061224489797e-06,
      "loss": 0.7233,
      "step": 4778
    },
    {
      "epoch": 2.1222,
      "grad_norm": 7.246288299560547,
      "learning_rate": 2.2551020408163267e-06,
      "loss": 0.3583,
      "step": 4779
    },
    {
      "epoch": 2.1224,
      "grad_norm": 8.743884086608887,
      "learning_rate": 2.2448979591836734e-06,
      "loss": 0.7548,
      "step": 4780
    },
    {
      "epoch": 2.1226,
      "grad_norm": 11.135361671447754,
      "learning_rate": 2.2346938775510204e-06,
      "loss": 1.4149,
      "step": 4781
    },
    {
      "epoch": 2.1228,
      "grad_norm": 1.941303014755249,
      "learning_rate": 2.2244897959183675e-06,
      "loss": 0.1728,
      "step": 4782
    },
    {
      "epoch": 2.123,
      "grad_norm": 8.995508193969727,
      "learning_rate": 2.214285714285714e-06,
      "loss": 0.7435,
      "step": 4783
    },
    {
      "epoch": 2.1232,
      "grad_norm": 2.2888870239257812,
      "learning_rate": 2.204081632653061e-06,
      "loss": 0.2329,
      "step": 4784
    },
    {
      "epoch": 2.1234,
      "grad_norm": 2.0693838596343994,
      "learning_rate": 2.1938775510204083e-06,
      "loss": 0.2532,
      "step": 4785
    },
    {
      "epoch": 2.1236,
      "grad_norm": 7.136364936828613,
      "learning_rate": 2.1836734693877553e-06,
      "loss": 0.1922,
      "step": 4786
    },
    {
      "epoch": 2.1238,
      "grad_norm": 5.488953590393066,
      "learning_rate": 2.1734693877551024e-06,
      "loss": 0.6459,
      "step": 4787
    },
    {
      "epoch": 2.124,
      "grad_norm": 2.1585514545440674,
      "learning_rate": 2.163265306122449e-06,
      "loss": 0.1159,
      "step": 4788
    },
    {
      "epoch": 2.1242,
      "grad_norm": 2.8502037525177,
      "learning_rate": 2.153061224489796e-06,
      "loss": 0.4822,
      "step": 4789
    },
    {
      "epoch": 2.1244,
      "grad_norm": 1.2102391719818115,
      "learning_rate": 2.142857142857143e-06,
      "loss": 0.0376,
      "step": 4790
    },
    {
      "epoch": 2.1246,
      "grad_norm": 2.719905376434326,
      "learning_rate": 2.13265306122449e-06,
      "loss": 0.1681,
      "step": 4791
    },
    {
      "epoch": 2.1248,
      "grad_norm": 3.575418472290039,
      "learning_rate": 2.122448979591837e-06,
      "loss": 0.5159,
      "step": 4792
    },
    {
      "epoch": 2.125,
      "grad_norm": 2.9629266262054443,
      "learning_rate": 2.112244897959184e-06,
      "loss": 0.2813,
      "step": 4793
    },
    {
      "epoch": 2.1252,
      "grad_norm": 5.93533182144165,
      "learning_rate": 2.1020408163265306e-06,
      "loss": 1.4368,
      "step": 4794
    },
    {
      "epoch": 2.1254,
      "grad_norm": 2.384918451309204,
      "learning_rate": 2.0918367346938776e-06,
      "loss": 0.143,
      "step": 4795
    },
    {
      "epoch": 2.1256,
      "grad_norm": 8.767003059387207,
      "learning_rate": 2.0816326530612247e-06,
      "loss": 1.633,
      "step": 4796
    },
    {
      "epoch": 2.1258,
      "grad_norm": 7.807423114776611,
      "learning_rate": 2.0714285714285713e-06,
      "loss": 1.032,
      "step": 4797
    },
    {
      "epoch": 2.126,
      "grad_norm": 10.32408618927002,
      "learning_rate": 2.0612244897959184e-06,
      "loss": 3.4824,
      "step": 4798
    },
    {
      "epoch": 2.1262,
      "grad_norm": 12.379114151000977,
      "learning_rate": 2.0510204081632654e-06,
      "loss": 1.1809,
      "step": 4799
    },
    {
      "epoch": 2.1264,
      "grad_norm": 11.27355670928955,
      "learning_rate": 2.040816326530612e-06,
      "loss": 0.5557,
      "step": 4800
    },
    {
      "epoch": 2.1266,
      "grad_norm": 1.92478609085083,
      "learning_rate": 2.030612244897959e-06,
      "loss": 0.1746,
      "step": 4801
    },
    {
      "epoch": 2.1268,
      "grad_norm": 6.682904243469238,
      "learning_rate": 2.020408163265306e-06,
      "loss": 0.8163,
      "step": 4802
    },
    {
      "epoch": 2.127,
      "grad_norm": 1.5593671798706055,
      "learning_rate": 2.0102040816326533e-06,
      "loss": 0.0997,
      "step": 4803
    },
    {
      "epoch": 2.1272,
      "grad_norm": 2.9079713821411133,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.275,
      "step": 4804
    },
    {
      "epoch": 2.1274,
      "grad_norm": 2.176316976547241,
      "learning_rate": 1.989795918367347e-06,
      "loss": 0.2564,
      "step": 4805
    },
    {
      "epoch": 2.1276,
      "grad_norm": 1.6099629402160645,
