{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0001,
      "grad_norm": 7.648532390594482,
      "learning_rate": 0.0,
      "loss": 12.4719,
      "step": 1
    },
    {
      "epoch": 0.0002,
      "grad_norm": 7.399874210357666,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 12.4717,
      "step": 2
    },
    {
      "epoch": 0.0003,
      "grad_norm": 8.38534927368164,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 12.463,
      "step": 3
    },
    {
      "epoch": 0.0004,
      "grad_norm": 6.242674350738525,
      "learning_rate": 6.000000000000001e-07,
      "loss": 12.4693,
      "step": 4
    },
    {
      "epoch": 0.0005,
      "grad_norm": 6.616119384765625,
      "learning_rate": 8.000000000000001e-07,
      "loss": 12.4745,
      "step": 5
    },
    {
      "epoch": 0.0006,
      "grad_norm": 7.008609294891357,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 12.4626,
      "step": 6
    },
    {
      "epoch": 0.0007,
      "grad_norm": 7.102725505828857,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 12.4548,
      "step": 7
    },
    {
      "epoch": 0.0008,
      "grad_norm": 7.628903388977051,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 12.4719,
      "step": 8
    },
    {
      "epoch": 0.0009,
      "grad_norm": 7.701151371002197,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 12.4634,
      "step": 9
    },
    {
      "epoch": 0.001,
      "grad_norm": 6.8943705558776855,
      "learning_rate": 1.8e-06,
      "loss": 12.4596,
      "step": 10
    },
    {
      "epoch": 0.0011,
      "grad_norm": 7.283838272094727,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 12.461,
      "step": 11
    },
    {
      "epoch": 0.0012,
      "grad_norm": 7.3759613037109375,
      "learning_rate": 2.2e-06,
      "loss": 12.4657,
      "step": 12
    },
    {
      "epoch": 0.0013,
      "grad_norm": 7.078854084014893,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 12.4519,
      "step": 13
    },
    {
      "epoch": 0.0014,
      "grad_norm": 6.9325995445251465,
      "learning_rate": 2.6e-06,
      "loss": 12.4557,
      "step": 14
    },
    {
      "epoch": 0.0015,
      "grad_norm": 7.351903438568115,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 12.4489,
      "step": 15
    },
    {
      "epoch": 0.0016,
      "grad_norm": 6.87741756439209,
      "learning_rate": 3e-06,
      "loss": 12.458,
      "step": 16
    },
    {
      "epoch": 0.0017,
      "grad_norm": 6.898277282714844,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 12.4505,
      "step": 17
    },
    {
      "epoch": 0.0018,
      "grad_norm": 8.604022026062012,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 12.4424,
      "step": 18
    },
    {
      "epoch": 0.0019,
      "grad_norm": 6.969699859619141,
      "learning_rate": 3.6e-06,
      "loss": 12.4416,
      "step": 19
    },
    {
      "epoch": 0.002,
      "grad_norm": 7.339034557342529,
      "learning_rate": 3.8e-06,
      "loss": 12.4437,
      "step": 20
    },
    {
      "epoch": 0.0021,
      "grad_norm": 8.565898895263672,
      "learning_rate": 4.000000000000001e-06,
      "loss": 12.4255,
      "step": 21
    },
    {
      "epoch": 0.0022,
      "grad_norm": 7.44255256652832,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 12.4478,
      "step": 22
    },
    {
      "epoch": 0.0023,
      "grad_norm": 7.609328746795654,
      "learning_rate": 4.4e-06,
      "loss": 12.4263,
      "step": 23
    },
    {
      "epoch": 0.0024,
      "grad_norm": 7.356610298156738,
      "learning_rate": 4.6e-06,
      "loss": 12.4531,
      "step": 24
    },
    {
      "epoch": 0.0025,
      "grad_norm": 7.821833610534668,
      "learning_rate": 4.800000000000001e-06,
      "loss": 12.4242,
      "step": 25
    },
    {
      "epoch": 0.0026,
      "grad_norm": 7.234078884124756,
      "learning_rate": 5e-06,
      "loss": 12.408,
      "step": 26
    },
    {
      "epoch": 0.0027,
      "grad_norm": 7.029654502868652,
      "learning_rate": 5.2e-06,
      "loss": 12.4093,
      "step": 27
    },
    {
      "epoch": 0.0028,
      "grad_norm": 7.329478740692139,
      "learning_rate": 5.4e-06,
      "loss": 12.4061,
      "step": 28
    },
    {
      "epoch": 0.0029,
      "grad_norm": 7.242101669311523,
      "learning_rate": 5.600000000000001e-06,
      "loss": 12.3999,
      "step": 29
    },
    {
      "epoch": 0.003,
      "grad_norm": 6.528043270111084,
      "learning_rate": 5.8e-06,
      "loss": 12.4024,
      "step": 30
    },
    {
      "epoch": 0.0031,
      "grad_norm": 6.992286205291748,
      "learning_rate": 6e-06,
      "loss": 12.4045,
      "step": 31
    },
    {
      "epoch": 0.0032,
      "grad_norm": 7.579302787780762,
      "learning_rate": 6.2e-06,
      "loss": 12.3834,
      "step": 32
    },
    {
      "epoch": 0.0033,
      "grad_norm": 6.1326003074646,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 12.3799,
      "step": 33
    },
    {
      "epoch": 0.0034,
      "grad_norm": 8.036105155944824,
      "learning_rate": 6.6e-06,
      "loss": 12.3871,
      "step": 34
    },
    {
      "epoch": 0.0035,
      "grad_norm": 7.351873397827148,
      "learning_rate": 6.800000000000001e-06,
      "loss": 12.3786,
      "step": 35
    },
    {
      "epoch": 0.0036,
      "grad_norm": 7.456817626953125,
      "learning_rate": 7.000000000000001e-06,
      "loss": 12.3528,
      "step": 36
    },
    {
      "epoch": 0.0037,
      "grad_norm": 7.327516078948975,
      "learning_rate": 7.2e-06,
      "loss": 12.3616,
      "step": 37
    },
    {
      "epoch": 0.0038,
      "grad_norm": 7.461891174316406,
      "learning_rate": 7.4e-06,
      "loss": 12.3858,
      "step": 38
    },
    {
      "epoch": 0.0039,
      "grad_norm": 7.308258056640625,
      "learning_rate": 7.6e-06,
      "loss": 12.3584,
      "step": 39
    },
    {
      "epoch": 0.004,
      "grad_norm": 6.686831951141357,
      "learning_rate": 7.8e-06,
      "loss": 12.3353,
      "step": 40
    },
    {
      "epoch": 0.0041,
      "grad_norm": 7.446435928344727,
      "learning_rate": 8.000000000000001e-06,
      "loss": 12.3636,
      "step": 41
    },
    {
      "epoch": 0.0042,
      "grad_norm": 5.901037216186523,
      "learning_rate": 8.200000000000001e-06,
      "loss": 12.3344,
      "step": 42
    },
    {
      "epoch": 0.0043,
      "grad_norm": 6.683596611022949,
      "learning_rate": 8.400000000000001e-06,
      "loss": 12.3282,
      "step": 43
    },
    {
      "epoch": 0.0044,
      "grad_norm": 6.163360595703125,
      "learning_rate": 8.599999999999999e-06,
      "loss": 12.3255,
      "step": 44
    },
    {
      "epoch": 0.0045,
      "grad_norm": 8.35460090637207,
      "learning_rate": 8.8e-06,
      "loss": 12.3432,
      "step": 45
    },
    {
      "epoch": 0.0046,
      "grad_norm": 6.052265644073486,
      "learning_rate": 9e-06,
      "loss": 12.2904,
      "step": 46
    },
    {
      "epoch": 0.0047,
      "grad_norm": 6.396187782287598,
      "learning_rate": 9.2e-06,
      "loss": 12.3139,
      "step": 47
    },
    {
      "epoch": 0.0048,
      "grad_norm": 5.875204563140869,
      "learning_rate": 9.4e-06,
      "loss": 12.2806,
      "step": 48
    },
    {
      "epoch": 0.0049,
      "grad_norm": 7.3826518058776855,
      "learning_rate": 9.600000000000001e-06,
      "loss": 12.2917,
      "step": 49
    },
    {
      "epoch": 0.005,
      "grad_norm": 8.302424430847168,
      "learning_rate": 9.800000000000001e-06,
      "loss": 12.3034,
      "step": 50
    },
    {
      "epoch": 0.0051,
      "grad_norm": 7.690797328948975,
      "learning_rate": 1e-05,
      "loss": 12.2796,
      "step": 51
    },
    {
      "epoch": 0.0052,
      "grad_norm": 5.916553497314453,
      "learning_rate": 1.02e-05,
      "loss": 12.2682,
      "step": 52
    },
    {
      "epoch": 0.0053,
      "grad_norm": 7.8484883308410645,
      "learning_rate": 1.04e-05,
      "loss": 12.2703,
      "step": 53
    },
    {
      "epoch": 0.0054,
      "grad_norm": 8.073691368103027,
      "learning_rate": 1.06e-05,
      "loss": 12.2913,
      "step": 54
    },
    {
      "epoch": 0.0055,
      "grad_norm": 7.065008640289307,
      "learning_rate": 1.08e-05,
      "loss": 12.2318,
      "step": 55
    },
    {
      "epoch": 0.0056,
      "grad_norm": 15.4788179397583,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 12.2669,
      "step": 56
    },
    {
      "epoch": 0.0057,
      "grad_norm": 9.56401252746582,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 12.2787,
      "step": 57
    },
    {
      "epoch": 0.0058,
      "grad_norm": 6.665317535400391,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 12.2142,
      "step": 58
    },
    {
      "epoch": 0.0059,
      "grad_norm": 6.9803900718688965,
      "learning_rate": 1.16e-05,
      "loss": 12.1902,
      "step": 59
    },
    {
      "epoch": 0.006,
      "grad_norm": 7.7879180908203125,
      "learning_rate": 1.18e-05,
      "loss": 12.1964,
      "step": 60
    },
    {
      "epoch": 0.0061,
      "grad_norm": 8.847668647766113,
      "learning_rate": 1.2e-05,
      "loss": 12.2158,
      "step": 61
    },
    {
      "epoch": 0.0062,
      "grad_norm": 10.124570846557617,
      "learning_rate": 1.22e-05,
      "loss": 12.2747,
      "step": 62
    },
    {
      "epoch": 0.0063,
      "grad_norm": 8.071097373962402,
      "learning_rate": 1.24e-05,
      "loss": 12.1667,
      "step": 63
    },
    {
      "epoch": 0.0064,
      "grad_norm": 7.75615930557251,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 12.1609,
      "step": 64
    },
    {
      "epoch": 0.0065,
      "grad_norm": 8.9764404296875,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 12.1833,
      "step": 65
    },
    {
      "epoch": 0.0066,
      "grad_norm": 7.0341691970825195,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 12.1549,
      "step": 66
    },
    {
      "epoch": 0.0067,
      "grad_norm": 7.779499530792236,
      "learning_rate": 1.32e-05,
      "loss": 12.1345,
      "step": 67
    },
    {
      "epoch": 0.0068,
      "grad_norm": 7.3287034034729,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 12.114,
      "step": 68
    },
    {
      "epoch": 0.0069,
      "grad_norm": 7.638967990875244,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 12.1131,
      "step": 69
    },
    {
      "epoch": 0.007,
      "grad_norm": 10.987247467041016,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 12.1624,
      "step": 70
    },
    {
      "epoch": 0.0071,
      "grad_norm": 7.008084774017334,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 12.0939,
      "step": 71
    },
    {
      "epoch": 0.0072,
      "grad_norm": 8.07899284362793,
      "learning_rate": 1.42e-05,
      "loss": 12.1024,
      "step": 72
    },
    {
      "epoch": 0.0073,
      "grad_norm": 10.491318702697754,
      "learning_rate": 1.44e-05,
      "loss": 12.1197,
      "step": 73
    },
    {
      "epoch": 0.0074,
      "grad_norm": 9.416706085205078,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 12.1392,
      "step": 74
    },
    {
      "epoch": 0.0075,
      "grad_norm": 6.171147346496582,
      "learning_rate": 1.48e-05,
      "loss": 12.0406,
      "step": 75
    },
    {
      "epoch": 0.0076,
      "grad_norm": 10.032532691955566,
      "learning_rate": 1.5e-05,
      "loss": 12.0803,
      "step": 76
    },
    {
      "epoch": 0.0077,
      "grad_norm": 10.748550415039062,
      "learning_rate": 1.52e-05,
      "loss": 12.0797,
      "step": 77
    },
    {
      "epoch": 0.0078,
      "grad_norm": 8.922002792358398,
      "learning_rate": 1.54e-05,
      "loss": 12.0321,
      "step": 78
    },
    {
      "epoch": 0.0079,
      "grad_norm": 8.929490089416504,
      "learning_rate": 1.56e-05,
      "loss": 12.0328,
      "step": 79
    },
    {
      "epoch": 0.008,
      "grad_norm": 11.472333908081055,
      "learning_rate": 1.58e-05,
      "loss": 12.0714,
      "step": 80
    },
    {
      "epoch": 0.0081,
      "grad_norm": 7.373025894165039,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 12.011,
      "step": 81
    },
    {
      "epoch": 0.0082,
      "grad_norm": 7.574748516082764,
      "learning_rate": 1.62e-05,
      "loss": 12.0097,
      "step": 82
    },
    {
      "epoch": 0.0083,
      "grad_norm": 8.725872039794922,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 11.9503,
      "step": 83
    },
    {
      "epoch": 0.0084,
      "grad_norm": 8.428218841552734,
      "learning_rate": 1.66e-05,
      "loss": 11.9692,
      "step": 84
    },
    {
      "epoch": 0.0085,
      "grad_norm": 8.9564847946167,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 11.9684,
      "step": 85
    },
    {
      "epoch": 0.0086,
      "grad_norm": 10.238028526306152,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 11.9161,
      "step": 86
    },
    {
      "epoch": 0.0087,
      "grad_norm": 7.63020658493042,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 11.8994,
      "step": 87
    },
    {
      "epoch": 0.0088,
      "grad_norm": 8.209653854370117,
      "learning_rate": 1.74e-05,
      "loss": 11.8777,
      "step": 88
    },
    {
      "epoch": 0.0089,
      "grad_norm": 9.631011962890625,
      "learning_rate": 1.76e-05,
      "loss": 11.9323,
      "step": 89
    },
    {
      "epoch": 0.009,
      "grad_norm": 6.879038333892822,
      "learning_rate": 1.78e-05,
      "loss": 11.8361,
      "step": 90
    },
    {
      "epoch": 0.0091,
      "grad_norm": 6.362052917480469,
      "learning_rate": 1.8e-05,
      "loss": 11.861,
      "step": 91
    },
    {
      "epoch": 0.0092,
      "grad_norm": 7.433073043823242,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 11.8832,
      "step": 92
    },
    {
      "epoch": 0.0093,
      "grad_norm": 12.823978424072266,
      "learning_rate": 1.84e-05,
      "loss": 11.8983,
      "step": 93
    },
    {
      "epoch": 0.0094,
      "grad_norm": 12.312000274658203,
      "learning_rate": 1.86e-05,
      "loss": 11.8803,
      "step": 94
    },
    {
      "epoch": 0.0095,
      "grad_norm": 6.221044063568115,
      "learning_rate": 1.88e-05,
      "loss": 11.7602,
      "step": 95
    },
    {
      "epoch": 0.0096,
      "grad_norm": 10.041823387145996,
      "learning_rate": 1.9e-05,
      "loss": 11.7846,
      "step": 96
    },
    {
      "epoch": 0.0097,
      "grad_norm": 16.665407180786133,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 11.9107,
      "step": 97
    },
    {
      "epoch": 0.0098,
      "grad_norm": 9.316551208496094,
      "learning_rate": 1.94e-05,
      "loss": 11.7606,
      "step": 98
    },
    {
      "epoch": 0.0099,
      "grad_norm": 6.824377536773682,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 11.7421,
      "step": 99
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.777750015258789,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 11.7062,
      "step": 100
    },
    {
      "epoch": 0.0101,
      "grad_norm": 17.080942153930664,
      "learning_rate": 2e-05,
      "loss": 11.9486,
      "step": 101
    },
    {
      "epoch": 0.0102,
      "grad_norm": 7.386170387268066,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 11.6885,
      "step": 102
    },
    {
      "epoch": 0.0103,
      "grad_norm": 10.094419479370117,
      "learning_rate": 2.04e-05,
      "loss": 11.6644,
      "step": 103
    },
    {
      "epoch": 0.0104,
      "grad_norm": 8.52861213684082,
      "learning_rate": 2.06e-05,
      "loss": 11.6578,
      "step": 104
    },
    {
      "epoch": 0.0105,
      "grad_norm": 11.234184265136719,
      "learning_rate": 2.08e-05,
      "loss": 11.6629,
      "step": 105
    },
    {
      "epoch": 0.0106,
      "grad_norm": 7.200984001159668,
      "learning_rate": 2.1e-05,
      "loss": 11.6469,
      "step": 106
    },
    {
      "epoch": 0.0107,
      "grad_norm": 14.434544563293457,
      "learning_rate": 2.12e-05,
      "loss": 11.7055,
      "step": 107
    },
    {
      "epoch": 0.0108,
      "grad_norm": 6.877614498138428,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 11.6206,
      "step": 108
    },
    {
      "epoch": 0.0109,
      "grad_norm": 7.947735786437988,
      "learning_rate": 2.16e-05,
      "loss": 11.5621,
      "step": 109
    },
    {
      "epoch": 0.011,
      "grad_norm": 9.640480041503906,
      "learning_rate": 2.18e-05,
      "loss": 11.5411,
      "step": 110
    },
    {
      "epoch": 0.0111,
      "grad_norm": 11.469850540161133,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 11.5296,
      "step": 111
    },
    {
      "epoch": 0.0112,
      "grad_norm": 6.347490310668945,
      "learning_rate": 2.22e-05,
      "loss": 11.5281,
      "step": 112
    },
    {
      "epoch": 0.0113,
      "grad_norm": 8.148683547973633,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 11.4665,
      "step": 113
    },
    {
      "epoch": 0.0114,
      "grad_norm": 8.383275032043457,
      "learning_rate": 2.26e-05,
      "loss": 11.4783,
      "step": 114
    },
    {
      "epoch": 0.0115,
      "grad_norm": 7.427790641784668,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 11.4595,
      "step": 115
    },
    {
      "epoch": 0.0116,
      "grad_norm": 10.230530738830566,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 11.4625,
      "step": 116
    },
    {
      "epoch": 0.0117,
      "grad_norm": 10.093222618103027,
      "learning_rate": 2.32e-05,
      "loss": 11.4083,
      "step": 117
    },
    {
      "epoch": 0.0118,
      "grad_norm": 11.360919952392578,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 11.4651,
      "step": 118
    },
    {
      "epoch": 0.0119,
      "grad_norm": 8.506665229797363,
      "learning_rate": 2.36e-05,
      "loss": 11.4134,
      "step": 119
    },
    {
      "epoch": 0.012,
      "grad_norm": 7.99984884262085,
      "learning_rate": 2.38e-05,
      "loss": 11.3842,
      "step": 120
    },
    {
      "epoch": 0.0121,
      "grad_norm": 9.04259967803955,
      "learning_rate": 2.4e-05,
      "loss": 11.3274,
      "step": 121
    },
    {
      "epoch": 0.0122,
      "grad_norm": 10.163827896118164,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 11.3296,
      "step": 122
    },
    {
      "epoch": 0.0123,
      "grad_norm": 8.472930908203125,
      "learning_rate": 2.44e-05,
      "loss": 11.2634,
      "step": 123
    },
    {
      "epoch": 0.0124,
      "grad_norm": 11.81704330444336,
      "learning_rate": 2.46e-05,
      "loss": 11.3658,
      "step": 124
    },
    {
      "epoch": 0.0125,
      "grad_norm": 15.789381980895996,
      "learning_rate": 2.48e-05,
      "loss": 11.3452,
      "step": 125
    },
    {
      "epoch": 0.0126,
      "grad_norm": 9.865291595458984,
      "learning_rate": 2.5e-05,
      "loss": 11.2314,
      "step": 126
    },
    {
      "epoch": 0.0127,
      "grad_norm": 9.963654518127441,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 11.2316,
      "step": 127
    },
    {
      "epoch": 0.0128,
      "grad_norm": 13.623619079589844,
      "learning_rate": 2.54e-05,
      "loss": 11.2711,
      "step": 128
    },
    {
      "epoch": 0.0129,
      "grad_norm": 7.889054775238037,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 11.1528,
      "step": 129
    },
    {
      "epoch": 0.013,
      "grad_norm": 6.869718074798584,
      "learning_rate": 2.58e-05,
      "loss": 11.2294,
      "step": 130
    },
    {
      "epoch": 0.0131,
      "grad_norm": 9.579012870788574,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 11.1817,
      "step": 131
    },
    {
      "epoch": 0.0132,
      "grad_norm": 8.201512336730957,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 11.1245,
      "step": 132
    },
    {
      "epoch": 0.0133,
      "grad_norm": 14.649398803710938,
      "learning_rate": 2.64e-05,
      "loss": 11.1965,
      "step": 133
    },
    {
      "epoch": 0.0134,
      "grad_norm": 9.028801918029785,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 11.0718,
      "step": 134
    },
    {
      "epoch": 0.0135,
      "grad_norm": 9.22400951385498,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 11.1044,
      "step": 135
    },
    {
      "epoch": 0.0136,
      "grad_norm": 11.695323944091797,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 11.1226,
      "step": 136
    },
    {
      "epoch": 0.0137,
      "grad_norm": 18.60801124572754,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 11.1517,
      "step": 137
    },
    {
      "epoch": 0.0138,
      "grad_norm": 7.96170711517334,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 10.9643,
      "step": 138
    },
    {
      "epoch": 0.0139,
      "grad_norm": 7.9706034660339355,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 11.0232,
      "step": 139
    },
    {
      "epoch": 0.014,
      "grad_norm": 8.123888969421387,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 11.0075,
      "step": 140
    },
    {
      "epoch": 0.0141,
      "grad_norm": 7.1840009689331055,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 11.0506,
      "step": 141
    },
    {
      "epoch": 0.0142,
      "grad_norm": 8.799921989440918,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 10.8777,
      "step": 142
    },
    {
      "epoch": 0.0143,
      "grad_norm": 12.570575714111328,
      "learning_rate": 2.84e-05,
      "loss": 11.0882,
      "step": 143
    },
    {
      "epoch": 0.0144,
      "grad_norm": 7.866396427154541,
      "learning_rate": 2.86e-05,
      "loss": 10.8932,
      "step": 144
    },
    {
      "epoch": 0.0145,
      "grad_norm": 7.065332889556885,
      "learning_rate": 2.88e-05,
      "loss": 10.8959,
      "step": 145
    },
    {
      "epoch": 0.0146,
      "grad_norm": 6.486983776092529,
      "learning_rate": 2.9e-05,
      "loss": 10.8528,
      "step": 146
    },
    {
      "epoch": 0.0147,
      "grad_norm": 8.932045936584473,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 10.9119,
      "step": 147
    },
    {
      "epoch": 0.0148,
      "grad_norm": 9.881388664245605,
      "learning_rate": 2.94e-05,
      "loss": 10.8656,
      "step": 148
    },
    {
      "epoch": 0.0149,
      "grad_norm": 8.467541694641113,
      "learning_rate": 2.96e-05,
      "loss": 10.7078,
      "step": 149
    },
    {
      "epoch": 0.015,
      "grad_norm": 10.275186538696289,
      "learning_rate": 2.98e-05,
      "loss": 10.7888,
      "step": 150
    },
    {
      "epoch": 0.0151,
      "grad_norm": 8.369752883911133,
      "learning_rate": 3e-05,
      "loss": 10.7455,
      "step": 151
    },
    {
      "epoch": 0.0152,
      "grad_norm": 7.553621768951416,
      "learning_rate": 3.02e-05,
      "loss": 10.6653,
      "step": 152
    },
    {
      "epoch": 0.0153,
      "grad_norm": 9.519147872924805,
      "learning_rate": 3.04e-05,
      "loss": 10.654,
      "step": 153
    },
    {
      "epoch": 0.0154,
      "grad_norm": 6.853105068206787,
      "learning_rate": 3.06e-05,
      "loss": 10.6551,
      "step": 154
    },
    {
      "epoch": 0.0155,
      "grad_norm": 6.395466327667236,
      "learning_rate": 3.08e-05,
      "loss": 10.6979,
      "step": 155
    },
    {
      "epoch": 0.0156,
      "grad_norm": 14.461307525634766,
      "learning_rate": 3.1e-05,
      "loss": 10.8024,
      "step": 156
    },
    {
      "epoch": 0.0157,
      "grad_norm": 10.691741943359375,
      "learning_rate": 3.12e-05,
      "loss": 10.5997,
      "step": 157
    },
    {
      "epoch": 0.0158,
      "grad_norm": 8.110105514526367,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 10.5545,
      "step": 158
    },
    {
      "epoch": 0.0159,
      "grad_norm": 9.46684741973877,
      "learning_rate": 3.16e-05,
      "loss": 10.6229,
      "step": 159
    },
    {
      "epoch": 0.016,
      "grad_norm": 9.2418794631958,
      "learning_rate": 3.18e-05,
      "loss": 10.5639,
      "step": 160
    },
    {
      "epoch": 0.0161,
      "grad_norm": 9.659248352050781,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 10.4789,
      "step": 161
    },
    {
      "epoch": 0.0162,
      "grad_norm": 9.653425216674805,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 10.4954,
      "step": 162
    },
    {
      "epoch": 0.0163,
      "grad_norm": 8.81847858428955,
      "learning_rate": 3.24e-05,
      "loss": 10.5362,
      "step": 163
    },
    {
      "epoch": 0.0164,
      "grad_norm": 6.982722759246826,
      "learning_rate": 3.26e-05,
      "loss": 10.4507,
      "step": 164
    },
    {
      "epoch": 0.0165,
      "grad_norm": 9.959033966064453,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 10.4462,
      "step": 165
    },
    {
      "epoch": 0.0166,
      "grad_norm": 6.076030731201172,
      "learning_rate": 3.3e-05,
      "loss": 10.3688,
      "step": 166
    },
    {
      "epoch": 0.0167,
      "grad_norm": 7.735734462738037,
      "learning_rate": 3.32e-05,
      "loss": 10.3691,
      "step": 167
    },
    {
      "epoch": 0.0168,
      "grad_norm": 6.742389678955078,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 10.3805,
      "step": 168
    },
    {
      "epoch": 0.0169,
      "grad_norm": 6.332170009613037,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 10.301,
      "step": 169
    },
    {
      "epoch": 0.017,
      "grad_norm": 14.718074798583984,
      "learning_rate": 3.38e-05,
      "loss": 10.2893,
      "step": 170
    },
    {
      "epoch": 0.0171,
      "grad_norm": 8.309751510620117,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 10.4881,
      "step": 171
    },
    {
      "epoch": 0.0172,
      "grad_norm": 14.965339660644531,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 10.3793,
      "step": 172
    },
    {
      "epoch": 0.0173,
      "grad_norm": 7.8455376625061035,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 10.3215,
      "step": 173
    },
    {
      "epoch": 0.0174,
      "grad_norm": 8.765157699584961,
      "learning_rate": 3.46e-05,
      "loss": 10.2315,
      "step": 174
    },
    {
      "epoch": 0.0175,
      "grad_norm": 9.7735013961792,
      "learning_rate": 3.48e-05,
      "loss": 10.2395,
      "step": 175
    },
    {
      "epoch": 0.0176,
      "grad_norm": 12.812241554260254,
      "learning_rate": 3.5e-05,
      "loss": 10.2745,
      "step": 176
    },
    {
      "epoch": 0.0177,
      "grad_norm": 10.647100448608398,
      "learning_rate": 3.52e-05,
      "loss": 10.1963,
      "step": 177
    },
    {
      "epoch": 0.0178,
      "grad_norm": 8.149502754211426,
      "learning_rate": 3.54e-05,
      "loss": 10.0515,
      "step": 178
    },
    {
      "epoch": 0.0179,
      "grad_norm": 13.465110778808594,
      "learning_rate": 3.56e-05,
      "loss": 10.264,
      "step": 179
    },
    {
      "epoch": 0.018,
      "grad_norm": 17.19188690185547,
      "learning_rate": 3.58e-05,
      "loss": 10.2873,
      "step": 180
    },
    {
      "epoch": 0.0181,
      "grad_norm": 7.953618049621582,
      "learning_rate": 3.6e-05,
      "loss": 10.1318,
      "step": 181
    },
    {
      "epoch": 0.0182,
      "grad_norm": 7.22465705871582,
      "learning_rate": 3.62e-05,
      "loss": 10.0534,
      "step": 182
    },
    {
      "epoch": 0.0183,
      "grad_norm": 8.814151763916016,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 10.1745,
      "step": 183
    },
    {
      "epoch": 0.0184,
      "grad_norm": 9.17065143585205,
      "learning_rate": 3.66e-05,
      "loss": 9.9745,
      "step": 184
    },
    {
      "epoch": 0.0185,
      "grad_norm": 9.2052001953125,
      "learning_rate": 3.68e-05,
      "loss": 10.0529,
      "step": 185
    },
    {
      "epoch": 0.0186,
      "grad_norm": 6.389113903045654,
      "learning_rate": 3.7e-05,
      "loss": 9.9924,
      "step": 186
    },
    {
      "epoch": 0.0187,
      "grad_norm": 11.894957542419434,
      "learning_rate": 3.72e-05,
      "loss": 10.0943,
      "step": 187
    },
    {
      "epoch": 0.0188,
      "grad_norm": 13.301338195800781,
      "learning_rate": 3.74e-05,
      "loss": 10.1091,
      "step": 188
    },
    {
      "epoch": 0.0189,
      "grad_norm": 6.577731132507324,
      "learning_rate": 3.76e-05,
      "loss": 9.9023,
      "step": 189
    },
    {
      "epoch": 0.019,
      "grad_norm": 5.2935004234313965,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 10.0472,
      "step": 190
    },
    {
      "epoch": 0.0191,
      "grad_norm": 8.200553894042969,
      "learning_rate": 3.8e-05,
      "loss": 9.9429,
      "step": 191
    },
    {
      "epoch": 0.0192,
      "grad_norm": 11.951884269714355,
      "learning_rate": 3.82e-05,
      "loss": 10.0683,
      "step": 192
    },
    {
      "epoch": 0.0193,
      "grad_norm": 5.687671184539795,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 10.0969,
      "step": 193
    },
    {
      "epoch": 0.0194,
      "grad_norm": 7.080389022827148,
      "learning_rate": 3.86e-05,
      "loss": 9.8782,
      "step": 194
    },
    {
      "epoch": 0.0195,
      "grad_norm": 8.110967636108398,
      "learning_rate": 3.88e-05,
      "loss": 9.8476,
      "step": 195
    },
    {
      "epoch": 0.0196,
      "grad_norm": 6.23143196105957,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 9.8479,
      "step": 196
    },
    {
      "epoch": 0.0197,
      "grad_norm": 5.930812358856201,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 9.8195,
      "step": 197
    },
    {
      "epoch": 0.0198,
      "grad_norm": 6.722044467926025,
      "learning_rate": 3.94e-05,
      "loss": 9.7734,
      "step": 198
    },
    {
      "epoch": 0.0199,
      "grad_norm": 16.948867797851562,
      "learning_rate": 3.960000000000001e-05,
      "loss": 9.8013,
      "step": 199
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.146793842315674,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 9.7604,
      "step": 200
    },
    {
      "epoch": 0.0201,
      "grad_norm": 8.838726043701172,
      "learning_rate": 4e-05,
      "loss": 9.8933,
      "step": 201
    },
    {
      "epoch": 0.0202,
      "grad_norm": 5.589120388031006,
      "learning_rate": 4.02e-05,
      "loss": 9.7714,
      "step": 202
    },
    {
      "epoch": 0.0203,
      "grad_norm": 7.171775817871094,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 9.6472,
      "step": 203
    },
    {
      "epoch": 0.0204,
      "grad_norm": 11.032878875732422,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 9.6872,
      "step": 204
    },
    {
      "epoch": 0.0205,
      "grad_norm": 6.188409805297852,
      "learning_rate": 4.08e-05,
      "loss": 10.0015,
      "step": 205
    },
    {
      "epoch": 0.0206,
      "grad_norm": 5.660280227661133,
      "learning_rate": 4.1e-05,
      "loss": 9.7019,
      "step": 206
    },
    {
      "epoch": 0.0207,
      "grad_norm": 8.25145149230957,
      "learning_rate": 4.12e-05,
      "loss": 9.7575,
      "step": 207
    },
    {
      "epoch": 0.0208,
      "grad_norm": 8.187992095947266,
      "learning_rate": 4.14e-05,
      "loss": 9.8082,
      "step": 208
    },
    {
      "epoch": 0.0209,
      "grad_norm": 5.932746887207031,
      "learning_rate": 4.16e-05,
      "loss": 9.4449,
      "step": 209
    },
    {
      "epoch": 0.021,
      "grad_norm": 6.103842735290527,
      "learning_rate": 4.18e-05,
      "loss": 9.6804,
      "step": 210
    },
    {
      "epoch": 0.0211,
      "grad_norm": 6.00308084487915,
      "learning_rate": 4.2e-05,
      "loss": 9.5816,
      "step": 211
    },
    {
      "epoch": 0.0212,
      "grad_norm": 6.490203380584717,
      "learning_rate": 4.22e-05,
      "loss": 9.6474,
      "step": 212
    },
    {
      "epoch": 0.0213,
      "grad_norm": 5.3530592918396,
      "learning_rate": 4.24e-05,
      "loss": 9.5447,
      "step": 213
    },
    {
      "epoch": 0.0214,
      "grad_norm": 9.077552795410156,
      "learning_rate": 4.26e-05,
      "loss": 9.5363,
      "step": 214
    },
    {
      "epoch": 0.0215,
      "grad_norm": 8.499333381652832,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 9.512,
      "step": 215
    },
    {
      "epoch": 0.0216,
      "grad_norm": 6.700957775115967,
      "learning_rate": 4.3e-05,
      "loss": 9.3819,
      "step": 216
    },
    {
      "epoch": 0.0217,
      "grad_norm": 5.201712608337402,
      "learning_rate": 4.32e-05,
      "loss": 9.4518,
      "step": 217
    },
    {
      "epoch": 0.0218,
      "grad_norm": 4.531014919281006,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 9.8229,
      "step": 218
    },
    {
      "epoch": 0.0219,
      "grad_norm": 5.866469860076904,
      "learning_rate": 4.36e-05,
      "loss": 9.2163,
      "step": 219
    },
    {
      "epoch": 0.022,
      "grad_norm": 17.529659271240234,
      "learning_rate": 4.38e-05,
      "loss": 9.9749,
      "step": 220
    },
    {
      "epoch": 0.0221,
      "grad_norm": 7.516631603240967,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 9.4473,
      "step": 221
    },
    {
      "epoch": 0.0222,
      "grad_norm": 8.116442680358887,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 9.4438,
      "step": 222
    },
    {
      "epoch": 0.0223,
      "grad_norm": 8.387486457824707,
      "learning_rate": 4.44e-05,
      "loss": 9.4132,
      "step": 223
    },
    {
      "epoch": 0.0224,
      "grad_norm": 7.383880138397217,
      "learning_rate": 4.46e-05,
      "loss": 9.3519,
      "step": 224
    },
    {
      "epoch": 0.0225,
      "grad_norm": 6.597602367401123,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 9.2682,
      "step": 225
    },
    {
      "epoch": 0.0226,
      "grad_norm": 11.548916816711426,
      "learning_rate": 4.5e-05,
      "loss": 9.3038,
      "step": 226
    },
    {
      "epoch": 0.0227,
      "grad_norm": 6.778423309326172,
      "learning_rate": 4.52e-05,
      "loss": 9.0375,
      "step": 227
    },
    {
      "epoch": 0.0228,
      "grad_norm": 6.946712017059326,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 9.304,
      "step": 228
    },
    {
      "epoch": 0.0229,
      "grad_norm": 7.708919525146484,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 9.3768,
      "step": 229
    },
    {
      "epoch": 0.023,
      "grad_norm": 6.78889799118042,
      "learning_rate": 4.58e-05,
      "loss": 9.3081,
      "step": 230
    },
    {
      "epoch": 0.0231,
      "grad_norm": 5.711223602294922,
      "learning_rate": 4.600000000000001e-05,
      "loss": 9.3642,
      "step": 231
    },
    {
      "epoch": 0.0232,
      "grad_norm": 6.024174690246582,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 9.3597,