      "learning_rate": 1.979591836734694e-06,
      "loss": 0.1026,
      "step": 4806
    },
    {
      "epoch": 2.1278,
      "grad_norm": 2.784722089767456,
      "learning_rate": 1.969387755102041e-06,
      "loss": 0.5327,
      "step": 4807
    },
    {
      "epoch": 2.128,
      "grad_norm": 1.4035965204238892,
      "learning_rate": 1.9591836734693877e-06,
      "loss": 0.0564,
      "step": 4808
    },
    {
      "epoch": 2.1282,
      "grad_norm": 2.1346659660339355,
      "learning_rate": 1.948979591836735e-06,
      "loss": 0.129,
      "step": 4809
    },
    {
      "epoch": 2.1284,
      "grad_norm": 1.2497564554214478,
      "learning_rate": 1.938775510204082e-06,
      "loss": 0.0352,
      "step": 4810
    },
    {
      "epoch": 2.1286,
      "grad_norm": 1.9146955013275146,
      "learning_rate": 1.9285714285714285e-06,
      "loss": 0.1686,
      "step": 4811
    },
    {
      "epoch": 2.1288,
      "grad_norm": 8.626206398010254,
      "learning_rate": 1.9183673469387756e-06,
      "loss": 0.6249,
      "step": 4812
    },
    {
      "epoch": 2.129,
      "grad_norm": 6.26157808303833,
      "learning_rate": 1.9081632653061226e-06,
      "loss": 1.314,
      "step": 4813
    },
    {
      "epoch": 2.1292,
      "grad_norm": 16.438100814819336,
      "learning_rate": 1.8979591836734693e-06,
      "loss": 2.1826,
      "step": 4814
    },
    {
      "epoch": 2.1294,
      "grad_norm": 9.846821784973145,
      "learning_rate": 1.8877551020408165e-06,
      "loss": 3.4003,
      "step": 4815
    },
    {
      "epoch": 2.1296,
      "grad_norm": 5.37025260925293,
      "learning_rate": 1.8775510204081634e-06,
      "loss": 0.7318,
      "step": 4816
    },
    {
      "epoch": 2.1298,
      "grad_norm": 4.28413724899292,
      "learning_rate": 1.86734693877551e-06,
      "loss": 0.7001,
      "step": 4817
    },
    {
      "epoch": 2.13,
      "grad_norm": 1.656635046005249,
      "learning_rate": 1.8571428571428573e-06,
      "loss": 0.1194,
      "step": 4818
    },
    {
      "epoch": 2.1302,
      "grad_norm": 2.847740411758423,
      "learning_rate": 1.8469387755102042e-06,
      "loss": 0.4024,
      "step": 4819
    },
    {
      "epoch": 2.1304,
      "grad_norm": 2.651296615600586,
      "learning_rate": 1.8367346938775512e-06,
      "loss": 0.4087,
      "step": 4820
    },
    {
      "epoch": 2.1306,
      "grad_norm": 5.253534317016602,
      "learning_rate": 1.826530612244898e-06,
      "loss": 1.5563,
      "step": 4821
    },
    {
      "epoch": 2.1308,
      "grad_norm": 5.369923114776611,
      "learning_rate": 1.816326530612245e-06,
      "loss": 1.2941,
      "step": 4822
    },
    {
      "epoch": 2.1310000000000002,
      "grad_norm": 4.879589080810547,
      "learning_rate": 1.806122448979592e-06,
      "loss": 0.801,
      "step": 4823
    },
    {
      "epoch": 2.1312,
      "grad_norm": 3.045949935913086,
      "learning_rate": 1.7959183673469388e-06,
      "loss": 0.1729,
      "step": 4824
    },
    {
      "epoch": 2.1314,
      "grad_norm": 5.828512668609619,
      "learning_rate": 1.7857142857142857e-06,
      "loss": 0.6962,
      "step": 4825
    },
    {
      "epoch": 2.1316,
      "grad_norm": 6.206547737121582,
      "learning_rate": 1.775510204081633e-06,
      "loss": 0.9211,
      "step": 4826
    },
    {
      "epoch": 2.1318,
      "grad_norm": 1.1685166358947754,
      "learning_rate": 1.7653061224489796e-06,
      "loss": 0.0379,
      "step": 4827
    },
    {
      "epoch": 2.132,
      "grad_norm": 5.419491767883301,
      "learning_rate": 1.7551020408163264e-06,
      "loss": 1.3864,
      "step": 4828
    },
    {
      "epoch": 2.1322,
      "grad_norm": 7.400469779968262,
      "learning_rate": 1.7448979591836737e-06,
      "loss": 0.7807,
      "step": 4829
    },
    {
      "epoch": 2.1324,
      "grad_norm": 2.7783303260803223,
      "learning_rate": 1.7346938775510206e-06,
      "loss": 0.5662,
      "step": 4830
    },
    {
      "epoch": 2.1326,
      "grad_norm": 9.906639099121094,
      "learning_rate": 1.7244897959183672e-06,
      "loss": 2.34,
      "step": 4831
    },
    {
      "epoch": 2.1328,
      "grad_norm": 4.272519111633301,
      "learning_rate": 1.7142857142857145e-06,
      "loss": 1.7126,
      "step": 4832
    },
    {
      "epoch": 2.133,
      "grad_norm": 2.4000024795532227,
      "learning_rate": 1.7040816326530613e-06,
      "loss": 0.2753,
      "step": 4833
    },
    {
      "epoch": 2.1332,
      "grad_norm": 6.222007751464844,