      "step": 232
    },
    {
      "epoch": 0.0233,
      "grad_norm": 6.322536945343018,
      "learning_rate": 4.64e-05,
      "loss": 9.0193,
      "step": 233
    },
    {
      "epoch": 0.0234,
      "grad_norm": 6.055264949798584,
      "learning_rate": 4.660000000000001e-05,
      "loss": 9.0955,
      "step": 234
    },
    {
      "epoch": 0.0235,
      "grad_norm": 5.709654331207275,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 9.1688,
      "step": 235
    },
    {
      "epoch": 0.0236,
      "grad_norm": 4.921345233917236,
      "learning_rate": 4.7e-05,
      "loss": 9.1013,
      "step": 236
    },
    {
      "epoch": 0.0237,
      "grad_norm": 5.05027961730957,
      "learning_rate": 4.72e-05,
      "loss": 9.127,
      "step": 237
    },
    {
      "epoch": 0.0238,
      "grad_norm": 5.0082807540893555,
      "learning_rate": 4.74e-05,
      "loss": 9.0422,
      "step": 238
    },
    {
      "epoch": 0.0239,
      "grad_norm": 6.666208267211914,
      "learning_rate": 4.76e-05,
      "loss": 9.1476,
      "step": 239
    },
    {
      "epoch": 0.024,
      "grad_norm": 6.276179790496826,
      "learning_rate": 4.78e-05,
      "loss": 8.9056,
      "step": 240
    },
    {
      "epoch": 0.0241,
      "grad_norm": 7.6879072189331055,
      "learning_rate": 4.8e-05,
      "loss": 8.8786,
      "step": 241
    },
    {
      "epoch": 0.0242,
      "grad_norm": 12.775483131408691,
      "learning_rate": 4.82e-05,
      "loss": 9.2336,
      "step": 242
    },
    {
      "epoch": 0.0243,
      "grad_norm": 6.869771957397461,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 9.1087,
      "step": 243
    },
    {
      "epoch": 0.0244,
      "grad_norm": 6.10952091217041,
      "learning_rate": 4.86e-05,
      "loss": 9.1639,
      "step": 244
    },
    {
      "epoch": 0.0245,
      "grad_norm": 8.490113258361816,
      "learning_rate": 4.88e-05,
      "loss": 9.1251,
      "step": 245
    },
    {
      "epoch": 0.0246,
      "grad_norm": 7.3618974685668945,
      "learning_rate": 4.9e-05,
      "loss": 8.6857,
      "step": 246
    },
    {
      "epoch": 0.0247,
      "grad_norm": 5.742708206176758,
      "learning_rate": 4.92e-05,
      "loss": 9.1922,
      "step": 247
    },
    {
      "epoch": 0.0248,
      "grad_norm": 6.574678421020508,
      "learning_rate": 4.94e-05,
      "loss": 9.1139,
      "step": 248
    },
    {
      "epoch": 0.0249,
      "grad_norm": 4.998880863189697,
      "learning_rate": 4.96e-05,
      "loss": 9.2472,
      "step": 249
    },
    {
      "epoch": 0.025,
      "grad_norm": 7.934985637664795,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 8.813,
      "step": 250
    },
    {
      "epoch": 0.0251,
      "grad_norm": 6.892264366149902,
      "learning_rate": 5e-05,
      "loss": 8.7634,
      "step": 251
    },
    {
      "epoch": 0.0252,
      "grad_norm": 4.708669662475586,
      "learning_rate": 5.02e-05,
      "loss": 9.018,
      "step": 252
    },
    {
      "epoch": 0.0253,
      "grad_norm": 6.535646915435791,
      "learning_rate": 5.0400000000000005e-05,
      "loss": 8.9307,
      "step": 253
    },
    {
      "epoch": 0.0254,
      "grad_norm": 5.925316333770752,
      "learning_rate": 5.0600000000000003e-05,
      "loss": 9.0097,
      "step": 254
    },
    {
      "epoch": 0.0255,
      "grad_norm": 5.313938617706299,
      "learning_rate": 5.08e-05,
      "loss": 8.9223,
      "step": 255
    },
    {
      "epoch": 0.0256,
      "grad_norm": 5.200139999389648,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 8.9244,
      "step": 256
    },
    {
      "epoch": 0.0257,
      "grad_norm": 7.309813022613525,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 8.8348,
      "step": 257
    },
    {
      "epoch": 0.0258,
      "grad_norm": 7.439190864562988,
      "learning_rate": 5.14e-05,
      "loss": 8.6878,
      "step": 258
    },
    {
      "epoch": 0.0259,
      "grad_norm": 14.85867977142334,
      "learning_rate": 5.16e-05,
      "loss": 9.1922,
      "step": 259
    },
    {
      "epoch": 0.026,
      "grad_norm": 5.545986175537109,
      "learning_rate": 5.1800000000000005e-05,
      "loss": 8.9029,
      "step": 260
    },
    {
      "epoch": 0.0261,
      "grad_norm": 6.241980075836182,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 8.6472,
      "step": 261
    },
    {
      "epoch": 0.0262,
      "grad_norm": 5.471828937530518,
      "learning_rate": 5.22e-05,
      "loss": 8.5644,
      "step": 262
    },
    {
      "epoch": 0.0263,
      "grad_norm": 6.046511173248291,
      "learning_rate": 5.2400000000000007e-05,
      "loss": 9.0201,
      "step": 263
    },
    {
      "epoch": 0.0264,
      "grad_norm": 5.293769836425781,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 8.7929,
      "step": 264
    },
    {
      "epoch": 0.0265,
      "grad_norm": 4.203950881958008,
      "learning_rate": 5.28e-05,
      "loss": 8.621,
      "step": 265
    },
    {
      "epoch": 0.0266,
      "grad_norm": 8.335159301757812,
      "learning_rate": 5.300000000000001e-05,
      "loss": 8.7987,
      "step": 266
    },
    {
      "epoch": 0.0267,
      "grad_norm": 6.446079730987549,
      "learning_rate": 5.3200000000000006e-05,
      "loss": 8.5557,
      "step": 267
    },
    {
      "epoch": 0.0268,
      "grad_norm": 6.423769950866699,
      "learning_rate": 5.3400000000000004e-05,
      "loss": 8.6886,
      "step": 268
    },
    {
      "epoch": 0.0269,
      "grad_norm": 7.578753471374512,
      "learning_rate": 5.360000000000001e-05,
      "loss": 8.3827,
      "step": 269
    },
    {
      "epoch": 0.027,
      "grad_norm": 6.533254623413086,
      "learning_rate": 5.380000000000001e-05,
      "loss": 8.7582,
      "step": 270
    },
    {
      "epoch": 0.0271,
      "grad_norm": 5.45490026473999,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 8.7272,
      "step": 271
    },
    {
      "epoch": 0.0272,
      "grad_norm": 3.8768043518066406,
      "learning_rate": 5.420000000000001e-05,
      "loss": 8.72,
      "step": 272
    },
    {
      "epoch": 0.0273,
      "grad_norm": 5.166446685791016,
      "learning_rate": 5.440000000000001e-05,
      "loss": 8.7967,
      "step": 273
    },
    {
      "epoch": 0.0274,
      "grad_norm": 4.8068695068359375,
      "learning_rate": 5.4600000000000006e-05,
      "loss": 8.5376,
      "step": 274
    },
    {
      "epoch": 0.0275,
      "grad_norm": 5.647087574005127,
      "learning_rate": 5.4800000000000004e-05,
      "loss": 8.5172,
      "step": 275
    },
    {
      "epoch": 0.0276,
      "grad_norm": 6.344799041748047,
      "learning_rate": 5.500000000000001e-05,
      "loss": 8.7516,
      "step": 276
    },
    {
      "epoch": 0.0277,
      "grad_norm": 5.906855583190918,
      "learning_rate": 5.520000000000001e-05,
      "loss": 8.2304,
      "step": 277
    },
    {
      "epoch": 0.0278,
      "grad_norm": 4.516210079193115,
      "learning_rate": 5.5400000000000005e-05,
      "loss": 8.8706,
      "step": 278
    },
    {
      "epoch": 0.0279,
      "grad_norm": 4.091322422027588,
      "learning_rate": 5.560000000000001e-05,
      "loss": 8.4931,
      "step": 279
    },
    {
      "epoch": 0.028,
      "grad_norm": 4.637244701385498,
      "learning_rate": 5.580000000000001e-05,
      "loss": 8.7217,
      "step": 280
    },
    {
      "epoch": 0.0281,
      "grad_norm": 5.489957332611084,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 8.2501,
      "step": 281
    },
    {
      "epoch": 0.0282,
      "grad_norm": 4.706979274749756,
      "learning_rate": 5.620000000000001e-05,
      "loss": 8.4724,
      "step": 282
    },
    {
      "epoch": 0.0283,
      "grad_norm": 6.563615322113037,
      "learning_rate": 5.6399999999999995e-05,
      "loss": 8.3902,
      "step": 283
    },
    {
      "epoch": 0.0284,
      "grad_norm": 3.5792253017425537,
      "learning_rate": 5.66e-05,
      "loss": 8.5265,
      "step": 284
    },
    {
      "epoch": 0.0285,
      "grad_norm": 4.14057731628418,
      "learning_rate": 5.68e-05,
      "loss": 8.4717,
      "step": 285
    },
    {
      "epoch": 0.0286,
      "grad_norm": 5.453494548797607,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 8.1893,
      "step": 286
    },
    {
      "epoch": 0.0287,
      "grad_norm": 5.685511112213135,
      "learning_rate": 5.72e-05,
      "loss": 8.4376,
      "step": 287
    },
    {
      "epoch": 0.0288,
      "grad_norm": 4.2895708084106445,
      "learning_rate": 5.74e-05,
      "loss": 8.499,
      "step": 288
    },
    {
      "epoch": 0.0289,
      "grad_norm": 4.025040626525879,
      "learning_rate": 5.76e-05,
      "loss": 8.2809,
      "step": 289
    },
    {
      "epoch": 0.029,
      "grad_norm": 5.4593377113342285,
      "learning_rate": 5.7799999999999995e-05,
      "loss": 8.4376,
      "step": 290
    },
    {
      "epoch": 0.0291,
      "grad_norm": 3.8860225677490234,
      "learning_rate": 5.8e-05,
      "loss": 8.5404,
      "step": 291
    },
    {
      "epoch": 0.0292,
      "grad_norm": 4.334435939788818,
      "learning_rate": 5.82e-05,
      "loss": 8.9367,
      "step": 292
    },
    {
      "epoch": 0.0293,
      "grad_norm": 4.778171062469482,
      "learning_rate": 5.8399999999999997e-05,
      "loss": 8.6325,
      "step": 293
    },
    {
      "epoch": 0.0294,
      "grad_norm": 6.4139251708984375,
      "learning_rate": 5.86e-05,
      "loss": 8.4343,
      "step": 294
    },
    {
      "epoch": 0.0295,
      "grad_norm": 6.983043193817139,
      "learning_rate": 5.88e-05,
      "loss": 8.1189,
      "step": 295
    },
    {
      "epoch": 0.0296,
      "grad_norm": 4.050938129425049,
      "learning_rate": 5.9e-05,
      "loss": 8.4417,
      "step": 296
    },
    {
      "epoch": 0.0297,
      "grad_norm": 2.8540124893188477,
      "learning_rate": 5.92e-05,
      "loss": 8.593,
      "step": 297
    },
    {
      "epoch": 0.0298,
      "grad_norm": 3.8843994140625,
      "learning_rate": 5.94e-05,
      "loss": 8.3702,
      "step": 298
    },
    {
      "epoch": 0.0299,
      "grad_norm": 5.090415954589844,
      "learning_rate": 5.96e-05,
      "loss": 8.5051,
      "step": 299
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.39723801612854,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 8.4562,
      "step": 300
    },
    {
      "epoch": 0.0301,
      "grad_norm": 3.2056491374969482,
      "learning_rate": 6e-05,
      "loss": 8.4768,
      "step": 301
    },
    {
      "epoch": 0.0302,
      "grad_norm": 3.8233134746551514,
      "learning_rate": 6.02e-05,
      "loss": 8.2285,
      "step": 302
    },
    {
      "epoch": 0.0303,
      "grad_norm": 3.8310599327087402,
      "learning_rate": 6.04e-05,
      "loss": 8.4285,
      "step": 303
    },
    {
      "epoch": 0.0304,
      "grad_norm": 4.808689594268799,
      "learning_rate": 6.06e-05,
      "loss": 8.2477,
      "step": 304
    },
    {
      "epoch": 0.0305,
      "grad_norm": 4.7010040283203125,
      "learning_rate": 6.08e-05,
      "loss": 8.2645,
      "step": 305
    },
    {
      "epoch": 0.0306,
      "grad_norm": 4.380141735076904,
      "learning_rate": 6.1e-05,
      "loss": 8.2772,
      "step": 306
    },
    {
      "epoch": 0.0307,
      "grad_norm": 4.442428112030029,
      "learning_rate": 6.12e-05,
      "loss": 8.0554,
      "step": 307
    },
    {
      "epoch": 0.0308,
      "grad_norm": 3.875422716140747,
      "learning_rate": 6.14e-05,
      "loss": 8.1014,
      "step": 308
    },
    {
      "epoch": 0.0309,
      "grad_norm": 3.48521089553833,
      "learning_rate": 6.16e-05,
      "loss": 8.3712,
      "step": 309
    },
    {
      "epoch": 0.031,
      "grad_norm": 5.262283802032471,
      "learning_rate": 6.18e-05,
      "loss": 7.8291,
      "step": 310
    },
    {
      "epoch": 0.0311,
      "grad_norm": 3.3017733097076416,
      "learning_rate": 6.2e-05,
      "loss": 8.2331,
      "step": 311
    },
    {
      "epoch": 0.0312,
      "grad_norm": 2.7119545936584473,
      "learning_rate": 6.220000000000001e-05,
      "loss": 8.2557,
      "step": 312
    },
    {
      "epoch": 0.0313,
      "grad_norm": 3.9138364791870117,
      "learning_rate": 6.24e-05,
      "loss": 7.8825,
      "step": 313
    },
    {
      "epoch": 0.0314,
      "grad_norm": 3.1505379676818848,
      "learning_rate": 6.26e-05,
      "loss": 8.1253,
      "step": 314
    },
    {
      "epoch": 0.0315,
      "grad_norm": 2.4292092323303223,
      "learning_rate": 6.280000000000001e-05,
      "loss": 8.5786,
      "step": 315
    },
    {
      "epoch": 0.0316,
      "grad_norm": 3.7229034900665283,
      "learning_rate": 6.3e-05,
      "loss": 7.7387,
      "step": 316
    },
    {
      "epoch": 0.0317,
      "grad_norm": 3.5769152641296387,
      "learning_rate": 6.32e-05,
      "loss": 7.8653,
      "step": 317
    },
    {
      "epoch": 0.0318,
      "grad_norm": 4.328640937805176,
      "learning_rate": 6.340000000000001e-05,
      "loss": 8.4092,
      "step": 318
    },
    {
      "epoch": 0.0319,
      "grad_norm": 3.600282907485962,
      "learning_rate": 6.36e-05,
      "loss": 7.9232,
      "step": 319
    },
    {
      "epoch": 0.032,
      "grad_norm": 2.1181747913360596,
      "learning_rate": 6.38e-05,
      "loss": 8.2072,
      "step": 320
    },
    {
      "epoch": 0.0321,
      "grad_norm": 3.467195987701416,
      "learning_rate": 6.400000000000001e-05,
      "loss": 8.306,
      "step": 321
    },
    {
      "epoch": 0.0322,
      "grad_norm": 2.462843179702759,
      "learning_rate": 6.42e-05,
      "loss": 8.3079,
      "step": 322
    },
    {
      "epoch": 0.0323,
      "grad_norm": 2.9807193279266357,
      "learning_rate": 6.440000000000001e-05,
      "loss": 7.6941,
      "step": 323
    },
    {
      "epoch": 0.0324,
      "grad_norm": 2.629209041595459,
      "learning_rate": 6.460000000000001e-05,
      "loss": 8.2518,
      "step": 324
    },
    {
      "epoch": 0.0325,
      "grad_norm": 2.9393324851989746,
      "learning_rate": 6.48e-05,
      "loss": 7.8852,
      "step": 325
    },
    {
      "epoch": 0.0326,
      "grad_norm": 3.363758087158203,
      "learning_rate": 6.500000000000001e-05,
      "loss": 8.2548,
      "step": 326
    },
    {
      "epoch": 0.0327,
      "grad_norm": 2.108536720275879,
      "learning_rate": 6.52e-05,
      "loss": 8.3784,
      "step": 327
    },
    {
      "epoch": 0.0328,
      "grad_norm": 2.9645400047302246,
      "learning_rate": 6.54e-05,
      "loss": 7.7275,
      "step": 328
    },
    {
      "epoch": 0.0329,
      "grad_norm": 3.1026885509490967,
      "learning_rate": 6.560000000000001e-05,
      "loss": 7.8511,
      "step": 329
    },
    {
      "epoch": 0.033,
      "grad_norm": 2.568486452102661,
      "learning_rate": 6.58e-05,
      "loss": 7.8486,
      "step": 330
    },
    {
      "epoch": 0.0331,
      "grad_norm": 2.4917151927948,
      "learning_rate": 6.6e-05,
      "loss": 7.7493,
      "step": 331
    },
    {
      "epoch": 0.0332,
      "grad_norm": 1.824112892150879,
      "learning_rate": 6.620000000000001e-05,
      "loss": 8.0865,
      "step": 332
    },
    {
      "epoch": 0.0333,
      "grad_norm": 1.7656618356704712,
      "learning_rate": 6.64e-05,
      "loss": 8.1276,
      "step": 333
    },
    {
      "epoch": 0.0334,
      "grad_norm": 2.609914779663086,
      "learning_rate": 6.66e-05,
      "loss": 7.6964,
      "step": 334
    },
    {
      "epoch": 0.0335,
      "grad_norm": 2.6417877674102783,
      "learning_rate": 6.680000000000001e-05,
      "loss": 7.9889,
      "step": 335
    },
    {
      "epoch": 0.0336,
      "grad_norm": 2.1247611045837402,
      "learning_rate": 6.7e-05,
      "loss": 7.9481,
      "step": 336
    },
    {
      "epoch": 0.0337,
      "grad_norm": 1.853927493095398,
      "learning_rate": 6.720000000000001e-05,
      "loss": 8.0781,
      "step": 337
    },
    {
      "epoch": 0.0338,
      "grad_norm": 4.823842525482178,
      "learning_rate": 6.740000000000001e-05,
      "loss": 7.8172,
      "step": 338
    },
    {
      "epoch": 0.0339,
      "grad_norm": 1.6947994232177734,
      "learning_rate": 6.76e-05,
      "loss": 8.4889,
      "step": 339
    },
    {
      "epoch": 0.034,
      "grad_norm": 2.0474886894226074,
      "learning_rate": 6.780000000000001e-05,
      "loss": 8.0308,
      "step": 340
    },
    {
      "epoch": 0.0341,
      "grad_norm": 2.0490152835845947,
      "learning_rate": 6.800000000000001e-05,
      "loss": 7.7597,
      "step": 341
    },
    {
      "epoch": 0.0342,
      "grad_norm": 2.6015026569366455,
      "learning_rate": 6.82e-05,
      "loss": 7.8971,
      "step": 342
    },
    {
      "epoch": 0.0343,
      "grad_norm": 2.5225143432617188,
      "learning_rate": 6.840000000000001e-05,
      "loss": 7.9424,
      "step": 343
    },
    {
      "epoch": 0.0344,
      "grad_norm": 1.486202359199524,
      "learning_rate": 6.860000000000001e-05,
      "loss": 8.1862,
      "step": 344
    },
    {
      "epoch": 0.0345,
      "grad_norm": 1.7903677225112915,
      "learning_rate": 6.879999999999999e-05,
      "loss": 8.279,
      "step": 345
    },
    {
      "epoch": 0.0346,
      "grad_norm": 1.3540446758270264,
      "learning_rate": 6.9e-05,
      "loss": 8.0958,
      "step": 346
    },
    {
      "epoch": 0.0347,
      "grad_norm": 1.6259207725524902,
      "learning_rate": 6.92e-05,
      "loss": 7.8419,
      "step": 347
    },
    {
      "epoch": 0.0348,
      "grad_norm": 1.3999181985855103,
      "learning_rate": 6.939999999999999e-05,
      "loss": 8.1124,
      "step": 348
    },
    {
      "epoch": 0.0349,
      "grad_norm": 1.6490485668182373,
      "learning_rate": 6.96e-05,
      "loss": 8.1367,
      "step": 349
    },
    {
      "epoch": 0.035,
      "grad_norm": 1.2689920663833618,
      "learning_rate": 6.98e-05,
      "loss": 7.9782,
      "step": 350
    },
    {
      "epoch": 0.0351,
      "grad_norm": 5.19437837600708,
      "learning_rate": 7e-05,
      "loss": 7.9573,
      "step": 351
    },
    {
      "epoch": 0.0352,
      "grad_norm": 2.0510149002075195,
      "learning_rate": 7.02e-05,
      "loss": 7.5297,
      "step": 352
    },
    {
      "epoch": 0.0353,
      "grad_norm": 1.3898564577102661,
      "learning_rate": 7.04e-05,
      "loss": 8.1872,
      "step": 353
    },
    {
      "epoch": 0.0354,
      "grad_norm": 1.2546610832214355,
      "learning_rate": 7.06e-05,
      "loss": 8.068,
      "step": 354
    },
    {
      "epoch": 0.0355,
      "grad_norm": 1.6638667583465576,
      "learning_rate": 7.08e-05,
      "loss": 8.2714,
      "step": 355
    },
    {
      "epoch": 0.0356,
      "grad_norm": 1.74191153049469,
      "learning_rate": 7.1e-05,
      "loss": 7.6859,
      "step": 356
    },
    {
      "epoch": 0.0357,
      "grad_norm": 1.7582992315292358,
      "learning_rate": 7.12e-05,
      "loss": 7.785,
      "step": 357
    },
    {
      "epoch": 0.0358,
      "grad_norm": 2.6550233364105225,
      "learning_rate": 7.14e-05,
      "loss": 7.4445,
      "step": 358
    },
    {
      "epoch": 0.0359,
      "grad_norm": 1.537835717201233,
      "learning_rate": 7.16e-05,
      "loss": 7.9782,
      "step": 359
    },
    {
      "epoch": 0.036,
      "grad_norm": 1.8184560537338257,
      "learning_rate": 7.18e-05,
      "loss": 7.7471,
      "step": 360
    },
    {
      "epoch": 0.0361,
      "grad_norm": 1.429017424583435,
      "learning_rate": 7.2e-05,
      "loss": 7.6018,
      "step": 361
    },
    {
      "epoch": 0.0362,
      "grad_norm": 1.3814053535461426,
      "learning_rate": 7.22e-05,
      "loss": 7.9557,
      "step": 362
    },
    {
      "epoch": 0.0363,
      "grad_norm": 1.416211724281311,
      "learning_rate": 7.24e-05,
      "loss": 7.7847,
      "step": 363
    },
    {
      "epoch": 0.0364,
      "grad_norm": 1.585835576057434,
      "learning_rate": 7.26e-05,
      "loss": 7.7955,
      "step": 364
    },
    {
      "epoch": 0.0365,
      "grad_norm": 1.79407799243927,
      "learning_rate": 7.280000000000001e-05,
      "loss": 8.2591,
      "step": 365
    },
    {
      "epoch": 0.0366,
      "grad_norm": 1.3257417678833008,
      "learning_rate": 7.3e-05,
      "loss": 7.8703,
      "step": 366
    },
    {
      "epoch": 0.0367,
      "grad_norm": 1.8655794858932495,
      "learning_rate": 7.32e-05,
      "loss": 7.6255,
      "step": 367
    },
    {
      "epoch": 0.0368,
      "grad_norm": 2.4462058544158936,
      "learning_rate": 7.340000000000001e-05,
      "loss": 8.3249,
      "step": 368
    },
    {
      "epoch": 0.0369,
      "grad_norm": 2.3074212074279785,
      "learning_rate": 7.36e-05,
      "loss": 7.7401,
      "step": 369
    },
    {
      "epoch": 0.037,
      "grad_norm": 1.5675281286239624,
      "learning_rate": 7.38e-05,
      "loss": 7.8272,
      "step": 370
    },
    {
      "epoch": 0.0371,
      "grad_norm": 1.6928905248641968,
      "learning_rate": 7.4e-05,
      "loss": 7.979,
      "step": 371
    },
    {
      "epoch": 0.0372,
      "grad_norm": 1.5424264669418335,
      "learning_rate": 7.42e-05,
      "loss": 7.684,
      "step": 372
    },
    {
      "epoch": 0.0373,
      "grad_norm": 5.877964973449707,
      "learning_rate": 7.44e-05,
      "loss": 7.6195,
      "step": 373
    },
    {
      "epoch": 0.0374,
      "grad_norm": 3.07493257522583,
      "learning_rate": 7.46e-05,
      "loss": 8.4745,
      "step": 374
    },
    {
      "epoch": 0.0375,
      "grad_norm": 1.8742576837539673,
      "learning_rate": 7.48e-05,
      "loss": 7.7241,
      "step": 375
    },
    {
      "epoch": 0.0376,
      "grad_norm": 2.5962796211242676,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.5812,
      "step": 376
    },
    {
      "epoch": 0.0377,
      "grad_norm": 2.886500120162964,
      "learning_rate": 7.52e-05,
      "loss": 7.3833,
      "step": 377
    },
    {
      "epoch": 0.0378,
      "grad_norm": 1.7347054481506348,
      "learning_rate": 7.54e-05,
      "loss": 7.74,
      "step": 378
    },
    {
      "epoch": 0.0379,
      "grad_norm": 1.4893149137496948,
      "learning_rate": 7.560000000000001e-05,
      "loss": 7.6916,
      "step": 379
    },
    {
      "epoch": 0.038,
      "grad_norm": 2.258767604827881,
      "learning_rate": 7.58e-05,
      "loss": 8.1802,
      "step": 380
    },
    {
      "epoch": 0.0381,
      "grad_norm": 5.084943771362305,
      "learning_rate": 7.6e-05,
      "loss": 8.1172,
      "step": 381
    },
    {
      "epoch": 0.0382,
      "grad_norm": 1.4305745363235474,
      "learning_rate": 7.620000000000001e-05,
      "loss": 7.8365,
      "step": 382
    },
    {
      "epoch": 0.0383,
      "grad_norm": 1.614778995513916,
      "learning_rate": 7.64e-05,
      "loss": 7.9732,
      "step": 383
    },
    {
      "epoch": 0.0384,
      "grad_norm": 1.7301764488220215,
      "learning_rate": 7.66e-05,
      "loss": 7.7794,
      "step": 384
    },
    {
      "epoch": 0.0385,
      "grad_norm": 2.5790152549743652,
      "learning_rate": 7.680000000000001e-05,
      "loss": 8.1802,
      "step": 385
    },
    {
      "epoch": 0.0386,
      "grad_norm": 1.459185242652893,
      "learning_rate": 7.7e-05,
      "loss": 7.9301,
      "step": 386
    },
    {
      "epoch": 0.0387,
      "grad_norm": 1.262910008430481,
      "learning_rate": 7.72e-05,
      "loss": 7.7208,
      "step": 387
    },
    {
      "epoch": 0.0388,
      "grad_norm": 2.266875982284546,
      "learning_rate": 7.740000000000001e-05,
      "loss": 8.1476,
      "step": 388
    },
    {
      "epoch": 0.0389,
      "grad_norm": 1.5800316333770752,
      "learning_rate": 7.76e-05,
      "loss": 8.0111,
      "step": 389
    },
    {
      "epoch": 0.039,
      "grad_norm": 6.056323051452637,
      "learning_rate": 7.780000000000001e-05,
      "loss": 7.9751,
      "step": 390
    },
    {
      "epoch": 0.0391,
      "grad_norm": 7.199629783630371,
      "learning_rate": 7.800000000000001e-05,
      "loss": 7.4985,
      "step": 391
    },
    {
      "epoch": 0.0392,
      "grad_norm": 3.046229600906372,
      "learning_rate": 7.82e-05,
      "loss": 7.917,
      "step": 392
    },
    {
      "epoch": 0.0393,
      "grad_norm": 2.2497661113739014,
      "learning_rate": 7.840000000000001e-05,
      "loss": 7.8299,
      "step": 393
    },
    {
      "epoch": 0.0394,
      "grad_norm": 1.2641912698745728,
      "learning_rate": 7.860000000000001e-05,
      "loss": 7.9026,
      "step": 394
    },
    {
      "epoch": 0.0395,
      "grad_norm": 1.6204124689102173,
      "learning_rate": 7.88e-05,
      "loss": 7.8586,
      "step": 395
    },
    {
      "epoch": 0.0396,
      "grad_norm": 1.7331560850143433,
      "learning_rate": 7.900000000000001e-05,
      "loss": 7.7666,
      "step": 396
    },
    {
      "epoch": 0.0397,
      "grad_norm": 1.9145573377609253,
      "learning_rate": 7.920000000000001e-05,
      "loss": 7.5323,
      "step": 397
    },
    {
      "epoch": 0.0398,
      "grad_norm": 1.4568936824798584,
      "learning_rate": 7.94e-05,
      "loss": 7.8926,
      "step": 398
    },
    {
      "epoch": 0.0399,
      "grad_norm": 1.4700785875320435,
      "learning_rate": 7.960000000000001e-05,
      "loss": 7.6927,
      "step": 399
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.7488411664962769,
      "learning_rate": 7.98e-05,
      "loss": 7.8641,
      "step": 400
    },
    {
      "epoch": 0.0401,
      "grad_norm": 1.8592464923858643,
      "learning_rate": 8e-05,
      "loss": 7.5426,
      "step": 401
    },
    {
      "epoch": 0.0402,
      "grad_norm": 2.35520339012146,
      "learning_rate": 8.020000000000001e-05,
      "loss": 7.905,
      "step": 402
    },
    {
      "epoch": 0.0403,
      "grad_norm": 3.192253351211548,
      "learning_rate": 8.04e-05,
      "loss": 8.1588,
      "step": 403
    },
    {
      "epoch": 0.0404,
      "grad_norm": 2.874979019165039,
      "learning_rate": 8.060000000000001e-05,
      "loss": 8.0778,
      "step": 404
    },
    {
      "epoch": 0.0405,
      "grad_norm": 2.1423099040985107,
      "learning_rate": 8.080000000000001e-05,
      "loss": 7.7887,
      "step": 405
    },
    {
      "epoch": 0.0406,
      "grad_norm": 2.482008934020996,
      "learning_rate": 8.1e-05,
      "loss": 8.1419,
      "step": 406
    },
    {
      "epoch": 0.0407,
      "grad_norm": 1.758724570274353,
      "learning_rate": 8.120000000000001e-05,
      "loss": 7.6581,
      "step": 407
    },
    {
      "epoch": 0.0408,
      "grad_norm": 3.2057316303253174,
      "learning_rate": 8.14e-05,
      "loss": 7.887,
      "step": 408
    },
    {
      "epoch": 0.0409,
      "grad_norm": 3.4203569889068604,
      "learning_rate": 8.16e-05,
      "loss": 7.9545,
      "step": 409
    },
    {
      "epoch": 0.041,
      "grad_norm": 3.9642443656921387,
      "learning_rate": 8.18e-05,
      "loss": 7.4032,
      "step": 410
    },
    {
      "epoch": 0.0411,
      "grad_norm": 7.907172203063965,
      "learning_rate": 8.2e-05,
      "loss": 8.0642,
      "step": 411
    },
    {
      "epoch": 0.0412,
      "grad_norm": 7.58804178237915,
      "learning_rate": 8.22e-05,
      "loss": 7.8052,
      "step": 412
    },
    {
      "epoch": 0.0413,
      "grad_norm": 3.774341583251953,
      "learning_rate": 8.24e-05,
      "loss": 8.5203,
      "step": 413
    },
    {
      "epoch": 0.0414,
      "grad_norm": 4.203991889953613,
      "learning_rate": 8.26e-05,
      "loss": 7.3312,
      "step": 414
    },
    {
      "epoch": 0.0415,
      "grad_norm": 1.7566256523132324,
      "learning_rate": 8.28e-05,
      "loss": 7.8657,
      "step": 415
    },
    {
      "epoch": 0.0416,
      "grad_norm": 3.299318790435791,
      "learning_rate": 8.3e-05,
      "loss": 7.3086,
      "step": 416
    },
    {
      "epoch": 0.0417,
      "grad_norm": 2.089462995529175,
      "learning_rate": 8.32e-05,
      "loss": 7.7588,
      "step": 417
    },
    {
      "epoch": 0.0418,
      "grad_norm": 1.630218267440796,
      "learning_rate": 8.34e-05,
      "loss": 7.649,
      "step": 418
    },
    {
      "epoch": 0.0419,
      "grad_norm": 15.750664710998535,
      "learning_rate": 8.36e-05,
      "loss": 8.4507,
      "step": 419
    },
    {
      "epoch": 0.042,
      "grad_norm": 2.2622852325439453,
      "learning_rate": 8.38e-05,
      "loss": 7.8153,
      "step": 420
    },
    {
      "epoch": 0.0421,
      "grad_norm": 2.7650928497314453,
      "learning_rate": 8.4e-05,
      "loss": 7.7618,
      "step": 421
    },
    {
      "epoch": 0.0422,
      "grad_norm": 2.07753586769104,
      "learning_rate": 8.42e-05,
      "loss": 7.846,
      "step": 422
    },
    {
      "epoch": 0.0423,
      "grad_norm": 2.4953413009643555,
      "learning_rate": 8.44e-05,
      "loss": 7.7367,
      "step": 423
    },
    {
      "epoch": 0.0424,
      "grad_norm": 3.1519718170166016,
      "learning_rate": 8.46e-05,
      "loss": 7.9783,
      "step": 424
    },
    {
      "epoch": 0.0425,
      "grad_norm": 5.165482521057129,
      "learning_rate": 8.48e-05,
      "loss": 8.7106,
      "step": 425
    },
    {
      "epoch": 0.0426,
      "grad_norm": 8.75490951538086,
      "learning_rate": 8.5e-05,
      "loss": 7.6009,
      "step": 426
    },
    {
      "epoch": 0.0427,
      "grad_norm": 1.7089295387268066,
      "learning_rate": 8.52e-05,
      "loss": 7.4589,
      "step": 427
    },
    {
      "epoch": 0.0428,
      "grad_norm": 1.9224222898483276,
      "learning_rate": 8.54e-05,
      "loss": 7.7311,
      "step": 428
    },
    {
      "epoch": 0.0429,
      "grad_norm": 1.6408371925354004,
      "learning_rate": 8.560000000000001e-05,
      "loss": 7.6714,
      "step": 429
    },
    {
      "epoch": 0.043,
      "grad_norm": 1.9524568319320679,
      "learning_rate": 8.58e-05,
      "loss": 7.7086,
      "step": 430
    },
    {
      "epoch": 0.0431,
      "grad_norm": 5.120546817779541,
      "learning_rate": 8.6e-05,
      "loss": 7.7735,
      "step": 431
    },
    {
      "epoch": 0.0432,
      "grad_norm": 2.198150634765625,
      "learning_rate": 8.620000000000001e-05,
      "loss": 7.6412,
      "step": 432
    },
    {
      "epoch": 0.0433,
      "grad_norm": 3.6643738746643066,
      "learning_rate": 8.64e-05,
      "loss": 8.1185,
      "step": 433
    },
    {
      "epoch": 0.0434,
      "grad_norm": 3.750804901123047,
      "learning_rate": 8.66e-05,
      "loss": 8.0373,
      "step": 434
    },
    {
      "epoch": 0.0435,
      "grad_norm": 3.636543035507202,
      "learning_rate": 8.680000000000001e-05,
      "loss": 7.6478,
      "step": 435
    },
    {
      "epoch": 0.0436,
      "grad_norm": 2.3060691356658936,
      "learning_rate": 8.7e-05,
      "loss": 7.9529,
      "step": 436
    },
    {
      "epoch": 0.0437,
      "grad_norm": 2.1625049114227295,
      "learning_rate": 8.72e-05,
      "loss": 7.8747,
      "step": 437
    },
    {
      "epoch": 0.0438,
      "grad_norm": 2.241903781890869,
      "learning_rate": 8.740000000000001e-05,
      "loss": 7.5337,
      "step": 438
    },
    {
      "epoch": 0.0439,
      "grad_norm": 1.8167463541030884,
      "learning_rate": 8.76e-05,
      "loss": 7.8329,
      "step": 439
    },
    {
      "epoch": 0.044,
      "grad_norm": 3.7932167053222656,
      "learning_rate": 8.78e-05,
      "loss": 7.7969,
      "step": 440
    },
    {
      "epoch": 0.0441,
      "grad_norm": 6.610228061676025,
      "learning_rate": 8.800000000000001e-05,
      "loss": 7.6369,
      "step": 441
    },
    {
      "epoch": 0.0442,
      "grad_norm": 2.5136971473693848,
      "learning_rate": 8.82e-05,
      "loss": 7.3177,
      "step": 442
    },
    {
      "epoch": 0.0443,
      "grad_norm": 4.004584789276123,
      "learning_rate": 8.840000000000001e-05,
      "loss": 7.7349,
      "step": 443
    },
    {
      "epoch": 0.0444,
      "grad_norm": 2.0512139797210693,
      "learning_rate": 8.86e-05,
      "loss": 7.9659,
      "step": 444
    },
    {
      "epoch": 0.0445,
      "grad_norm": 1.6350209712982178,
      "learning_rate": 8.88e-05,
      "loss": 7.5226,
      "step": 445
    },
    {
      "epoch": 0.0446,
      "grad_norm": 4.66409969329834,
      "learning_rate": 8.900000000000001e-05,
      "loss": 8.2333,
      "step": 446
    },
    {
      "epoch": 0.0447,
      "grad_norm": 2.6796064376831055,
      "learning_rate": 8.92e-05,
      "loss": 7.8904,
      "step": 447
    },
    {
      "epoch": 0.0448,
      "grad_norm": 15.26917552947998,
      "learning_rate": 8.94e-05,
      "loss": 8.079,
      "step": 448
    },
    {
      "epoch": 0.0449,
      "grad_norm": 2.830918788909912,
      "learning_rate": 8.960000000000001e-05,
      "loss": 7.7752,
      "step": 449
    },
    {
      "epoch": 0.045,
      "grad_norm": 1.9573915004730225,
      "learning_rate": 8.98e-05,
      "loss": 7.476,
      "step": 450
    },
    {
      "epoch": 0.0451,
      "grad_norm": 5.311532020568848,
      "learning_rate": 9e-05,
      "loss": 7.7611,
      "step": 451
    },
    {
      "epoch": 0.0452,
      "grad_norm": 2.304410219192505,
      "learning_rate": 9.020000000000001e-05,
      "loss": 7.7267,
      "step": 452
    },
    {
      "epoch": 0.0453,
      "grad_norm": 1.8590316772460938,
      "learning_rate": 9.04e-05,
      "loss": 7.6861,
      "step": 453
    },
    {
      "epoch": 0.0454,
      "grad_norm": 4.056859493255615,
      "learning_rate": 9.06e-05,
      "loss": 7.3399,
      "step": 454
    },
    {
      "epoch": 0.0455,
      "grad_norm": 5.388113498687744,
      "learning_rate": 9.080000000000001e-05,
      "loss": 8.3073,
      "step": 455
    },
    {
      "epoch": 0.0456,
      "grad_norm": 11.814997673034668,
      "learning_rate": 9.1e-05,
      "loss": 7.796,
      "step": 456
    },
    {
      "epoch": 0.0457,
      "grad_norm": 4.4363250732421875,
      "learning_rate": 9.120000000000001e-05,
      "loss": 8.1154,
      "step": 457
    },
    {
      "epoch": 0.0458,
      "grad_norm": 4.067905426025391,
      "learning_rate": 9.140000000000001e-05,
      "loss": 8.0501,
      "step": 458
    },
    {
      "epoch": 0.0459,
      "grad_norm": 2.5782341957092285,
      "learning_rate": 9.16e-05,
      "loss": 7.4018,
      "step": 459
    },
    {
      "epoch": 0.046,
      "grad_norm": 5.880455017089844,
      "learning_rate": 9.180000000000001e-05,
      "loss": 8.8547,
      "step": 460
    },
    {
      "epoch": 0.0461,
      "grad_norm": 9.010140419006348,
      "learning_rate": 9.200000000000001e-05,
      "loss": 8.1722,
      "step": 461
    },
    {