      "learning_rate": 1.6938775510204082e-06,
      "loss": 0.6795,
      "step": 4834
    },
    {
      "epoch": 2.1334,
      "grad_norm": 3.4705138206481934,
      "learning_rate": 1.6836734693877552e-06,
      "loss": 0.6273,
      "step": 4835
    },
    {
      "epoch": 2.1336,
      "grad_norm": 1.5016515254974365,
      "learning_rate": 1.673469387755102e-06,
      "loss": 0.1128,
      "step": 4836
    },
    {
      "epoch": 2.1338,
      "grad_norm": 2.719780921936035,
      "learning_rate": 1.6632653061224492e-06,
      "loss": 0.5315,
      "step": 4837
    },
    {
      "epoch": 2.134,
      "grad_norm": 1.00386381149292,
      "learning_rate": 1.653061224489796e-06,
      "loss": 0.0354,
      "step": 4838
    },
    {
      "epoch": 2.1342,
      "grad_norm": 1.8480079174041748,
      "learning_rate": 1.6428571428571429e-06,
      "loss": 0.1274,
      "step": 4839
    },
    {
      "epoch": 2.1344,
      "grad_norm": 2.8276331424713135,
      "learning_rate": 1.63265306122449e-06,
      "loss": 0.2234,
      "step": 4840
    },
    {
      "epoch": 2.1346,
      "grad_norm": 1.1041818857192993,
      "learning_rate": 1.6224489795918368e-06,
      "loss": 0.0323,
      "step": 4841
    },
    {
      "epoch": 2.1348,
      "grad_norm": 11.41901969909668,
      "learning_rate": 1.6122448979591836e-06,
      "loss": 0.7364,
      "step": 4842
    },
    {
      "epoch": 2.135,
      "grad_norm": 2.2846317291259766,
      "learning_rate": 1.602040816326531e-06,
      "loss": 0.2361,
      "step": 4843
    },
    {
      "epoch": 2.1352,
      "grad_norm": 6.557143688201904,
      "learning_rate": 1.5918367346938775e-06,
      "loss": 0.273,
      "step": 4844
    },
    {
      "epoch": 2.1354,
      "grad_norm": 2.2313144207000732,
      "learning_rate": 1.5816326530612244e-06,
      "loss": 0.1999,
      "step": 4845
    },
    {
      "epoch": 2.1356,
      "grad_norm": 0.8703025579452515,
      "learning_rate": 1.5714285714285717e-06,
      "loss": 0.0131,
      "step": 4846
    },
    {
      "epoch": 2.1358,
      "grad_norm": 3.4729299545288086,
      "learning_rate": 1.5612244897959185e-06,
      "loss": 0.3988,
      "step": 4847
    },
    {
      "epoch": 2.136,
      "grad_norm": 2.4414684772491455,
      "learning_rate": 1.5510204081632654e-06,
      "loss": 0.1224,
      "step": 4848
    },
    {
      "epoch": 2.1362,
      "grad_norm": 5.00660514831543,
      "learning_rate": 1.5408163265306122e-06,
      "loss": 0.7455,
      "step": 4849
    },
    {
      "epoch": 2.1364,
      "grad_norm": 5.4602580070495605,
      "learning_rate": 1.5306122448979593e-06,
      "loss": 0.6444,
      "step": 4850
    },
    {
      "epoch": 2.1366,
      "grad_norm": 1.9978864192962646,
      "learning_rate": 1.5204081632653061e-06,
      "loss": 0.2164,
      "step": 4851
    },
    {
      "epoch": 2.1368,
      "grad_norm": 1.0184167623519897,
      "learning_rate": 1.5102040816326532e-06,
      "loss": 0.029,
      "step": 4852
    },
    {
      "epoch": 2.137,
      "grad_norm": 2.7992334365844727,
      "learning_rate": 1.5e-06,
      "loss": 0.4683,
      "step": 4853
    },
    {
      "epoch": 2.1372,
      "grad_norm": 2.517106533050537,
      "learning_rate": 1.489795918367347e-06,
      "loss": 0.1445,
      "step": 4854
    },
    {
      "epoch": 2.1374,
      "grad_norm": 1.8562113046646118,
      "learning_rate": 1.479591836734694e-06,
      "loss": 0.1674,
      "step": 4855
    },
    {
      "epoch": 2.1376,
      "grad_norm": 1.714816689491272,
      "learning_rate": 1.4693877551020408e-06,
      "loss": 0.1415,
      "step": 4856
    },
    {
      "epoch": 2.1378,
      "grad_norm": 0.8883129954338074,
      "learning_rate": 1.4591836734693879e-06,
      "loss": 0.0156,
      "step": 4857
    },
    {
      "epoch": 2.138,
      "grad_norm": 2.0923383235931396,
      "learning_rate": 1.4489795918367347e-06,
      "loss": 0.2406,
      "step": 4858
    },
    {
      "epoch": 2.1382,
      "grad_norm": 2.276726484298706,
      "learning_rate": 1.4387755102040818e-06,
      "loss": 0.2007,
      "step": 4859
    },
    {
      "epoch": 2.1384,
      "grad_norm": 2.1536126136779785,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 0.2267,
      "step": 4860
    },
    {
      "epoch": 2.1386,
      "grad_norm": 4.831217288970947,
      "learning_rate": 1.4183673469387757e-06,
      "loss": 0.264,
      "step": 4861
    },
    {
      "epoch": 2.1388,
      "grad_norm": 0.8061777949333191,
      "learning_rate": 1.4081632653061225e-06,
      "loss": 0.0162,
      "step": 4862
    },
    {
      "epoch": 2.1390000000000002,
      "grad_norm": 2.4018330574035645,
      "learning_rate": 1.3979591836734694e-06,
      "loss": 0.2791,
      "step": 4863
    },
    {
      "epoch": 2.1391999999999998,
      "grad_norm": 3.1337363719940186,
      "learning_rate": 1.3877551020408165e-06,
      "loss": 0.3743,
      "step": 4864
    },
    {
      "epoch": 2.1394,
      "grad_norm": 6.414299488067627,
      "learning_rate": 1.3775510204081633e-06,
      "loss": 0.284,
      "step": 4865
    },
    {
      "epoch": 2.1396,
      "grad_norm": 7.778178691864014,
      "learning_rate": 1.3673469387755102e-06,
      "loss": 1.4856,
      "step": 4866
    },
    {
      "epoch": 2.1398,
      "grad_norm": 1.8873628377914429,
      "learning_rate": 1.3571428571428572e-06,
      "loss": 0.1629,
      "step": 4867
    },
    {
      "epoch": 2.14,
      "grad_norm": 9.496567726135254,
      "learning_rate": 1.346938775510204e-06,
      "loss": 3.3572,
      "step": 4868
    },
    {
      "epoch": 2.1402,
      "grad_norm": 1.016174077987671,
      "learning_rate": 1.3367346938775511e-06,
      "loss": 0.0316,
      "step": 4869
    },
    {
      "epoch": 2.1404,
      "grad_norm": 0.9162819981575012,
      "learning_rate": 1.326530612244898e-06,
      "loss": 0.0095,
      "step": 4870
    },
    {
      "epoch": 2.1406,
      "grad_norm": 1.6717957258224487,
      "learning_rate": 1.316326530612245e-06,
      "loss": 0.1213,
      "step": 4871
    },
    {
      "epoch": 2.1408,
      "grad_norm": 1.7956033945083618,
      "learning_rate": 1.306122448979592e-06,
      "loss": 0.1451,
      "step": 4872
    },
    {
      "epoch": 2.141,
      "grad_norm": 1.9850627183914185,
      "learning_rate": 1.2959183673469387e-06,
      "loss": 0.0518,
      "step": 4873
    },
    {
      "epoch": 2.1412,
      "grad_norm": 2.4787042140960693,
      "learning_rate": 1.2857142857142858e-06,
      "loss": 0.132,
      "step": 4874
    },
    {
      "epoch": 2.1414,
      "grad_norm": 2.5726118087768555,
      "learning_rate": 1.2755102040816327e-06,
      "loss": 0.144,
      "step": 4875
    },
    {
      "epoch": 2.1416,
      "grad_norm": 4.900381565093994,
      "learning_rate": 1.2653061224489797e-06,
      "loss": 0.5928,
      "step": 4876
    },
    {
      "epoch": 2.1418,
      "grad_norm": 2.545435667037964,
      "learning_rate": 1.2551020408163266e-06,
      "loss": 0.3196,
      "step": 4877
    },
    {
      "epoch": 2.142,
      "grad_norm": 1.777809739112854,
      "learning_rate": 1.2448979591836736e-06,
      "loss": 0.0613,
      "step": 4878
    },
    {
      "epoch": 2.1422,
      "grad_norm": 5.111660003662109,
      "learning_rate": 1.2346938775510205e-06,
      "loss": 1.4366,
      "step": 4879
    },
    {
      "epoch": 2.1424,
      "grad_norm": 3.8646697998046875,
      "learning_rate": 1.2244897959183673e-06,
      "loss": 0.4065,
      "step": 4880
    },
    {
      "epoch": 2.1426,
      "grad_norm": 3.850189447402954,
      "learning_rate": 1.2142857142857144e-06,
      "loss": 0.0971,
      "step": 4881
    },
    {
      "epoch": 2.1428,
      "grad_norm": 3.2342708110809326,
      "learning_rate": 1.2040816326530612e-06,
      "loss": 0.5958,
      "step": 4882
    },
    {
      "epoch": 2.143,
      "grad_norm": 4.939568519592285,
      "learning_rate": 1.193877551020408e-06,
      "loss": 0.4098,
      "step": 4883
    },
    {
      "epoch": 2.1432,
      "grad_norm": 6.049392223358154,
      "learning_rate": 1.1836734693877552e-06,
      "loss": 0.8668,
      "step": 4884
    },
    {
      "epoch": 2.1434,
      "grad_norm": 2.739666700363159,
      "learning_rate": 1.173469387755102e-06,
      "loss": 0.2615,
      "step": 4885
    },
    {
      "epoch": 2.1436,
      "grad_norm": 9.856095314025879,
      "learning_rate": 1.163265306122449e-06,
      "loss": 2.2432,
      "step": 4886
    },
    {
      "epoch": 2.1438,
      "grad_norm": 5.986268043518066,
      "learning_rate": 1.153061224489796e-06,
      "loss": 0.6472,
      "step": 4887
    },
    {
      "epoch": 2.144,
      "grad_norm": 4.728291034698486,
      "learning_rate": 1.142857142857143e-06,
      "loss": 0.6578,
      "step": 4888
    },
    {
      "epoch": 2.1442,
      "grad_norm": 2.241907835006714,
      "learning_rate": 1.1326530612244898e-06,
      "loss": 0.1339,
      "step": 4889