      "epoch": 0.0462,
      "grad_norm": 3.3707854747772217,
      "learning_rate": 9.22e-05,
      "loss": 7.991,
      "step": 462
    },
    {
      "epoch": 0.0463,
      "grad_norm": 3.655522108078003,
      "learning_rate": 9.240000000000001e-05,
      "loss": 7.4126,
      "step": 463
    },
    {
      "epoch": 0.0464,
      "grad_norm": 7.509081840515137,
      "learning_rate": 9.260000000000001e-05,
      "loss": 8.1257,
      "step": 464
    },
    {
      "epoch": 0.0465,
      "grad_norm": 4.1848554611206055,
      "learning_rate": 9.28e-05,
      "loss": 7.596,
      "step": 465
    },
    {
      "epoch": 0.0466,
      "grad_norm": 2.8866262435913086,
      "learning_rate": 9.300000000000001e-05,
      "loss": 7.6585,
      "step": 466
    },
    {
      "epoch": 0.0467,
      "grad_norm": 2.725951671600342,
      "learning_rate": 9.320000000000002e-05,
      "loss": 7.8162,
      "step": 467
    },
    {
      "epoch": 0.0468,
      "grad_norm": 3.998549699783325,
      "learning_rate": 9.340000000000001e-05,
      "loss": 7.6896,
      "step": 468
    },
    {
      "epoch": 0.0469,
      "grad_norm": 3.4922304153442383,
      "learning_rate": 9.360000000000001e-05,
      "loss": 7.5591,
      "step": 469
    },
    {
      "epoch": 0.047,
      "grad_norm": 2.2627651691436768,
      "learning_rate": 9.38e-05,
      "loss": 7.5697,
      "step": 470
    },
    {
      "epoch": 0.0471,
      "grad_norm": 2.0302317142486572,
      "learning_rate": 9.4e-05,
      "loss": 7.7378,
      "step": 471
    },
    {
      "epoch": 0.0472,
      "grad_norm": 2.607848882675171,
      "learning_rate": 9.42e-05,
      "loss": 8.0941,
      "step": 472
    },
    {
      "epoch": 0.0473,
      "grad_norm": 3.7337183952331543,
      "learning_rate": 9.44e-05,
      "loss": 8.2049,
      "step": 473
    },
    {
      "epoch": 0.0474,
      "grad_norm": 14.654979705810547,
      "learning_rate": 9.46e-05,
      "loss": 7.4512,
      "step": 474
    },
    {
      "epoch": 0.0475,
      "grad_norm": 9.912879943847656,
      "learning_rate": 9.48e-05,
      "loss": 7.5748,
      "step": 475
    },
    {
      "epoch": 0.0476,
      "grad_norm": 11.240350723266602,
      "learning_rate": 9.5e-05,
      "loss": 7.9222,
      "step": 476
    },
    {
      "epoch": 0.0477,
      "grad_norm": 2.3065080642700195,
      "learning_rate": 9.52e-05,
      "loss": 7.945,
      "step": 477
    },
    {
      "epoch": 0.0478,
      "grad_norm": 8.254630088806152,
      "learning_rate": 9.54e-05,
      "loss": 7.2009,
      "step": 478
    },
    {
      "epoch": 0.0479,
      "grad_norm": 6.359790802001953,
      "learning_rate": 9.56e-05,
      "loss": 7.9087,
      "step": 479
    },
    {
      "epoch": 0.048,
      "grad_norm": 21.677230834960938,
      "learning_rate": 9.58e-05,
      "loss": 8.5065,
      "step": 480
    },
    {
      "epoch": 0.0481,
      "grad_norm": 2.3348782062530518,
      "learning_rate": 9.6e-05,
      "loss": 7.3291,
      "step": 481
    },
    {
      "epoch": 0.0482,
      "grad_norm": 2.951018810272217,
      "learning_rate": 9.620000000000001e-05,
      "loss": 7.7617,
      "step": 482
    },
    {
      "epoch": 0.0483,
      "grad_norm": 4.57310152053833,
      "learning_rate": 9.64e-05,
      "loss": 7.9715,
      "step": 483
    },
    {
      "epoch": 0.0484,
      "grad_norm": 3.4184410572052,
      "learning_rate": 9.66e-05,
      "loss": 7.7978,
      "step": 484
    },
    {
      "epoch": 0.0485,
      "grad_norm": 29.830482482910156,
      "learning_rate": 9.680000000000001e-05,
      "loss": 8.62,
      "step": 485
    },
    {
      "epoch": 0.0486,
      "grad_norm": 3.7488598823547363,
      "learning_rate": 9.7e-05,
      "loss": 7.7133,
      "step": 486
    },
    {
      "epoch": 0.0487,
      "grad_norm": 4.461332321166992,
      "learning_rate": 9.72e-05,
      "loss": 7.7807,
      "step": 487
    },
    {
      "epoch": 0.0488,
      "grad_norm": 2.432406187057495,
      "learning_rate": 9.74e-05,
      "loss": 7.4042,
      "step": 488
    },
    {
      "epoch": 0.0489,
      "grad_norm": 2.5470128059387207,
      "learning_rate": 9.76e-05,
      "loss": 7.6515,
      "step": 489
    },
    {
      "epoch": 0.049,
      "grad_norm": 2.567387342453003,
      "learning_rate": 9.78e-05,
      "loss": 7.2721,
      "step": 490
    },
    {
      "epoch": 0.0491,
      "grad_norm": 16.364469528198242,
      "learning_rate": 9.8e-05,
      "loss": 7.5728,
      "step": 491
    },
    {
      "epoch": 0.0492,
      "grad_norm": 3.220630645751953,
      "learning_rate": 9.82e-05,
      "loss": 7.4521,
      "step": 492
    },
    {
      "epoch": 0.0493,
      "grad_norm": 10.75374984741211,
      "learning_rate": 9.84e-05,
      "loss": 7.3663,
      "step": 493
    },
    {
      "epoch": 0.0494,
      "grad_norm": 3.690920829772949,
      "learning_rate": 9.86e-05,
      "loss": 7.4356,
      "step": 494
    },
    {
      "epoch": 0.0495,
      "grad_norm": 4.411746025085449,
      "learning_rate": 9.88e-05,
      "loss": 8.1105,
      "step": 495
    },
    {
      "epoch": 0.0496,
      "grad_norm": 5.082192420959473,
      "learning_rate": 9.900000000000001e-05,
      "loss": 7.8781,
      "step": 496
    },
    {
      "epoch": 0.0497,
      "grad_norm": 2.7750840187072754,
      "learning_rate": 9.92e-05,
      "loss": 7.6926,
      "step": 497
    },
    {
      "epoch": 0.0498,
      "grad_norm": 9.13431167602539,
      "learning_rate": 9.94e-05,
      "loss": 7.31,
      "step": 498
    },
    {
      "epoch": 0.0499,
      "grad_norm": 49.51872253417969,
      "learning_rate": 9.960000000000001e-05,
      "loss": 9.862,
      "step": 499
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.465449094772339,
      "learning_rate": 9.98e-05,
      "loss": 7.7101,
      "step": 500
    },
    {
      "epoch": 0.0501,
      "grad_norm": 3.9321649074554443,
      "learning_rate": 0.0001,
      "loss": 8.0274,
      "step": 501
    },
    {
      "epoch": 0.0502,
      "grad_norm": 2.989752769470215,
      "learning_rate": 0.00010020000000000001,
      "loss": 7.1404,
      "step": 502
    },
    {
      "epoch": 0.0503,
      "grad_norm": 3.7616686820983887,
      "learning_rate": 0.0001004,
      "loss": 7.5392,
      "step": 503
    },
    {
      "epoch": 0.0504,
      "grad_norm": 6.97094202041626,
      "learning_rate": 0.0001006,
      "loss": 7.4858,
      "step": 504
    },
    {
      "epoch": 0.0505,
      "grad_norm": 5.6057448387146,
      "learning_rate": 0.00010080000000000001,
      "loss": 8.4251,
      "step": 505
    },
    {
      "epoch": 0.0506,
      "grad_norm": 3.59206223487854,
      "learning_rate": 0.000101,
      "loss": 7.8242,
      "step": 506
    },
    {
      "epoch": 0.0507,
      "grad_norm": 3.227590560913086,
      "learning_rate": 0.00010120000000000001,
      "loss": 7.7995,
      "step": 507
    },
    {
      "epoch": 0.0508,
      "grad_norm": 4.5499982833862305,
      "learning_rate": 0.00010140000000000001,
      "loss": 7.3939,
      "step": 508
    },
    {
      "epoch": 0.0509,
      "grad_norm": 6.309101104736328,
      "learning_rate": 0.0001016,
      "loss": 8.0702,
      "step": 509
    },
    {
      "epoch": 0.051,
      "grad_norm": 5.605581760406494,
      "learning_rate": 0.00010180000000000001,
      "loss": 8.1138,
      "step": 510
    },
    {
      "epoch": 0.0511,
      "grad_norm": 5.263078212738037,
      "learning_rate": 0.00010200000000000001,
      "loss": 8.08,
      "step": 511
    },
    {
      "epoch": 0.0512,
      "grad_norm": 8.090388298034668,
      "learning_rate": 0.0001022,
      "loss": 8.1824,
      "step": 512
    },
    {
      "epoch": 0.0513,
      "grad_norm": 4.091671466827393,
      "learning_rate": 0.00010240000000000001,
      "loss": 7.9785,
      "step": 513
    },
    {
      "epoch": 0.0514,
      "grad_norm": 3.1390414237976074,
      "learning_rate": 0.00010260000000000001,
      "loss": 7.3697,
      "step": 514
    },
    {
      "epoch": 0.0515,
      "grad_norm": 4.750081539154053,
      "learning_rate": 0.0001028,
      "loss": 7.9339,
      "step": 515
    },
    {
      "epoch": 0.0516,
      "grad_norm": 3.8537280559539795,
      "learning_rate": 0.00010300000000000001,
      "loss": 7.2287,
      "step": 516
    },
    {
      "epoch": 0.0517,
      "grad_norm": 4.524524688720703,
      "learning_rate": 0.0001032,
      "loss": 7.3352,
      "step": 517
    },
    {
      "epoch": 0.0518,
      "grad_norm": 3.5437495708465576,
      "learning_rate": 0.0001034,
      "loss": 7.5483,
      "step": 518
    },
    {
      "epoch": 0.0519,
      "grad_norm": 4.475898265838623,
      "learning_rate": 0.00010360000000000001,
      "loss": 7.9989,
      "step": 519
    },
    {
      "epoch": 0.052,
      "grad_norm": 3.92375111579895,
      "learning_rate": 0.0001038,
      "loss": 8.1354,
      "step": 520
    },
    {
      "epoch": 0.0521,
      "grad_norm": 2.5535964965820312,
      "learning_rate": 0.00010400000000000001,
      "loss": 7.3992,
      "step": 521
    },
    {
      "epoch": 0.0522,
      "grad_norm": 4.474313259124756,
      "learning_rate": 0.00010420000000000001,
      "loss": 8.3157,
      "step": 522
    },
    {
      "epoch": 0.0523,
      "grad_norm": 2.6088762283325195,
      "learning_rate": 0.0001044,
      "loss": 7.3862,
      "step": 523
    },
    {
      "epoch": 0.0524,
      "grad_norm": 2.7977139949798584,
      "learning_rate": 0.00010460000000000001,
      "loss": 7.2494,
      "step": 524
    },
    {
      "epoch": 0.0525,
      "grad_norm": 5.415034770965576,
      "learning_rate": 0.00010480000000000001,
      "loss": 7.3867,
      "step": 525
    },
    {
      "epoch": 0.0526,
      "grad_norm": 13.567598342895508,
      "learning_rate": 0.000105,
      "loss": 7.7845,
      "step": 526
    },
    {
      "epoch": 0.0527,
      "grad_norm": 23.320627212524414,
      "learning_rate": 0.00010520000000000001,
      "loss": 7.4953,
      "step": 527
    },
    {
      "epoch": 0.0528,
      "grad_norm": 9.794024467468262,
      "learning_rate": 0.00010540000000000001,
      "loss": 7.4898,
      "step": 528
    },
    {
      "epoch": 0.0529,
      "grad_norm": 17.020992279052734,
      "learning_rate": 0.0001056,
      "loss": 7.6507,
      "step": 529
    },
    {
      "epoch": 0.053,
      "grad_norm": 34.73387908935547,
      "learning_rate": 0.00010580000000000001,
      "loss": 7.6394,
      "step": 530
    },
    {
      "epoch": 0.0531,
      "grad_norm": 3.2606236934661865,
      "learning_rate": 0.00010600000000000002,
      "loss": 7.4593,
      "step": 531
    },
    {
      "epoch": 0.0532,
      "grad_norm": 18.698745727539062,
      "learning_rate": 0.0001062,
      "loss": 7.6577,
      "step": 532
    },
    {
      "epoch": 0.0533,
      "grad_norm": 2.277480125427246,
      "learning_rate": 0.00010640000000000001,
      "loss": 7.1977,
      "step": 533
    },
    {
      "epoch": 0.0534,
      "grad_norm": 2.786043643951416,
      "learning_rate": 0.00010660000000000002,
      "loss": 7.3202,
      "step": 534
    },
    {
      "epoch": 0.0535,
      "grad_norm": 2.765519142150879,
      "learning_rate": 0.00010680000000000001,
      "loss": 7.601,
      "step": 535
    },
    {
      "epoch": 0.0536,
      "grad_norm": 5.368846893310547,
      "learning_rate": 0.00010700000000000001,
      "loss": 8.2631,
      "step": 536
    },
    {
      "epoch": 0.0537,
      "grad_norm": 5.326869010925293,
      "learning_rate": 0.00010720000000000002,
      "loss": 8.1688,
      "step": 537
    },
    {
      "epoch": 0.0538,
      "grad_norm": 3.1491808891296387,
      "learning_rate": 0.00010740000000000001,
      "loss": 7.5258,
      "step": 538
    },
    {
      "epoch": 0.0539,
      "grad_norm": 3.476951837539673,
      "learning_rate": 0.00010760000000000001,
      "loss": 7.2833,
      "step": 539
    },
    {
      "epoch": 0.054,
      "grad_norm": 3.5136775970458984,
      "learning_rate": 0.00010780000000000002,
      "loss": 7.594,
      "step": 540
    },
    {
      "epoch": 0.0541,
      "grad_norm": 3.443000078201294,
      "learning_rate": 0.00010800000000000001,
      "loss": 7.6905,
      "step": 541
    },
    {
      "epoch": 0.0542,
      "grad_norm": 4.559494495391846,
      "learning_rate": 0.00010820000000000001,
      "loss": 7.6548,
      "step": 542
    },
    {
      "epoch": 0.0543,
      "grad_norm": 3.6187198162078857,
      "learning_rate": 0.00010840000000000002,
      "loss": 7.392,
      "step": 543
    },
    {
      "epoch": 0.0544,
      "grad_norm": 4.149837970733643,
      "learning_rate": 0.00010860000000000001,
      "loss": 8.0862,
      "step": 544
    },
    {
      "epoch": 0.0545,
      "grad_norm": 17.721710205078125,
      "learning_rate": 0.00010880000000000002,
      "loss": 7.7056,
      "step": 545
    },
    {
      "epoch": 0.0546,
      "grad_norm": 9.764822006225586,
      "learning_rate": 0.000109,
      "loss": 8.0363,
      "step": 546
    },
    {
      "epoch": 0.0547,
      "grad_norm": 6.650572776794434,
      "learning_rate": 0.00010920000000000001,
      "loss": 7.0568,
      "step": 547
    },
    {
      "epoch": 0.0548,
      "grad_norm": 29.418264389038086,
      "learning_rate": 0.00010940000000000002,
      "loss": 7.6771,
      "step": 548
    },
    {
      "epoch": 0.0549,
      "grad_norm": 3.718080759048462,
      "learning_rate": 0.00010960000000000001,
      "loss": 8.056,
      "step": 549
    },
    {
      "epoch": 0.055,
      "grad_norm": 2.1358180046081543,
      "learning_rate": 0.00010980000000000001,
      "loss": 6.98,
      "step": 550
    },
    {
      "epoch": 0.0551,
      "grad_norm": 2.625967502593994,
      "learning_rate": 0.00011000000000000002,
      "loss": 7.4088,
      "step": 551
    },
    {
      "epoch": 0.0552,
      "grad_norm": 9.880970001220703,
      "learning_rate": 0.00011020000000000001,
      "loss": 7.5329,
      "step": 552
    },
    {
      "epoch": 0.0553,
      "grad_norm": 3.790710926055908,
      "learning_rate": 0.00011040000000000001,
      "loss": 7.519,
      "step": 553
    },
    {
      "epoch": 0.0554,
      "grad_norm": 3.4173836708068848,
      "learning_rate": 0.00011060000000000002,
      "loss": 7.6954,
      "step": 554
    },
    {
      "epoch": 0.0555,
      "grad_norm": 10.466639518737793,
      "learning_rate": 0.00011080000000000001,
      "loss": 7.3902,
      "step": 555
    },
    {
      "epoch": 0.0556,
      "grad_norm": 3.416452407836914,
      "learning_rate": 0.00011100000000000001,
      "loss": 8.3436,
      "step": 556
    },
    {
      "epoch": 0.0557,
      "grad_norm": 4.269401550292969,
      "learning_rate": 0.00011120000000000002,
      "loss": 6.4806,
      "step": 557
    },
    {
      "epoch": 0.0558,
      "grad_norm": 10.728877067565918,
      "learning_rate": 0.00011140000000000001,
      "loss": 7.8837,
      "step": 558
    },
    {
      "epoch": 0.0559,
      "grad_norm": 3.680208206176758,
      "learning_rate": 0.00011160000000000002,
      "loss": 7.2494,
      "step": 559
    },
    {
      "epoch": 0.056,
      "grad_norm": 2.645761489868164,
      "learning_rate": 0.00011180000000000002,
      "loss": 7.7881,
      "step": 560
    },
    {
      "epoch": 0.0561,
      "grad_norm": 5.909026145935059,
      "learning_rate": 0.00011200000000000001,
      "loss": 7.5205,
      "step": 561
    },
    {
      "epoch": 0.0562,
      "grad_norm": 3.3205010890960693,
      "learning_rate": 0.00011220000000000002,
      "loss": 7.6675,
      "step": 562
    },
    {
      "epoch": 0.0563,
      "grad_norm": 3.2637898921966553,
      "learning_rate": 0.00011240000000000002,
      "loss": 7.5838,
      "step": 563
    },
    {
      "epoch": 0.0564,
      "grad_norm": 2.8143157958984375,
      "learning_rate": 0.0001126,
      "loss": 7.8138,
      "step": 564
    },
    {
      "epoch": 0.0565,
      "grad_norm": 2.212270975112915,
      "learning_rate": 0.00011279999999999999,
      "loss": 7.5257,
      "step": 565
    },
    {
      "epoch": 0.0566,
      "grad_norm": 2.8903121948242188,
      "learning_rate": 0.000113,
      "loss": 7.3513,
      "step": 566
    },
    {
      "epoch": 0.0567,
      "grad_norm": 2.4863414764404297,
      "learning_rate": 0.0001132,
      "loss": 7.5685,
      "step": 567
    },
    {
      "epoch": 0.0568,
      "grad_norm": 3.7077078819274902,
      "learning_rate": 0.00011339999999999999,
      "loss": 7.3639,
      "step": 568
    },
    {
      "epoch": 0.0569,
      "grad_norm": 9.903769493103027,
      "learning_rate": 0.0001136,
      "loss": 8.1807,
      "step": 569
    },
    {
      "epoch": 0.057,
      "grad_norm": 4.127170562744141,
      "learning_rate": 0.0001138,
      "loss": 7.1916,
      "step": 570
    },
    {
      "epoch": 0.0571,
      "grad_norm": 2.902012348175049,
      "learning_rate": 0.00011399999999999999,
      "loss": 7.6223,
      "step": 571
    },
    {
      "epoch": 0.0572,
      "grad_norm": 4.90426778793335,
      "learning_rate": 0.0001142,
      "loss": 7.3652,
      "step": 572
    },
    {
      "epoch": 0.0573,
      "grad_norm": 29.687532424926758,
      "learning_rate": 0.0001144,
      "loss": 8.005,
      "step": 573
    },
    {
      "epoch": 0.0574,
      "grad_norm": 5.367135047912598,
      "learning_rate": 0.0001146,
      "loss": 7.674,
      "step": 574
    },
    {
      "epoch": 0.0575,
      "grad_norm": 4.485906600952148,
      "learning_rate": 0.0001148,
      "loss": 7.5814,
      "step": 575
    },
    {
      "epoch": 0.0576,
      "grad_norm": 3.0028045177459717,
      "learning_rate": 0.00011499999999999999,
      "loss": 7.4997,
      "step": 576
    },
    {
      "epoch": 0.0577,
      "grad_norm": 9.661340713500977,
      "learning_rate": 0.0001152,
      "loss": 7.6184,
      "step": 577
    },
    {
      "epoch": 0.0578,
      "grad_norm": 4.276668071746826,
      "learning_rate": 0.0001154,
      "loss": 7.4119,
      "step": 578
    },
    {
      "epoch": 0.0579,
      "grad_norm": 3.7751481533050537,
      "learning_rate": 0.00011559999999999999,
      "loss": 7.6893,
      "step": 579
    },
    {
      "epoch": 0.058,
      "grad_norm": 3.0032336711883545,
      "learning_rate": 0.0001158,
      "loss": 7.8895,
      "step": 580
    },
    {
      "epoch": 0.0581,
      "grad_norm": 4.864978790283203,
      "learning_rate": 0.000116,
      "loss": 7.4743,
      "step": 581
    },
    {
      "epoch": 0.0582,
      "grad_norm": 2.807108163833618,
      "learning_rate": 0.00011619999999999999,
      "loss": 7.4233,
      "step": 582
    },
    {
      "epoch": 0.0583,
      "grad_norm": 5.854761123657227,
      "learning_rate": 0.0001164,
      "loss": 7.4431,
      "step": 583
    },
    {
      "epoch": 0.0584,
      "grad_norm": 11.614725112915039,
      "learning_rate": 0.0001166,
      "loss": 7.6496,
      "step": 584
    },
    {
      "epoch": 0.0585,
      "grad_norm": 22.37323570251465,
      "learning_rate": 0.00011679999999999999,
      "loss": 7.0508,
      "step": 585
    },
    {
      "epoch": 0.0586,
      "grad_norm": 4.820290565490723,
      "learning_rate": 0.000117,
      "loss": 7.5696,
      "step": 586
    },
    {
      "epoch": 0.0587,
      "grad_norm": 6.611201286315918,
      "learning_rate": 0.0001172,
      "loss": 7.1164,
      "step": 587
    },
    {
      "epoch": 0.0588,
      "grad_norm": 9.740876197814941,
      "learning_rate": 0.0001174,
      "loss": 7.337,
      "step": 588
    },
    {
      "epoch": 0.0589,
      "grad_norm": 3.032153367996216,
      "learning_rate": 0.0001176,
      "loss": 7.2941,
      "step": 589
    },
    {
      "epoch": 0.059,
      "grad_norm": 3.159376859664917,
      "learning_rate": 0.0001178,
      "loss": 7.5472,
      "step": 590
    },
    {
      "epoch": 0.0591,
      "grad_norm": 7.690126419067383,
      "learning_rate": 0.000118,
      "loss": 7.8923,
      "step": 591
    },
    {
      "epoch": 0.0592,
      "grad_norm": 4.062870502471924,
      "learning_rate": 0.0001182,
      "loss": 7.0907,
      "step": 592
    },
    {
      "epoch": 0.0593,
      "grad_norm": 3.1193575859069824,
      "learning_rate": 0.0001184,
      "loss": 7.6031,
      "step": 593
    },
    {
      "epoch": 0.0594,
      "grad_norm": 4.304540157318115,
      "learning_rate": 0.0001186,
      "loss": 7.0959,
      "step": 594
    },
    {
      "epoch": 0.0595,
      "grad_norm": 5.97119140625,
      "learning_rate": 0.0001188,
      "loss": 8.0271,
      "step": 595
    },
    {
      "epoch": 0.0596,
      "grad_norm": 4.90590238571167,
      "learning_rate": 0.000119,
      "loss": 7.8689,
      "step": 596
    },
    {
      "epoch": 0.0597,
      "grad_norm": 5.338168621063232,
      "learning_rate": 0.0001192,
      "loss": 7.0206,
      "step": 597
    },
    {
      "epoch": 0.0598,
      "grad_norm": 3.595289707183838,
      "learning_rate": 0.0001194,
      "loss": 7.6956,
      "step": 598
    },
    {
      "epoch": 0.0599,
      "grad_norm": 8.530720710754395,
      "learning_rate": 0.00011960000000000001,
      "loss": 7.3465,
      "step": 599
    },
    {
      "epoch": 0.06,
      "grad_norm": 20.937349319458008,
      "learning_rate": 0.0001198,
      "loss": 7.4845,
      "step": 600
    },
    {
      "epoch": 0.0601,
      "grad_norm": 12.101160049438477,
      "learning_rate": 0.00012,
      "loss": 7.7097,
      "step": 601
    },
    {
      "epoch": 0.0602,
      "grad_norm": 27.44350242614746,
      "learning_rate": 0.00012020000000000001,
      "loss": 7.8613,
      "step": 602
    },
    {
      "epoch": 0.0603,
      "grad_norm": 4.412171363830566,
      "learning_rate": 0.0001204,
      "loss": 7.5161,
      "step": 603
    },
    {
      "epoch": 0.0604,
      "grad_norm": 3.170563220977783,
      "learning_rate": 0.0001206,
      "loss": 7.5561,
      "step": 604
    },
    {
      "epoch": 0.0605,
      "grad_norm": 9.883574485778809,
      "learning_rate": 0.0001208,
      "loss": 7.5901,
      "step": 605
    },
    {
      "epoch": 0.0606,
      "grad_norm": 12.101638793945312,
      "learning_rate": 0.000121,
      "loss": 7.8672,
      "step": 606
    },
    {
      "epoch": 0.0607,
      "grad_norm": 2.936558485031128,
      "learning_rate": 0.0001212,
      "loss": 7.5458,
      "step": 607
    },
    {
      "epoch": 0.0608,
      "grad_norm": 21.87476921081543,
      "learning_rate": 0.0001214,
      "loss": 8.2634,
      "step": 608
    },
    {
      "epoch": 0.0609,
      "grad_norm": 2.3735198974609375,
      "learning_rate": 0.0001216,
      "loss": 7.2868,
      "step": 609
    },
    {
      "epoch": 0.061,
      "grad_norm": 8.860730171203613,
      "learning_rate": 0.0001218,
      "loss": 6.9221,
      "step": 610
    },
    {
      "epoch": 0.0611,
      "grad_norm": 3.3877577781677246,
      "learning_rate": 0.000122,
      "loss": 7.2506,
      "step": 611
    },
    {
      "epoch": 0.0612,
      "grad_norm": 6.921363353729248,
      "learning_rate": 0.00012220000000000002,
      "loss": 7.2051,
      "step": 612
    },
    {
      "epoch": 0.0613,
      "grad_norm": 4.0446014404296875,
      "learning_rate": 0.0001224,
      "loss": 7.4834,
      "step": 613
    },
    {
      "epoch": 0.0614,
      "grad_norm": 5.87910795211792,
      "learning_rate": 0.0001226,
      "loss": 7.6412,
      "step": 614
    },
    {
      "epoch": 0.0615,
      "grad_norm": 3.8759493827819824,
      "learning_rate": 0.0001228,
      "loss": 7.7234,
      "step": 615
    },
    {
      "epoch": 0.0616,
      "grad_norm": 3.6695828437805176,
      "learning_rate": 0.000123,
      "loss": 6.6545,
      "step": 616
    },
    {
      "epoch": 0.0617,
      "grad_norm": 2.354377269744873,
      "learning_rate": 0.0001232,
      "loss": 7.4495,
      "step": 617
    },
    {
      "epoch": 0.0618,
      "grad_norm": 2.714322805404663,
      "learning_rate": 0.00012340000000000002,
      "loss": 7.6452,
      "step": 618
    },
    {
      "epoch": 0.0619,
      "grad_norm": 6.526347637176514,
      "learning_rate": 0.0001236,
      "loss": 7.7143,
      "step": 619
    },
    {
      "epoch": 0.062,
      "grad_norm": 12.68463134765625,
      "learning_rate": 0.0001238,
      "loss": 7.6001,
      "step": 620
    },
    {
      "epoch": 0.0621,
      "grad_norm": 10.26412582397461,
      "learning_rate": 0.000124,
      "loss": 7.4774,
      "step": 621
    },
    {
      "epoch": 0.0622,
      "grad_norm": 4.952575206756592,
      "learning_rate": 0.0001242,
      "loss": 6.9735,
      "step": 622
    },
    {
      "epoch": 0.0623,
      "grad_norm": 3.915154218673706,
      "learning_rate": 0.00012440000000000002,
      "loss": 7.1628,
      "step": 623
    },
    {
      "epoch": 0.0624,
      "grad_norm": 4.831092357635498,
      "learning_rate": 0.0001246,
      "loss": 7.4012,
      "step": 624
    },
    {
      "epoch": 0.0625,
      "grad_norm": 6.616661548614502,
      "learning_rate": 0.0001248,
      "loss": 7.5984,
      "step": 625
    },
    {
      "epoch": 0.0626,
      "grad_norm": 11.426172256469727,
      "learning_rate": 0.000125,
      "loss": 7.3538,
      "step": 626
    },
    {
      "epoch": 0.0627,
      "grad_norm": 6.71727180480957,
      "learning_rate": 0.0001252,
      "loss": 7.0426,
      "step": 627
    },
    {
      "epoch": 0.0628,
      "grad_norm": 2.998314619064331,
      "learning_rate": 0.0001254,
      "loss": 7.2739,
      "step": 628
    },
    {
      "epoch": 0.0629,
      "grad_norm": 6.025814533233643,
      "learning_rate": 0.00012560000000000002,
      "loss": 7.2631,
      "step": 629
    },
    {
      "epoch": 0.063,
      "grad_norm": 6.870246887207031,
      "learning_rate": 0.0001258,
      "loss": 7.4198,
      "step": 630
    },
    {
      "epoch": 0.0631,
      "grad_norm": 12.944926261901855,
      "learning_rate": 0.000126,
      "loss": 7.6141,
      "step": 631
    },
    {
      "epoch": 0.0632,
      "grad_norm": 6.726109504699707,
      "learning_rate": 0.0001262,
      "loss": 7.5954,
      "step": 632
    },
    {
      "epoch": 0.0633,
      "grad_norm": 5.518542289733887,
      "learning_rate": 0.0001264,
      "loss": 7.5461,
      "step": 633
    },
    {
      "epoch": 0.0634,
      "grad_norm": 6.447314262390137,
      "learning_rate": 0.00012660000000000001,
      "loss": 7.5932,
      "step": 634
    },
    {
      "epoch": 0.0635,
      "grad_norm": 5.3773956298828125,
      "learning_rate": 0.00012680000000000002,
      "loss": 7.8841,
      "step": 635
    },
    {
      "epoch": 0.0636,
      "grad_norm": 3.483095169067383,
      "learning_rate": 0.000127,
      "loss": 7.5443,
      "step": 636
    },
    {
      "epoch": 0.0637,
      "grad_norm": 3.4145708084106445,
      "learning_rate": 0.0001272,
      "loss": 7.1447,
      "step": 637
    },
    {
      "epoch": 0.0638,
      "grad_norm": 3.5379257202148438,
      "learning_rate": 0.0001274,
      "loss": 7.2418,
      "step": 638
    },
    {
      "epoch": 0.0639,
      "grad_norm": 21.006671905517578,
      "learning_rate": 0.0001276,
      "loss": 7.9715,
      "step": 639
    },
    {
      "epoch": 0.064,
      "grad_norm": 16.85822296142578,
      "learning_rate": 0.00012780000000000002,
      "loss": 7.3095,
      "step": 640
    },
    {
      "epoch": 0.0641,
      "grad_norm": 25.980436325073242,
      "learning_rate": 0.00012800000000000002,
      "loss": 8.0585,
      "step": 641
    },
    {
      "epoch": 0.0642,
      "grad_norm": 21.544918060302734,
      "learning_rate": 0.0001282,
      "loss": 7.0738,
      "step": 642
    },
    {
      "epoch": 0.0643,
      "grad_norm": 18.340700149536133,
      "learning_rate": 0.0001284,
      "loss": 7.785,
      "step": 643
    },
    {
      "epoch": 0.0644,
      "grad_norm": 5.291457653045654,
      "learning_rate": 0.0001286,
      "loss": 7.9362,
      "step": 644
    },
    {
      "epoch": 0.0645,
      "grad_norm": 14.493172645568848,
      "learning_rate": 0.00012880000000000001,
      "loss": 7.3248,
      "step": 645
    },
    {
      "epoch": 0.0646,
      "grad_norm": 3.264024019241333,
      "learning_rate": 0.00012900000000000002,
      "loss": 7.0801,
      "step": 646
    },
    {
      "epoch": 0.0647,
      "grad_norm": 4.025453090667725,
      "learning_rate": 0.00012920000000000002,
      "loss": 6.9778,
      "step": 647
    },
    {
      "epoch": 0.0648,
      "grad_norm": 5.402032375335693,
      "learning_rate": 0.0001294,
      "loss": 8.0097,
      "step": 648
    },
    {
      "epoch": 0.0649,
      "grad_norm": 4.241204738616943,
      "learning_rate": 0.0001296,
      "loss": 7.5296,
      "step": 649
    },
    {
      "epoch": 0.065,
      "grad_norm": 29.66694450378418,
      "learning_rate": 0.0001298,
      "loss": 7.7962,
      "step": 650
    },
    {
      "epoch": 0.0651,
      "grad_norm": 4.19865608215332,
      "learning_rate": 0.00013000000000000002,
      "loss": 7.3957,
      "step": 651
    },
    {
      "epoch": 0.0652,
      "grad_norm": 3.148063898086548,
      "learning_rate": 0.00013020000000000002,
      "loss": 7.2748,
      "step": 652
    },
    {
      "epoch": 0.0653,
      "grad_norm": 6.733952522277832,
      "learning_rate": 0.0001304,
      "loss": 7.2863,
      "step": 653
    },
    {
      "epoch": 0.0654,
      "grad_norm": 4.642484188079834,
      "learning_rate": 0.0001306,
      "loss": 7.0082,
      "step": 654
    },
    {
      "epoch": 0.0655,
      "grad_norm": 2.6257758140563965,
      "learning_rate": 0.0001308,
      "loss": 7.2831,
      "step": 655
    },
    {
      "epoch": 0.0656,
      "grad_norm": 3.375138282775879,
      "learning_rate": 0.000131,
      "loss": 7.9514,
      "step": 656
    },
    {
      "epoch": 0.0657,
      "grad_norm": 5.123089790344238,
      "learning_rate": 0.00013120000000000002,
      "loss": 6.8574,
      "step": 657
    },
    {
      "epoch": 0.0658,
      "grad_norm": 2.721933364868164,
      "learning_rate": 0.00013140000000000002,
      "loss": 6.9774,
      "step": 658
    },
    {
      "epoch": 0.0659,
      "grad_norm": 3.3738181591033936,
      "learning_rate": 0.0001316,
      "loss": 7.0064,
      "step": 659
    },
    {
      "epoch": 0.066,
      "grad_norm": 8.242439270019531,
      "learning_rate": 0.0001318,
      "loss": 7.3652,
      "step": 660
    },
    {
      "epoch": 0.0661,
      "grad_norm": 5.507571220397949,
      "learning_rate": 0.000132,
      "loss": 6.8426,
      "step": 661
    },
    {
      "epoch": 0.0662,
      "grad_norm": 7.2990570068359375,
      "learning_rate": 0.00013220000000000001,
      "loss": 7.2632,
      "step": 662
    },
    {
      "epoch": 0.0663,
      "grad_norm": 5.467508316040039,
      "learning_rate": 0.00013240000000000002,
      "loss": 7.6845,
      "step": 663
    },
    {
      "epoch": 0.0664,
      "grad_norm": 5.045270919799805,
      "learning_rate": 0.00013260000000000002,
      "loss": 7.5371,
      "step": 664
    },
    {
      "epoch": 0.0665,
      "grad_norm": 5.577767848968506,
      "learning_rate": 0.0001328,
      "loss": 7.3325,
      "step": 665
    },
    {
      "epoch": 0.0666,
      "grad_norm": 5.451627731323242,
      "learning_rate": 0.000133,
      "loss": 7.3851,
      "step": 666
    },
    {
      "epoch": 0.0667,
      "grad_norm": 4.757533073425293,
      "learning_rate": 0.0001332,
      "loss": 7.1115,
      "step": 667
    },
    {
      "epoch": 0.0668,
      "grad_norm": 4.863078594207764,
      "learning_rate": 0.00013340000000000002,
      "loss": 6.6278,
      "step": 668
    },
    {
      "epoch": 0.0669,
      "grad_norm": 4.463343143463135,
      "learning_rate": 0.00013360000000000002,
      "loss": 7.3518,
      "step": 669
    },
    {
      "epoch": 0.067,
      "grad_norm": 4.601968288421631,
      "learning_rate": 0.00013380000000000003,
      "loss": 7.1191,
      "step": 670
    },
    {
      "epoch": 0.0671,
      "grad_norm": 4.651142597198486,
      "learning_rate": 0.000134,
      "loss": 6.9611,
      "step": 671
    },
    {
      "epoch": 0.0672,
      "grad_norm": 3.1579580307006836,
      "learning_rate": 0.0001342,
      "loss": 6.8957,
      "step": 672
    },
    {
      "epoch": 0.0673,
      "grad_norm": 29.961605072021484,
      "learning_rate": 0.00013440000000000001,
      "loss": 7.608,
      "step": 673
    },
    {
      "epoch": 0.0674,
      "grad_norm": 19.359527587890625,
      "learning_rate": 0.00013460000000000002,
      "loss": 7.2747,
      "step": 674
    },
    {
      "epoch": 0.0675,
      "grad_norm": 30.543346405029297,
      "learning_rate": 0.00013480000000000002,
      "loss": 8.0874,
      "step": 675
    },
    {
      "epoch": 0.0676,
      "grad_norm": 4.152474880218506,
      "learning_rate": 0.00013500000000000003,
      "loss": 7.165,
      "step": 676
    },
    {
      "epoch": 0.0677,
      "grad_norm": 52.30223846435547,
      "learning_rate": 0.0001352,
      "loss": 7.6365,
      "step": 677
    },
    {
      "epoch": 0.0678,
      "grad_norm": 8.166813850402832,
      "learning_rate": 0.0001354,
      "loss": 6.9223,
      "step": 678
    },
    {
      "epoch": 0.0679,
      "grad_norm": 15.366547584533691,
      "learning_rate": 0.00013560000000000002,
      "loss": 7.2623,
      "step": 679
    },
    {
      "epoch": 0.068,
      "grad_norm": 4.8240647315979,
      "learning_rate": 0.00013580000000000002,
      "loss": 6.8875,
      "step": 680
    },
    {
      "epoch": 0.0681,
      "grad_norm": 61.314449310302734,
      "learning_rate": 0.00013600000000000003,
      "loss": 8.5438,
      "step": 681
    },
    {
      "epoch": 0.0682,
      "grad_norm": 4.371884822845459,
      "learning_rate": 0.0001362,
      "loss": 7.0378,
      "step": 682
    },
    {
      "epoch": 0.0683,
      "grad_norm": 8.791536331176758,
      "learning_rate": 0.0001364,
      "loss": 7.5944,
      "step": 683
    },
    {
      "epoch": 0.0684,
      "grad_norm": 13.262563705444336,
      "learning_rate": 0.0001366,
      "loss": 6.9474,
      "step": 684
    },
    {
      "epoch": 0.0685,
      "grad_norm": 3.744182825088501,
      "learning_rate": 0.00013680000000000002,
      "loss": 7.2388,
      "step": 685
    },
    {
      "epoch": 0.0686,
      "grad_norm": 6.575530529022217,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.8945,
      "step": 686
    },
    {
      "epoch": 0.0687,
      "grad_norm": 3.6469550132751465,
      "learning_rate": 0.00013720000000000003,
      "loss": 7.2215,
      "step": 687
    },
    {
      "epoch": 0.0688,
      "grad_norm": 6.964643955230713,
      "learning_rate": 0.0001374,
      "loss": 7.0936,
      "step": 688
    },
    {
      "epoch": 0.0689,
      "grad_norm": 7.719574451446533,
      "learning_rate": 0.00013759999999999998,
      "loss": 7.2398,
      "step": 689
    },
    {
      "epoch": 0.069,
      "grad_norm": 4.1688618659973145,
      "learning_rate": 0.0001378,
      "loss": 7.0156,
      "step": 690
    },
    {
      "epoch": 0.0691,
      "grad_norm": 4.87483549118042,
      "learning_rate": 0.000138,
      "loss": 7.0137,
      "step": 691
    },
    {
      "epoch": 0.0692,
      "grad_norm": 4.838281154632568,
      "learning_rate": 0.0001382,
      "loss": 6.8534,
      "step": 692
    },
    {
      "epoch": 0.0693,
      "grad_norm": 4.840050220489502,
      "learning_rate": 0.0001384,
      "loss": 6.8991,
      "step": 693
    },
    {
      "epoch": 0.0694,
      "grad_norm": 7.077585697174072,
      "learning_rate": 0.0001386,
      "loss": 7.0046,
      "step": 694
    },
    {
      "epoch": 0.0695,
      "grad_norm": 3.8322794437408447,
      "learning_rate": 0.00013879999999999999,
      "loss": 7.1144,
      "step": 695
    },
    {
      "epoch": 0.0696,
      "grad_norm": 19.375001907348633,
      "learning_rate": 0.000139,
      "loss": 7.3403,
      "step": 696
    },
    {
      "epoch": 0.0697,
      "grad_norm": 4.469961643218994,
      "learning_rate": 0.0001392,
      "loss": 8.0547,
      "step": 697
    },
    {
      "epoch": 0.0698,
      "grad_norm": 4.1195387840271,
      "learning_rate": 0.0001394,
      "loss": 7.6468,
      "step": 698
    },
    {
      "epoch": 0.0699,
      "grad_norm": 13.412221908569336,
      "learning_rate": 0.0001396,
      "loss": 6.7323,
      "step": 699
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.992982864379883,
      "learning_rate": 0.0001398,
      "loss": 7.0787,
      "step": 700
    },
    {
      "epoch": 0.0701,
      "grad_norm": 16.758211135864258,
      "learning_rate": 0.00014,
      "loss": 7.3083,
      "step": 701
    },
    {
      "epoch": 0.0702,
      "grad_norm": 22.177547454833984,
      "learning_rate": 0.0001402,
      "loss": 7.5409,
      "step": 702
    },
    {
      "epoch": 0.0703,
      "grad_norm": 6.197084426879883,
      "learning_rate": 0.0001404,
      "loss": 7.627,
      "step": 703
    },
    {
      "epoch": 0.0704,
      "grad_norm": 3.1823089122772217,
      "learning_rate": 0.0001406,
      "loss": 7.1331,
      "step": 704
    },
    {
      "epoch": 0.0705,
      "grad_norm": 4.422345161437988,
      "learning_rate": 0.0001408,
      "loss": 7.3769,
      "step": 705
    },
    {
      "epoch": 0.0706,
      "grad_norm": 6.225074768066406,
      "learning_rate": 0.000141,
      "loss": 7.4541,
      "step": 706
    },
    {
      "epoch": 0.0707,
      "grad_norm": 2.9992876052856445,
      "learning_rate": 0.0001412,
      "loss": 7.2511,
      "step": 707
    },
    {
      "epoch": 0.0708,
      "grad_norm": 4.677290916442871,
      "learning_rate": 0.0001414,
      "loss": 6.9179,
      "step": 708
    },
    {
      "epoch": 0.0709,
      "grad_norm": 7.120545387268066,
      "learning_rate": 0.0001416,
      "loss": 7.2053,
      "step": 709
    },
    {
      "epoch": 0.071,
      "grad_norm": 7.875759601593018,
      "learning_rate": 0.0001418,
      "loss": 7.344,
      "step": 710
    },
    {
      "epoch": 0.0711,
      "grad_norm": 3.741600751876831,
      "learning_rate": 0.000142,
      "loss": 6.8226,
      "step": 711
    },
    {
      "epoch": 0.0712,
      "grad_norm": 4.170473098754883,
      "learning_rate": 0.0001422,
      "loss": 7.0018,
      "step": 712
    },
    {
      "epoch": 0.0713,
      "grad_norm": 3.511518955230713,
      "learning_rate": 0.0001424,
      "loss": 7.4372,
      "step": 713
    },
    {
      "epoch": 0.0714,
      "grad_norm": 3.5072054862976074,
      "learning_rate": 0.0001426,
      "loss": 7.0726,
      "step": 714
    },
    {
      "epoch": 0.0715,
      "grad_norm": 6.041854381561279,
      "learning_rate": 0.0001428,
      "loss": 7.1256,
      "step": 715
    },
    {
      "epoch": 0.0716,
      "grad_norm": 2.9688358306884766,
      "learning_rate": 0.000143,
      "loss": 7.5836,
      "step": 716
    },
    {
      "epoch": 0.0717,
      "grad_norm": 25.155399322509766,
      "learning_rate": 0.0001432,
      "loss": 7.8188,
      "step": 717
    },
    {
      "epoch": 0.0718,
      "grad_norm": 6.106550216674805,
      "learning_rate": 0.0001434,
      "loss": 6.9338,
      "step": 718
    },
    {
      "epoch": 0.0719,
      "grad_norm": 6.470298767089844,
      "learning_rate": 0.0001436,
      "loss": 6.9327,
      "step": 719
    },
    {
      "epoch": 0.072,
      "grad_norm": 10.117387771606445,
      "learning_rate": 0.0001438,
      "loss": 7.7681,
      "step": 720
    },
    {
      "epoch": 0.0721,
      "grad_norm": 32.549293518066406,
      "learning_rate": 0.000144,
      "loss": 7.5878,
      "step": 721
    },
    {
      "epoch": 0.0722,
      "grad_norm": 6.842835426330566,
      "learning_rate": 0.0001442,
      "loss": 6.6821,
      "step": 722
    },
    {
      "epoch": 0.0723,
      "grad_norm": 5.548720836639404,
      "learning_rate": 0.0001444,
      "loss": 7.4643,
      "step": 723
    },
    {
      "epoch": 0.0724,
      "grad_norm": 2.9626762866973877,
      "learning_rate": 0.0001446,
      "loss": 6.8786,
      "step": 724
    },
    {
      "epoch": 0.0725,
      "grad_norm": 5.541468143463135,
      "learning_rate": 0.0001448,
      "loss": 7.1186,
      "step": 725
    },
    {
      "epoch": 0.0726,
      "grad_norm": 4.904905319213867,
      "learning_rate": 0.000145,
      "loss": 6.9763,
      "step": 726
    },
    {
      "epoch": 0.0727,
      "grad_norm": 11.856684684753418,
      "learning_rate": 0.0001452,
      "loss": 7.2995,
      "step": 727
    },
    {
      "epoch": 0.0728,
      "grad_norm": 5.324658393859863,
      "learning_rate": 0.0001454,
      "loss": 6.971,
      "step": 728
    },
    {
      "epoch": 0.0729,
      "grad_norm": 8.075563430786133,
      "learning_rate": 0.00014560000000000002,
      "loss": 7.1455,
      "step": 729
    },
    {
      "epoch": 0.073,
      "grad_norm": 14.151449203491211,
      "learning_rate": 0.0001458,
      "loss": 7.5627,
      "step": 730
    },
    {
      "epoch": 0.0731,
      "grad_norm": 4.019474506378174,
      "learning_rate": 0.000146,
      "loss": 7.1634,
      "step": 731
    },
    {
      "epoch": 0.0732,
      "grad_norm": 9.517170906066895,
      "learning_rate": 0.0001462,
      "loss": 7.2093,
      "step": 732
    },
    {
      "epoch": 0.0733,
      "grad_norm": 10.162856101989746,
      "learning_rate": 0.0001464,
      "loss": 7.2644,
      "step": 733
    },
    {
      "epoch": 0.0734,
      "grad_norm": 3.8313047885894775,
      "learning_rate": 0.0001466,
      "loss": 7.0877,
      "step": 734
    },
    {
      "epoch": 0.0735,
      "grad_norm": 16.92441749572754,
      "learning_rate": 0.00014680000000000002,
      "loss": 8.1864,
      "step": 735
    },
    {
      "epoch": 0.0736,
      "grad_norm": 5.420156002044678,
      "learning_rate": 0.000147,
      "loss": 7.3511,
      "step": 736
    },
    {
      "epoch": 0.0737,
      "grad_norm": 4.650784015655518,
      "learning_rate": 0.0001472,
      "loss": 7.2157,
      "step": 737
    },
    {
      "epoch": 0.0738,
      "grad_norm": 2.9390485286712646,
      "learning_rate": 0.0001474,
      "loss": 7.3112,
      "step": 738
    },
    {
      "epoch": 0.0739,
      "grad_norm": 6.099313735961914,
      "learning_rate": 0.0001476,
      "loss": 7.2161,
      "step": 739
    },
    {
      "epoch": 0.074,
      "grad_norm": 8.140044212341309,
      "learning_rate": 0.00014780000000000001,
      "loss": 7.0026,
      "step": 740
    },
    {
      "epoch": 0.0741,
      "grad_norm": 4.674846649169922,
      "learning_rate": 0.000148,
      "loss": 6.974,
      "step": 741
    },
    {
      "epoch": 0.0742,
      "grad_norm": 6.648042678833008,
      "learning_rate": 0.0001482,
      "loss": 7.8925,
      "step": 742
    },
    {
      "epoch": 0.0743,
      "grad_norm": 9.798699378967285,
      "learning_rate": 0.0001484,
      "loss": 7.4155,
      "step": 743
    },
    {
      "epoch": 0.0744,
      "grad_norm": 6.849429607391357,
      "learning_rate": 0.0001486,
      "loss": 8.2364,
      "step": 744
    },
    {
      "epoch": 0.0745,
      "grad_norm": 5.99636173248291,
      "learning_rate": 0.0001488,
      "loss": 7.4622,
      "step": 745
    },
    {
      "epoch": 0.0746,
      "grad_norm": 5.3183064460754395,
      "learning_rate": 0.00014900000000000002,
      "loss": 7.3286,
      "step": 746
    },
    {
      "epoch": 0.0747,
      "grad_norm": 15.126444816589355,
      "learning_rate": 0.0001492,
      "loss": 7.5559,
      "step": 747
    },
    {
      "epoch": 0.0748,
      "grad_norm": 7.203035354614258,
      "learning_rate": 0.0001494,
      "loss": 7.1522,
      "step": 748
    },
    {
      "epoch": 0.0749,
      "grad_norm": 5.465102672576904,
      "learning_rate": 0.0001496,
      "loss": 7.0516,
      "step": 749
    },
    {
      "epoch": 0.075,
      "grad_norm": 7.611809730529785,
      "learning_rate": 0.0001498,
      "loss": 6.5472,
      "step": 750
    },
    {
      "epoch": 0.0751,
      "grad_norm": 8.898043632507324,
      "learning_rate": 0.00015000000000000001,
      "loss": 7.2243,
      "step": 751
    },
    {
      "epoch": 0.0752,
      "grad_norm": 3.7406883239746094,
      "learning_rate": 0.00015020000000000002,
      "loss": 7.386,
      "step": 752
    },
    {
      "epoch": 0.0753,
      "grad_norm": 6.3689961433410645,
      "learning_rate": 0.0001504,
      "loss": 6.9269,
      "step": 753
    },
    {
      "epoch": 0.0754,
      "grad_norm": 5.388195991516113,
      "learning_rate": 0.0001506,
      "loss": 6.8626,
      "step": 754
    },
    {
      "epoch": 0.0755,
      "grad_norm": 3.798511028289795,
      "learning_rate": 0.0001508,
      "loss": 7.7169,
      "step": 755
    },
    {
      "epoch": 0.0756,
      "grad_norm": 9.113856315612793,
      "learning_rate": 0.000151,
      "loss": 6.814,
      "step": 756
    },
    {
      "epoch": 0.0757,
      "grad_norm": 5.23146915435791,
      "learning_rate": 0.00015120000000000002,
      "loss": 7.2766,
      "step": 757
    },
    {
      "epoch": 0.0758,
      "grad_norm": 4.530202865600586,
      "learning_rate": 0.00015140000000000002,
      "loss": 7.0688,
      "step": 758
    },
    {
      "epoch": 0.0759,
      "grad_norm": 4.636505126953125,
      "learning_rate": 0.0001516,
      "loss": 7.4736,
      "step": 759
    },
    {
      "epoch": 0.076,
      "grad_norm": 4.124752998352051,
      "learning_rate": 0.0001518,
      "loss": 6.7041,
      "step": 760
    },
    {
      "epoch": 0.0761,
      "grad_norm": 12.439164161682129,
      "learning_rate": 0.000152,
      "loss": 6.6151,
      "step": 761
    },
    {
      "epoch": 0.0762,
      "grad_norm": 6.845168590545654,
      "learning_rate": 0.0001522,
      "loss": 6.9107,
      "step": 762
    },
    {
      "epoch": 0.0763,
      "grad_norm": 4.113043785095215,
      "learning_rate": 0.00015240000000000002,
      "loss": 6.9361,
      "step": 763
    },
    {
      "epoch": 0.0764,
      "grad_norm": 4.439939022064209,
      "learning_rate": 0.00015260000000000002,
      "loss": 7.4537,
      "step": 764
    },
    {
      "epoch": 0.0765,
      "grad_norm": 32.75420379638672,
      "learning_rate": 0.0001528,
      "loss": 7.7879,
      "step": 765
    },
    {
      "epoch": 0.0766,
      "grad_norm": 4.298772811889648,
      "learning_rate": 0.000153,
      "loss": 7.2707,
      "step": 766
    },
    {
      "epoch": 0.0767,
      "grad_norm": 5.283163070678711,
      "learning_rate": 0.0001532,
      "loss": 7.1269,
      "step": 767
    },
    {
      "epoch": 0.0768,
      "grad_norm": 10.612497329711914,
      "learning_rate": 0.00015340000000000002,
      "loss": 7.5158,
      "step": 768
    },
    {
      "epoch": 0.0769,
      "grad_norm": 8.533021926879883,
      "learning_rate": 0.00015360000000000002,
      "loss": 7.4827,
      "step": 769
    },
    {
      "epoch": 0.077,
      "grad_norm": 4.981786727905273,
      "learning_rate": 0.0001538,
      "loss": 6.802,
      "step": 770
    },
    {
      "epoch": 0.0771,
      "grad_norm": 3.7310562133789062,
      "learning_rate": 0.000154,
      "loss": 6.7169,
      "step": 771
    },
    {
      "epoch": 0.0772,
      "grad_norm": 4.525305271148682,
      "learning_rate": 0.0001542,
      "loss": 7.2565,
      "step": 772
    },
    {
      "epoch": 0.0773,
      "grad_norm": 3.999262809753418,
      "learning_rate": 0.0001544,
      "loss": 7.0843,
      "step": 773
    },
    {
      "epoch": 0.0774,
      "grad_norm": 3.6029906272888184,
      "learning_rate": 0.00015460000000000002,
      "loss": 7.1662,
      "step": 774
    },
    {
      "epoch": 0.0775,
      "grad_norm": 24.661516189575195,
      "learning_rate": 0.00015480000000000002,
      "loss": 7.3346,
      "step": 775
    },
    {
      "epoch": 0.0776,
      "grad_norm": 4.347546577453613,
      "learning_rate": 0.000155,
      "loss": 6.5019,
      "step": 776
    },
    {
      "epoch": 0.0777,
      "grad_norm": 5.066408157348633,
      "learning_rate": 0.0001552,
      "loss": 7.4025,
      "step": 777
    },
    {
      "epoch": 0.0778,
      "grad_norm": 48.751670837402344,
      "learning_rate": 0.0001554,
      "loss": 7.8933,
      "step": 778
    },
    {
      "epoch": 0.0779,
      "grad_norm": 5.6344404220581055,
      "learning_rate": 0.00015560000000000001,
      "loss": 6.9854,
      "step": 779
    },
    {
      "epoch": 0.078,
      "grad_norm": 5.292802333831787,
      "learning_rate": 0.00015580000000000002,
      "loss": 7.4937,
      "step": 780
    },
    {
      "epoch": 0.0781,
      "grad_norm": 5.217813014984131,
      "learning_rate": 0.00015600000000000002,
      "loss": 7.2425,
      "step": 781
    },
    {
      "epoch": 0.0782,
      "grad_norm": 5.441041946411133,
      "learning_rate": 0.0001562,
      "loss": 7.2137,
      "step": 782
    },
    {
      "epoch": 0.0783,
      "grad_norm": 18.92131996154785,
      "learning_rate": 0.0001564,
      "loss": 6.9637,
      "step": 783
    },
    {
      "epoch": 0.0784,
      "grad_norm": 8.150403022766113,
      "learning_rate": 0.0001566,
      "loss": 7.026,
      "step": 784
    },
    {
      "epoch": 0.0785,
      "grad_norm": 4.343782901763916,
      "learning_rate": 0.00015680000000000002,
      "loss": 6.3588,
      "step": 785
    },
    {
      "epoch": 0.0786,
      "grad_norm": 13.904641151428223,
      "learning_rate": 0.00015700000000000002,
      "loss": 7.0372,
      "step": 786
    },
    {
      "epoch": 0.0787,
      "grad_norm": 14.935260772705078,
      "learning_rate": 0.00015720000000000003,
      "loss": 7.2977,
      "step": 787
    },
    {
      "epoch": 0.0788,
      "grad_norm": 6.490274429321289,
      "learning_rate": 0.0001574,
      "loss": 7.9423,
      "step": 788
    },
    {
      "epoch": 0.0789,
      "grad_norm": 17.131515502929688,
      "learning_rate": 0.0001576,
      "loss": 7.2146,
      "step": 789
    },
    {
      "epoch": 0.079,
      "grad_norm": 4.4295430183410645,
      "learning_rate": 0.00015780000000000001,
      "loss": 7.0728,
      "step": 790
    },
    {
      "epoch": 0.0791,
      "grad_norm": 4.62942361831665,
      "learning_rate": 0.00015800000000000002,
      "loss": 7.4751,
      "step": 791
    },
    {
      "epoch": 0.0792,
      "grad_norm": 3.4293460845947266,
      "learning_rate": 0.00015820000000000002,
      "loss": 7.2049,
      "step": 792
    },
    {
      "epoch": 0.0793,
      "grad_norm": 3.490250587463379,
      "learning_rate": 0.00015840000000000003,
      "loss": 6.6937,
      "step": 793
    },
    {
      "epoch": 0.0794,
      "grad_norm": 4.507073879241943,
      "learning_rate": 0.0001586,
      "loss": 7.3623,
      "step": 794
    },
    {
      "epoch": 0.0795,
      "grad_norm": 3.2151615619659424,
      "learning_rate": 0.0001588,
      "loss": 7.2808,
      "step": 795
    },
    {
      "epoch": 0.0796,
      "grad_norm": 2.460549831390381,
      "learning_rate": 0.00015900000000000002,
      "loss": 7.2708,
      "step": 796
    },
    {
      "epoch": 0.0797,
      "grad_norm": 3.827099561691284,
      "learning_rate": 0.00015920000000000002,
      "loss": 7.58,
      "step": 797
    },
    {
      "epoch": 0.0798,
      "grad_norm": 4.03687047958374,
      "learning_rate": 0.00015940000000000003,
      "loss": 6.82,
      "step": 798
    },
    {
      "epoch": 0.0799,
      "grad_norm": 18.460262298583984,
      "learning_rate": 0.0001596,
      "loss": 6.8489,
      "step": 799
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.3828349113464355,
      "learning_rate": 0.0001598,
      "loss": 7.048,
      "step": 800
    },
    {
      "epoch": 0.0801,
      "grad_norm": 8.131593704223633,
      "learning_rate": 0.00016,
      "loss": 8.0731,
      "step": 801
    },
    {
      "epoch": 0.0802,
      "grad_norm": 3.199291706085205,
      "learning_rate": 0.00016020000000000002,
      "loss": 7.2305,
      "step": 802
    },
    {
      "epoch": 0.0803,
      "grad_norm": 7.3391876220703125,
      "learning_rate": 0.00016040000000000002,
      "loss": 7.2057,
      "step": 803
    },
    {
      "epoch": 0.0804,
      "grad_norm": 33.594947814941406,
      "learning_rate": 0.00016060000000000003,
      "loss": 6.9462,
      "step": 804
    },
    {
      "epoch": 0.0805,
      "grad_norm": 12.214316368103027,
      "learning_rate": 0.0001608,
      "loss": 7.7696,
      "step": 805
    },
    {
      "epoch": 0.0806,
      "grad_norm": 7.877429485321045,
      "learning_rate": 0.000161,
      "loss": 7.1598,
      "step": 806
    },
    {
      "epoch": 0.0807,
      "grad_norm": 6.0569634437561035,
      "learning_rate": 0.00016120000000000002,
      "loss": 7.3528,
      "step": 807
    },
    {
      "epoch": 0.0808,
      "grad_norm": 25.502315521240234,
      "learning_rate": 0.00016140000000000002,
      "loss": 8.0207,
      "step": 808
    },
    {
      "epoch": 0.0809,
      "grad_norm": 15.811558723449707,
      "learning_rate": 0.00016160000000000002,
      "loss": 7.4122,
      "step": 809
    },
    {
      "epoch": 0.081,
      "grad_norm": 5.46779727935791,
      "learning_rate": 0.00016180000000000003,
      "loss": 7.0173,
      "step": 810
    },
    {
      "epoch": 0.0811,
      "grad_norm": 31.307207107543945,
      "learning_rate": 0.000162,
      "loss": 6.8314,
      "step": 811
    },
    {
      "epoch": 0.0812,
      "grad_norm": 14.116832733154297,
      "learning_rate": 0.0001622,
      "loss": 6.8362,
      "step": 812
    },
    {
      "epoch": 0.0813,
      "grad_norm": 16.04140281677246,
      "learning_rate": 0.00016240000000000002,
      "loss": 7.1674,
      "step": 813
    },
    {
      "epoch": 0.0814,
      "grad_norm": 4.74333381652832,
      "learning_rate": 0.0001626,
      "loss": 6.876,
      "step": 814
    },
    {
      "epoch": 0.0815,
      "grad_norm": 4.89963960647583,
      "learning_rate": 0.0001628,
      "loss": 6.7041,
      "step": 815
    },
    {
      "epoch": 0.0816,
      "grad_norm": 4.8412766456604,
      "learning_rate": 0.000163,
      "loss": 7.3875,
      "step": 816
    },
    {
      "epoch": 0.0817,
      "grad_norm": 4.565502643585205,
      "learning_rate": 0.0001632,
      "loss": 7.084,
      "step": 817
    },
    {
      "epoch": 0.0818,
      "grad_norm": 5.439000606536865,
      "learning_rate": 0.0001634,
      "loss": 6.5941,
      "step": 818
    },
    {
      "epoch": 0.0819,
      "grad_norm": 5.494342803955078,
      "learning_rate": 0.0001636,
      "loss": 6.8743,
      "step": 819
    },
    {
      "epoch": 0.082,
      "grad_norm": 4.809220790863037,
      "learning_rate": 0.0001638,
      "loss": 6.5915,
      "step": 820
    },
    {
      "epoch": 0.0821,
      "grad_norm": 10.969738006591797,
      "learning_rate": 0.000164,
      "loss": 7.0664,
      "step": 821
    },
    {
      "epoch": 0.0822,
      "grad_norm": 4.008328914642334,
      "learning_rate": 0.0001642,
      "loss": 7.0348,
      "step": 822
    },
    {
      "epoch": 0.0823,
      "grad_norm": 7.907390594482422,
      "learning_rate": 0.0001644,
      "loss": 7.1436,
      "step": 823
    },
    {
      "epoch": 0.0824,
      "grad_norm": 11.380629539489746,
      "learning_rate": 0.0001646,
      "loss": 7.4018,
      "step": 824
    },
    {
      "epoch": 0.0825,
      "grad_norm": 9.390121459960938,
      "learning_rate": 0.0001648,
      "loss": 6.8081,
      "step": 825
    },
    {
      "epoch": 0.0826,
      "grad_norm": 5.918392181396484,
      "learning_rate": 0.000165,
      "loss": 6.7154,
      "step": 826
    },
    {
      "epoch": 0.0827,
      "grad_norm": 3.92728328704834,
      "learning_rate": 0.0001652,
      "loss": 6.7915,
      "step": 827
    },
    {
      "epoch": 0.0828,
      "grad_norm": 13.540765762329102,
      "learning_rate": 0.0001654,
      "loss": 7.0725,
      "step": 828
    },
    {
      "epoch": 0.0829,
      "grad_norm": 3.5492749214172363,
      "learning_rate": 0.0001656,
      "loss": 6.8579,
      "step": 829
    },
    {
      "epoch": 0.083,
      "grad_norm": 3.5679306983947754,
      "learning_rate": 0.0001658,
      "loss": 6.6405,
      "step": 830
    },
    {
      "epoch": 0.0831,
      "grad_norm": 3.9305801391601562,
      "learning_rate": 0.000166,
      "loss": 7.5462,
      "step": 831
    },
    {
      "epoch": 0.0832,
      "grad_norm": 3.889373302459717,
      "learning_rate": 0.0001662,
      "loss": 7.5024,
      "step": 832
    },
    {
      "epoch": 0.0833,
      "grad_norm": 5.9754838943481445,
      "learning_rate": 0.0001664,
      "loss": 7.2243,
      "step": 833
    },
    {
      "epoch": 0.0834,
      "grad_norm": 6.250219345092773,
      "learning_rate": 0.0001666,
      "loss": 7.2148,
      "step": 834
    },
    {
      "epoch": 0.0835,
      "grad_norm": 3.597778797149658,
      "learning_rate": 0.0001668,
      "loss": 7.1077,
      "step": 835
    },
    {
      "epoch": 0.0836,
      "grad_norm": 2.986114501953125,
      "learning_rate": 0.000167,
      "loss": 7.0213,
      "step": 836
    },
    {
      "epoch": 0.0837,
      "grad_norm": 3.1118268966674805,
      "learning_rate": 0.0001672,
      "loss": 6.8926,
      "step": 837
    },
    {
      "epoch": 0.0838,
      "grad_norm": 3.736161708831787,
      "learning_rate": 0.0001674,
      "loss": 7.1398,
      "step": 838
    },
    {
      "epoch": 0.0839,
      "grad_norm": 12.72769832611084,
      "learning_rate": 0.0001676,
      "loss": 6.9002,
      "step": 839
    },
    {
      "epoch": 0.084,
      "grad_norm": 5.461266040802002,
      "learning_rate": 0.0001678,
      "loss": 6.9845,
      "step": 840
    },
    {
      "epoch": 0.0841,
      "grad_norm": 7.33488655090332,
      "learning_rate": 0.000168,
      "loss": 7.1802,
      "step": 841
    },
    {
      "epoch": 0.0842,
      "grad_norm": 17.176977157592773,
      "learning_rate": 0.0001682,
      "loss": 7.4854,
      "step": 842
    },
    {
      "epoch": 0.0843,
      "grad_norm": 4.9659600257873535,
      "learning_rate": 0.0001684,
      "loss": 6.805,
      "step": 843
    },
    {
      "epoch": 0.0844,
      "grad_norm": 6.806047439575195,
      "learning_rate": 0.0001686,
      "loss": 7.6618,
      "step": 844
    },
    {
      "epoch": 0.0845,
      "grad_norm": 3.9508183002471924,
      "learning_rate": 0.0001688,
      "loss": 6.4685,
      "step": 845
    },
    {
      "epoch": 0.0846,
      "grad_norm": 3.905745029449463,
      "learning_rate": 0.00016900000000000002,
      "loss": 7.0511,
      "step": 846
    },
    {
      "epoch": 0.0847,
      "grad_norm": 3.6470706462860107,
      "learning_rate": 0.0001692,
      "loss": 6.8347,
      "step": 847
    },
    {
      "epoch": 0.0848,
      "grad_norm": 4.804910182952881,
      "learning_rate": 0.0001694,
      "loss": 7.1789,
      "step": 848
    },
    {
      "epoch": 0.0849,
      "grad_norm": 10.311763763427734,
      "learning_rate": 0.0001696,
      "loss": 7.2153,
      "step": 849
    },
    {
      "epoch": 0.085,
      "grad_norm": 8.795881271362305,
      "learning_rate": 0.0001698,
      "loss": 6.8401,
      "step": 850
    },
    {
      "epoch": 0.0851,
      "grad_norm": 4.028997898101807,
      "learning_rate": 0.00017,
      "loss": 6.5504,
      "step": 851
    },
    {
      "epoch": 0.0852,
      "grad_norm": 5.391139507293701,
      "learning_rate": 0.00017020000000000002,
      "loss": 7.5991,
      "step": 852
    },
    {
      "epoch": 0.0853,
      "grad_norm": 25.018274307250977,
      "learning_rate": 0.0001704,
      "loss": 7.2749,
      "step": 853
    },
    {
      "epoch": 0.0854,
      "grad_norm": 5.6703901290893555,
      "learning_rate": 0.0001706,
      "loss": 7.2565,
      "step": 854
    },
    {
      "epoch": 0.0855,
      "grad_norm": 5.789599418640137,
      "learning_rate": 0.0001708,
      "loss": 6.9274,
      "step": 855
    },
    {
      "epoch": 0.0856,
      "grad_norm": 10.444012641906738,
      "learning_rate": 0.000171,
      "loss": 7.3317,
      "step": 856
    },
    {
      "epoch": 0.0857,
      "grad_norm": 3.6762800216674805,
      "learning_rate": 0.00017120000000000001,
      "loss": 7.2356,
      "step": 857
    },
    {
      "epoch": 0.0858,
      "grad_norm": 3.880971670150757,
      "learning_rate": 0.0001714,
      "loss": 6.3548,
      "step": 858
    },
    {
      "epoch": 0.0859,
      "grad_norm": 3.7576401233673096,
      "learning_rate": 0.0001716,
      "loss": 6.5808,
      "step": 859
    },
    {
      "epoch": 0.086,
      "grad_norm": 6.138988971710205,
      "learning_rate": 0.0001718,
      "loss": 6.8576,
      "step": 860
    },
    {
      "epoch": 0.0861,
      "grad_norm": 73.68218231201172,
      "learning_rate": 0.000172,
      "loss": 8.2504,
      "step": 861
    },
    {
      "epoch": 0.0862,
      "grad_norm": 15.407068252563477,
      "learning_rate": 0.0001722,
      "loss": 6.7586,
      "step": 862
    },
    {
      "epoch": 0.0863,
      "grad_norm": 12.2488431930542,
      "learning_rate": 0.00017240000000000002,
      "loss": 7.6605,
      "step": 863
    },
    {
      "epoch": 0.0864,
      "grad_norm": 5.4976911544799805,
      "learning_rate": 0.0001726,
      "loss": 7.1972,
      "step": 864
    },
    {
      "epoch": 0.0865,
      "grad_norm": 4.018867492675781,
      "learning_rate": 0.0001728,
      "loss": 6.8955,
      "step": 865
    },
    {
      "epoch": 0.0866,
      "grad_norm": 4.687366962432861,
      "learning_rate": 0.000173,
      "loss": 6.5933,
      "step": 866
    },
    {
      "epoch": 0.0867,
      "grad_norm": 12.344051361083984,
      "learning_rate": 0.0001732,
      "loss": 6.6364,
      "step": 867
    },
    {
      "epoch": 0.0868,
      "grad_norm": 6.03907585144043,
      "learning_rate": 0.0001734,
      "loss": 7.2256,
      "step": 868
    },
    {
      "epoch": 0.0869,
      "grad_norm": 3.8393728733062744,
      "learning_rate": 0.00017360000000000002,
      "loss": 7.1169,
      "step": 869
    },
    {
      "epoch": 0.087,
      "grad_norm": 2.536062479019165,
      "learning_rate": 0.0001738,
      "loss": 7.0526,
      "step": 870
    },
    {
      "epoch": 0.0871,
      "grad_norm": 4.98984432220459,
      "learning_rate": 0.000174,
      "loss": 7.1075,
      "step": 871
    },
    {
      "epoch": 0.0872,
      "grad_norm": 3.463768243789673,
      "learning_rate": 0.0001742,
      "loss": 7.0315,
      "step": 872
    },
    {
      "epoch": 0.0873,
      "grad_norm": 5.810747146606445,
      "learning_rate": 0.0001744,
      "loss": 7.6817,
      "step": 873
    },
    {
      "epoch": 0.0874,
      "grad_norm": 14.575197219848633,
      "learning_rate": 0.00017460000000000002,
      "loss": 7.4933,
      "step": 874
    },
    {
      "epoch": 0.0875,
      "grad_norm": 7.615855693817139,
      "learning_rate": 0.00017480000000000002,
      "loss": 6.6253,
      "step": 875
    },
    {
      "epoch": 0.0876,
      "grad_norm": 4.181806564331055,
      "learning_rate": 0.000175,
      "loss": 7.0101,
      "step": 876
    },
    {
      "epoch": 0.0877,
      "grad_norm": 6.206484794616699,
      "learning_rate": 0.0001752,
      "loss": 7.1807,
      "step": 877
    },
    {
      "epoch": 0.0878,
      "grad_norm": 4.070094108581543,
      "learning_rate": 0.0001754,
      "loss": 7.0565,
      "step": 878
    },
    {
      "epoch": 0.0879,
      "grad_norm": 4.0230207443237305,
      "learning_rate": 0.0001756,
      "loss": 7.0842,
      "step": 879
    },
    {
      "epoch": 0.088,
      "grad_norm": 23.659271240234375,
      "learning_rate": 0.00017580000000000002,
      "loss": 7.5441,
      "step": 880
    },
    {
      "epoch": 0.0881,
      "grad_norm": 9.729623794555664,
      "learning_rate": 0.00017600000000000002,
      "loss": 7.0748,
      "step": 881
    },
    {
      "epoch": 0.0882,
      "grad_norm": 8.202091217041016,
      "learning_rate": 0.0001762,
      "loss": 6.7353,
      "step": 882
    },
    {
      "epoch": 0.0883,
      "grad_norm": 4.450812816619873,
      "learning_rate": 0.0001764,
      "loss": 6.7851,
      "step": 883
    },
    {
      "epoch": 0.0884,
      "grad_norm": 9.16330623626709,
      "learning_rate": 0.0001766,
      "loss": 7.4784,
      "step": 884
    },
    {
      "epoch": 0.0885,
      "grad_norm": 9.906651496887207,
      "learning_rate": 0.00017680000000000001,
      "loss": 7.217,
      "step": 885
    },
    {
      "epoch": 0.0886,
      "grad_norm": 6.146594047546387,
      "learning_rate": 0.00017700000000000002,
      "loss": 6.7052,
      "step": 886
    },
    {
      "epoch": 0.0887,
      "grad_norm": 5.286199569702148,
      "learning_rate": 0.0001772,
      "loss": 7.1987,
      "step": 887
    },
    {
      "epoch": 0.0888,
      "grad_norm": 4.827280521392822,
      "learning_rate": 0.0001774,
      "loss": 6.3226,
      "step": 888
    },
    {
      "epoch": 0.0889,
      "grad_norm": 6.787034511566162,
      "learning_rate": 0.0001776,
      "loss": 7.1258,
      "step": 889
    },
    {
      "epoch": 0.089,
      "grad_norm": 3.879135847091675,
      "learning_rate": 0.0001778,
      "loss": 6.9543,
      "step": 890
    },
    {
      "epoch": 0.0891,
      "grad_norm": 23.666730880737305,
      "learning_rate": 0.00017800000000000002,
      "loss": 6.8733,
      "step": 891
    },
    {
      "epoch": 0.0892,
      "grad_norm": 7.074818134307861,
      "learning_rate": 0.00017820000000000002,
      "loss": 6.8929,
      "step": 892
    },
    {
      "epoch": 0.0893,
      "grad_norm": 11.122230529785156,
      "learning_rate": 0.0001784,
      "loss": 7.2024,
      "step": 893
    },
    {
      "epoch": 0.0894,
      "grad_norm": 36.56306457519531,
      "learning_rate": 0.0001786,
      "loss": 7.9083,
      "step": 894
    },
    {
      "epoch": 0.0895,
      "grad_norm": 15.13006591796875,
      "learning_rate": 0.0001788,
      "loss": 7.0631,
      "step": 895
    },
    {
      "epoch": 0.0896,
      "grad_norm": 3.2422385215759277,
      "learning_rate": 0.00017900000000000001,
      "loss": 6.6744,
      "step": 896
    },
    {
      "epoch": 0.0897,
      "grad_norm": 3.6930792331695557,
      "learning_rate": 0.00017920000000000002,
      "loss": 7.6176,
      "step": 897
    },
    {
      "epoch": 0.0898,
      "grad_norm": 12.0908203125,
      "learning_rate": 0.00017940000000000002,
      "loss": 6.842,
      "step": 898
    },
    {
      "epoch": 0.0899,
      "grad_norm": 4.322810649871826,
      "learning_rate": 0.0001796,
      "loss": 6.8417,
      "step": 899
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.608077526092529,
      "learning_rate": 0.0001798,
      "loss": 6.8053,
      "step": 900
    },
    {
      "epoch": 0.0901,
      "grad_norm": 44.53501892089844,
      "learning_rate": 0.00018,
      "loss": 7.8115,
      "step": 901
    },
    {
      "epoch": 0.0902,
      "grad_norm": 6.269430637359619,
      "learning_rate": 0.00018020000000000002,
      "loss": 6.9088,
      "step": 902
    },
    {
      "epoch": 0.0903,
      "grad_norm": 6.087910175323486,
      "learning_rate": 0.00018040000000000002,
      "loss": 7.2507,
      "step": 903
    },
    {
      "epoch": 0.0904,
      "grad_norm": 5.649314880371094,
      "learning_rate": 0.00018060000000000003,
      "loss": 7.1106,
      "step": 904
    },
    {
      "epoch": 0.0905,
      "grad_norm": 5.124192714691162,
      "learning_rate": 0.0001808,
      "loss": 7.348,
      "step": 905
    },
    {
      "epoch": 0.0906,
      "grad_norm": 5.5590105056762695,
      "learning_rate": 0.000181,
      "loss": 6.686,
      "step": 906
    },
    {
      "epoch": 0.0907,
      "grad_norm": 4.107783317565918,
      "learning_rate": 0.0001812,
      "loss": 6.9153,
      "step": 907
    },
    {
      "epoch": 0.0908,
      "grad_norm": 13.860953330993652,
      "learning_rate": 0.00018140000000000002,
      "loss": 7.0023,
      "step": 908
    },
    {
      "epoch": 0.0909,
      "grad_norm": 4.05351448059082,
      "learning_rate": 0.00018160000000000002,
      "loss": 6.9399,
      "step": 909
    },
    {
      "epoch": 0.091,
      "grad_norm": 7.015347480773926,
      "learning_rate": 0.00018180000000000003,
      "loss": 6.5182,
      "step": 910
    },
    {
      "epoch": 0.0911,
      "grad_norm": 3.55946946144104,
      "learning_rate": 0.000182,
      "loss": 6.4981,
      "step": 911
    },
    {
      "epoch": 0.0912,
      "grad_norm": 17.292724609375,
      "learning_rate": 0.0001822,
      "loss": 8.317,
      "step": 912
    },
    {
      "epoch": 0.0913,
      "grad_norm": 6.074021339416504,
      "learning_rate": 0.00018240000000000002,
      "loss": 7.3428,
      "step": 913
    },
    {
      "epoch": 0.0914,
      "grad_norm": 13.51147174835205,
      "learning_rate": 0.00018260000000000002,
      "loss": 6.9635,
      "step": 914
    },
    {
      "epoch": 0.0915,
      "grad_norm": 4.390360355377197,
      "learning_rate": 0.00018280000000000003,
      "loss": 6.7597,
      "step": 915
    },
    {
      "epoch": 0.0916,
      "grad_norm": 5.239279747009277,
      "learning_rate": 0.000183,
      "loss": 6.5374,
      "step": 916
    },
    {
      "epoch": 0.0917,
      "grad_norm": 16.581058502197266,
      "learning_rate": 0.0001832,
      "loss": 7.0988,
      "step": 917
    },
    {
      "epoch": 0.0918,