    },
    {
      "epoch": 2.1444,
      "grad_norm": 3.052717447280884,
      "learning_rate": 1.1224489795918367e-06,
      "loss": 0.8341,
      "step": 4890
    },
    {
      "epoch": 2.1446,
      "grad_norm": 5.318640232086182,
      "learning_rate": 1.1122448979591838e-06,
      "loss": 0.364,
      "step": 4891
    },
    {
      "epoch": 2.1448,
      "grad_norm": 3.5043251514434814,
      "learning_rate": 1.1020408163265306e-06,
      "loss": 1.6492,
      "step": 4892
    },
    {
      "epoch": 2.145,
      "grad_norm": 9.889541625976562,
      "learning_rate": 1.0918367346938777e-06,
      "loss": 1.1815,
      "step": 4893
    },
    {
      "epoch": 2.1452,
      "grad_norm": 2.5116753578186035,
      "learning_rate": 1.0816326530612245e-06,
      "loss": 0.5012,
      "step": 4894
    },
    {
      "epoch": 2.1454,
      "grad_norm": 1.443432092666626,
      "learning_rate": 1.0714285714285716e-06,
      "loss": 0.0875,
      "step": 4895
    },
    {
      "epoch": 2.1456,
      "grad_norm": 1.9345141649246216,
      "learning_rate": 1.0612244897959184e-06,
      "loss": 0.1529,
      "step": 4896
    },
    {
      "epoch": 2.1458,
      "grad_norm": 1.706526517868042,
      "learning_rate": 1.0510204081632653e-06,
      "loss": 0.1565,
      "step": 4897
    },
    {
      "epoch": 2.146,
      "grad_norm": 1.3103222846984863,
      "learning_rate": 1.0408163265306123e-06,
      "loss": 0.0434,
      "step": 4898
    },
    {
      "epoch": 2.1462,
      "grad_norm": 6.11220645904541,
      "learning_rate": 1.0306122448979592e-06,
      "loss": 0.3047,
      "step": 4899
    },
    {
      "epoch": 2.1464,
      "grad_norm": 2.7306883335113525,
      "learning_rate": 1.020408163265306e-06,
      "loss": 0.341,
      "step": 4900
    },
    {
      "epoch": 2.1466,
      "grad_norm": 15.686260223388672,
      "learning_rate": 1.010204081632653e-06,
      "loss": 1.5644,
      "step": 4901
    },
    {
      "epoch": 2.1468,
      "grad_norm": 5.284987926483154,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.4679,
      "step": 4902
    },
    {
      "epoch": 2.147,
      "grad_norm": 6.0595574378967285,
      "learning_rate": 9.89795918367347e-07,
      "loss": 0.4288,
      "step": 4903
    },
    {
      "epoch": 2.1471999999999998,
      "grad_norm": 2.0993542671203613,
      "learning_rate": 9.795918367346939e-07,
      "loss": 0.2359,
      "step": 4904
    },
    {
      "epoch": 2.1474,
      "grad_norm": 4.383464336395264,
      "learning_rate": 9.69387755102041e-07,
      "loss": 0.8041,
      "step": 4905
    },
    {
      "epoch": 2.1476,
      "grad_norm": 2.5457303524017334,
      "learning_rate": 9.591836734693878e-07,
      "loss": 0.5014,
      "step": 4906
    },
    {
      "epoch": 2.1478,
      "grad_norm": 2.813450336456299,
      "learning_rate": 9.489795918367346e-07,
      "loss": 0.2483,
      "step": 4907
    },
    {
      "epoch": 2.148,
      "grad_norm": 1.4544942378997803,
      "learning_rate": 9.387755102040817e-07,
      "loss": 0.1044,
      "step": 4908
    },
    {
      "epoch": 2.1482,
      "grad_norm": 1.8861054182052612,
      "learning_rate": 9.285714285714287e-07,
      "loss": 0.1481,
      "step": 4909
    },
    {
      "epoch": 2.1484,
      "grad_norm": 0.4566737413406372,
      "learning_rate": 9.183673469387756e-07,
      "loss": 0.0085,
      "step": 4910
    },
    {
      "epoch": 2.1486,
      "grad_norm": 2.440262794494629,
      "learning_rate": 9.081632653061225e-07,
      "loss": 0.2129,
      "step": 4911
    },
    {
      "epoch": 2.1488,
      "grad_norm": 1.998534917831421,
      "learning_rate": 8.979591836734694e-07,
      "loss": 0.1949,
      "step": 4912
    },
    {
      "epoch": 2.149,
      "grad_norm": 7.0681047439575195,
      "learning_rate": 8.877551020408165e-07,
      "loss": 0.8244,
      "step": 4913
    },
    {
      "epoch": 2.1492,
      "grad_norm": 1.2315425872802734,
      "learning_rate": 8.775510204081632e-07,
      "loss": 0.0363,
      "step": 4914
    },
    {
      "epoch": 2.1494,
      "grad_norm": 11.267387390136719,
      "learning_rate": 8.673469387755103e-07,
      "loss": 1.4418,
      "step": 4915
    },
    {
      "epoch": 2.1496,
      "grad_norm": 1.8728132247924805,
      "learning_rate": 8.571428571428572e-07,
      "loss": 0.1121,
      "step": 4916
    },
    {
      "epoch": 2.1498,
      "grad_norm": 9.476338386535645,
      "learning_rate": 8.469387755102041e-07,
      "loss": 0.2879,