      "grad_norm": 3.3401846885681152,
      "learning_rate": 0.0001834,
      "loss": 6.8736,
      "step": 918
    },
    {
      "epoch": 0.0919,
      "grad_norm": 5.49683141708374,
      "learning_rate": 0.00018360000000000002,
      "loss": 7.131,
      "step": 919
    },
    {
      "epoch": 0.092,
      "grad_norm": 3.2990808486938477,
      "learning_rate": 0.00018380000000000002,
      "loss": 6.8398,
      "step": 920
    },
    {
      "epoch": 0.0921,
      "grad_norm": 4.386711120605469,
      "learning_rate": 0.00018400000000000003,
      "loss": 6.6627,
      "step": 921
    },
    {
      "epoch": 0.0922,
      "grad_norm": 3.2821545600891113,
      "learning_rate": 0.0001842,
      "loss": 6.4804,
      "step": 922
    },
    {
      "epoch": 0.0923,
      "grad_norm": 2.779437780380249,
      "learning_rate": 0.0001844,
      "loss": 7.0331,
      "step": 923
    },
    {
      "epoch": 0.0924,
      "grad_norm": 8.51917839050293,
      "learning_rate": 0.00018460000000000001,
      "loss": 6.492,
      "step": 924
    },
    {
      "epoch": 0.0925,
      "grad_norm": 12.202203750610352,
      "learning_rate": 0.00018480000000000002,
      "loss": 6.5362,
      "step": 925
    },
    {
      "epoch": 0.0926,
      "grad_norm": 3.168071746826172,
      "learning_rate": 0.00018500000000000002,
      "loss": 6.9651,
      "step": 926
    },
    {
      "epoch": 0.0927,
      "grad_norm": 27.488046646118164,
      "learning_rate": 0.00018520000000000003,
      "loss": 6.9554,
      "step": 927
    },
    {
      "epoch": 0.0928,
      "grad_norm": 16.298906326293945,
      "learning_rate": 0.0001854,
      "loss": 6.7621,
      "step": 928
    },
    {
      "epoch": 0.0929,
      "grad_norm": 2.6459529399871826,
      "learning_rate": 0.0001856,
      "loss": 6.843,
      "step": 929
    },
    {
      "epoch": 0.093,
      "grad_norm": 4.886815071105957,
      "learning_rate": 0.00018580000000000002,
      "loss": 7.401,
      "step": 930
    },
    {
      "epoch": 0.0931,
      "grad_norm": 3.3537356853485107,
      "learning_rate": 0.00018600000000000002,
      "loss": 6.8189,
      "step": 931
    },
    {
      "epoch": 0.0932,
      "grad_norm": 2.9334847927093506,
      "learning_rate": 0.00018620000000000003,
      "loss": 7.4391,
      "step": 932
    },
    {
      "epoch": 0.0933,
      "grad_norm": 3.183530807495117,
      "learning_rate": 0.00018640000000000003,
      "loss": 7.1284,
      "step": 933
    },
    {
      "epoch": 0.0934,
      "grad_norm": 3.073129177093506,
      "learning_rate": 0.0001866,
      "loss": 6.841,
      "step": 934
    },
    {
      "epoch": 0.0935,
      "grad_norm": 4.384754180908203,
      "learning_rate": 0.00018680000000000001,
      "loss": 6.5508,
      "step": 935
    },
    {
      "epoch": 0.0936,
      "grad_norm": 4.908419609069824,
      "learning_rate": 0.00018700000000000002,
      "loss": 6.7967,
      "step": 936
    },
    {
      "epoch": 0.0937,
      "grad_norm": 21.680166244506836,
      "learning_rate": 0.00018720000000000002,
      "loss": 7.1579,
      "step": 937
    },
    {
      "epoch": 0.0938,
      "grad_norm": 5.752041816711426,
      "learning_rate": 0.00018740000000000003,
      "loss": 7.1384,
      "step": 938
    },
    {
      "epoch": 0.0939,
      "grad_norm": 4.529350757598877,
      "learning_rate": 0.0001876,
      "loss": 6.9556,
      "step": 939
    },
    {
      "epoch": 0.094,
      "grad_norm": 2.650780200958252,
      "learning_rate": 0.0001878,
      "loss": 6.7766,
      "step": 940
    },
    {
      "epoch": 0.0941,
      "grad_norm": 10.476943969726562,
      "learning_rate": 0.000188,
      "loss": 6.5742,
      "step": 941
    },
    {
      "epoch": 0.0942,
      "grad_norm": 2.7315735816955566,
      "learning_rate": 0.0001882,
      "loss": 7.4329,
      "step": 942
    },
    {
      "epoch": 0.0943,
      "grad_norm": 3.3913047313690186,
      "learning_rate": 0.0001884,
      "loss": 6.5802,
      "step": 943
    },
    {
      "epoch": 0.0944,
      "grad_norm": 3.998079299926758,
      "learning_rate": 0.0001886,
      "loss": 6.8132,
      "step": 944
    },
    {
      "epoch": 0.0945,
      "grad_norm": 24.37404441833496,
      "learning_rate": 0.0001888,
      "loss": 7.3873,
      "step": 945
    },
    {
      "epoch": 0.0946,
      "grad_norm": 6.77671480178833,
      "learning_rate": 0.00018899999999999999,
      "loss": 6.7382,
      "step": 946
    },
    {
      "epoch": 0.0947,
      "grad_norm": 26.45954132080078,
      "learning_rate": 0.0001892,
      "loss": 7.5704,
      "step": 947
    },
    {
      "epoch": 0.0948,
      "grad_norm": 10.62767505645752,
      "learning_rate": 0.0001894,
      "loss": 6.9731,
      "step": 948
    },
    {
      "epoch": 0.0949,
      "grad_norm": 3.9411978721618652,
      "learning_rate": 0.0001896,
      "loss": 6.6512,
      "step": 949
    },
    {
      "epoch": 0.095,
      "grad_norm": 4.751908779144287,
      "learning_rate": 0.0001898,
      "loss": 7.4046,
      "step": 950
    },
    {
      "epoch": 0.0951,
      "grad_norm": 3.6539804935455322,
      "learning_rate": 0.00019,
      "loss": 6.9363,
      "step": 951
    },
    {
      "epoch": 0.0952,
      "grad_norm": 4.5021233558654785,
      "learning_rate": 0.0001902,
      "loss": 6.4668,
      "step": 952
    },
    {
      "epoch": 0.0953,
      "grad_norm": 3.8840153217315674,
      "learning_rate": 0.0001904,
      "loss": 6.5611,
      "step": 953
    },
    {
      "epoch": 0.0954,
      "grad_norm": 2.821376323699951,
      "learning_rate": 0.0001906,
      "loss": 6.3473,
      "step": 954
    },
    {
      "epoch": 0.0955,
      "grad_norm": 4.995748519897461,
      "learning_rate": 0.0001908,
      "loss": 6.6011,
      "step": 955
    },
    {
      "epoch": 0.0956,
      "grad_norm": 9.564947128295898,
      "learning_rate": 0.000191,
      "loss": 6.6515,
      "step": 956
    },
    {
      "epoch": 0.0957,
      "grad_norm": 4.911442279815674,
      "learning_rate": 0.0001912,
      "loss": 6.6692,
      "step": 957
    },
    {
      "epoch": 0.0958,
      "grad_norm": 5.140913009643555,
      "learning_rate": 0.0001914,
      "loss": 6.7417,
      "step": 958
    },
    {
      "epoch": 0.0959,
      "grad_norm": 30.7878475189209,
      "learning_rate": 0.0001916,
      "loss": 7.9211,
      "step": 959
    },
    {
      "epoch": 0.096,
      "grad_norm": 8.359962463378906,
      "learning_rate": 0.0001918,
      "loss": 6.8004,
      "step": 960
    },
    {
      "epoch": 0.0961,
      "grad_norm": 7.747732162475586,
      "learning_rate": 0.000192,
      "loss": 6.5995,
      "step": 961
    },
    {
      "epoch": 0.0962,
      "grad_norm": 13.89732551574707,
      "learning_rate": 0.0001922,
      "loss": 7.0838,
      "step": 962
    },
    {
      "epoch": 0.0963,
      "grad_norm": 8.05943775177002,
      "learning_rate": 0.00019240000000000001,
      "loss": 6.5927,
      "step": 963
    },
    {
      "epoch": 0.0964,
      "grad_norm": 5.030289649963379,
      "learning_rate": 0.0001926,
      "loss": 6.2992,
      "step": 964
    },
    {
      "epoch": 0.0965,
      "grad_norm": 6.593816757202148,
      "learning_rate": 0.0001928,
      "loss": 7.3807,
      "step": 965
    },
    {
      "epoch": 0.0966,
      "grad_norm": 4.377501964569092,
      "learning_rate": 0.000193,
      "loss": 6.8549,
      "step": 966
    },
    {
      "epoch": 0.0967,
      "grad_norm": 9.2803955078125,
      "learning_rate": 0.0001932,
      "loss": 6.6688,
      "step": 967
    },
    {
      "epoch": 0.0968,
      "grad_norm": 4.252203464508057,
      "learning_rate": 0.0001934,
      "loss": 6.7368,
      "step": 968
    },
    {
      "epoch": 0.0969,
      "grad_norm": 3.791898488998413,
      "learning_rate": 0.00019360000000000002,
      "loss": 6.4377,
      "step": 969
    },
    {
      "epoch": 0.097,
      "grad_norm": 5.707334518432617,
      "learning_rate": 0.0001938,
      "loss": 6.7234,
      "step": 970
    },
    {
      "epoch": 0.0971,
      "grad_norm": 3.758333683013916,
      "learning_rate": 0.000194,
      "loss": 7.0147,
      "step": 971
    },
    {
      "epoch": 0.0972,
      "grad_norm": 5.861918926239014,
      "learning_rate": 0.0001942,
      "loss": 7.2715,
      "step": 972
    },
    {
      "epoch": 0.0973,
      "grad_norm": 5.559395790100098,
      "learning_rate": 0.0001944,
      "loss": 7.4734,
      "step": 973
    },
    {
      "epoch": 0.0974,
      "grad_norm": 6.025390625,
      "learning_rate": 0.00019460000000000001,
      "loss": 6.7906,
      "step": 974
    },
    {
      "epoch": 0.0975,
      "grad_norm": 8.598268508911133,
      "learning_rate": 0.0001948,
      "loss": 6.7395,
      "step": 975
    },
    {
      "epoch": 0.0976,
      "grad_norm": 11.242158889770508,
      "learning_rate": 0.000195,
      "loss": 7.2326,
      "step": 976
    },
    {
      "epoch": 0.0977,
      "grad_norm": 51.80280303955078,
      "learning_rate": 0.0001952,
      "loss": 7.4723,
      "step": 977
    },
    {
      "epoch": 0.0978,
      "grad_norm": 9.123377799987793,
      "learning_rate": 0.0001954,
      "loss": 6.2191,
      "step": 978
    },
    {
      "epoch": 0.0979,
      "grad_norm": 14.366161346435547,
      "learning_rate": 0.0001956,
      "loss": 6.6751,
      "step": 979
    },
    {
      "epoch": 0.098,
      "grad_norm": 12.64511489868164,
      "learning_rate": 0.00019580000000000002,
      "loss": 7.0948,
      "step": 980
    },
    {
      "epoch": 0.0981,
      "grad_norm": 2.768404483795166,
      "learning_rate": 0.000196,
      "loss": 6.4178,
      "step": 981
    },
    {
      "epoch": 0.0982,
      "grad_norm": 2.9697253704071045,
      "learning_rate": 0.0001962,
      "loss": 6.2331,
      "step": 982
    },
    {
      "epoch": 0.0983,
      "grad_norm": 8.066802978515625,
      "learning_rate": 0.0001964,
      "loss": 7.8872,
      "step": 983
    },
    {
      "epoch": 0.0984,
      "grad_norm": 4.5120086669921875,
      "learning_rate": 0.0001966,
      "loss": 7.729,
      "step": 984
    },
    {
      "epoch": 0.0985,
      "grad_norm": 4.40260124206543,
      "learning_rate": 0.0001968,
      "loss": 6.5297,
      "step": 985
    },
    {
      "epoch": 0.0986,
      "grad_norm": 3.735461711883545,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.0165,
      "step": 986
    },
    {
      "epoch": 0.0987,
      "grad_norm": 3.2970645427703857,
      "learning_rate": 0.0001972,
      "loss": 6.1731,
      "step": 987
    },
    {
      "epoch": 0.0988,
      "grad_norm": 4.396055698394775,
      "learning_rate": 0.0001974,
      "loss": 6.6618,
      "step": 988
    },
    {
      "epoch": 0.0989,
      "grad_norm": 12.071581840515137,
      "learning_rate": 0.0001976,
      "loss": 7.426,
      "step": 989
    },
    {
      "epoch": 0.099,
      "grad_norm": 5.920867919921875,
      "learning_rate": 0.0001978,
      "loss": 7.0802,
      "step": 990
    },
    {
      "epoch": 0.0991,
      "grad_norm": 17.44622802734375,
      "learning_rate": 0.00019800000000000002,
      "loss": 7.1817,
      "step": 991
    },
    {
      "epoch": 0.0992,
      "grad_norm": 7.952199935913086,
      "learning_rate": 0.00019820000000000002,
      "loss": 6.3687,
      "step": 992
    },
    {
      "epoch": 0.0993,
      "grad_norm": 3.872014284133911,
      "learning_rate": 0.0001984,
      "loss": 6.4622,
      "step": 993
    },
    {
      "epoch": 0.0994,
      "grad_norm": 5.992473125457764,
      "learning_rate": 0.0001986,
      "loss": 6.5439,
      "step": 994
    },
    {
      "epoch": 0.0995,
      "grad_norm": 18.409996032714844,
      "learning_rate": 0.0001988,
      "loss": 6.3908,
      "step": 995
    },
    {
      "epoch": 0.0996,
      "grad_norm": 5.229300498962402,
      "learning_rate": 0.000199,
      "loss": 6.7831,
      "step": 996
    },
    {
      "epoch": 0.0997,
      "grad_norm": 5.22222900390625,
      "learning_rate": 0.00019920000000000002,
      "loss": 6.8288,
      "step": 997
    },
    {
      "epoch": 0.0998,
      "grad_norm": 3.559264659881592,
      "learning_rate": 0.00019940000000000002,
      "loss": 6.6633,
      "step": 998
    },
    {
      "epoch": 0.0999,
      "grad_norm": 7.369091510772705,
      "learning_rate": 0.0001996,
      "loss": 6.9141,
      "step": 999
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.3985090255737305,
      "learning_rate": 0.0001998,
      "loss": 6.6734,
      "step": 1000
    },
    {
      "epoch": 0.1001,
      "grad_norm": 3.169312000274658,
      "learning_rate": 0.0002,
      "loss": 6.7748,
      "step": 1001
    },
    {
      "epoch": 0.1002,
      "grad_norm": 3.785557985305786,
      "learning_rate": 0.00019999999390765168,
      "loss": 6.6729,
      "step": 1002
    },
    {
      "epoch": 0.1003,
      "grad_norm": 3.1260640621185303,
      "learning_rate": 0.0001999999756306074,
      "loss": 6.2233,
      "step": 1003
    },
    {
      "epoch": 0.1004,
      "grad_norm": 4.252016544342041,
      "learning_rate": 0.00019999994516886946,
      "loss": 6.2151,
      "step": 1004
    },
    {
      "epoch": 0.1005,
      "grad_norm": 7.778120994567871,
      "learning_rate": 0.00019999990252244151,
      "loss": 6.6237,
      "step": 1005
    },
    {
      "epoch": 0.1006,
      "grad_norm": 6.606187343597412,
      "learning_rate": 0.00019999984769132877,
      "loss": 6.5476,
      "step": 1006
    },
    {
      "epoch": 0.1007,
      "grad_norm": 2.909498453140259,
      "learning_rate": 0.00019999978067553796,
      "loss": 6.8122,
      "step": 1007
    },
    {
      "epoch": 0.1008,
      "grad_norm": 4.9978837966918945,
      "learning_rate": 0.00019999970147507713,
      "loss": 6.5877,
      "step": 1008
    },
    {
      "epoch": 0.1009,
      "grad_norm": 26.579387664794922,
      "learning_rate": 0.00019999961008995606,
      "loss": 6.9334,
      "step": 1009
    },
    {
      "epoch": 0.101,
      "grad_norm": 4.350188732147217,
      "learning_rate": 0.00019999950652018584,
      "loss": 6.629,
      "step": 1010
    },
    {
      "epoch": 0.1011,
      "grad_norm": 3.717897653579712,
      "learning_rate": 0.00019999939076577905,
      "loss": 6.8265,
      "step": 1011
    },
    {
      "epoch": 0.1012,
      "grad_norm": 24.673341751098633,
      "learning_rate": 0.00019999926282674983,
      "loss": 6.583,
      "step": 1012
    },
    {
      "epoch": 0.1013,
      "grad_norm": 2.669795274734497,
      "learning_rate": 0.00019999912270311375,
      "loss": 6.3128,
      "step": 1013
    },
    {
      "epoch": 0.1014,
      "grad_norm": 8.960829734802246,
      "learning_rate": 0.00019999897039488796,
      "loss": 7.3029,
      "step": 1014
    },
    {
      "epoch": 0.1015,
      "grad_norm": 3.3488595485687256,
      "learning_rate": 0.0001999988059020909,
      "loss": 6.6912,
      "step": 1015
    },
    {
      "epoch": 0.1016,
      "grad_norm": 7.115213871002197,
      "learning_rate": 0.00019999862922474268,
      "loss": 7.3214,
      "step": 1016
    },
    {
      "epoch": 0.1017,
      "grad_norm": 3.025301933288574,
      "learning_rate": 0.00019999844036286483,
      "loss": 6.4481,
      "step": 1017
    },
    {
      "epoch": 0.1018,
      "grad_norm": 3.3842365741729736,
      "learning_rate": 0.00019999823931648035,
      "loss": 6.8868,
      "step": 1018
    },
    {
      "epoch": 0.1019,
      "grad_norm": 3.692653179168701,
      "learning_rate": 0.0001999980260856137,
      "loss": 6.4051,
      "step": 1019
    },
    {
      "epoch": 0.102,
      "grad_norm": 5.28713846206665,
      "learning_rate": 0.00019999780067029094,
      "loss": 7.2462,
      "step": 1020
    },
    {
      "epoch": 0.1021,
      "grad_norm": 5.584995746612549,
      "learning_rate": 0.00019999756307053948,
      "loss": 6.4813,
      "step": 1021
    },
    {
      "epoch": 0.1022,
      "grad_norm": 3.4761526584625244,
      "learning_rate": 0.00019999731328638828,
      "loss": 6.8233,
      "step": 1022
    },
    {
      "epoch": 0.1023,
      "grad_norm": 6.6064653396606445,
      "learning_rate": 0.0001999970513178678,
      "loss": 7.184,
      "step": 1023
    },
    {
      "epoch": 0.1024,
      "grad_norm": 15.039304733276367,
      "learning_rate": 0.00019999677716500992,
      "loss": 7.1382,
      "step": 1024
    },
    {
      "epoch": 0.1025,
      "grad_norm": 4.661031246185303,
      "learning_rate": 0.0001999964908278481,
      "loss": 6.588,
      "step": 1025
    },
    {
      "epoch": 0.1026,
      "grad_norm": 2.912132978439331,
      "learning_rate": 0.00019999619230641713,
      "loss": 6.9387,
      "step": 1026
    },
    {
      "epoch": 0.1027,
      "grad_norm": 26.85057830810547,
      "learning_rate": 0.00019999588160075348,
      "loss": 6.6621,
      "step": 1027
    },
    {
      "epoch": 0.1028,
      "grad_norm": 14.351490020751953,
      "learning_rate": 0.000199995558710895,
      "loss": 6.4792,
      "step": 1028
    },
    {
      "epoch": 0.1029,
      "grad_norm": 3.5657410621643066,
      "learning_rate": 0.000199995223636881,
      "loss": 7.0638,
      "step": 1029
    },
    {
      "epoch": 0.103,
      "grad_norm": 22.641542434692383,
      "learning_rate": 0.0001999948763787523,
      "loss": 6.9559,
      "step": 1030
    },
    {
      "epoch": 0.1031,
      "grad_norm": 3.2278096675872803,
      "learning_rate": 0.00019999451693655123,
      "loss": 6.5849,
      "step": 1031
    },
    {
      "epoch": 0.1032,
      "grad_norm": 5.481954097747803,
      "learning_rate": 0.00019999414531032158,
      "loss": 6.3999,
      "step": 1032
    },
    {
      "epoch": 0.1033,
      "grad_norm": 7.523856163024902,
      "learning_rate": 0.00019999376150010866,
      "loss": 7.5632,
      "step": 1033
    },
    {
      "epoch": 0.1034,
      "grad_norm": 6.282206058502197,
      "learning_rate": 0.0001999933655059592,
      "loss": 7.1803,
      "step": 1034
    },
    {
      "epoch": 0.1035,
      "grad_norm": 3.2850961685180664,
      "learning_rate": 0.00019999295732792146,
      "loss": 7.0991,
      "step": 1035
    },
    {
      "epoch": 0.1036,
      "grad_norm": 5.966711044311523,
      "learning_rate": 0.00019999253696604522,
      "loss": 6.6306,
      "step": 1036
    },
    {
      "epoch": 0.1037,
      "grad_norm": 7.291424751281738,
      "learning_rate": 0.00019999210442038162,
      "loss": 6.1395,
      "step": 1037
    },
    {
      "epoch": 0.1038,
      "grad_norm": 5.062778472900391,
      "learning_rate": 0.00019999165969098343,
      "loss": 6.6293,
      "step": 1038
    },
    {
      "epoch": 0.1039,
      "grad_norm": 11.080771446228027,
      "learning_rate": 0.00019999120277790477,
      "loss": 6.8236,
      "step": 1039
    },
    {
      "epoch": 0.104,
      "grad_norm": 5.832948684692383,
      "learning_rate": 0.0001999907336812014,
      "loss": 6.4577,
      "step": 1040
    },
    {
      "epoch": 0.1041,
      "grad_norm": 4.672344207763672,
      "learning_rate": 0.00019999025240093044,
      "loss": 6.1673,
      "step": 1041
    },
    {
      "epoch": 0.1042,
      "grad_norm": 3.8682305812835693,
      "learning_rate": 0.0001999897589371505,
      "loss": 7.0268,
      "step": 1042
    },
    {
      "epoch": 0.1043,
      "grad_norm": 9.401063919067383,
      "learning_rate": 0.00019998925328992175,
      "loss": 6.2307,
      "step": 1043
    },
    {
      "epoch": 0.1044,
      "grad_norm": 9.493770599365234,
      "learning_rate": 0.0001999887354593058,
      "loss": 6.0288,
      "step": 1044
    },
    {
      "epoch": 0.1045,
      "grad_norm": 18.447874069213867,
      "learning_rate": 0.0001999882054453657,
      "loss": 6.3418,
      "step": 1045
    },
    {
      "epoch": 0.1046,
      "grad_norm": 18.309114456176758,
      "learning_rate": 0.00019998766324816607,
      "loss": 6.3903,
      "step": 1046
    },
    {
      "epoch": 0.1047,
      "grad_norm": 5.922817230224609,
      "learning_rate": 0.000199987108867773,
      "loss": 6.0086,
      "step": 1047
    },
    {
      "epoch": 0.1048,
      "grad_norm": 3.1558761596679688,
      "learning_rate": 0.00019998654230425395,
      "loss": 6.0804,
      "step": 1048
    },
    {
      "epoch": 0.1049,
      "grad_norm": 4.870303153991699,
      "learning_rate": 0.00019998596355767805,
      "loss": 7.2387,
      "step": 1049
    },
    {
      "epoch": 0.105,
      "grad_norm": 5.648087978363037,
      "learning_rate": 0.00019998537262811577,
      "loss": 6.7228,
      "step": 1050
    },
    {
      "epoch": 0.1051,
      "grad_norm": 3.177678108215332,
      "learning_rate": 0.00019998476951563915,
      "loss": 6.6176,
      "step": 1051
    },
    {
      "epoch": 0.1052,
      "grad_norm": 3.40899658203125,
      "learning_rate": 0.00019998415422032163,
      "loss": 6.4249,
      "step": 1052
    },
    {
      "epoch": 0.1053,
      "grad_norm": 9.390474319458008,
      "learning_rate": 0.00019998352674223816,
      "loss": 7.1079,
      "step": 1053
    },
    {
      "epoch": 0.1054,
      "grad_norm": 4.2464494705200195,
      "learning_rate": 0.00019998288708146527,
      "loss": 6.4778,
      "step": 1054
    },
    {
      "epoch": 0.1055,
      "grad_norm": 3.4427030086517334,
      "learning_rate": 0.0001999822352380809,
      "loss": 7.0306,
      "step": 1055
    },
    {
      "epoch": 0.1056,
      "grad_norm": 5.064059734344482,
      "learning_rate": 0.00019998157121216444,
      "loss": 6.4489,
      "step": 1056
    },
    {
      "epoch": 0.1057,
      "grad_norm": 11.473577499389648,
      "learning_rate": 0.00019998089500379676,
      "loss": 6.833,
      "step": 1057
    },
    {
      "epoch": 0.1058,
      "grad_norm": 5.269840240478516,
      "learning_rate": 0.00019998020661306037,
      "loss": 7.0804,
      "step": 1058
    },
    {
      "epoch": 0.1059,
      "grad_norm": 4.0801615715026855,
      "learning_rate": 0.000199979506040039,
      "loss": 6.6843,
      "step": 1059
    },
    {
      "epoch": 0.106,
      "grad_norm": 3.838423490524292,
      "learning_rate": 0.00019997879328481814,
      "loss": 6.4041,
      "step": 1060
    },
    {
      "epoch": 0.1061,
      "grad_norm": 5.302448749542236,
      "learning_rate": 0.00019997806834748456,
      "loss": 6.3194,
      "step": 1061
    },
    {
      "epoch": 0.1062,
      "grad_norm": 5.858061790466309,
      "learning_rate": 0.0001999773312281266,
      "loss": 7.6092,
      "step": 1062
    },
    {
      "epoch": 0.1063,
      "grad_norm": 12.390033721923828,
      "learning_rate": 0.00019997658192683412,
      "loss": 6.1418,
      "step": 1063
    },
    {
      "epoch": 0.1064,
      "grad_norm": 3.584667921066284,
      "learning_rate": 0.00019997582044369843,
      "loss": 6.4433,
      "step": 1064
    },
    {
      "epoch": 0.1065,
      "grad_norm": 27.005020141601562,
      "learning_rate": 0.00019997504677881224,
      "loss": 7.2888,
      "step": 1065
    },
    {
      "epoch": 0.1066,
      "grad_norm": 3.234863758087158,
      "learning_rate": 0.00019997426093226986,
      "loss": 7.1486,
      "step": 1066
    },
    {
      "epoch": 0.1067,
      "grad_norm": 4.169369697570801,
      "learning_rate": 0.000199973462904167,
      "loss": 6.2409,
      "step": 1067
    },
    {
      "epoch": 0.1068,
      "grad_norm": 9.032537460327148,
      "learning_rate": 0.000199972652694601,
      "loss": 7.7925,
      "step": 1068
    },
    {
      "epoch": 0.1069,
      "grad_norm": 4.498800754547119,
      "learning_rate": 0.00019997183030367048,
      "loss": 6.4675,
      "step": 1069
    },
    {
      "epoch": 0.107,
      "grad_norm": 21.4664363861084,
      "learning_rate": 0.0001999709957314757,
      "loss": 6.8851,
      "step": 1070
    },
    {
      "epoch": 0.1071,
      "grad_norm": 10.487557411193848,
      "learning_rate": 0.00019997014897811833,
      "loss": 5.7454,
      "step": 1071
    },
    {
      "epoch": 0.1072,
      "grad_norm": 8.970891952514648,
      "learning_rate": 0.00019996929004370152,
      "loss": 6.4483,
      "step": 1072
    },
    {
      "epoch": 0.1073,
      "grad_norm": 7.498667240142822,
      "learning_rate": 0.00019996841892833,
      "loss": 5.9314,
      "step": 1073
    },
    {
      "epoch": 0.1074,
      "grad_norm": 5.338964462280273,
      "learning_rate": 0.00019996753563210985,
      "loss": 6.3224,
      "step": 1074
    },
    {
      "epoch": 0.1075,
      "grad_norm": 6.098053455352783,
      "learning_rate": 0.00019996664015514871,
      "loss": 6.7343,
      "step": 1075
    },
    {
      "epoch": 0.1076,
      "grad_norm": 3.5250587463378906,
      "learning_rate": 0.00019996573249755572,
      "loss": 6.5742,
      "step": 1076
    },
    {
      "epoch": 0.1077,
      "grad_norm": 6.356448650360107,
      "learning_rate": 0.00019996481265944146,
      "loss": 6.6041,
      "step": 1077
    },
    {
      "epoch": 0.1078,
      "grad_norm": 3.726581335067749,
      "learning_rate": 0.000199963880640918,
      "loss": 6.1126,
      "step": 1078
    },
    {
      "epoch": 0.1079,
      "grad_norm": 32.01093292236328,
      "learning_rate": 0.00019996293644209887,
      "loss": 7.6463,
      "step": 1079
    },
    {
      "epoch": 0.108,
      "grad_norm": 5.77610445022583,
      "learning_rate": 0.0001999619800630992,
      "loss": 6.5466,
      "step": 1080
    },
    {
      "epoch": 0.1081,
      "grad_norm": 4.847800254821777,
      "learning_rate": 0.00019996101150403543,
      "loss": 7.2129,
      "step": 1081
    },
    {
      "epoch": 0.1082,
      "grad_norm": 4.566919326782227,
      "learning_rate": 0.00019996003076502565,
      "loss": 6.9735,
      "step": 1082
    },
    {
      "epoch": 0.1083,
      "grad_norm": 8.850497245788574,
      "learning_rate": 0.00019995903784618936,
      "loss": 6.7078,
      "step": 1083
    },
    {
      "epoch": 0.1084,
      "grad_norm": 5.056117534637451,
      "learning_rate": 0.00019995803274764747,
      "loss": 6.286,
      "step": 1084
    },
    {
      "epoch": 0.1085,
      "grad_norm": 3.434666872024536,
      "learning_rate": 0.0001999570154695225,
      "loss": 6.5413,
      "step": 1085
    },
    {
      "epoch": 0.1086,
      "grad_norm": 3.4922497272491455,
      "learning_rate": 0.00019995598601193842,
      "loss": 6.2886,
      "step": 1086
    },
    {
      "epoch": 0.1087,
      "grad_norm": 4.667288780212402,
      "learning_rate": 0.00019995494437502064,
      "loss": 6.4164,
      "step": 1087
    },
    {
      "epoch": 0.1088,
      "grad_norm": 16.6750545501709,
      "learning_rate": 0.00019995389055889605,
      "loss": 6.7872,
      "step": 1088
    },
    {
      "epoch": 0.1089,
      "grad_norm": 11.693888664245605,
      "learning_rate": 0.0001999528245636931,
      "loss": 7.9,
      "step": 1089
    },
    {
      "epoch": 0.109,
      "grad_norm": 5.272411346435547,
      "learning_rate": 0.0001999517463895417,
      "loss": 6.781,
      "step": 1090
    },
    {
      "epoch": 0.1091,
      "grad_norm": 7.613633155822754,
      "learning_rate": 0.00019995065603657316,
      "loss": 6.5098,
      "step": 1091
    },
    {
      "epoch": 0.1092,
      "grad_norm": 7.862076759338379,
      "learning_rate": 0.00019994955350492038,
      "loss": 6.8075,
      "step": 1092
    },
    {
      "epoch": 0.1093,
      "grad_norm": 4.835395336151123,
      "learning_rate": 0.0001999484387947177,
      "loss": 6.2161,
      "step": 1093
    },
    {
      "epoch": 0.1094,
      "grad_norm": 15.160221099853516,
      "learning_rate": 0.00019994731190610087,
      "loss": 7.6424,
      "step": 1094
    },
    {
      "epoch": 0.1095,
      "grad_norm": 2.797943353652954,
      "learning_rate": 0.0001999461728392073,
      "loss": 6.9697,
      "step": 1095
    },
    {
      "epoch": 0.1096,
      "grad_norm": 2.698169708251953,
      "learning_rate": 0.00019994502159417573,
      "loss": 6.5683,
      "step": 1096
    },
    {
      "epoch": 0.1097,
      "grad_norm": 2.71466326713562,
      "learning_rate": 0.00019994385817114646,
      "loss": 6.6556,
      "step": 1097
    },
    {
      "epoch": 0.1098,
      "grad_norm": 3.394883871078491,
      "learning_rate": 0.00019994268257026118,
      "loss": 6.9972,
      "step": 1098
    },
    {
      "epoch": 0.1099,
      "grad_norm": 4.022953033447266,
      "learning_rate": 0.00019994149479166324,
      "loss": 6.7853,
      "step": 1099
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.2173123359680176,
      "learning_rate": 0.0001999402948354973,
      "loss": 6.5411,
      "step": 1100
    },
    {
      "epoch": 0.1101,
      "grad_norm": 4.639122486114502,
      "learning_rate": 0.0001999390827019096,
      "loss": 6.2914,
      "step": 1101
    },
    {
      "epoch": 0.1102,
      "grad_norm": 4.974206447601318,
      "learning_rate": 0.0001999378583910478,
      "loss": 7.3334,
      "step": 1102
    },
    {
      "epoch": 0.1103,
      "grad_norm": 5.360745906829834,
      "learning_rate": 0.0001999366219030611,
      "loss": 7.3682,
      "step": 1103
    },
    {
      "epoch": 0.1104,
      "grad_norm": 5.81308126449585,
      "learning_rate": 0.00019993537323810014,
      "loss": 7.4544,
      "step": 1104
    },
    {
      "epoch": 0.1105,
      "grad_norm": 33.99951171875,
      "learning_rate": 0.0001999341123963171,
      "loss": 9.161,
      "step": 1105
    },
    {
      "epoch": 0.1106,
      "grad_norm": 4.052433967590332,
      "learning_rate": 0.00019993283937786563,
      "loss": 6.1665,
      "step": 1106
    },
    {
      "epoch": 0.1107,
      "grad_norm": 45.020904541015625,
      "learning_rate": 0.0001999315541829008,
      "loss": 7.8416,
      "step": 1107
    },
    {
      "epoch": 0.1108,
      "grad_norm": 20.70722007751465,
      "learning_rate": 0.0001999302568115792,
      "loss": 6.9499,
      "step": 1108
    },
    {
      "epoch": 0.1109,
      "grad_norm": 12.2205810546875,
      "learning_rate": 0.00019992894726405893,
      "loss": 6.4111,
      "step": 1109
    },
    {
      "epoch": 0.111,
      "grad_norm": 6.233757972717285,
      "learning_rate": 0.00019992762554049955,
      "loss": 6.4043,
      "step": 1110
    },
    {
      "epoch": 0.1111,
      "grad_norm": 4.076844215393066,
      "learning_rate": 0.0001999262916410621,
      "loss": 6.734,
      "step": 1111
    },
    {
      "epoch": 0.1112,
      "grad_norm": 5.168939590454102,
      "learning_rate": 0.00019992494556590916,
      "loss": 6.1598,
      "step": 1112
    },
    {
      "epoch": 0.1113,
      "grad_norm": 17.323007583618164,
      "learning_rate": 0.00019992358731520468,
      "loss": 6.7368,
      "step": 1113
    },
    {
      "epoch": 0.1114,
      "grad_norm": 2.8590612411499023,
      "learning_rate": 0.0001999222168891142,
      "loss": 6.6757,
      "step": 1114
    },
    {
      "epoch": 0.1115,
      "grad_norm": 2.928380012512207,
      "learning_rate": 0.0001999208342878047,
      "loss": 6.4173,
      "step": 1115
    },
    {
      "epoch": 0.1116,
      "grad_norm": 22.616880416870117,
      "learning_rate": 0.0001999194395114446,
      "loss": 7.1044,
      "step": 1116
    },
    {
      "epoch": 0.1117,
      "grad_norm": 7.736636638641357,
      "learning_rate": 0.00019991803256020393,
      "loss": 7.3104,
      "step": 1117
    },
    {
      "epoch": 0.1118,
      "grad_norm": 2.883592367172241,
      "learning_rate": 0.000199916613434254,
      "loss": 6.3382,
      "step": 1118
    },
    {
      "epoch": 0.1119,
      "grad_norm": 5.109788417816162,
      "learning_rate": 0.00019991518213376787,
      "loss": 6.7044,
      "step": 1119
    },
    {
      "epoch": 0.112,
      "grad_norm": 6.539107799530029,
      "learning_rate": 0.00019991373865891986,
      "loss": 6.5314,
      "step": 1120
    },
    {
      "epoch": 0.1121,
      "grad_norm": 4.676784515380859,
      "learning_rate": 0.00019991228300988585,
      "loss": 6.6096,
      "step": 1121
    },
    {
      "epoch": 0.1122,
      "grad_norm": 3.319533586502075,
      "learning_rate": 0.00019991081518684321,
      "loss": 5.9054,
      "step": 1122
    },
    {
      "epoch": 0.1123,
      "grad_norm": 17.18437957763672,
      "learning_rate": 0.00019990933518997084,
      "loss": 7.215,
      "step": 1123
    },
    {
      "epoch": 0.1124,
      "grad_norm": 38.266624450683594,
      "learning_rate": 0.00019990784301944902,
      "loss": 7.1628,
      "step": 1124
    },
    {
      "epoch": 0.1125,
      "grad_norm": 9.529886245727539,
      "learning_rate": 0.00019990633867545955,
      "loss": 6.2558,
      "step": 1125
    },
    {
      "epoch": 0.1126,
      "grad_norm": 6.535031795501709,
      "learning_rate": 0.0001999048221581858,
      "loss": 6.5227,
      "step": 1126
    },
    {
      "epoch": 0.1127,
      "grad_norm": 19.795207977294922,
      "learning_rate": 0.0001999032934678125,
      "loss": 6.6508,
      "step": 1127
    },
    {
      "epoch": 0.1128,
      "grad_norm": 4.617257118225098,
      "learning_rate": 0.0001999017526045259,
      "loss": 6.6884,
      "step": 1128
    },
    {
      "epoch": 0.1129,
      "grad_norm": 3.962411880493164,
      "learning_rate": 0.00019990019956851382,
      "loss": 6.3713,
      "step": 1129
    },
    {
      "epoch": 0.113,
      "grad_norm": 8.001884460449219,
      "learning_rate": 0.00019989863435996544,
      "loss": 6.4933,
      "step": 1130
    },
    {
      "epoch": 0.1131,
      "grad_norm": 9.043047904968262,
      "learning_rate": 0.00019989705697907149,
      "loss": 6.8698,
      "step": 1131
    },
    {
      "epoch": 0.1132,
      "grad_norm": 6.793410778045654,
      "learning_rate": 0.00019989546742602414,
      "loss": 7.7401,
      "step": 1132
    },
    {
      "epoch": 0.1133,
      "grad_norm": 7.655816555023193,
      "learning_rate": 0.00019989386570101714,
      "loss": 7.2839,
      "step": 1133
    },
    {
      "epoch": 0.1134,
      "grad_norm": 3.847797393798828,
      "learning_rate": 0.0001998922518042456,
      "loss": 6.7597,
      "step": 1134
    },
    {
      "epoch": 0.1135,
      "grad_norm": 6.726383686065674,
      "learning_rate": 0.00019989062573590616,
      "loss": 6.8539,
      "step": 1135
    },
    {
      "epoch": 0.1136,
      "grad_norm": 2.928148031234741,
      "learning_rate": 0.00019988898749619702,
      "loss": 6.6756,
      "step": 1136
    },
    {
      "epoch": 0.1137,
      "grad_norm": 3.3017468452453613,
      "learning_rate": 0.0001998873370853177,
      "loss": 6.2514,
      "step": 1137
    },
    {
      "epoch": 0.1138,
      "grad_norm": 4.54594087600708,
      "learning_rate": 0.00019988567450346937,
      "loss": 6.4343,
      "step": 1138
    },
    {
      "epoch": 0.1139,
      "grad_norm": 8.352383613586426,
      "learning_rate": 0.0001998839997508546,
      "loss": 6.7195,
      "step": 1139
    },
    {
      "epoch": 0.114,
      "grad_norm": 4.371987819671631,
      "learning_rate": 0.00019988231282767744,
      "loss": 6.309,
      "step": 1140
    },
    {
      "epoch": 0.1141,
      "grad_norm": 4.878377437591553,
      "learning_rate": 0.0001998806137341434,
      "loss": 6.2552,
      "step": 1141
    },
    {
      "epoch": 0.1142,
      "grad_norm": 3.417438268661499,
      "learning_rate": 0.00019987890247045957,
      "loss": 6.717,
      "step": 1142
    },
    {
      "epoch": 0.1143,
      "grad_norm": 3.891733407974243,
      "learning_rate": 0.00019987717903683448,
      "loss": 6.6382,
      "step": 1143
    },
    {
      "epoch": 0.1144,
      "grad_norm": 5.455896377563477,
      "learning_rate": 0.00019987544343347803,
      "loss": 6.1887,
      "step": 1144
    },
    {
      "epoch": 0.1145,
      "grad_norm": 2.8694088459014893,
      "learning_rate": 0.00019987369566060176,
      "loss": 6.5412,
      "step": 1145
    },
    {
      "epoch": 0.1146,
      "grad_norm": 6.299776554107666,
      "learning_rate": 0.0001998719357184186,
      "loss": 6.128,
      "step": 1146
    },
    {
      "epoch": 0.1147,
      "grad_norm": 4.127383232116699,
      "learning_rate": 0.00019987016360714307,
      "loss": 6.1161,
      "step": 1147
    },
    {
      "epoch": 0.1148,
      "grad_norm": 3.947369337081909,
      "learning_rate": 0.00019986837932699103,
      "loss": 6.2351,
      "step": 1148
    },
    {
      "epoch": 0.1149,
      "grad_norm": 2.6175525188446045,
      "learning_rate": 0.00019986658287817987,
      "loss": 6.9125,
      "step": 1149
    },
    {
      "epoch": 0.115,
      "grad_norm": 10.173328399658203,
      "learning_rate": 0.00019986477426092855,
      "loss": 6.3433,
      "step": 1150
    },
    {
      "epoch": 0.1151,
      "grad_norm": 3.780080795288086,
      "learning_rate": 0.0001998629534754574,
      "loss": 6.527,
      "step": 1151
    },
    {
      "epoch": 0.1152,
      "grad_norm": 3.3589093685150146,
      "learning_rate": 0.0001998611205219883,
      "loss": 6.6455,
      "step": 1152
    },
    {
      "epoch": 0.1153,
      "grad_norm": 18.18379783630371,
      "learning_rate": 0.00019985927540074454,
      "loss": 6.8286,
      "step": 1153
    },
    {
      "epoch": 0.1154,
      "grad_norm": 16.248607635498047,
      "learning_rate": 0.00019985741811195097,
      "loss": 7.5453,
      "step": 1154
    },
    {
      "epoch": 0.1155,
      "grad_norm": 3.8799681663513184,
      "learning_rate": 0.00019985554865583393,
      "loss": 6.4323,
      "step": 1155
    },
    {
      "epoch": 0.1156,
      "grad_norm": 2.689540386199951,
      "learning_rate": 0.0001998536670326212,
      "loss": 6.3941,
      "step": 1156
    },
    {
      "epoch": 0.1157,
      "grad_norm": 5.277322292327881,
      "learning_rate": 0.000199851773242542,
      "loss": 6.3075,
      "step": 1157
    },
    {
      "epoch": 0.1158,
      "grad_norm": 5.751522541046143,
      "learning_rate": 0.0001998498672858271,
      "loss": 7.0612,
      "step": 1158
    },
    {
      "epoch": 0.1159,
      "grad_norm": 3.3003900051116943,
      "learning_rate": 0.00019984794916270876,
      "loss": 6.3685,
      "step": 1159
    },
    {
      "epoch": 0.116,
      "grad_norm": 5.697538375854492,
      "learning_rate": 0.00019984601887342073,
      "loss": 6.9968,
      "step": 1160
    },
    {
      "epoch": 0.1161,
      "grad_norm": 4.963936805725098,
      "learning_rate": 0.00019984407641819812,
      "loss": 7.2615,
      "step": 1161
    },
    {
      "epoch": 0.1162,
      "grad_norm": 4.509936332702637,
      "learning_rate": 0.00019984212179727766,
      "loss": 6.359,
      "step": 1162
    },
    {
      "epoch": 0.1163,
      "grad_norm": 5.287754535675049,
      "learning_rate": 0.00019984015501089752,
      "loss": 6.63,
      "step": 1163
    },
    {
      "epoch": 0.1164,
      "grad_norm": 3.288304090499878,
      "learning_rate": 0.00019983817605929733,
      "loss": 6.4356,
      "step": 1164
    },
    {
      "epoch": 0.1165,
      "grad_norm": 8.058497428894043,
      "learning_rate": 0.00019983618494271826,
      "loss": 7.3249,
      "step": 1165
    },
    {
      "epoch": 0.1166,
      "grad_norm": 5.701746463775635,
      "learning_rate": 0.00019983418166140285,
      "loss": 6.8946,
      "step": 1166
    },
    {
      "epoch": 0.1167,
      "grad_norm": 14.103972434997559,
      "learning_rate": 0.00019983216621559524,
      "loss": 7.1948,
      "step": 1167
    },
    {
      "epoch": 0.1168,
      "grad_norm": 7.511901378631592,
      "learning_rate": 0.00019983013860554101,
      "loss": 6.2834,
      "step": 1168
    },
    {
      "epoch": 0.1169,
      "grad_norm": 2.808946132659912,
      "learning_rate": 0.00019982809883148722,
      "loss": 6.7888,
      "step": 1169
    },
    {
      "epoch": 0.117,
      "grad_norm": 21.78084373474121,
      "learning_rate": 0.00019982604689368239,
      "loss": 7.2793,
      "step": 1170
    },
    {
      "epoch": 0.1171,
      "grad_norm": 4.261631488800049,
      "learning_rate": 0.00019982398279237655,
      "loss": 6.2727,
      "step": 1171
    },
    {
      "epoch": 0.1172,
      "grad_norm": 4.971683979034424,
      "learning_rate": 0.0001998219065278212,
      "loss": 6.7709,
      "step": 1172
    },
    {
      "epoch": 0.1173,
      "grad_norm": 5.302138328552246,
      "learning_rate": 0.00019981981810026934,
      "loss": 7.0352,
      "step": 1173
    },
    {
      "epoch": 0.1174,
      "grad_norm": 4.785744667053223,
      "learning_rate": 0.0001998177175099754,
      "loss": 6.7422,
      "step": 1174
    },
    {
      "epoch": 0.1175,
      "grad_norm": 3.920616865158081,
      "learning_rate": 0.00019981560475719538,
      "loss": 5.9316,
      "step": 1175
    },
    {
      "epoch": 0.1176,
      "grad_norm": 6.0425262451171875,
      "learning_rate": 0.0001998134798421867,
      "loss": 5.9095,
      "step": 1176
    },
    {
      "epoch": 0.1177,
      "grad_norm": 5.35383415222168,
      "learning_rate": 0.00019981134276520827,
      "loss": 6.4431,
      "step": 1177
    },
    {
      "epoch": 0.1178,
      "grad_norm": 2.643869161605835,
      "learning_rate": 0.00019980919352652048,
      "loss": 6.0826,
      "step": 1178
    },
    {
      "epoch": 0.1179,
      "grad_norm": 3.5103583335876465,
      "learning_rate": 0.00019980703212638522,
      "loss": 6.867,
      "step": 1179
    },
    {
      "epoch": 0.118,
      "grad_norm": 3.9272263050079346,
      "learning_rate": 0.00019980485856506582,
      "loss": 6.4106,
      "step": 1180
    },
    {
      "epoch": 0.1181,
      "grad_norm": 9.571043014526367,
      "learning_rate": 0.00019980267284282717,
      "loss": 6.6414,
      "step": 1181
    },
    {
      "epoch": 0.1182,
      "grad_norm": 2.9557905197143555,
      "learning_rate": 0.00019980047495993554,
      "loss": 6.1178,
      "step": 1182
    },
    {
      "epoch": 0.1183,
      "grad_norm": 4.971242427825928,
      "learning_rate": 0.00019979826491665881,
      "loss": 6.6556,
      "step": 1183
    },
    {
      "epoch": 0.1184,
      "grad_norm": 4.229152202606201,
      "learning_rate": 0.00019979604271326616,
      "loss": 6.2789,
      "step": 1184
    },
    {
      "epoch": 0.1185,
      "grad_norm": 3.563152551651001,
      "learning_rate": 0.00019979380835002846,
      "loss": 6.755,
      "step": 1185
    },
    {
      "epoch": 0.1186,
      "grad_norm": 21.083065032958984,
      "learning_rate": 0.0001997915618272179,
      "loss": 6.2952,
      "step": 1186
    },
    {
      "epoch": 0.1187,
      "grad_norm": 10.38369083404541,
      "learning_rate": 0.00019978930314510824,
      "loss": 7.0015,
      "step": 1187
    },
    {
      "epoch": 0.1188,
      "grad_norm": 4.350432395935059,
      "learning_rate": 0.0001997870323039747,
      "loss": 6.8537,
      "step": 1188
    },
    {
      "epoch": 0.1189,
      "grad_norm": 7.179705619812012,
      "learning_rate": 0.00019978474930409395,
      "loss": 6.0022,
      "step": 1189
    },
    {
      "epoch": 0.119,
      "grad_norm": 4.343725204467773,
      "learning_rate": 0.00019978245414574417,
      "loss": 6.5898,
      "step": 1190
    },
    {
      "epoch": 0.1191,
      "grad_norm": 8.492365837097168,
      "learning_rate": 0.000199780146829205,
      "loss": 5.7696,
      "step": 1191
    },
    {
      "epoch": 0.1192,
      "grad_norm": 4.202782154083252,
      "learning_rate": 0.00019977782735475764,
      "loss": 6.1872,
      "step": 1192
    },
    {
      "epoch": 0.1193,
      "grad_norm": 3.4761199951171875,
      "learning_rate": 0.00019977549572268468,
      "loss": 7.1533,
      "step": 1193
    },
    {
      "epoch": 0.1194,
      "grad_norm": 5.098177433013916,
      "learning_rate": 0.00019977315193327018,
      "loss": 7.2223,
      "step": 1194
    },
    {
      "epoch": 0.1195,
      "grad_norm": 3.865321159362793,
      "learning_rate": 0.00019977079598679977,
      "loss": 6.489,
      "step": 1195
    },
    {
      "epoch": 0.1196,
      "grad_norm": 11.583856582641602,
      "learning_rate": 0.00019976842788356055,
      "loss": 6.9505,
      "step": 1196
    },
    {
      "epoch": 0.1197,
      "grad_norm": 5.30537223815918,
      "learning_rate": 0.000199766047623841,
      "loss": 6.3977,
      "step": 1197
    },
    {
      "epoch": 0.1198,
      "grad_norm": 5.085717678070068,
      "learning_rate": 0.00019976365520793114,
      "loss": 6.421,
      "step": 1198
    },
    {
      "epoch": 0.1199,
      "grad_norm": 3.086678981781006,
      "learning_rate": 0.00019976125063612252,
      "loss": 6.3968,
      "step": 1199
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.5635972023010254,
      "learning_rate": 0.00019975883390870817,
      "loss": 6.1616,
      "step": 1200
    },
    {
      "epoch": 0.1201,
      "grad_norm": 7.286343574523926,
      "learning_rate": 0.00019975640502598244,
      "loss": 5.5693,
      "step": 1201
    },
    {
      "epoch": 0.1202,
      "grad_norm": 4.626665115356445,
      "learning_rate": 0.0001997539639882414,
      "loss": 6.4054,
      "step": 1202
    },
    {
      "epoch": 0.1203,
      "grad_norm": 4.949975490570068,
      "learning_rate": 0.00019975151079578237,
      "loss": 6.8366,
      "step": 1203
    },
    {
      "epoch": 0.1204,
      "grad_norm": 11.917214393615723,
      "learning_rate": 0.0001997490454489044,
      "loss": 5.9743,
      "step": 1204
    },
    {
      "epoch": 0.1205,
      "grad_norm": 2.6019840240478516,
      "learning_rate": 0.00019974656794790775,
      "loss": 6.6644,
      "step": 1205
    },
    {
      "epoch": 0.1206,
      "grad_norm": 5.627778053283691,
      "learning_rate": 0.0001997440782930944,
      "loss": 7.1525,
      "step": 1206
    },
    {
      "epoch": 0.1207,
      "grad_norm": 13.305916786193848,
      "learning_rate": 0.00019974157648476766,
      "loss": 6.717,
      "step": 1207
    },
    {
      "epoch": 0.1208,
      "grad_norm": 3.5155088901519775,
      "learning_rate": 0.00019973906252323238,
      "loss": 6.0594,
      "step": 1208
    },
    {
      "epoch": 0.1209,
      "grad_norm": 8.534208297729492,
      "learning_rate": 0.00019973653640879485,
      "loss": 6.3324,
      "step": 1209
    },
    {
      "epoch": 0.121,
      "grad_norm": 18.715084075927734,
      "learning_rate": 0.00019973399814176293,
      "loss": 7.6782,
      "step": 1210
    },
    {
      "epoch": 0.1211,
      "grad_norm": 4.902933120727539,
      "learning_rate": 0.00019973144772244582,
      "loss": 6.33,
      "step": 1211
    },
    {
      "epoch": 0.1212,
      "grad_norm": 9.316807746887207,
      "learning_rate": 0.00019972888515115434,
      "loss": 6.3463,
      "step": 1212
    },
    {
      "epoch": 0.1213,
      "grad_norm": 2.9979586601257324,
      "learning_rate": 0.0001997263104282007,
      "loss": 6.4855,
      "step": 1213
    },
    {
      "epoch": 0.1214,
      "grad_norm": 4.421186447143555,
      "learning_rate": 0.00019972372355389867,
      "loss": 6.5991,
      "step": 1214
    },
    {
      "epoch": 0.1215,
      "grad_norm": 7.093933582305908,
      "learning_rate": 0.00019972112452856339,
      "loss": 6.5393,
      "step": 1215
    },
    {
      "epoch": 0.1216,
      "grad_norm": 3.462965726852417,
      "learning_rate": 0.00019971851335251158,
      "loss": 5.827,
      "step": 1216
    },
    {
      "epoch": 0.1217,
      "grad_norm": 4.021228790283203,
      "learning_rate": 0.0001997158900260614,
      "loss": 5.8162,
      "step": 1217
    },
    {
      "epoch": 0.1218,
      "grad_norm": 4.407204627990723,
      "learning_rate": 0.00019971325454953248,
      "loss": 5.908,
      "step": 1218
    },
    {
      "epoch": 0.1219,
      "grad_norm": 4.932379245758057,
      "learning_rate": 0.00019971060692324598,
      "loss": 6.1749,
      "step": 1219
    },
    {
      "epoch": 0.122,
      "grad_norm": 27.511734008789062,
      "learning_rate": 0.00019970794714752445,
      "loss": 7.3882,
      "step": 1220
    },
    {
      "epoch": 0.1221,
      "grad_norm": 3.2979865074157715,
      "learning_rate": 0.00019970527522269205,
      "loss": 6.3925,
      "step": 1221
    },
    {
      "epoch": 0.1222,
      "grad_norm": 4.525108337402344,
      "learning_rate": 0.00019970259114907425,
      "loss": 6.3677,
      "step": 1222
    },
    {
      "epoch": 0.1223,
      "grad_norm": 9.11737060546875,
      "learning_rate": 0.0001996998949269982,
      "loss": 6.0709,
      "step": 1223
    },
    {
      "epoch": 0.1224,
      "grad_norm": 6.620458126068115,
      "learning_rate": 0.00019969718655679232,
      "loss": 6.8831,
      "step": 1224
    },
    {
      "epoch": 0.1225,
      "grad_norm": 9.47276782989502,
      "learning_rate": 0.00019969446603878673,
      "loss": 6.0502,
      "step": 1225
    },
    {
      "epoch": 0.1226,
      "grad_norm": 2.938058376312256,
      "learning_rate": 0.0001996917333733128,
      "loss": 6.676,
      "step": 1226
    },
    {
      "epoch": 0.1227,
      "grad_norm": 3.314056158065796,
      "learning_rate": 0.0001996889885607036,
      "loss": 6.509,
      "step": 1227
    },
    {
      "epoch": 0.1228,
      "grad_norm": 23.433948516845703,
      "learning_rate": 0.0001996862316012935,
      "loss": 6.4566,
      "step": 1228
    },
    {
      "epoch": 0.1229,
      "grad_norm": 22.97016143798828,
      "learning_rate": 0.00019968346249541846,
      "loss": 6.9096,
      "step": 1229
    },
    {
      "epoch": 0.123,
      "grad_norm": 4.376377105712891,
      "learning_rate": 0.0001996806812434159,
      "loss": 6.1791,
      "step": 1230
    },
    {
      "epoch": 0.1231,
      "grad_norm": 28.546232223510742,
      "learning_rate": 0.00019967788784562473,
      "loss": 6.8142,
      "step": 1231
    },
    {
      "epoch": 0.1232,
      "grad_norm": 2.9523799419403076,
      "learning_rate": 0.00019967508230238522,
      "loss": 6.3485,
      "step": 1232
    },
    {
      "epoch": 0.1233,
      "grad_norm": 37.53013229370117,
      "learning_rate": 0.00019967226461403933,
      "loss": 7.97,
      "step": 1233
    },
    {
      "epoch": 0.1234,
      "grad_norm": 5.648580551147461,
      "learning_rate": 0.0001996694347809303,
      "loss": 7.0033,
      "step": 1234
    },
    {
      "epoch": 0.1235,
      "grad_norm": 6.023107528686523,
      "learning_rate": 0.00019966659280340297,
      "loss": 6.1792,
      "step": 1235
    },
    {
      "epoch": 0.1236,
      "grad_norm": 3.158841609954834,
      "learning_rate": 0.00019966373868180365,
      "loss": 6.0766,
      "step": 1236
    },
    {
      "epoch": 0.1237,
      "grad_norm": 5.0534281730651855,
      "learning_rate": 0.0001996608724164801,
      "loss": 5.8044,
      "step": 1237
    },
    {
      "epoch": 0.1238,
      "grad_norm": 6.993600368499756,
      "learning_rate": 0.00019965799400778152,
      "loss": 6.4811,
      "step": 1238
    },
    {
      "epoch": 0.1239,
      "grad_norm": 3.1193156242370605,
      "learning_rate": 0.00019965510345605866,
      "loss": 7.1116,
      "step": 1239
    },
    {
      "epoch": 0.124,
      "grad_norm": 4.141943454742432,
      "learning_rate": 0.00019965220076166376,
      "loss": 6.9942,
      "step": 1240
    },
    {
      "epoch": 0.1241,
      "grad_norm": 6.945807456970215,
      "learning_rate": 0.00019964928592495045,
      "loss": 6.3691,
      "step": 1241
    },
    {
      "epoch": 0.1242,
      "grad_norm": 3.6951539516448975,
      "learning_rate": 0.0001996463589462739,
      "loss": 6.57,
      "step": 1242
    },
    {
      "epoch": 0.1243,
      "grad_norm": 3.5448975563049316,
      "learning_rate": 0.00019964341982599078,
      "loss": 5.9593,
      "step": 1243
    },
    {
      "epoch": 0.1244,
      "grad_norm": 3.7808022499084473,
      "learning_rate": 0.00019964046856445924,
      "loss": 6.552,
      "step": 1244
    },
    {
      "epoch": 0.1245,
      "grad_norm": 15.486818313598633,
      "learning_rate": 0.00019963750516203884,
      "loss": 6.7635,
      "step": 1245
    },
    {
      "epoch": 0.1246,
      "grad_norm": 7.203108310699463,
      "learning_rate": 0.00019963452961909063,
      "loss": 7.3002,
      "step": 1246
    },
    {
      "epoch": 0.1247,
      "grad_norm": 26.34151268005371,
      "learning_rate": 0.00019963154193597727,
      "loss": 7.3027,
      "step": 1247
    },
    {
      "epoch": 0.1248,
      "grad_norm": 23.025920867919922,
      "learning_rate": 0.00019962854211306267,
      "loss": 7.1196,
      "step": 1248
    },
    {
      "epoch": 0.1249,
      "grad_norm": 4.456088542938232,
      "learning_rate": 0.0001996255301507125,
      "loss": 7.2229,
      "step": 1249
    },
    {
      "epoch": 0.125,
      "grad_norm": 6.5024590492248535,
      "learning_rate": 0.0001996225060492936,
      "loss": 6.1175,
      "step": 1250
    },
    {
      "epoch": 0.1251,
      "grad_norm": 2.7996158599853516,
      "learning_rate": 0.00019961946980917456,
      "loss": 6.1923,
      "step": 1251
    },
    {
      "epoch": 0.1252,
      "grad_norm": 18.677892684936523,
      "learning_rate": 0.00019961642143072529,
      "loss": 6.764,
      "step": 1252
    },
    {
      "epoch": 0.1253,
      "grad_norm": 8.214037895202637,
      "learning_rate": 0.00019961336091431727,
      "loss": 6.7991,
      "step": 1253
    },
    {
      "epoch": 0.1254,
      "grad_norm": 2.6039607524871826,
      "learning_rate": 0.00019961028826032332,
      "loss": 6.2956,
      "step": 1254
    },
    {
      "epoch": 0.1255,
      "grad_norm": 12.79025650024414,
      "learning_rate": 0.00019960720346911797,
      "loss": 6.8,
      "step": 1255
    },
    {
      "epoch": 0.1256,
      "grad_norm": 3.083996534347534,
      "learning_rate": 0.00019960410654107697,
      "loss": 6.7031,
      "step": 1256
    },
    {
      "epoch": 0.1257,
      "grad_norm": 8.85771369934082,
      "learning_rate": 0.00019960099747657774,
      "loss": 7.3458,
      "step": 1257
    },
    {
      "epoch": 0.1258,
      "grad_norm": 3.297135829925537,
      "learning_rate": 0.00019959787627599906,
      "loss": 6.59,
      "step": 1258
    },
    {
      "epoch": 0.1259,
      "grad_norm": 3.8552956581115723,
      "learning_rate": 0.00019959474293972129,
      "loss": 6.0705,
      "step": 1259
    },
    {
      "epoch": 0.126,
      "grad_norm": 6.250372409820557,
      "learning_rate": 0.0001995915974681262,
      "loss": 6.6215,
      "step": 1260
    },
    {
      "epoch": 0.1261,
      "grad_norm": 5.76372766494751,
      "learning_rate": 0.00019958843986159704,
      "loss": 6.6012,
      "step": 1261
    },
    {
      "epoch": 0.1262,
      "grad_norm": 5.4000935554504395,
      "learning_rate": 0.00019958527012051857,
      "loss": 7.678,
      "step": 1262
    },
    {
      "epoch": 0.1263,
      "grad_norm": 5.426791667938232,
      "learning_rate": 0.000199582088245277,
      "loss": 6.4717,
      "step": 1263
    },
    {
      "epoch": 0.1264,
      "grad_norm": 4.794586658477783,
      "learning_rate": 0.00019957889423626005,
      "loss": 6.0562,
      "step": 1264
    },
    {
      "epoch": 0.1265,
      "grad_norm": 2.476564645767212,
      "learning_rate": 0.00019957568809385694,
      "loss": 6.3723,
      "step": 1265
    },
    {
      "epoch": 0.1266,
      "grad_norm": 4.300589084625244,
      "learning_rate": 0.00019957246981845822,
      "loss": 6.6223,
      "step": 1266
    },
    {
      "epoch": 0.1267,
      "grad_norm": 3.688951015472412,
      "learning_rate": 0.0001995692394104561,
      "loss": 6.615,
      "step": 1267
    },
    {
      "epoch": 0.1268,
      "grad_norm": 4.689548492431641,
      "learning_rate": 0.0001995659968702442,
      "loss": 8.4474,
      "step": 1268
    },
    {
      "epoch": 0.1269,
      "grad_norm": 7.535949230194092,
      "learning_rate": 0.00019956274219821757,
      "loss": 6.5523,
      "step": 1269
    },
    {
      "epoch": 0.127,
      "grad_norm": 5.51613712310791,
      "learning_rate": 0.00019955947539477284,
      "loss": 5.8356,
      "step": 1270
    },
    {
      "epoch": 0.1271,
      "grad_norm": 33.16654968261719,
      "learning_rate": 0.00019955619646030802,
      "loss": 7.0581,
      "step": 1271
    },
    {
      "epoch": 0.1272,
      "grad_norm": 37.21415328979492,
      "learning_rate": 0.0001995529053952226,
      "loss": 7.2874,
      "step": 1272
    },
    {
      "epoch": 0.1273,
      "grad_norm": 2.8793556690216064,
      "learning_rate": 0.0001995496021999177,
      "loss": 6.2414,
      "step": 1273
    },
    {
      "epoch": 0.1274,
      "grad_norm": 3.5289831161499023,
      "learning_rate": 0.00019954628687479572,
      "loss": 6.5813,
      "step": 1274
    },
    {
      "epoch": 0.1275,
      "grad_norm": 4.586955547332764,
      "learning_rate": 0.00019954295942026064,
      "loss": 5.9707,
      "step": 1275
    },
    {
      "epoch": 0.1276,
      "grad_norm": 6.076490879058838,
      "learning_rate": 0.00019953961983671788,
      "loss": 6.2262,
      "step": 1276
    },
    {
      "epoch": 0.1277,
      "grad_norm": 3.840427875518799,
      "learning_rate": 0.0001995362681245744,
      "loss": 6.4822,
      "step": 1277
    },
    {
      "epoch": 0.1278,
      "grad_norm": 4.270594120025635,
      "learning_rate": 0.00019953290428423857,
      "loss": 5.672,
      "step": 1278
    },
    {
      "epoch": 0.1279,
      "grad_norm": 5.2021284103393555,
      "learning_rate": 0.00019952952831612026,
      "loss": 6.449,
      "step": 1279
    },
    {
      "epoch": 0.128,
      "grad_norm": 3.5048468112945557,
      "learning_rate": 0.00019952614022063084,
      "loss": 6.3551,
      "step": 1280
    },
    {
      "epoch": 0.1281,
      "grad_norm": 6.398503303527832,
      "learning_rate": 0.0001995227399981831,
      "loss": 6.8283,
      "step": 1281
    },
    {
      "epoch": 0.1282,
      "grad_norm": 8.101448059082031,
      "learning_rate": 0.00019951932764919144,
      "loss": 6.3548,
      "step": 1282
    },
    {
      "epoch": 0.1283,
      "grad_norm": 2.873020887374878,
      "learning_rate": 0.00019951590317407152,
      "loss": 6.3214,
      "step": 1283
    },
    {
      "epoch": 0.1284,
      "grad_norm": 3.5268895626068115,
      "learning_rate": 0.0001995124665732407,
      "loss": 6.3021,
      "step": 1284
    },
    {
      "epoch": 0.1285,
      "grad_norm": 5.360505104064941,
      "learning_rate": 0.00019950901784711764,
      "loss": 6.2365,
      "step": 1285
    },
    {
      "epoch": 0.1286,
      "grad_norm": 12.167865753173828,
      "learning_rate": 0.00019950555699612267,
      "loss": 6.7652,
      "step": 1286
    },
    {
      "epoch": 0.1287,
      "grad_norm": 4.20221471786499,
      "learning_rate": 0.00019950208402067733,
      "loss": 6.5302,
      "step": 1287
    },
    {
      "epoch": 0.1288,
      "grad_norm": 7.991434574127197,
      "learning_rate": 0.00019949859892120491,
      "loss": 6.4571,
      "step": 1288
    },
    {
      "epoch": 0.1289,
      "grad_norm": 5.133400917053223,
      "learning_rate": 0.00019949510169813003,
      "loss": 7.2776,
      "step": 1289
    },
    {
      "epoch": 0.129,
      "grad_norm": 4.094973564147949,
      "learning_rate": 0.0001994915923518788,
      "loss": 6.5069,
      "step": 1290
    },
    {
      "epoch": 0.1291,
      "grad_norm": 3.346757411956787,
      "learning_rate": 0.00019948807088287883,
      "loss": 5.5505,
      "step": 1291
    },
    {
      "epoch": 0.1292,
      "grad_norm": 10.845585823059082,
      "learning_rate": 0.00019948453729155922,
      "loss": 6.662,
      "step": 1292
    },
    {
      "epoch": 0.1293,
      "grad_norm": 7.266964912414551,
      "learning_rate": 0.00019948099157835047,
      "loss": 6.6472,
      "step": 1293
    },
    {
      "epoch": 0.1294,
      "grad_norm": 6.721320152282715,
      "learning_rate": 0.00019947743374368467,
      "loss": 6.2515,
      "step": 1294
    },
    {
      "epoch": 0.1295,
      "grad_norm": 2.5288281440734863,
      "learning_rate": 0.00019947386378799532,
      "loss": 6.4069,
      "step": 1295
    },
    {
      "epoch": 0.1296,
      "grad_norm": 16.47856903076172,
      "learning_rate": 0.00019947028171171742,
      "loss": 6.8276,
      "step": 1296
    },
    {
      "epoch": 0.1297,
      "grad_norm": 6.577784538269043,
      "learning_rate": 0.00019946668751528744,
      "loss": 6.375,
      "step": 1297
    },
    {
      "epoch": 0.1298,
      "grad_norm": 5.8585076332092285,
      "learning_rate": 0.00019946308119914323,
      "loss": 6.1028,
      "step": 1298
    },
    {
      "epoch": 0.1299,
      "grad_norm": 11.06511116027832,
      "learning_rate": 0.00019945946276372434,
      "loss": 5.711,
      "step": 1299
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.5989737510681152,
      "learning_rate": 0.00019945583220947158,
      "loss": 6.1355,
      "step": 1300
    },
    {
      "epoch": 0.1301,
      "grad_norm": 2.8581960201263428,
      "learning_rate": 0.00019945218953682734,
      "loss": 6.5545,
      "step": 1301
    },
    {
      "epoch": 0.1302,
      "grad_norm": 4.240442752838135,
      "learning_rate": 0.00019944853474623548,
      "loss": 6.2685,
      "step": 1302
    },
    {
      "epoch": 0.1303,
      "grad_norm": 4.814721584320068,
      "learning_rate": 0.00019944486783814134,
      "loss": 6.0837,
      "step": 1303
    },
    {
      "epoch": 0.1304,
      "grad_norm": 3.45200514793396,
      "learning_rate": 0.00019944118881299168,
      "loss": 6.5097,
      "step": 1304
    },
    {
      "epoch": 0.1305,
      "grad_norm": 11.079977989196777,
      "learning_rate": 0.0001994374976712348,
      "loss": 6.4797,
      "step": 1305
    },
    {
      "epoch": 0.1306,
      "grad_norm": 3.1817235946655273,
      "learning_rate": 0.00019943379441332047,
      "loss": 6.9824,
      "step": 1306
    },
    {
      "epoch": 0.1307,
      "grad_norm": 4.462839126586914,
      "learning_rate": 0.0001994300790396999,
      "loss": 6.2941,
      "step": 1307
    },
    {
      "epoch": 0.1308,
      "grad_norm": 3.692131280899048,
      "learning_rate": 0.0001994263515508258,
      "loss": 6.1023,
      "step": 1308
    },
    {
      "epoch": 0.1309,
      "grad_norm": 3.216075897216797,
      "learning_rate": 0.00019942261194715236,
      "loss": 6.4982,
      "step": 1309
    },
    {
      "epoch": 0.131,
      "grad_norm": 6.136855125427246,
      "learning_rate": 0.00019941886022913522,
      "loss": 6.2452,
      "step": 1310
    },
    {
      "epoch": 0.1311,
      "grad_norm": 7.975552082061768,
      "learning_rate": 0.00019941509639723155,
      "loss": 7.3243,
      "step": 1311
    },
    {
      "epoch": 0.1312,
      "grad_norm": 5.026971340179443,
      "learning_rate": 0.0001994113204518999,
      "loss": 5.9646,
      "step": 1312
    },
    {
      "epoch": 0.1313,
      "grad_norm": 3.346031427383423,
      "learning_rate": 0.00019940753239360047,
      "loss": 6.4776,
      "step": 1313
    },
    {
      "epoch": 0.1314,
      "grad_norm": 4.569288730621338,
      "learning_rate": 0.00019940373222279473,
      "loss": 7.2119,
      "step": 1314
    },
    {
      "epoch": 0.1315,
      "grad_norm": 4.3197784423828125,
      "learning_rate": 0.0001993999199399457,
      "loss": 6.1004,
      "step": 1315
    },
    {
      "epoch": 0.1316,
      "grad_norm": 5.219241619110107,
      "learning_rate": 0.000199396095545518,
      "loss": 6.8364,
      "step": 1316
    },
    {
      "epoch": 0.1317,
      "grad_norm": 17.420940399169922,
      "learning_rate": 0.0001993922590399775,
      "loss": 6.2743,
      "step": 1317
    },
    {
      "epoch": 0.1318,
      "grad_norm": 4.402424335479736,
      "learning_rate": 0.00019938841042379174,
      "loss": 6.1231,
      "step": 1318
    },
    {
      "epoch": 0.1319,
      "grad_norm": 17.533578872680664,
      "learning_rate": 0.00019938454969742968,
      "loss": 7.5912,
      "step": 1319
    },
    {
      "epoch": 0.132,
      "grad_norm": 4.176123142242432,
      "learning_rate": 0.00019938067686136167,
      "loss": 6.3557,
      "step": 1320
    },
    {
      "epoch": 0.1321,
      "grad_norm": 2.858701467514038,
      "learning_rate": 0.00019937679191605963,
      "loss": 6.1861,
      "step": 1321
    },
    {
      "epoch": 0.1322,
      "grad_norm": 2.7702693939208984,
      "learning_rate": 0.00019937289486199696,
      "loss": 6.7124,
      "step": 1322
    },
    {
      "epoch": 0.1323,
      "grad_norm": 10.585866928100586,
      "learning_rate": 0.00019936898569964848,
      "loss": 5.8652,
      "step": 1323
    },
    {
      "epoch": 0.1324,
      "grad_norm": 3.052502393722534,
      "learning_rate": 0.0001993650644294905,
      "loss": 6.3275,
      "step": 1324
    },
    {
      "epoch": 0.1325,
      "grad_norm": 4.068549156188965,
      "learning_rate": 0.00019936113105200085,
      "loss": 6.6695,
      "step": 1325
    },
    {
      "epoch": 0.1326,
      "grad_norm": 12.86643123626709,
      "learning_rate": 0.00019935718556765876,
      "loss": 6.1713,
      "step": 1326
    },
    {
      "epoch": 0.1327,
      "grad_norm": 2.2531979084014893,
      "learning_rate": 0.000199353227976945,
      "loss": 6.5233,
      "step": 1327
    },
    {
      "epoch": 0.1328,
      "grad_norm": 2.996063470840454,
      "learning_rate": 0.00019934925828034175,
      "loss": 7.016,
      "step": 1328
    },
    {
      "epoch": 0.1329,
      "grad_norm": 3.94227933883667,
      "learning_rate": 0.00019934527647833276,
      "loss": 6.0408,
      "step": 1329
    },
    {
      "epoch": 0.133,
      "grad_norm": 3.822396993637085,
      "learning_rate": 0.0001993412825714032,
      "loss": 6.1522,
      "step": 1330
    },
    {
      "epoch": 0.1331,
      "grad_norm": 6.308868408203125,
      "learning_rate": 0.00019933727656003963,
      "loss": 6.3243,
      "step": 1331
    },
    {
      "epoch": 0.1332,
      "grad_norm": 2.6109886169433594,
      "learning_rate": 0.0001993332584447303,
      "loss": 6.3054,
      "step": 1332
    },
    {
      "epoch": 0.1333,
      "grad_norm": 30.921567916870117,
      "learning_rate": 0.00019932922822596473,
      "loss": 6.7413,
      "step": 1333
    },
    {
      "epoch": 0.1334,
      "grad_norm": 12.135939598083496,
      "learning_rate": 0.00019932518590423394,
      "loss": 6.458,
      "step": 1334
    },
    {
      "epoch": 0.1335,
      "grad_norm": 2.8504276275634766,
      "learning_rate": 0.00019932113148003058,
      "loss": 6.4438,
      "step": 1335
    },
    {
      "epoch": 0.1336,
      "grad_norm": 23.3405704498291,
      "learning_rate": 0.00019931706495384863,
      "loss": 6.3243,
      "step": 1336
    },
    {
      "epoch": 0.1337,
      "grad_norm": 3.6833534240722656,
      "learning_rate": 0.00019931298632618356,
      "loss": 6.5173,
      "step": 1337
    },
    {
      "epoch": 0.1338,
      "grad_norm": 3.4211742877960205,
      "learning_rate": 0.00019930889559753234,
      "loss": 6.7067,
      "step": 1338
    },
    {
      "epoch": 0.1339,
      "grad_norm": 11.009700775146484,
      "learning_rate": 0.00019930479276839344,
      "loss": 6.5176,
      "step": 1339
    },
    {
      "epoch": 0.134,
      "grad_norm": 2.4424262046813965,
      "learning_rate": 0.00019930067783926675,
      "loss": 6.3363,
      "step": 1340
    },
    {
      "epoch": 0.1341,
      "grad_norm": 16.638147354125977,
      "learning_rate": 0.0001992965508106537,
      "loss": 6.78,
      "step": 1341
    },
    {
      "epoch": 0.1342,
      "grad_norm": 3.868515968322754,
      "learning_rate": 0.00019929241168305714,
      "loss": 6.7968,
      "step": 1342
    },
    {
      "epoch": 0.1343,
      "grad_norm": 3.540454864501953,
      "learning_rate": 0.00019928826045698136,
      "loss": 5.5251,
      "step": 1343
    },
    {
      "epoch": 0.1344,
      "grad_norm": 3.39860200881958,
      "learning_rate": 0.00019928409713293227,
      "loss": 6.0524,
      "step": 1344
    },
    {
      "epoch": 0.1345,
      "grad_norm": 4.949613571166992,
      "learning_rate": 0.00019927992171141708,
      "loss": 6.7757,
      "step": 1345
    },
    {
      "epoch": 0.1346,
      "grad_norm": 5.908738613128662,
      "learning_rate": 0.00019927573419294456,
      "loss": 6.7889,
      "step": 1346
    },
    {
      "epoch": 0.1347,
      "grad_norm": 27.755475997924805,
      "learning_rate": 0.000199271534578025,
      "loss": 7.1517,
      "step": 1347
    },
    {
      "epoch": 0.1348,
      "grad_norm": 7.899183750152588,
      "learning_rate": 0.00019926732286717003,
      "loss": 6.0513,
      "step": 1348
    },
    {
      "epoch": 0.1349,
      "grad_norm": 2.5348196029663086,
      "learning_rate": 0.0001992630990608929,
      "loss": 6.4269,
      "step": 1349
    },
    {
      "epoch": 0.135,
      "grad_norm": 4.06800651550293,
      "learning_rate": 0.00019925886315970824,
      "loss": 6.5875,
      "step": 1350
    },
    {
      "epoch": 0.1351,
      "grad_norm": 3.434058427810669,
      "learning_rate": 0.00019925461516413223,
      "loss": 5.9552,
      "step": 1351
    },
    {
      "epoch": 0.1352,
      "grad_norm": 14.005159378051758,
      "learning_rate": 0.0001992503550746824,
      "loss": 6.9582,
      "step": 1352
    },
    {
      "epoch": 0.1353,