      "step": 4917
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.17253041267395,
      "learning_rate": 8.36734693877551e-07,
      "loss": 0.2077,
      "step": 4918
    },
    {
      "epoch": 2.1502,
      "grad_norm": 2.146371603012085,
      "learning_rate": 8.26530612244898e-07,
      "loss": 0.2161,
      "step": 4919
    },
    {
      "epoch": 2.1504,
      "grad_norm": 4.625229835510254,
      "learning_rate": 8.16326530612245e-07,
      "loss": 0.5836,
      "step": 4920
    },
    {
      "epoch": 2.1506,
      "grad_norm": 3.1853513717651367,
      "learning_rate": 8.061224489795918e-07,
      "loss": 0.5375,
      "step": 4921
    },
    {
      "epoch": 2.1508,
      "grad_norm": 1.9648888111114502,
      "learning_rate": 7.959183673469388e-07,
      "loss": 0.2261,
      "step": 4922
    },
    {
      "epoch": 2.151,
      "grad_norm": 2.480511426925659,
      "learning_rate": 7.857142857142858e-07,
      "loss": 0.3026,
      "step": 4923
    },
    {
      "epoch": 2.1512000000000002,
      "grad_norm": 3.7170333862304688,
      "learning_rate": 7.755102040816327e-07,
      "loss": 0.6515,
      "step": 4924
    },
    {
      "epoch": 2.1514,
      "grad_norm": 3.0362749099731445,
      "learning_rate": 7.653061224489796e-07,
      "loss": 0.4871,
      "step": 4925
    },
    {
      "epoch": 2.1516,
      "grad_norm": 3.072406768798828,
      "learning_rate": 7.551020408163266e-07,
      "loss": 0.2954,
      "step": 4926
    },
    {
      "epoch": 2.1518,
      "grad_norm": 1.0148082971572876,
      "learning_rate": 7.448979591836736e-07,
      "loss": 0.0305,
      "step": 4927
    },
    {
      "epoch": 2.152,
      "grad_norm": 2.6846325397491455,
      "learning_rate": 7.346938775510204e-07,
      "loss": 0.4181,
      "step": 4928
    },
    {
      "epoch": 2.1522,
      "grad_norm": 3.4255530834198,
      "learning_rate": 7.244897959183674e-07,
      "loss": 0.4283,
      "step": 4929
    },
    {
      "epoch": 2.1524,
      "grad_norm": 5.002836227416992,
      "learning_rate": 7.142857142857143e-07,
      "loss": 1.4199,
      "step": 4930
    },
    {
      "epoch": 2.1526,
      "grad_norm": 5.339587688446045,
      "learning_rate": 7.040816326530613e-07,
      "loss": 1.3885,
      "step": 4931
    },
    {
      "epoch": 2.1528,
      "grad_norm": 2.046875476837158,
      "learning_rate": 6.938775510204082e-07,
      "loss": 0.1375,
      "step": 4932
    },
    {
      "epoch": 2.153,
      "grad_norm": 2.331223964691162,
      "learning_rate": 6.836734693877551e-07,
      "loss": 0.1424,
      "step": 4933
    },
    {
      "epoch": 2.1532,
      "grad_norm": 4.11829948425293,
      "learning_rate": 6.73469387755102e-07,
      "loss": 0.825,
      "step": 4934
    },
    {
      "epoch": 2.1534,
      "grad_norm": 10.498679161071777,
      "learning_rate": 6.63265306122449e-07,
      "loss": 1.3925,
      "step": 4935
    },
    {
      "epoch": 2.1536,
      "grad_norm": 4.633434772491455,
      "learning_rate": 6.53061224489796e-07,
      "loss": 0.7704,
      "step": 4936
    },
    {
      "epoch": 2.1538,
      "grad_norm": 9.199163436889648,
      "learning_rate": 6.428571428571429e-07,
      "loss": 3.3639,
      "step": 4937
    },
    {
      "epoch": 2.154,
      "grad_norm": 8.057214736938477,
      "learning_rate": 6.326530612244899e-07,
      "loss": 1.2202,
      "step": 4938
    },
    {
      "epoch": 2.1542,
      "grad_norm": 1.9962745904922485,
      "learning_rate": 6.224489795918368e-07,
      "loss": 0.1147,
      "step": 4939
    },
    {
      "epoch": 2.1544,
      "grad_norm": 1.3766710758209229,
      "learning_rate": 6.122448979591837e-07,
      "loss": 0.0292,
      "step": 4940
    },
    {
      "epoch": 2.1546,
      "grad_norm": 4.408767223358154,
      "learning_rate": 6.020408163265306e-07,
      "loss": 0.5736,
      "step": 4941
    },
    {
      "epoch": 2.1548,
      "grad_norm": 0.9604965448379517,
      "learning_rate": 5.918367346938776e-07,
      "loss": 0.0322,
      "step": 4942
    },
    {
      "epoch": 2.155,
      "grad_norm": 2.2919974327087402,
      "learning_rate": 5.816326530612245e-07,
      "loss": 0.2447,
      "step": 4943
    },
    {
      "epoch": 2.1552,
      "grad_norm": 2.1274476051330566,
      "learning_rate": 5.714285714285715e-07,
      "loss": 0.2329,
      "step": 4944
    },
    {
      "epoch": 2.1554,
      "grad_norm": 3.189530849456787,
      "learning_rate": 5.612244897959183e-07,