      "grad_norm": 2.8106627464294434,
      "learning_rate": 0.00019924608289187786,
      "loss": 6.6771,
      "step": 1353
    },
    {
      "epoch": 0.1354,
      "grad_norm": 3.9993860721588135,
      "learning_rate": 0.00019924179861623915,
      "loss": 6.5374,
      "step": 1354
    },
    {
      "epoch": 0.1355,
      "grad_norm": 7.936709403991699,
      "learning_rate": 0.00019923750224828832,
      "loss": 6.0953,
      "step": 1355
    },
    {
      "epoch": 0.1356,
      "grad_norm": 7.096683025360107,
      "learning_rate": 0.00019923319378854887,
      "loss": 5.8566,
      "step": 1356
    },
    {
      "epoch": 0.1357,
      "grad_norm": 4.494136810302734,
      "learning_rate": 0.00019922887323754577,
      "loss": 6.4158,
      "step": 1357
    },
    {
      "epoch": 0.1358,
      "grad_norm": 2.567351818084717,
      "learning_rate": 0.00019922454059580544,
      "loss": 6.1294,
      "step": 1358
    },
    {
      "epoch": 0.1359,
      "grad_norm": 3.473238945007324,
      "learning_rate": 0.00019922019586385585,
      "loss": 6.1998,
      "step": 1359
    },
    {
      "epoch": 0.136,
      "grad_norm": 5.265823841094971,
      "learning_rate": 0.00019921583904222633,
      "loss": 6.4273,
      "step": 1360
    },
    {
      "epoch": 0.1361,
      "grad_norm": 3.723710775375366,
      "learning_rate": 0.0001992114701314478,
      "loss": 6.8428,
      "step": 1361
    },
    {
      "epoch": 0.1362,
      "grad_norm": 2.6500344276428223,
      "learning_rate": 0.00019920708913205256,
      "loss": 5.343,
      "step": 1362
    },
    {
      "epoch": 0.1363,
      "grad_norm": 5.611306190490723,
      "learning_rate": 0.00019920269604457446,
      "loss": 7.2919,
      "step": 1363
    },
    {
      "epoch": 0.1364,
      "grad_norm": 5.109810829162598,
      "learning_rate": 0.0001991982908695487,
      "loss": 6.0545,
      "step": 1364
    },
    {
      "epoch": 0.1365,
      "grad_norm": 3.2115113735198975,
      "learning_rate": 0.00019919387360751217,
      "loss": 6.6558,
      "step": 1365
    },
    {
      "epoch": 0.1366,
      "grad_norm": 8.60498332977295,
      "learning_rate": 0.000199189444259003,
      "loss": 6.3807,
      "step": 1366
    },
    {
      "epoch": 0.1367,
      "grad_norm": 2.9771347045898438,
      "learning_rate": 0.0001991850028245609,
      "loss": 6.1566,
      "step": 1367
    },
    {
      "epoch": 0.1368,
      "grad_norm": 17.837268829345703,
      "learning_rate": 0.00019918054930472706,
      "loss": 5.9537,
      "step": 1368
    },
    {
      "epoch": 0.1369,
      "grad_norm": 5.413870811462402,
      "learning_rate": 0.00019917608370004417,
      "loss": 6.7876,
      "step": 1369
    },
    {
      "epoch": 0.137,
      "grad_norm": 9.183757781982422,
      "learning_rate": 0.0001991716060110563,
      "loss": 6.246,
      "step": 1370
    },
    {
      "epoch": 0.1371,
      "grad_norm": 13.569846153259277,
      "learning_rate": 0.00019916711623830903,
      "loss": 6.1863,
      "step": 1371
    },
    {
      "epoch": 0.1372,
      "grad_norm": 14.655313491821289,
      "learning_rate": 0.0001991626143823495,
      "loss": 7.28,
      "step": 1372
    },
    {
      "epoch": 0.1373,
      "grad_norm": 4.010460376739502,
      "learning_rate": 0.00019915810044372618,
      "loss": 6.6698,
      "step": 1373
    },
    {
      "epoch": 0.1374,
      "grad_norm": 2.5924267768859863,
      "learning_rate": 0.0001991535744229891,
      "loss": 5.9555,
      "step": 1374
    },
    {
      "epoch": 0.1375,
      "grad_norm": 8.512405395507812,
      "learning_rate": 0.00019914903632068973,
      "loss": 6.0173,
      "step": 1375
    },
    {
      "epoch": 0.1376,
      "grad_norm": 5.430909156799316,
      "learning_rate": 0.00019914448613738106,
      "loss": 5.7243,
      "step": 1376
    },
    {
      "epoch": 0.1377,
      "grad_norm": 3.9521844387054443,
      "learning_rate": 0.00019913992387361745,
      "loss": 5.8009,
      "step": 1377
    },
    {
      "epoch": 0.1378,
      "grad_norm": 18.037199020385742,
      "learning_rate": 0.00019913534952995486,
      "loss": 6.0484,
      "step": 1378
    },
    {
      "epoch": 0.1379,
      "grad_norm": 9.69560432434082,
      "learning_rate": 0.00019913076310695068,
      "loss": 7.3806,
      "step": 1379
    },
    {
      "epoch": 0.138,
      "grad_norm": 5.404945373535156,
      "learning_rate": 0.00019912616460516364,
      "loss": 6.366,
      "step": 1380
    },
    {
      "epoch": 0.1381,
      "grad_norm": 4.960036754608154,
      "learning_rate": 0.00019912155402515417,
      "loss": 6.0718,
      "step": 1381
    },
    {
      "epoch": 0.1382,
      "grad_norm": 6.2083659172058105,
      "learning_rate": 0.00019911693136748403,
      "loss": 5.7435,
      "step": 1382
    },
    {
      "epoch": 0.1383,
      "grad_norm": 5.784119129180908,
      "learning_rate": 0.0001991122966327164,
      "loss": 7.1655,
      "step": 1383
    },
    {
      "epoch": 0.1384,
      "grad_norm": 4.985413551330566,
      "learning_rate": 0.0001991076498214161,
      "loss": 5.8731,
      "step": 1384
    },
    {
      "epoch": 0.1385,
      "grad_norm": 5.365553855895996,
      "learning_rate": 0.0001991029909341493,
      "loss": 6.5143,
      "step": 1385
    },
    {
      "epoch": 0.1386,
      "grad_norm": 4.7905049324035645,
      "learning_rate": 0.00019909831997148362,
      "loss": 6.4534,
      "step": 1386
    },
    {
      "epoch": 0.1387,
      "grad_norm": 5.010308742523193,
      "learning_rate": 0.00019909363693398828,
      "loss": 6.3556,
      "step": 1387
    },
    {
      "epoch": 0.1388,
      "grad_norm": 4.567235946655273,
      "learning_rate": 0.00019908894182223388,
      "loss": 7.5104,
      "step": 1388
    },
    {
      "epoch": 0.1389,
      "grad_norm": 3.771850347518921,
      "learning_rate": 0.00019908423463679248,
      "loss": 6.9983,
      "step": 1389
    },
    {
      "epoch": 0.139,
      "grad_norm": 13.039653778076172,
      "learning_rate": 0.0001990795153782376,
      "loss": 5.8345,
      "step": 1390
    },
    {
      "epoch": 0.1391,
      "grad_norm": 20.022335052490234,
      "learning_rate": 0.00019907478404714436,
      "loss": 7.2073,
      "step": 1391
    },
    {
      "epoch": 0.1392,
      "grad_norm": 23.92179298400879,
      "learning_rate": 0.0001990700406440892,
      "loss": 6.4649,
      "step": 1392
    },
    {
      "epoch": 0.1393,
      "grad_norm": 12.315613746643066,
      "learning_rate": 0.00019906528516965008,
      "loss": 6.8883,
      "step": 1393
    },
    {
      "epoch": 0.1394,
      "grad_norm": 9.916460990905762,
      "learning_rate": 0.0001990605176244065,
      "loss": 6.3732,
      "step": 1394
    },
    {
      "epoch": 0.1395,
      "grad_norm": 3.5285940170288086,
      "learning_rate": 0.0001990557380089393,
      "loss": 6.7412,
      "step": 1395
    },
    {
      "epoch": 0.1396,
      "grad_norm": 15.474716186523438,
      "learning_rate": 0.0001990509463238309,
      "loss": 6.4012,
      "step": 1396
    },
    {
      "epoch": 0.1397,
      "grad_norm": 17.405101776123047,
      "learning_rate": 0.00019904614256966512,
      "loss": 6.4248,
      "step": 1397
    },
    {
      "epoch": 0.1398,
      "grad_norm": 3.7369120121002197,
      "learning_rate": 0.00019904132674702734,
      "loss": 6.4636,
      "step": 1398
    },
    {
      "epoch": 0.1399,
      "grad_norm": 4.112911224365234,
      "learning_rate": 0.0001990364988565043,
      "loss": 6.0119,
      "step": 1399
    },
    {
      "epoch": 0.14,
      "grad_norm": 5.82694149017334,
      "learning_rate": 0.0001990316588986843,
      "loss": 6.7427,
      "step": 1400
    },
    {
      "epoch": 0.1401,
      "grad_norm": 5.715142250061035,
      "learning_rate": 0.00019902680687415705,
      "loss": 6.4647,
      "step": 1401
    },
    {
      "epoch": 0.1402,
      "grad_norm": 5.792476654052734,
      "learning_rate": 0.00019902194278351374,
      "loss": 6.1107,
      "step": 1402
    },
    {
      "epoch": 0.1403,
      "grad_norm": 3.470132350921631,
      "learning_rate": 0.00019901706662734712,
      "loss": 6.0504,
      "step": 1403
    },
    {
      "epoch": 0.1404,
      "grad_norm": 5.023730754852295,
      "learning_rate": 0.0001990121784062512,
      "loss": 5.8437,
      "step": 1404
    },
    {
      "epoch": 0.1405,
      "grad_norm": 4.175595760345459,
      "learning_rate": 0.00019900727812082177,
      "loss": 6.4541,
      "step": 1405
    },
    {
      "epoch": 0.1406,
      "grad_norm": 3.7531445026397705,
      "learning_rate": 0.00019900236577165576,
      "loss": 5.7811,
      "step": 1406
    },
    {
      "epoch": 0.1407,
      "grad_norm": 11.103231430053711,
      "learning_rate": 0.0001989974413593518,
      "loss": 6.9594,
      "step": 1407
    },
    {
      "epoch": 0.1408,
      "grad_norm": 3.580679178237915,
      "learning_rate": 0.0001989925048845099,
      "loss": 7.0534,
      "step": 1408
    },
    {
      "epoch": 0.1409,
      "grad_norm": 11.431265830993652,
      "learning_rate": 0.00019898755634773158,
      "loss": 7.2328,
      "step": 1409
    },
    {
      "epoch": 0.141,
      "grad_norm": 3.824816942214966,
      "learning_rate": 0.00019898259574961978,
      "loss": 5.649,
      "step": 1410
    },
    {
      "epoch": 0.1411,
      "grad_norm": 3.3451790809631348,
      "learning_rate": 0.0001989776230907789,
      "loss": 6.985,
      "step": 1411
    },
    {
      "epoch": 0.1412,
      "grad_norm": 6.523676872253418,
      "learning_rate": 0.00019897263837181491,
      "loss": 5.7281,
      "step": 1412
    },
    {
      "epoch": 0.1413,
      "grad_norm": 6.93057107925415,
      "learning_rate": 0.0001989676415933351,
      "loss": 7.6129,
      "step": 1413
    },
    {
      "epoch": 0.1414,
      "grad_norm": 3.308619499206543,
      "learning_rate": 0.00019896263275594842,
      "loss": 6.1504,
      "step": 1414
    },
    {
      "epoch": 0.1415,
      "grad_norm": 5.92955207824707,
      "learning_rate": 0.0001989576118602651,
      "loss": 6.2166,
      "step": 1415
    },
    {
      "epoch": 0.1416,
      "grad_norm": 3.30224609375,
      "learning_rate": 0.00019895257890689696,
      "loss": 5.8701,
      "step": 1416
    },
    {
      "epoch": 0.1417,
      "grad_norm": 5.595959663391113,
      "learning_rate": 0.00019894753389645723,
      "loss": 5.6471,
      "step": 1417
    },
    {
      "epoch": 0.1418,
      "grad_norm": 3.7319538593292236,
      "learning_rate": 0.0001989424768295606,
      "loss": 6.2624,
      "step": 1418
    },
    {
      "epoch": 0.1419,
      "grad_norm": 3.034482479095459,
      "learning_rate": 0.00019893740770682335,
      "loss": 5.9573,
      "step": 1419
    },
    {
      "epoch": 0.142,
      "grad_norm": 4.285284519195557,
      "learning_rate": 0.00019893232652886306,
      "loss": 6.9362,
      "step": 1420
    },
    {
      "epoch": 0.1421,
      "grad_norm": 3.782917022705078,
      "learning_rate": 0.00019892723329629887,
      "loss": 5.8848,
      "step": 1421
    },
    {
      "epoch": 0.1422,
      "grad_norm": 2.777339220046997,
      "learning_rate": 0.00019892212800975135,
      "loss": 6.2184,
      "step": 1422
    },
    {
      "epoch": 0.1423,
      "grad_norm": 10.942635536193848,
      "learning_rate": 0.00019891701066984262,
      "loss": 5.9683,
      "step": 1423
    },
    {
      "epoch": 0.1424,
      "grad_norm": 5.816223621368408,
      "learning_rate": 0.00019891188127719618,
      "loss": 6.3865,
      "step": 1424
    },
    {
      "epoch": 0.1425,
      "grad_norm": 3.3964524269104004,
      "learning_rate": 0.00019890673983243706,
      "loss": 6.3667,
      "step": 1425
    },
    {
      "epoch": 0.1426,
      "grad_norm": 4.653342247009277,
      "learning_rate": 0.0001989015863361917,
      "loss": 6.6341,
      "step": 1426
    },
    {
      "epoch": 0.1427,
      "grad_norm": 14.952950477600098,
      "learning_rate": 0.00019889642078908804,
      "loss": 6.0839,
      "step": 1427
    },
    {
      "epoch": 0.1428,
      "grad_norm": 16.79621696472168,
      "learning_rate": 0.00019889124319175547,
      "loss": 6.2924,
      "step": 1428
    },
    {
      "epoch": 0.1429,
      "grad_norm": 6.144786834716797,
      "learning_rate": 0.0001988860535448249,
      "loss": 6.7485,
      "step": 1429
    },
    {
      "epoch": 0.143,
      "grad_norm": 5.633111476898193,
      "learning_rate": 0.00019888085184892868,
      "loss": 6.1384,
      "step": 1430
    },
    {
      "epoch": 0.1431,
      "grad_norm": 3.2314867973327637,
      "learning_rate": 0.0001988756381047006,
      "loss": 6.1331,
      "step": 1431
    },
    {
      "epoch": 0.1432,
      "grad_norm": 2.8328423500061035,
      "learning_rate": 0.00019887041231277593,
      "loss": 6.0577,
      "step": 1432
    },
    {
      "epoch": 0.1433,
      "grad_norm": 7.474536418914795,
      "learning_rate": 0.0001988651744737914,
      "loss": 6.6866,
      "step": 1433
    },
    {
      "epoch": 0.1434,
      "grad_norm": 3.011575698852539,
      "learning_rate": 0.00019885992458838528,
      "loss": 5.9278,
      "step": 1434
    },
    {
      "epoch": 0.1435,
      "grad_norm": 5.305382251739502,
      "learning_rate": 0.0001988546626571972,
      "loss": 6.1204,
      "step": 1435
    },
    {
      "epoch": 0.1436,
      "grad_norm": 3.854435682296753,
      "learning_rate": 0.00019884938868086835,
      "loss": 5.9899,
      "step": 1436
    },
    {
      "epoch": 0.1437,
      "grad_norm": 5.586604118347168,
      "learning_rate": 0.00019884410266004135,
      "loss": 6.0136,
      "step": 1437
    },
    {
      "epoch": 0.1438,
      "grad_norm": 5.480003833770752,
      "learning_rate": 0.00019883880459536024,
      "loss": 6.4441,
      "step": 1438
    },
    {
      "epoch": 0.1439,
      "grad_norm": 2.928748607635498,
      "learning_rate": 0.00019883349448747062,
      "loss": 6.451,
      "step": 1439
    },
    {
      "epoch": 0.144,
      "grad_norm": 3.3310482501983643,
      "learning_rate": 0.00019882817233701948,
      "loss": 6.1891,
      "step": 1440
    },
    {
      "epoch": 0.1441,
      "grad_norm": 8.793728828430176,
      "learning_rate": 0.0001988228381446553,
      "loss": 5.6101,
      "step": 1441
    },
    {
      "epoch": 0.1442,
      "grad_norm": 17.750041961669922,
      "learning_rate": 0.00019881749191102808,
      "loss": 6.2156,
      "step": 1442
    },
    {
      "epoch": 0.1443,
      "grad_norm": 5.058167457580566,
      "learning_rate": 0.0001988121336367892,
      "loss": 6.0673,
      "step": 1443
    },
    {
      "epoch": 0.1444,
      "grad_norm": 4.037942886352539,
      "learning_rate": 0.00019880676332259154,
      "loss": 6.7652,
      "step": 1444
    },
    {
      "epoch": 0.1445,
      "grad_norm": 16.098041534423828,
      "learning_rate": 0.00019880138096908952,
      "loss": 6.8519,
      "step": 1445
    },
    {
      "epoch": 0.1446,
      "grad_norm": 7.843860149383545,
      "learning_rate": 0.00019879598657693891,
      "loss": 6.8988,
      "step": 1446
    },
    {
      "epoch": 0.1447,
      "grad_norm": 5.607309341430664,
      "learning_rate": 0.00019879058014679704,
      "loss": 6.523,
      "step": 1447
    },
    {
      "epoch": 0.1448,
      "grad_norm": 5.621946334838867,
      "learning_rate": 0.00019878516167932261,
      "loss": 5.63,
      "step": 1448
    },
    {
      "epoch": 0.1449,
      "grad_norm": 22.03716468811035,
      "learning_rate": 0.0001987797311751759,
      "loss": 6.7694,
      "step": 1449
    },
    {
      "epoch": 0.145,
      "grad_norm": 5.422957897186279,
      "learning_rate": 0.00019877428863501856,
      "loss": 5.7306,
      "step": 1450
    },
    {
      "epoch": 0.1451,
      "grad_norm": 3.5506598949432373,
      "learning_rate": 0.00019876883405951377,
      "loss": 6.2785,
      "step": 1451
    },
    {
      "epoch": 0.1452,
      "grad_norm": 4.203254222869873,
      "learning_rate": 0.00019876336744932614,
      "loss": 6.3767,
      "step": 1452
    },
    {
      "epoch": 0.1453,
      "grad_norm": 6.975799560546875,
      "learning_rate": 0.0001987578888051218,
      "loss": 6.6991,
      "step": 1453
    },
    {
      "epoch": 0.1454,
      "grad_norm": 3.5477805137634277,
      "learning_rate": 0.00019875239812756825,
      "loss": 5.8142,
      "step": 1454
    },
    {
      "epoch": 0.1455,
      "grad_norm": 4.335143566131592,
      "learning_rate": 0.00019874689541733457,
      "loss": 6.0938,
      "step": 1455
    },
    {
      "epoch": 0.1456,
      "grad_norm": 3.7003626823425293,
      "learning_rate": 0.00019874138067509117,
      "loss": 5.9813,
      "step": 1456
    },
    {
      "epoch": 0.1457,
      "grad_norm": 13.3601713180542,
      "learning_rate": 0.00019873585390151003,
      "loss": 6.4686,
      "step": 1457
    },
    {
      "epoch": 0.1458,
      "grad_norm": 3.8584108352661133,
      "learning_rate": 0.00019873031509726462,
      "loss": 5.8883,
      "step": 1458
    },
    {
      "epoch": 0.1459,
      "grad_norm": 5.075403213500977,
      "learning_rate": 0.00019872476426302982,
      "loss": 6.5476,
      "step": 1459
    },
    {
      "epoch": 0.146,
      "grad_norm": 4.31878137588501,
      "learning_rate": 0.00019871920139948192,
      "loss": 6.4399,
      "step": 1460
    },
    {
      "epoch": 0.1461,
      "grad_norm": 3.560209035873413,
      "learning_rate": 0.0001987136265072988,
      "loss": 6.0298,
      "step": 1461
    },
    {
      "epoch": 0.1462,
      "grad_norm": 3.5443663597106934,
      "learning_rate": 0.00019870803958715972,
      "loss": 5.9017,
      "step": 1462
    },
    {
      "epoch": 0.1463,
      "grad_norm": 10.941161155700684,
      "learning_rate": 0.0001987024406397454,
      "loss": 5.7007,
      "step": 1463
    },
    {
      "epoch": 0.1464,
      "grad_norm": 8.254899978637695,
      "learning_rate": 0.00019869682966573813,
      "loss": 6.6231,
      "step": 1464
    },
    {
      "epoch": 0.1465,
      "grad_norm": 20.43697166442871,
      "learning_rate": 0.00019869120666582153,
      "loss": 6.4515,
      "step": 1465
    },
    {
      "epoch": 0.1466,
      "grad_norm": 3.5105202198028564,
      "learning_rate": 0.00019868557164068074,
      "loss": 6.6117,
      "step": 1466
    },
    {
      "epoch": 0.1467,
      "grad_norm": 2.602978467941284,
      "learning_rate": 0.0001986799245910024,
      "loss": 7.1826,
      "step": 1467
    },
    {
      "epoch": 0.1468,
      "grad_norm": 2.946438789367676,
      "learning_rate": 0.00019867426551747457,
      "loss": 6.1644,
      "step": 1468
    },
    {
      "epoch": 0.1469,
      "grad_norm": 4.012312412261963,
      "learning_rate": 0.0001986685944207868,
      "loss": 5.726,
      "step": 1469
    },
    {
      "epoch": 0.147,
      "grad_norm": 4.76703405380249,
      "learning_rate": 0.0001986629113016301,
      "loss": 6.4226,
      "step": 1470
    },
    {
      "epoch": 0.1471,
      "grad_norm": 13.068105697631836,
      "learning_rate": 0.00019865721616069696,
      "loss": 5.976,
      "step": 1471
    },
    {
      "epoch": 0.1472,
      "grad_norm": 4.860212802886963,
      "learning_rate": 0.00019865150899868125,
      "loss": 6.2931,
      "step": 1472
    },
    {
      "epoch": 0.1473,
      "grad_norm": 29.89514923095703,
      "learning_rate": 0.00019864578981627844,
      "loss": 7.0604,
      "step": 1473
    },
    {
      "epoch": 0.1474,
      "grad_norm": 5.654849529266357,
      "learning_rate": 0.00019864005861418535,
      "loss": 5.8409,
      "step": 1474
    },
    {
      "epoch": 0.1475,
      "grad_norm": 4.882045745849609,
      "learning_rate": 0.0001986343153931003,
      "loss": 6.2488,
      "step": 1475
    },
    {
      "epoch": 0.1476,
      "grad_norm": 8.88980770111084,
      "learning_rate": 0.00019862856015372317,
      "loss": 6.7347,
      "step": 1476
    },
    {
      "epoch": 0.1477,
      "grad_norm": 6.1968183517456055,
      "learning_rate": 0.00019862279289675509,
      "loss": 5.8379,
      "step": 1477
    },
    {
      "epoch": 0.1478,
      "grad_norm": 2.697148084640503,
      "learning_rate": 0.0001986170136228989,
      "loss": 6.1113,
      "step": 1478
    },
    {
      "epoch": 0.1479,
      "grad_norm": 32.48728942871094,
      "learning_rate": 0.0001986112223328587,
      "loss": 7.7204,
      "step": 1479
    },
    {
      "epoch": 0.148,
      "grad_norm": 27.91783332824707,
      "learning_rate": 0.00019860541902734022,
      "loss": 8.0789,
      "step": 1480
    },
    {
      "epoch": 0.1481,
      "grad_norm": 3.921055316925049,
      "learning_rate": 0.0001985996037070505,
      "loss": 6.4839,
      "step": 1481
    },
    {
      "epoch": 0.1482,
      "grad_norm": 4.076384544372559,
      "learning_rate": 0.00019859377637269815,
      "loss": 6.1701,
      "step": 1482
    },
    {
      "epoch": 0.1483,
      "grad_norm": 3.069115400314331,
      "learning_rate": 0.00019858793702499323,
      "loss": 6.2109,
      "step": 1483
    },
    {
      "epoch": 0.1484,
      "grad_norm": 2.3846874237060547,
      "learning_rate": 0.00019858208566464724,
      "loss": 6.4423,
      "step": 1484
    },
    {
      "epoch": 0.1485,
      "grad_norm": 3.824368476867676,
      "learning_rate": 0.00019857622229237313,
      "loss": 6.0903,
      "step": 1485
    },
    {
      "epoch": 0.1486,
      "grad_norm": 6.7495036125183105,
      "learning_rate": 0.00019857034690888537,
      "loss": 7.4153,
      "step": 1486
    },
    {
      "epoch": 0.1487,
      "grad_norm": 3.90415096282959,
      "learning_rate": 0.00019856445951489982,
      "loss": 5.5949,
      "step": 1487
    },
    {
      "epoch": 0.1488,
      "grad_norm": 4.187326908111572,
      "learning_rate": 0.00019855856011113384,
      "loss": 6.4236,
      "step": 1488
    },
    {
      "epoch": 0.1489,
      "grad_norm": 2.845231056213379,
      "learning_rate": 0.00019855264869830629,
      "loss": 6.3131,
      "step": 1489
    },
    {
      "epoch": 0.149,
      "grad_norm": 4.594104766845703,
      "learning_rate": 0.00019854672527713744,
      "loss": 5.9447,
      "step": 1490
    },
    {
      "epoch": 0.1491,
      "grad_norm": 2.8727729320526123,
      "learning_rate": 0.00019854078984834903,
      "loss": 5.8562,
      "step": 1491
    },
    {
      "epoch": 0.1492,
      "grad_norm": 3.8906025886535645,
      "learning_rate": 0.00019853484241266428,
      "loss": 5.9232,
      "step": 1492
    },
    {
      "epoch": 0.1493,
      "grad_norm": 7.498710632324219,
      "learning_rate": 0.00019852888297080786,
      "loss": 5.7993,
      "step": 1493
    },
    {
      "epoch": 0.1494,
      "grad_norm": 8.324222564697266,
      "learning_rate": 0.00019852291152350592,
      "loss": 5.9244,
      "step": 1494
    },
    {
      "epoch": 0.1495,
      "grad_norm": 12.446283340454102,
      "learning_rate": 0.0001985169280714861,
      "loss": 6.1901,
      "step": 1495
    },
    {
      "epoch": 0.1496,
      "grad_norm": 18.07090950012207,
      "learning_rate": 0.0001985109326154774,
      "loss": 6.4329,
      "step": 1496
    },
    {
      "epoch": 0.1497,
      "grad_norm": 2.594386577606201,
      "learning_rate": 0.00019850492515621038,
      "loss": 5.7845,
      "step": 1497
    },
    {
      "epoch": 0.1498,
      "grad_norm": 8.959917068481445,
      "learning_rate": 0.00019849890569441703,
      "loss": 5.8325,
      "step": 1498
    },
    {
      "epoch": 0.1499,
      "grad_norm": 7.499693393707275,
      "learning_rate": 0.00019849287423083078,
      "loss": 7.7219,
      "step": 1499
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.1004981994628906,
      "learning_rate": 0.00019848683076618658,
      "loss": 5.7371,
      "step": 1500
    },
    {
      "epoch": 0.1501,
      "grad_norm": 3.459298610687256,
      "learning_rate": 0.00019848077530122083,
      "loss": 5.7896,
      "step": 1501
    },
    {
      "epoch": 0.1502,
      "grad_norm": 5.823365688323975,
      "learning_rate": 0.00019847470783667127,
      "loss": 6.7645,
      "step": 1502
    },
    {
      "epoch": 0.1503,
      "grad_norm": 3.151432991027832,
      "learning_rate": 0.0001984686283732773,
      "loss": 6.1176,
      "step": 1503
    },
    {
      "epoch": 0.1504,
      "grad_norm": 3.4135518074035645,
      "learning_rate": 0.00019846253691177966,
      "loss": 6.1582,
      "step": 1504
    },
    {
      "epoch": 0.1505,
      "grad_norm": 4.000507354736328,
      "learning_rate": 0.00019845643345292054,
      "loss": 6.2319,
      "step": 1505
    },
    {
      "epoch": 0.1506,
      "grad_norm": 27.338428497314453,
      "learning_rate": 0.00019845031799744367,
      "loss": 6.5127,
      "step": 1506
    },
    {
      "epoch": 0.1507,
      "grad_norm": 2.712653875350952,
      "learning_rate": 0.0001984441905460942,
      "loss": 6.6966,
      "step": 1507
    },
    {
      "epoch": 0.1508,
      "grad_norm": 12.803380966186523,
      "learning_rate": 0.00019843805109961868,
      "loss": 6.2504,
      "step": 1508
    },
    {
      "epoch": 0.1509,
      "grad_norm": 2.900704860687256,
      "learning_rate": 0.00019843189965876526,
      "loss": 6.5554,
      "step": 1509
    },
    {
      "epoch": 0.151,
      "grad_norm": 14.524004936218262,
      "learning_rate": 0.00019842573622428345,
      "loss": 5.7478,
      "step": 1510
    },
    {
      "epoch": 0.1511,
      "grad_norm": 4.3778910636901855,
      "learning_rate": 0.0001984195607969242,
      "loss": 5.6818,
      "step": 1511
    },
    {
      "epoch": 0.1512,
      "grad_norm": 24.40483283996582,
      "learning_rate": 0.00019841337337744004,
      "loss": 6.3612,
      "step": 1512
    },
    {
      "epoch": 0.1513,
      "grad_norm": 8.543416023254395,
      "learning_rate": 0.00019840717396658484,
      "loss": 6.0479,
      "step": 1513
    },
    {
      "epoch": 0.1514,
      "grad_norm": 22.04160499572754,
      "learning_rate": 0.00019840096256511398,
      "loss": 7.1893,
      "step": 1514
    },
    {
      "epoch": 0.1515,
      "grad_norm": 6.460065841674805,
      "learning_rate": 0.00019839473917378434,
      "loss": 6.1182,
      "step": 1515
    },
    {
      "epoch": 0.1516,
      "grad_norm": 6.316243648529053,
      "learning_rate": 0.00019838850379335417,
      "loss": 6.3078,
      "step": 1516
    },
    {
      "epoch": 0.1517,
      "grad_norm": 3.568207263946533,
      "learning_rate": 0.00019838225642458327,
      "loss": 5.7707,
      "step": 1517
    },
    {
      "epoch": 0.1518,
      "grad_norm": 5.437285900115967,
      "learning_rate": 0.00019837599706823284,
      "loss": 6.5071,
      "step": 1518
    },
    {
      "epoch": 0.1519,
      "grad_norm": 4.808868885040283,
      "learning_rate": 0.00019836972572506557,
      "loss": 6.191,
      "step": 1519
    },
    {
      "epoch": 0.152,
      "grad_norm": 4.332277297973633,
      "learning_rate": 0.00019836344239584564,
      "loss": 5.9919,
      "step": 1520
    },
    {
      "epoch": 0.1521,
      "grad_norm": 8.278935432434082,
      "learning_rate": 0.00019835714708133862,
      "loss": 6.2549,
      "step": 1521
    },
    {
      "epoch": 0.1522,
      "grad_norm": 7.206933498382568,
      "learning_rate": 0.00019835083978231156,
      "loss": 6.9454,
      "step": 1522
    },
    {
      "epoch": 0.1523,
      "grad_norm": 3.386845350265503,
      "learning_rate": 0.00019834452049953297,
      "loss": 5.9527,
      "step": 1523
    },
    {
      "epoch": 0.1524,
      "grad_norm": 2.846667766571045,
      "learning_rate": 0.0001983381892337729,
      "loss": 6.0791,
      "step": 1524
    },
    {
      "epoch": 0.1525,
      "grad_norm": 4.292538642883301,
      "learning_rate": 0.00019833184598580276,
      "loss": 6.4015,
      "step": 1525
    },
    {
      "epoch": 0.1526,
      "grad_norm": 5.919532775878906,
      "learning_rate": 0.0001983254907563955,
      "loss": 6.1401,
      "step": 1526
    },
    {
      "epoch": 0.1527,
      "grad_norm": 5.008467674255371,
      "learning_rate": 0.00019831912354632535,
      "loss": 6.4535,
      "step": 1527
    },
    {
      "epoch": 0.1528,
      "grad_norm": 5.122410297393799,
      "learning_rate": 0.0001983127443563683,
      "loss": 6.7536,
      "step": 1528
    },
    {
      "epoch": 0.1529,
      "grad_norm": 3.031797409057617,
      "learning_rate": 0.00019830635318730154,
      "loss": 6.1229,
      "step": 1529
    },
    {
      "epoch": 0.153,
      "grad_norm": 33.59365463256836,
      "learning_rate": 0.00019829995003990388,
      "loss": 8.083,
      "step": 1530
    },
    {
      "epoch": 0.1531,
      "grad_norm": 13.721955299377441,
      "learning_rate": 0.00019829353491495545,
      "loss": 6.1038,
      "step": 1531
    },
    {
      "epoch": 0.1532,
      "grad_norm": 9.163947105407715,
      "learning_rate": 0.00019828710781323792,
      "loss": 5.6243,
      "step": 1532
    },
    {
      "epoch": 0.1533,
      "grad_norm": 22.69277572631836,
      "learning_rate": 0.00019828066873553448,
      "loss": 6.556,
      "step": 1533
    },
    {
      "epoch": 0.1534,
      "grad_norm": 18.138051986694336,
      "learning_rate": 0.00019827421768262967,
      "loss": 6.6381,
      "step": 1534
    },
    {
      "epoch": 0.1535,
      "grad_norm": 5.5928473472595215,
      "learning_rate": 0.0001982677546553095,
      "loss": 5.8243,
      "step": 1535
    },
    {
      "epoch": 0.1536,
      "grad_norm": 2.326807737350464,
      "learning_rate": 0.00019826127965436152,
      "loss": 6.0193,
      "step": 1536
    },
    {
      "epoch": 0.1537,
      "grad_norm": 3.499063491821289,
      "learning_rate": 0.00019825479268057467,
      "loss": 5.9065,
      "step": 1537
    },
    {
      "epoch": 0.1538,
      "grad_norm": 6.936647415161133,
      "learning_rate": 0.0001982482937347394,
      "loss": 6.3481,
      "step": 1538
    },
    {
      "epoch": 0.1539,
      "grad_norm": 2.650195837020874,
      "learning_rate": 0.00019824178281764753,
      "loss": 6.5946,
      "step": 1539
    },
    {
      "epoch": 0.154,
      "grad_norm": 4.636447906494141,
      "learning_rate": 0.00019823525993009243,
      "loss": 5.7455,
      "step": 1540
    },
    {
      "epoch": 0.1541,
      "grad_norm": 4.741785049438477,
      "learning_rate": 0.0001982287250728689,
      "loss": 6.1886,
      "step": 1541
    },
    {
      "epoch": 0.1542,
      "grad_norm": 8.141077041625977,
      "learning_rate": 0.00019822217824677315,
      "loss": 7.2414,
      "step": 1542
    },
    {
      "epoch": 0.1543,
      "grad_norm": 3.0670690536499023,
      "learning_rate": 0.0001982156194526029,
      "loss": 5.5023,
      "step": 1543
    },
    {
      "epoch": 0.1544,
      "grad_norm": 4.374127388000488,
      "learning_rate": 0.0001982090486911574,
      "loss": 6.0073,
      "step": 1544
    },
    {
      "epoch": 0.1545,
      "grad_norm": 4.687154769897461,
      "learning_rate": 0.0001982024659632372,
      "loss": 6.9501,
      "step": 1545
    },
    {
      "epoch": 0.1546,
      "grad_norm": 2.3160223960876465,
      "learning_rate": 0.00019819587126964437,
      "loss": 5.7467,
      "step": 1546
    },
    {
      "epoch": 0.1547,
      "grad_norm": 8.60235595703125,
      "learning_rate": 0.00019818926461118253,
      "loss": 5.8646,
      "step": 1547
    },
    {
      "epoch": 0.1548,
      "grad_norm": 4.081870079040527,
      "learning_rate": 0.0001981826459886566,
      "loss": 6.2451,
      "step": 1548
    },
    {
      "epoch": 0.1549,
      "grad_norm": 13.51298999786377,
      "learning_rate": 0.00019817601540287306,
      "loss": 5.6537,
      "step": 1549
    },
    {
      "epoch": 0.155,
      "grad_norm": 6.785743713378906,
      "learning_rate": 0.0001981693728546399,
      "loss": 6.31,
      "step": 1550
    },
    {
      "epoch": 0.1551,
      "grad_norm": 2.9385159015655518,
      "learning_rate": 0.00019816271834476642,
      "loss": 6.2,
      "step": 1551
    },
    {
      "epoch": 0.1552,
      "grad_norm": 5.322979927062988,
      "learning_rate": 0.00019815605187406345,
      "loss": 6.0935,
      "step": 1552
    },
    {
      "epoch": 0.1553,
      "grad_norm": 3.096200466156006,
      "learning_rate": 0.0001981493734433433,
      "loss": 6.1902,
      "step": 1553
    },
    {
      "epoch": 0.1554,
      "grad_norm": 3.750493049621582,
      "learning_rate": 0.0001981426830534197,
      "loss": 6.2069,
      "step": 1554
    },
    {
      "epoch": 0.1555,
      "grad_norm": 4.590871810913086,
      "learning_rate": 0.00019813598070510792,
      "loss": 5.8919,
      "step": 1555
    },
    {
      "epoch": 0.1556,
      "grad_norm": 7.3258056640625,
      "learning_rate": 0.0001981292663992245,
      "loss": 5.5865,
      "step": 1556
    },
    {
      "epoch": 0.1557,
      "grad_norm": 3.21292781829834,
      "learning_rate": 0.00019812254013658768,
      "loss": 5.8689,
      "step": 1557
    },
    {
      "epoch": 0.1558,
      "grad_norm": 2.7371909618377686,
      "learning_rate": 0.00019811580191801697,
      "loss": 6.3014,
      "step": 1558
    },
    {
      "epoch": 0.1559,
      "grad_norm": 4.3314008712768555,
      "learning_rate": 0.0001981090517443334,
      "loss": 5.7967,
      "step": 1559
    },
    {
      "epoch": 0.156,
      "grad_norm": 3.1336276531219482,
      "learning_rate": 0.0001981022896163595,
      "loss": 6.0419,
      "step": 1560
    },
    {
      "epoch": 0.1561,
      "grad_norm": 7.648438453674316,
      "learning_rate": 0.00019809551553491916,
      "loss": 6.4768,
      "step": 1561
    },
    {
      "epoch": 0.1562,
      "grad_norm": 6.874436378479004,
      "learning_rate": 0.00019808872950083782,
      "loss": 5.9018,
      "step": 1562
    },
    {
      "epoch": 0.1563,
      "grad_norm": 3.956057071685791,
      "learning_rate": 0.00019808193151494232,
      "loss": 6.345,
      "step": 1563
    },
    {
      "epoch": 0.1564,
      "grad_norm": 4.152203559875488,
      "learning_rate": 0.000198075121578061,
      "loss": 6.7544,
      "step": 1564
    },
    {
      "epoch": 0.1565,
      "grad_norm": 3.0950779914855957,
      "learning_rate": 0.00019806829969102357,
      "loss": 5.9786,
      "step": 1565
    },
    {
      "epoch": 0.1566,
      "grad_norm": 3.1351404190063477,
      "learning_rate": 0.0001980614658546613,
      "loss": 6.0386,
      "step": 1566
    },
    {
      "epoch": 0.1567,
      "grad_norm": 4.378595352172852,
      "learning_rate": 0.00019805462006980689,
      "loss": 5.9761,
      "step": 1567
    },
    {