      "loss": 0.3092,
      "step": 4945
    },
    {
      "epoch": 2.1556,
      "grad_norm": 1.0164053440093994,
      "learning_rate": 5.510204081632653e-07,
      "loss": 0.0263,
      "step": 4946
    },
    {
      "epoch": 2.1558,
      "grad_norm": 1.5905165672302246,
      "learning_rate": 5.408163265306123e-07,
      "loss": 0.1089,
      "step": 4947
    },
    {
      "epoch": 2.156,
      "grad_norm": 2.1196746826171875,
      "learning_rate": 5.306122448979592e-07,
      "loss": 0.2421,
      "step": 4948
    },
    {
      "epoch": 2.1562,
      "grad_norm": 1.8495643138885498,
      "learning_rate": 5.204081632653062e-07,
      "loss": 0.1581,
      "step": 4949
    },
    {
      "epoch": 2.1564,
      "grad_norm": 3.3096840381622314,
      "learning_rate": 5.10204081632653e-07,
      "loss": 0.8094,
      "step": 4950
    },
    {
      "epoch": 2.1566,
      "grad_norm": 2.5738346576690674,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.1388,
      "step": 4951
    },
    {
      "epoch": 2.1568,
      "grad_norm": 3.5386643409729004,
      "learning_rate": 4.897959183673469e-07,
      "loss": 1.5187,
      "step": 4952
    },
    {
      "epoch": 2.157,
      "grad_norm": 8.421414375305176,
      "learning_rate": 4.795918367346939e-07,
      "loss": 1.4135,
      "step": 4953
    },
    {
      "epoch": 2.1572,
      "grad_norm": 7.643795967102051,
      "learning_rate": 4.6938775510204085e-07,
      "loss": 1.1508,
      "step": 4954
    },
    {
      "epoch": 2.1574,
      "grad_norm": 9.643633842468262,
      "learning_rate": 4.591836734693878e-07,
      "loss": 2.5107,
      "step": 4955
    },
    {
      "epoch": 2.1576,
      "grad_norm": 11.657377243041992,
      "learning_rate": 4.489795918367347e-07,
      "loss": 3.1162,
      "step": 4956
    },
    {
      "epoch": 2.1578,
      "grad_norm": 5.536891460418701,
      "learning_rate": 4.387755102040816e-07,
      "loss": 1.4635,
      "step": 4957
    },
    {
      "epoch": 2.158,
      "grad_norm": 2.9929583072662354,
      "learning_rate": 4.285714285714286e-07,
      "loss": 0.2885,
      "step": 4958
    },
    {
      "epoch": 2.1582,
      "grad_norm": 0.9444872736930847,
      "learning_rate": 4.183673469387755e-07,
      "loss": 0.0164,
      "step": 4959
    },
    {
      "epoch": 2.1584,
      "grad_norm": 1.915464997291565,
      "learning_rate": 4.081632653061225e-07,
      "loss": 0.3739,
      "step": 4960
    },
    {
      "epoch": 2.1586,
      "grad_norm": 1.9991296529769897,
      "learning_rate": 3.979591836734694e-07,
      "loss": 0.1583,
      "step": 4961
    },
    {
      "epoch": 2.1588,
      "grad_norm": 4.060037612915039,
      "learning_rate": 3.8775510204081634e-07,
      "loss": 1.5909,
      "step": 4962
    },
    {
      "epoch": 2.159,
      "grad_norm": 1.268680453300476,
      "learning_rate": 3.775510204081633e-07,
      "loss": 0.044,
      "step": 4963
    },
    {
      "epoch": 2.1592000000000002,
      "grad_norm": 4.501353740692139,
      "learning_rate": 3.673469387755102e-07,
      "loss": 0.5581,
      "step": 4964
    },
    {
      "epoch": 2.1593999999999998,
      "grad_norm": 1.8425730466842651,
      "learning_rate": 3.5714285714285716e-07,
      "loss": 0.1389,
      "step": 4965
    },
    {
      "epoch": 2.1596,
      "grad_norm": 2.0111634731292725,
      "learning_rate": 3.469387755102041e-07,
      "loss": 0.1959,
      "step": 4966
    },
    {
      "epoch": 2.1598,
      "grad_norm": 2.507601737976074,
      "learning_rate": 3.36734693877551e-07,
      "loss": 0.494,
      "step": 4967
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.3330937623977661,
      "learning_rate": 3.26530612244898e-07,
      "loss": 0.0433,
      "step": 4968
    },
    {
      "epoch": 2.1602,
      "grad_norm": 3.633925676345825,
      "learning_rate": 3.1632653061224493e-07,
      "loss": 1.6239,
      "step": 4969
    },
    {
      "epoch": 2.1604,
      "grad_norm": 0.8505246043205261,
      "learning_rate": 3.0612244897959183e-07,
      "loss": 0.0187,
      "step": 4970
    },
    {
      "epoch": 2.1606,
      "grad_norm": 1.1604036092758179,
      "learning_rate": 2.959183673469388e-07,
      "loss": 0.0253,
      "step": 4971
    },
    {
      "epoch": 2.1608,
      "grad_norm": 1.7584632635116577,
      "learning_rate": 2.8571428571428575e-07,
      "loss": 0.1514,
      "step": 4972
    },
    {
      "epoch": 2.161,
      "grad_norm": 1.6417275667190552,
      "learning_rate": 2.7551020408163265e-07,
      "loss": 0.025,
      "step": 4973
    },
    {
      "epoch": 2.1612,
      "grad_norm": 2.554945707321167,
      "learning_rate": 2.653061224489796e-07,
      "loss": 0.5624,
      "step": 4974
    },
    {
      "epoch": 2.1614,
      "grad_norm": 2.8897547721862793,
      "learning_rate": 2.551020408163265e-07,
      "loss": 0.3113,
      "step": 4975
    },
    {
      "epoch": 2.1616,
      "grad_norm": 4.395506381988525,
      "learning_rate": 2.4489795918367347e-07,
      "loss": 1.6157,
      "step": 4976
    },
    {
      "epoch": 2.1618,
      "grad_norm": 5.517244338989258,
      "learning_rate": 2.3469387755102042e-07,
      "loss": 1.743,
      "step": 4977
    },
    {
      "epoch": 2.162,
      "grad_norm": 5.6711015701293945,
      "learning_rate": 2.2448979591836735e-07,
      "loss": 0.5785,
      "step": 4978
    },
    {
      "epoch": 2.1622,
      "grad_norm": 3.8922955989837646,
      "learning_rate": 2.142857142857143e-07,
      "loss": 2.177,
      "step": 4979
    },
    {
      "epoch": 2.1624,
      "grad_norm": 5.796224594116211,
      "learning_rate": 2.0408163265306124e-07,
      "loss": 2.4335,
      "step": 4980
    },
    {
      "epoch": 2.1626,
      "grad_norm": 3.8742527961730957,
      "learning_rate": 1.9387755102040817e-07,
      "loss": 1.4992,
      "step": 4981
    },
    {
      "epoch": 2.1628,
      "grad_norm": 11.639328002929688,
      "learning_rate": 1.836734693877551e-07,
      "loss": 3.339,
      "step": 4982
    },
    {
      "epoch": 2.163,
      "grad_norm": 5.487311840057373,
      "learning_rate": 1.7346938775510206e-07,
      "loss": 0.5887,
      "step": 4983
    },
    {
      "epoch": 2.1632,
      "grad_norm": 1.668665885925293,
      "learning_rate": 1.63265306122449e-07,
      "loss": 0.1145,
      "step": 4984
    },
    {
      "epoch": 2.1634,
      "grad_norm": 2.9366931915283203,
      "learning_rate": 1.5306122448979592e-07,
      "loss": 0.6556,
      "step": 4985
    },
    {
      "epoch": 2.1636,
      "grad_norm": 4.772943496704102,
      "learning_rate": 1.4285714285714287e-07,
      "loss": 0.8894,
      "step": 4986
    },
    {
      "epoch": 2.1638,
      "grad_norm": 5.80991792678833,
      "learning_rate": 1.326530612244898e-07,
      "loss": 2.3163,
      "step": 4987
    },
    {
      "epoch": 2.164,
      "grad_norm": 6.079235553741455,
      "learning_rate": 1.2244897959183673e-07,
      "loss": 2.0411,
      "step": 4988
    },
    {
      "epoch": 2.1642,
      "grad_norm": 11.773283004760742,
      "learning_rate": 1.1224489795918368e-07,
      "loss": 3.309,
      "step": 4989
    },
    {
      "epoch": 2.1644,
      "grad_norm": 5.10943078994751,
      "learning_rate": 1.0204081632653062e-07,
      "loss": 1.5065,
      "step": 4990
    },
    {
      "epoch": 2.1646,
      "grad_norm": 9.724143981933594,
      "learning_rate": 9.183673469387755e-08,
      "loss": 2.3739,
      "step": 4991
    },
    {
      "epoch": 2.1648,
      "grad_norm": 1.881788730621338,
      "learning_rate": 8.16326530612245e-08,
      "loss": 0.0728,
      "step": 4992
    },
    {
      "epoch": 2.165,
      "grad_norm": 5.462317943572998,
      "learning_rate": 7.142857142857144e-08,
      "loss": 0.6102,
      "step": 4993
    },
    {
      "epoch": 2.1652,
      "grad_norm": 1.4766038656234741,
      "learning_rate": 6.122448979591837e-08,
      "loss": 0.0714,
      "step": 4994
    },
    {
      "epoch": 2.1654,
      "grad_norm": 2.4013559818267822,
      "learning_rate": 5.102040816326531e-08,
      "loss": 0.4663,
      "step": 4995
    },
    {
      "epoch": 2.1656,
      "grad_norm": 4.605373859405518,
      "learning_rate": 4.081632653061225e-08,
      "loss": 0.8669,
      "step": 4996
    },
    {
      "epoch": 2.1658,
      "grad_norm": 1.9811826944351196,
      "learning_rate": 3.0612244897959183e-08,
      "loss": 0.0347,
      "step": 4997
    },
    {
      "epoch": 2.166,
      "grad_norm": 9.743058204650879,
      "learning_rate": 2.0408163265306123e-08,
      "loss": 3.2979,
      "step": 4998
    },
    {
      "epoch": 2.1662,
      "grad_norm": 1.0800037384033203,
      "learning_rate": 1.0204081632653062e-08,
      "loss": 0.024,
      "step": 4999
    },
    {
      "epoch": 2.1664,
      "grad_norm": 1.7821890115737915,
      "learning_rate": 0.0,
      "loss": 0.1462,
      "step": 5000
    }
  ],
  "logging_steps": 1,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}