      "epoch": 0.1568,
      "grad_norm": 3.6324026584625244,
      "learning_rate": 0.00019804776233729444,
      "loss": 6.1768,
      "step": 1568
    },
    {
      "epoch": 0.1569,
      "grad_norm": 3.914083957672119,
      "learning_rate": 0.0001980408926579596,
      "loss": 6.4058,
      "step": 1569
    },
    {
      "epoch": 0.157,
      "grad_norm": 2.9790897369384766,
      "learning_rate": 0.00019803401103263933,
      "loss": 5.4266,
      "step": 1570
    },
    {
      "epoch": 0.1571,
      "grad_norm": 5.98195219039917,
      "learning_rate": 0.00019802711746217218,
      "loss": 6.9734,
      "step": 1571
    },
    {
      "epoch": 0.1572,
      "grad_norm": 3.4176979064941406,
      "learning_rate": 0.00019802021194739814,
      "loss": 6.5187,
      "step": 1572
    },
    {
      "epoch": 0.1573,
      "grad_norm": 4.543689250946045,
      "learning_rate": 0.00019801329448915862,
      "loss": 5.9767,
      "step": 1573
    },
    {
      "epoch": 0.1574,
      "grad_norm": 8.349156379699707,
      "learning_rate": 0.00019800636508829643,
      "loss": 6.6527,
      "step": 1574
    },
    {
      "epoch": 0.1575,
      "grad_norm": 5.307377815246582,
      "learning_rate": 0.00019799942374565597,
      "loss": 5.8214,
      "step": 1575
    },
    {
      "epoch": 0.1576,
      "grad_norm": 5.636430740356445,
      "learning_rate": 0.00019799247046208297,
      "loss": 6.9537,
      "step": 1576
    },
    {
      "epoch": 0.1577,
      "grad_norm": 8.706647872924805,
      "learning_rate": 0.0001979855052384247,
      "loss": 6.0177,
      "step": 1577
    },
    {
      "epoch": 0.1578,
      "grad_norm": 4.717194557189941,
      "learning_rate": 0.00019797852807552983,
      "loss": 6.2671,
      "step": 1578
    },
    {
      "epoch": 0.1579,
      "grad_norm": 9.34829330444336,
      "learning_rate": 0.00019797153897424852,
      "loss": 7.4442,
      "step": 1579
    },
    {
      "epoch": 0.158,
      "grad_norm": 3.1785640716552734,
      "learning_rate": 0.00019796453793543238,
      "loss": 5.9889,
      "step": 1580
    },
    {
      "epoch": 0.1581,
      "grad_norm": 5.352281093597412,
      "learning_rate": 0.0001979575249599344,
      "loss": 6.0439,
      "step": 1581
    },
    {
      "epoch": 0.1582,
      "grad_norm": 5.774744987487793,
      "learning_rate": 0.00019795050004860917,
      "loss": 6.822,
      "step": 1582
    },
    {
      "epoch": 0.1583,
      "grad_norm": 2.4030251502990723,
      "learning_rate": 0.00019794346320231265,
      "loss": 6.2823,
      "step": 1583
    },
    {
      "epoch": 0.1584,
      "grad_norm": 11.258435249328613,
      "learning_rate": 0.00019793641442190221,
      "loss": 6.2815,
      "step": 1584
    },
    {
      "epoch": 0.1585,
      "grad_norm": 6.004825115203857,
      "learning_rate": 0.00019792935370823675,
      "loss": 6.3627,
      "step": 1585
    },
    {
      "epoch": 0.1586,
      "grad_norm": 16.16559410095215,
      "learning_rate": 0.00019792228106217658,
      "loss": 6.7931,
      "step": 1586
    },
    {
      "epoch": 0.1587,
      "grad_norm": 2.992741823196411,
      "learning_rate": 0.00019791519648458352,
      "loss": 5.8892,
      "step": 1587
    },
    {
      "epoch": 0.1588,
      "grad_norm": 3.030811309814453,
      "learning_rate": 0.00019790809997632076,
      "loss": 5.6811,
      "step": 1588
    },
    {
      "epoch": 0.1589,
      "grad_norm": 7.701308727264404,
      "learning_rate": 0.000197900991538253,
      "loss": 5.9741,
      "step": 1589
    },
    {
      "epoch": 0.159,
      "grad_norm": 4.401525497436523,
      "learning_rate": 0.00019789387117124637,
      "loss": 6.1451,
      "step": 1590
    },
    {
      "epoch": 0.1591,
      "grad_norm": 3.969043254852295,
      "learning_rate": 0.0001978867388761685,
      "loss": 5.7228,
      "step": 1591
    },
    {
      "epoch": 0.1592,
      "grad_norm": 3.67364764213562,
      "learning_rate": 0.00019787959465388842,
      "loss": 5.7507,
      "step": 1592
    },
    {
      "epoch": 0.1593,
      "grad_norm": 4.679394721984863,
      "learning_rate": 0.00019787243850527664,
      "loss": 6.1197,
      "step": 1593
    },
    {
      "epoch": 0.1594,
      "grad_norm": 2.877375364303589,
      "learning_rate": 0.0001978652704312051,
      "loss": 6.128,
      "step": 1594
    },
    {
      "epoch": 0.1595,
      "grad_norm": 5.089925289154053,
      "learning_rate": 0.00019785809043254722,
      "loss": 6.2634,
      "step": 1595
    },
    {
      "epoch": 0.1596,
      "grad_norm": 2.3162341117858887,
      "learning_rate": 0.00019785089851017787,
      "loss": 5.9112,
      "step": 1596
    },
    {
      "epoch": 0.1597,
      "grad_norm": 3.273693323135376,
      "learning_rate": 0.0001978436946649733,
      "loss": 6.4035,
      "step": 1597
    },
    {
      "epoch": 0.1598,
      "grad_norm": 4.453209400177002,
      "learning_rate": 0.00019783647889781136,
      "loss": 5.9532,
      "step": 1598
    },
    {
      "epoch": 0.1599,
      "grad_norm": 15.431026458740234,
      "learning_rate": 0.00019782925120957124,
      "loss": 6.2718,
      "step": 1599
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.754751682281494,
      "learning_rate": 0.0001978220116011336,
      "loss": 6.5255,
      "step": 1600
    },
    {
      "epoch": 0.1601,
      "grad_norm": 3.376430034637451,
      "learning_rate": 0.00019781476007338058,
      "loss": 5.7392,
      "step": 1601
    },
    {
      "epoch": 0.1602,
      "grad_norm": 4.911788463592529,
      "learning_rate": 0.00019780749662719573,
      "loss": 5.9304,
      "step": 1602
    },
    {
      "epoch": 0.1603,
      "grad_norm": 16.590452194213867,
      "learning_rate": 0.0001978002212634641,
      "loss": 7.2373,
      "step": 1603
    },
    {
      "epoch": 0.1604,
      "grad_norm": 7.351429462432861,
      "learning_rate": 0.0001977929339830722,
      "loss": 5.9722,
      "step": 1604
    },
    {
      "epoch": 0.1605,
      "grad_norm": 4.058887958526611,
      "learning_rate": 0.0001977856347869079,
      "loss": 5.948,
      "step": 1605
    },
    {
      "epoch": 0.1606,
      "grad_norm": 3.998201847076416,
      "learning_rate": 0.00019777832367586063,
      "loss": 6.3358,
      "step": 1606
    },
    {
      "epoch": 0.1607,
      "grad_norm": 3.2245588302612305,
      "learning_rate": 0.00019777100065082118,
      "loss": 5.8844,
      "step": 1607
    },
    {
      "epoch": 0.1608,
      "grad_norm": 3.947009325027466,
      "learning_rate": 0.00019776366571268192,
      "loss": 5.8417,
      "step": 1608
    },
    {
      "epoch": 0.1609,
      "grad_norm": 4.296567916870117,
      "learning_rate": 0.00019775631886233654,
      "loss": 5.6319,
      "step": 1609
    },
    {
      "epoch": 0.161,
      "grad_norm": 3.8585739135742188,
      "learning_rate": 0.0001977489601006802,
      "loss": 6.208,
      "step": 1610
    },
    {
      "epoch": 0.1611,
      "grad_norm": 2.29195237159729,
      "learning_rate": 0.0001977415894286096,
      "loss": 5.7449,
      "step": 1611
    },
    {
      "epoch": 0.1612,
      "grad_norm": 3.251027822494507,
      "learning_rate": 0.0001977342068470228,
      "loss": 6.3473,
      "step": 1612
    },
    {
      "epoch": 0.1613,
      "grad_norm": 5.671850681304932,
      "learning_rate": 0.00019772681235681936,
      "loss": 6.0903,
      "step": 1613
    },
    {
      "epoch": 0.1614,
      "grad_norm": 4.480466842651367,
      "learning_rate": 0.00019771940595890027,
      "loss": 6.9286,
      "step": 1614
    },
    {
      "epoch": 0.1615,
      "grad_norm": 3.3003499507904053,
      "learning_rate": 0.000197711987654168,
      "loss": 5.4616,
      "step": 1615
    },
    {
      "epoch": 0.1616,
      "grad_norm": 3.4173426628112793,
      "learning_rate": 0.0001977045574435264,
      "loss": 6.022,
      "step": 1616
    },
    {
      "epoch": 0.1617,
      "grad_norm": 10.710733413696289,
      "learning_rate": 0.00019769711532788083,
      "loss": 6.1207,
      "step": 1617
    },
    {
      "epoch": 0.1618,
      "grad_norm": 5.66557502746582,
      "learning_rate": 0.0001976896613081381,
      "loss": 5.8925,
      "step": 1618
    },
    {
      "epoch": 0.1619,
      "grad_norm": 3.407240867614746,
      "learning_rate": 0.0001976821953852065,
      "loss": 5.5554,
      "step": 1619
    },
    {
      "epoch": 0.162,
      "grad_norm": 3.6636109352111816,
      "learning_rate": 0.0001976747175599957,
      "loss": 5.9656,
      "step": 1620
    },
    {
      "epoch": 0.1621,
      "grad_norm": 3.0673422813415527,
      "learning_rate": 0.0001976672278334168,
      "loss": 6.0484,
      "step": 1621
    },
    {
      "epoch": 0.1622,
      "grad_norm": 3.081266403198242,
      "learning_rate": 0.00019765972620638248,
      "loss": 6.0532,
      "step": 1622
    },
    {
      "epoch": 0.1623,
      "grad_norm": 9.27580451965332,
      "learning_rate": 0.00019765221267980675,
      "loss": 6.3982,
      "step": 1623
    },
    {
      "epoch": 0.1624,
      "grad_norm": 4.433178424835205,
      "learning_rate": 0.00019764468725460508,
      "loss": 6.6283,
      "step": 1624
    },
    {
      "epoch": 0.1625,
      "grad_norm": 45.95431137084961,
      "learning_rate": 0.00019763714993169452,
      "loss": 7.5439,
      "step": 1625
    },
    {
      "epoch": 0.1626,
      "grad_norm": 3.2304697036743164,
      "learning_rate": 0.00019762960071199333,
      "loss": 6.0515,
      "step": 1626
    },
    {
      "epoch": 0.1627,
      "grad_norm": 9.72169303894043,
      "learning_rate": 0.0001976220395964215,
      "loss": 6.3116,
      "step": 1627
    },
    {
      "epoch": 0.1628,
      "grad_norm": 7.3086981773376465,
      "learning_rate": 0.00019761446658590024,
      "loss": 5.6387,
      "step": 1628
    },
    {
      "epoch": 0.1629,
      "grad_norm": 5.588435173034668,
      "learning_rate": 0.00019760688168135232,
      "loss": 6.5557,
      "step": 1629
    },
    {
      "epoch": 0.163,
      "grad_norm": 2.3366096019744873,
      "learning_rate": 0.00019759928488370193,
      "loss": 5.8439,
      "step": 1630
    },
    {
      "epoch": 0.1631,
      "grad_norm": 3.838951826095581,
      "learning_rate": 0.00019759167619387476,
      "loss": 6.2408,
      "step": 1631
    },
    {
      "epoch": 0.1632,
      "grad_norm": 7.105139255523682,
      "learning_rate": 0.00019758405561279784,
      "loss": 6.6988,
      "step": 1632
    },
    {
      "epoch": 0.1633,
      "grad_norm": 6.190372467041016,
      "learning_rate": 0.00019757642314139977,
      "loss": 6.2683,
      "step": 1633
    },
    {
      "epoch": 0.1634,
      "grad_norm": 7.042876720428467,
      "learning_rate": 0.00019756877878061052,
      "loss": 6.5115,
      "step": 1634
    },
    {
      "epoch": 0.1635,
      "grad_norm": 3.6987123489379883,
      "learning_rate": 0.0001975611225313615,
      "loss": 5.472,
      "step": 1635
    },
    {
      "epoch": 0.1636,
      "grad_norm": 3.925579786300659,
      "learning_rate": 0.00019755345439458565,
      "loss": 6.3838,
      "step": 1636
    },
    {
      "epoch": 0.1637,
      "grad_norm": 6.031131744384766,
      "learning_rate": 0.00019754577437121733,
      "loss": 6.5048,
      "step": 1637
    },
    {
      "epoch": 0.1638,
      "grad_norm": 4.565425395965576,
      "learning_rate": 0.00019753808246219224,
      "loss": 6.4723,
      "step": 1638
    },
    {
      "epoch": 0.1639,
      "grad_norm": 2.740828275680542,
      "learning_rate": 0.00019753037866844771,
      "loss": 6.1886,
      "step": 1639
    },
    {
      "epoch": 0.164,
      "grad_norm": 6.146812915802002,
      "learning_rate": 0.00019752266299092236,
      "loss": 6.2213,
      "step": 1640
    },
    {
      "epoch": 0.1641,
      "grad_norm": 7.787671089172363,
      "learning_rate": 0.00019751493543055632,
      "loss": 5.8215,
      "step": 1641
    },
    {
      "epoch": 0.1642,
      "grad_norm": 3.1370863914489746,
      "learning_rate": 0.00019750719598829122,
      "loss": 5.9606,
      "step": 1642
    },
    {
      "epoch": 0.1643,
      "grad_norm": 7.119648456573486,
      "learning_rate": 0.00019749944466507008,
      "loss": 5.8083,
      "step": 1643
    },
    {
      "epoch": 0.1644,
      "grad_norm": 2.087402582168579,
      "learning_rate": 0.00019749168146183731,
      "loss": 5.7254,
      "step": 1644
    },
    {
      "epoch": 0.1645,
      "grad_norm": 4.9123687744140625,
      "learning_rate": 0.0001974839063795389,
      "loss": 5.8686,
      "step": 1645
    },
    {
      "epoch": 0.1646,
      "grad_norm": 3.666273355484009,
      "learning_rate": 0.0001974761194191222,
      "loss": 6.6011,
      "step": 1646
    },
    {
      "epoch": 0.1647,
      "grad_norm": 10.289949417114258,
      "learning_rate": 0.00019746832058153602,
      "loss": 5.5122,
      "step": 1647
    },
    {
      "epoch": 0.1648,
      "grad_norm": 17.32869529724121,
      "learning_rate": 0.0001974605098677306,
      "loss": 5.3807,
      "step": 1648
    },
    {
      "epoch": 0.1649,
      "grad_norm": 8.930936813354492,
      "learning_rate": 0.00019745268727865774,
      "loss": 6.2029,
      "step": 1649
    },
    {
      "epoch": 0.165,
      "grad_norm": 8.50529956817627,
      "learning_rate": 0.00019744485281527049,
      "loss": 6.6637,
      "step": 1650
    },
    {
      "epoch": 0.1651,
      "grad_norm": 3.8397767543792725,
      "learning_rate": 0.00019743700647852354,
      "loss": 6.0909,
      "step": 1651
    },
    {
      "epoch": 0.1652,
      "grad_norm": 17.237743377685547,
      "learning_rate": 0.00019742914826937288,
      "loss": 6.6919,
      "step": 1652
    },
    {
      "epoch": 0.1653,
      "grad_norm": 8.28303050994873,
      "learning_rate": 0.00019742127818877606,
      "loss": 5.8433,
      "step": 1653
    },
    {
      "epoch": 0.1654,
      "grad_norm": 4.375149250030518,
      "learning_rate": 0.000197413396237692,
      "loss": 5.3857,
      "step": 1654
    },
    {
      "epoch": 0.1655,
      "grad_norm": 5.751855850219727,
      "learning_rate": 0.00019740550241708108,
      "loss": 6.0302,
      "step": 1655
    },
    {
      "epoch": 0.1656,
      "grad_norm": 6.237917423248291,
      "learning_rate": 0.0001973975967279052,
      "loss": 5.9806,
      "step": 1656
    },
    {
      "epoch": 0.1657,
      "grad_norm": 5.004090309143066,
      "learning_rate": 0.0001973896791711275,
      "loss": 6.2828,
      "step": 1657
    },
    {
      "epoch": 0.1658,
      "grad_norm": 4.924671173095703,
      "learning_rate": 0.0001973817497477129,
      "loss": 5.655,
      "step": 1658
    },
    {
      "epoch": 0.1659,
      "grad_norm": 5.3838090896606445,
      "learning_rate": 0.00019737380845862745,
      "loss": 6.0247,
      "step": 1659
    },
    {
      "epoch": 0.166,
      "grad_norm": 6.031036853790283,
      "learning_rate": 0.0001973658553048388,
      "loss": 6.5652,
      "step": 1660
    },
    {
      "epoch": 0.1661,
      "grad_norm": 4.628446578979492,
      "learning_rate": 0.00019735789028731604,
      "loss": 6.6366,
      "step": 1661
    },
    {
      "epoch": 0.1662,
      "grad_norm": 2.7642483711242676,
      "learning_rate": 0.00019734991340702966,
      "loss": 5.7292,
      "step": 1662
    },
    {
      "epoch": 0.1663,
      "grad_norm": 18.35847282409668,
      "learning_rate": 0.00019734192466495162,
      "loss": 6.5445,
      "step": 1663
    },
    {
      "epoch": 0.1664,
      "grad_norm": 5.721039772033691,
      "learning_rate": 0.0001973339240620553,
      "loss": 5.338,
      "step": 1664
    },
    {
      "epoch": 0.1665,
      "grad_norm": 5.393744468688965,
      "learning_rate": 0.0001973259115993156,
      "loss": 7.18,
      "step": 1665
    },
    {
      "epoch": 0.1666,
      "grad_norm": 5.730160713195801,
      "learning_rate": 0.00019731788727770885,
      "loss": 6.245,
      "step": 1666
    },
    {
      "epoch": 0.1667,
      "grad_norm": 4.636256694793701,
      "learning_rate": 0.00019730985109821266,
      "loss": 5.8002,
      "step": 1667
    },
    {
      "epoch": 0.1668,
      "grad_norm": 13.22574520111084,
      "learning_rate": 0.0001973018030618063,
      "loss": 6.3404,
      "step": 1668
    },
    {
      "epoch": 0.1669,
      "grad_norm": 9.577740669250488,
      "learning_rate": 0.0001972937431694704,
      "loss": 6.3045,
      "step": 1669
    },
    {
      "epoch": 0.167,
      "grad_norm": 2.5999999046325684,
      "learning_rate": 0.00019728567142218703,
      "loss": 5.9881,
      "step": 1670
    },
    {
      "epoch": 0.1671,
      "grad_norm": 3.448861598968506,
      "learning_rate": 0.00019727758782093967,
      "loss": 5.6796,
      "step": 1671
    },
    {
      "epoch": 0.1672,
      "grad_norm": 20.695524215698242,
      "learning_rate": 0.00019726949236671332,
      "loss": 6.0615,
      "step": 1672
    },
    {
      "epoch": 0.1673,
      "grad_norm": 2.815276861190796,
      "learning_rate": 0.00019726138506049438,
      "loss": 5.8418,
      "step": 1673
    },
    {
      "epoch": 0.1674,
      "grad_norm": 12.619145393371582,
      "learning_rate": 0.00019725326590327066,
      "loss": 6.1396,
      "step": 1674
    },
    {
      "epoch": 0.1675,
      "grad_norm": 2.595052719116211,
      "learning_rate": 0.00019724513489603155,
      "loss": 6.1225,
      "step": 1675
    },
    {
      "epoch": 0.1676,
      "grad_norm": 4.288460731506348,
      "learning_rate": 0.00019723699203976766,
      "loss": 6.8082,
      "step": 1676
    },
    {
      "epoch": 0.1677,
      "grad_norm": 8.159684181213379,
      "learning_rate": 0.00019722883733547128,
      "loss": 6.6036,
      "step": 1677
    },
    {
      "epoch": 0.1678,
      "grad_norm": 2.652388334274292,
      "learning_rate": 0.00019722067078413599,
      "loss": 6.775,
      "step": 1678
    },
    {
      "epoch": 0.1679,
      "grad_norm": 9.71923828125,
      "learning_rate": 0.00019721249238675688,
      "loss": 5.7811,
      "step": 1679
    },
    {
      "epoch": 0.168,
      "grad_norm": 4.92319393157959,
      "learning_rate": 0.00019720430214433042,
      "loss": 5.9421,
      "step": 1680
    },
    {
      "epoch": 0.1681,
      "grad_norm": 2.759002923965454,
      "learning_rate": 0.00019719610005785465,
      "loss": 5.6899,
      "step": 1681
    },
    {
      "epoch": 0.1682,
      "grad_norm": 2.7991034984588623,
      "learning_rate": 0.00019718788612832887,
      "loss": 6.5069,
      "step": 1682
    },
    {
      "epoch": 0.1683,
      "grad_norm": 7.929409503936768,
      "learning_rate": 0.00019717966035675397,
      "loss": 6.4674,
      "step": 1683
    },
    {
      "epoch": 0.1684,
      "grad_norm": 4.950623989105225,
      "learning_rate": 0.00019717142274413223,
      "loss": 6.346,
      "step": 1684
    },
    {
      "epoch": 0.1685,
      "grad_norm": 2.597456455230713,
      "learning_rate": 0.0001971631732914674,
      "loss": 6.4167,
      "step": 1685
    },
    {
      "epoch": 0.1686,
      "grad_norm": 5.129065990447998,
      "learning_rate": 0.0001971549119997646,
      "loss": 6.7607,
      "step": 1686
    },
    {
      "epoch": 0.1687,
      "grad_norm": 2.733593225479126,
      "learning_rate": 0.00019714663887003054,
      "loss": 6.5966,
      "step": 1687
    },
    {
      "epoch": 0.1688,
      "grad_norm": 3.53849720954895,
      "learning_rate": 0.00019713835390327316,
      "loss": 6.5528,
      "step": 1688
    },
    {
      "epoch": 0.1689,
      "grad_norm": 7.747496604919434,
      "learning_rate": 0.000197130057100502,
      "loss": 6.2795,
      "step": 1689
    },
    {
      "epoch": 0.169,
      "grad_norm": 3.6908981800079346,
      "learning_rate": 0.00019712174846272805,
      "loss": 5.4851,
      "step": 1690
    },
    {
      "epoch": 0.1691,
      "grad_norm": 6.637197017669678,
      "learning_rate": 0.00019711342799096361,
      "loss": 6.3603,
      "step": 1691
    },
    {
      "epoch": 0.1692,
      "grad_norm": 13.245962142944336,
      "learning_rate": 0.00019710509568622258,
      "loss": 6.7264,
      "step": 1692
    },
    {
      "epoch": 0.1693,
      "grad_norm": 3.4783310890197754,
      "learning_rate": 0.00019709675154952017,
      "loss": 5.7559,
      "step": 1693
    },
    {
      "epoch": 0.1694,
      "grad_norm": 3.297973394393921,
      "learning_rate": 0.0001970883955818731,
      "loss": 5.9229,
      "step": 1694
    },
    {
      "epoch": 0.1695,
      "grad_norm": 7.694060325622559,
      "learning_rate": 0.00019708002778429955,
      "loss": 5.9189,
      "step": 1695
    },
    {
      "epoch": 0.1696,
      "grad_norm": 5.675607204437256,
      "learning_rate": 0.00019707164815781908,
      "loss": 6.4181,
      "step": 1696
    },
    {
      "epoch": 0.1697,
      "grad_norm": 7.144732475280762,
      "learning_rate": 0.00019706325670345275,
      "loss": 6.8189,
      "step": 1697
    },
    {
      "epoch": 0.1698,
      "grad_norm": 5.6223464012146,
      "learning_rate": 0.000197054853422223,
      "loss": 6.1878,
      "step": 1698
    },
    {
      "epoch": 0.1699,
      "grad_norm": 4.400817394256592,
      "learning_rate": 0.00019704643831515374,
      "loss": 6.1198,
      "step": 1699
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.0815608501434326,
      "learning_rate": 0.00019703801138327038,
      "loss": 5.9085,
      "step": 1700
    },
    {
      "epoch": 0.1701,
      "grad_norm": 3.826268434524536,
      "learning_rate": 0.00019702957262759965,
      "loss": 6.2175,
      "step": 1701
    },
    {
      "epoch": 0.1702,
      "grad_norm": 4.122454643249512,
      "learning_rate": 0.00019702112204916984,
      "loss": 5.7439,
      "step": 1702
    },
    {
      "epoch": 0.1703,
      "grad_norm": 8.064776420593262,
      "learning_rate": 0.0001970126596490106,
      "loss": 6.436,
      "step": 1703
    },
    {
      "epoch": 0.1704,
      "grad_norm": 6.785343647003174,
      "learning_rate": 0.00019700418542815306,
      "loss": 6.2931,
      "step": 1704
    },
    {
      "epoch": 0.1705,
      "grad_norm": 4.7347517013549805,
      "learning_rate": 0.00019699569938762973,
      "loss": 6.1518,
      "step": 1705
    },
    {
      "epoch": 0.1706,
      "grad_norm": 2.2616629600524902,
      "learning_rate": 0.00019698720152847468,
      "loss": 6.0067,
      "step": 1706
    },
    {
      "epoch": 0.1707,
      "grad_norm": 3.566413164138794,
      "learning_rate": 0.00019697869185172331,
      "loss": 5.7964,
      "step": 1707
    },
    {
      "epoch": 0.1708,
      "grad_norm": 3.678370952606201,
      "learning_rate": 0.00019697017035841252,
      "loss": 5.9383,
      "step": 1708
    },
    {
      "epoch": 0.1709,
      "grad_norm": 2.8759851455688477,
      "learning_rate": 0.0001969616370495806,
      "loss": 6.023,
      "step": 1709
    },
    {
      "epoch": 0.171,
      "grad_norm": 3.447256565093994,
      "learning_rate": 0.00019695309192626734,
      "loss": 5.8743,
      "step": 1710
    },
    {
      "epoch": 0.1711,
      "grad_norm": 14.110333442687988,
      "learning_rate": 0.0001969445349895139,
      "loss": 5.9004,
      "step": 1711
    },
    {
      "epoch": 0.1712,
      "grad_norm": 6.073970794677734,
      "learning_rate": 0.00019693596624036292,
      "loss": 5.9612,
      "step": 1712
    },
    {
      "epoch": 0.1713,
      "grad_norm": 3.668555498123169,
      "learning_rate": 0.00019692738567985853,
      "loss": 5.9416,
      "step": 1713
    },
    {
      "epoch": 0.1714,
      "grad_norm": 5.8636674880981445,
      "learning_rate": 0.0001969187933090462,
      "loss": 5.3956,
      "step": 1714
    },
    {
      "epoch": 0.1715,
      "grad_norm": 5.271744251251221,
      "learning_rate": 0.00019691018912897286,
      "loss": 6.322,
      "step": 1715
    },
    {
      "epoch": 0.1716,
      "grad_norm": 20.9029598236084,
      "learning_rate": 0.00019690157314068696,
      "loss": 5.8993,
      "step": 1716
    },
    {
      "epoch": 0.1717,
      "grad_norm": 2.49660325050354,
      "learning_rate": 0.0001968929453452383,
      "loss": 5.716,
      "step": 1717
    },
    {
      "epoch": 0.1718,
      "grad_norm": 2.7432658672332764,
      "learning_rate": 0.00019688430574367819,
      "loss": 6.255,
      "step": 1718
    },
    {
      "epoch": 0.1719,
      "grad_norm": 3.144899606704712,
      "learning_rate": 0.00019687565433705926,
      "loss": 5.4331,
      "step": 1719
    },
    {
      "epoch": 0.172,
      "grad_norm": 4.954991817474365,
      "learning_rate": 0.00019686699112643572,
      "loss": 6.3986,
      "step": 1720
    },
    {
      "epoch": 0.1721,
      "grad_norm": 2.8653640747070312,
      "learning_rate": 0.0001968583161128631,
      "loss": 5.7573,
      "step": 1721
    },
    {
      "epoch": 0.1722,
      "grad_norm": 11.386573791503906,
      "learning_rate": 0.00019684962929739853,
      "loss": 5.9898,
      "step": 1722
    },
    {
      "epoch": 0.1723,
      "grad_norm": 3.826537847518921,
      "learning_rate": 0.00019684093068110038,
      "loss": 5.524,
      "step": 1723
    },
    {
      "epoch": 0.1724,
      "grad_norm": 4.638415813446045,
      "learning_rate": 0.00019683222026502858,
      "loss": 6.7938,
      "step": 1724
    },
    {
      "epoch": 0.1725,
      "grad_norm": 4.866546630859375,
      "learning_rate": 0.00019682349805024446,
      "loss": 6.5119,
      "step": 1725
    },
    {
      "epoch": 0.1726,
      "grad_norm": 3.6870741844177246,
      "learning_rate": 0.0001968147640378108,
      "loss": 6.0626,
      "step": 1726
    },
    {
      "epoch": 0.1727,
      "grad_norm": 5.90501594543457,
      "learning_rate": 0.00019680601822879182,
      "loss": 6.4801,
      "step": 1727
    },
    {
      "epoch": 0.1728,
      "grad_norm": 2.800372362136841,
      "learning_rate": 0.00019679726062425316,
      "loss": 5.5985,
      "step": 1728
    },
    {
      "epoch": 0.1729,
      "grad_norm": 7.771866321563721,
      "learning_rate": 0.00019678849122526187,
      "loss": 8.2333,
      "step": 1729
    },
    {
      "epoch": 0.173,
      "grad_norm": 3.258699893951416,
      "learning_rate": 0.00019677971003288655,
      "loss": 5.7101,
      "step": 1730
    },
    {
      "epoch": 0.1731,
      "grad_norm": 7.840193271636963,
      "learning_rate": 0.00019677091704819715,
      "loss": 5.984,
      "step": 1731
    },
    {
      "epoch": 0.1732,
      "grad_norm": 3.136154890060425,
      "learning_rate": 0.000196762112272265,
      "loss": 6.221,
      "step": 1732
    },
    {
      "epoch": 0.1733,
      "grad_norm": 3.183802604675293,
      "learning_rate": 0.00019675329570616298,
      "loss": 5.9113,
      "step": 1733
    },
    {
      "epoch": 0.1734,
      "grad_norm": 6.5895185470581055,
      "learning_rate": 0.0001967444673509654,
      "loss": 6.4387,
      "step": 1734
    },
    {
      "epoch": 0.1735,
      "grad_norm": 2.8669931888580322,
      "learning_rate": 0.00019673562720774792,
      "loss": 5.7965,
      "step": 1735
    },
    {
      "epoch": 0.1736,
      "grad_norm": 2.2173919677734375,
      "learning_rate": 0.0001967267752775877,
      "loss": 5.8811,
      "step": 1736
    },
    {
      "epoch": 0.1737,
      "grad_norm": 6.472240447998047,
      "learning_rate": 0.0001967179115615633,
      "loss": 5.9218,
      "step": 1737
    },
    {
      "epoch": 0.1738,
      "grad_norm": 9.446806907653809,
      "learning_rate": 0.00019670903606075474,
      "loss": 5.7381,
      "step": 1738
    },
    {
      "epoch": 0.1739,
      "grad_norm": 17.03736686706543,
      "learning_rate": 0.00019670014877624353,
      "loss": 5.7711,
      "step": 1739
    },
    {
      "epoch": 0.174,
      "grad_norm": 9.264403343200684,
      "learning_rate": 0.00019669124970911247,
      "loss": 7.6732,
      "step": 1740
    },
    {
      "epoch": 0.1741,
      "grad_norm": 11.823596954345703,
      "learning_rate": 0.00019668233886044597,
      "loss": 6.1182,
      "step": 1741
    },
    {
      "epoch": 0.1742,
      "grad_norm": 3.84055233001709,
      "learning_rate": 0.0001966734162313297,
      "loss": 6.133,
      "step": 1742
    },
    {
      "epoch": 0.1743,
      "grad_norm": 2.978358507156372,
      "learning_rate": 0.00019666448182285094,
      "loss": 5.8545,
      "step": 1743
    },
    {
      "epoch": 0.1744,
      "grad_norm": 4.596590042114258,
      "learning_rate": 0.00019665553563609825,
      "loss": 6.1756,
      "step": 1744
    },
    {
      "epoch": 0.1745,
      "grad_norm": 2.488898992538452,
      "learning_rate": 0.00019664657767216176,
      "loss": 6.0226,
      "step": 1745
    },
    {
      "epoch": 0.1746,
      "grad_norm": 12.84595012664795,
      "learning_rate": 0.00019663760793213296,
      "loss": 5.8614,
      "step": 1746
    },
    {
      "epoch": 0.1747,
      "grad_norm": 5.358705520629883,
      "learning_rate": 0.0001966286264171047,
      "loss": 6.2001,
      "step": 1747
    },
    {
      "epoch": 0.1748,
      "grad_norm": 4.480710029602051,
      "learning_rate": 0.00019661963312817148,
      "loss": 5.6806,
      "step": 1748
    },
    {
      "epoch": 0.1749,
      "grad_norm": 2.479482412338257,
      "learning_rate": 0.00019661062806642903,
      "loss": 5.4744,
      "step": 1749
    },
    {
      "epoch": 0.175,
      "grad_norm": 2.903468370437622,
      "learning_rate": 0.00019660161123297458,
      "loss": 6.4886,
      "step": 1750
    },
    {
      "epoch": 0.1751,
      "grad_norm": 3.1048455238342285,
      "learning_rate": 0.00019659258262890683,
      "loss": 6.2749,
      "step": 1751
    },
    {
      "epoch": 0.1752,
      "grad_norm": 5.646725654602051,
      "learning_rate": 0.00019658354225532589,
      "loss": 5.7987,
      "step": 1752
    },
    {
      "epoch": 0.1753,
      "grad_norm": 6.540036201477051,
      "learning_rate": 0.00019657449011333328,
      "loss": 6.9478,
      "step": 1753
    },
    {
      "epoch": 0.1754,
      "grad_norm": 5.778735160827637,
      "learning_rate": 0.00019656542620403203,
      "loss": 6.3518,
      "step": 1754
    },
    {
      "epoch": 0.1755,
      "grad_norm": 2.989036798477173,
      "learning_rate": 0.00019655635052852647,
      "loss": 5.8749,
      "step": 1755
    },
    {
      "epoch": 0.1756,
      "grad_norm": 4.034169673919678,
      "learning_rate": 0.0001965472630879225,
      "loss": 6.1252,
      "step": 1756
    },
    {
      "epoch": 0.1757,
      "grad_norm": 4.534963130950928,
      "learning_rate": 0.0001965381638833274,
      "loss": 5.4714,
      "step": 1757
    },
    {
      "epoch": 0.1758,
      "grad_norm": 5.523154258728027,
      "learning_rate": 0.00019652905291584984,
      "loss": 5.9617,
      "step": 1758
    },
    {
      "epoch": 0.1759,
      "grad_norm": 5.336920261383057,
      "learning_rate": 0.0001965199301866,
      "loss": 6.2244,
      "step": 1759
    },
    {
      "epoch": 0.176,
      "grad_norm": 4.042906761169434,
      "learning_rate": 0.00019651079569668945,
      "loss": 6.5173,
      "step": 1760
    },
    {
      "epoch": 0.1761,
      "grad_norm": 9.107699394226074,
      "learning_rate": 0.00019650164944723115,
      "loss": 5.884,
      "step": 1761
    },
    {
      "epoch": 0.1762,
      "grad_norm": 24.44321632385254,
      "learning_rate": 0.00019649249143933962,
      "loss": 6.445,
      "step": 1762
    },
    {
      "epoch": 0.1763,
      "grad_norm": 11.313589096069336,
      "learning_rate": 0.00019648332167413067,
      "loss": 5.3304,
      "step": 1763
    },
    {
      "epoch": 0.1764,
      "grad_norm": 12.472349166870117,
      "learning_rate": 0.0001964741401527217,
      "loss": 6.2141,
      "step": 1764
    },
    {
      "epoch": 0.1765,
      "grad_norm": 8.564106941223145,
      "learning_rate": 0.00019646494687623135,
      "loss": 6.4469,
      "step": 1765
    },
    {
      "epoch": 0.1766,
      "grad_norm": 18.460866928100586,
      "learning_rate": 0.00019645574184577982,
      "loss": 6.0025,
      "step": 1766
    },
    {
      "epoch": 0.1767,
      "grad_norm": 2.190577983856201,
      "learning_rate": 0.00019644652506248874,
      "loss": 6.1895,
      "step": 1767
    },
    {
      "epoch": 0.1768,
      "grad_norm": 3.651763439178467,
      "learning_rate": 0.00019643729652748113,
      "loss": 5.7478,
      "step": 1768
    },
    {
      "epoch": 0.1769,
      "grad_norm": 3.248133897781372,
      "learning_rate": 0.00019642805624188147,
      "loss": 5.8585,
      "step": 1769
    },
    {
      "epoch": 0.177,
      "grad_norm": 2.428518533706665,
      "learning_rate": 0.00019641880420681566,
      "loss": 5.4357,
      "step": 1770
    },
    {
      "epoch": 0.1771,
      "grad_norm": 5.414050102233887,
      "learning_rate": 0.00019640954042341103,
      "loss": 6.0659,
      "step": 1771
    },
    {
      "epoch": 0.1772,
      "grad_norm": 3.235907793045044,
      "learning_rate": 0.0001964002648927963,
      "loss": 5.5982,
      "step": 1772
    },
    {
      "epoch": 0.1773,
      "grad_norm": 5.609501361846924,
      "learning_rate": 0.00019639097761610174,
      "loss": 5.896,
      "step": 1773
    },
    {
      "epoch": 0.1774,
      "grad_norm": 14.26980209350586,
      "learning_rate": 0.00019638167859445895,
      "loss": 5.8491,
      "step": 1774
    },
    {
      "epoch": 0.1775,
      "grad_norm": 11.399382591247559,
      "learning_rate": 0.000196372367829001,
      "loss": 6.6284,
      "step": 1775
    },
    {
      "epoch": 0.1776,
      "grad_norm": 3.82888126373291,
      "learning_rate": 0.0001963630453208623,
      "loss": 6.142,
      "step": 1776
    },
    {
      "epoch": 0.1777,
      "grad_norm": 6.922585487365723,
      "learning_rate": 0.00019635371107117888,
      "loss": 5.9771,
      "step": 1777
    },
    {
      "epoch": 0.1778,
      "grad_norm": 9.215946197509766,
      "learning_rate": 0.000196344365081088,
      "loss": 5.7482,
      "step": 1778
    },
    {
      "epoch": 0.1779,
      "grad_norm": 4.526270389556885,
      "learning_rate": 0.0001963350073517285,
      "loss": 6.0522,
      "step": 1779
    },
    {
      "epoch": 0.178,
      "grad_norm": 3.323275089263916,
      "learning_rate": 0.00019632563788424053,
      "loss": 6.0508,
      "step": 1780
    },
    {
      "epoch": 0.1781,
      "grad_norm": 3.143465518951416,
      "learning_rate": 0.00019631625667976583,
      "loss": 5.954,
      "step": 1781
    },
    {
      "epoch": 0.1782,
      "grad_norm": 2.9758384227752686,
      "learning_rate": 0.00019630686373944738,
      "loss": 6.1417,
      "step": 1782
    },
    {
      "epoch": 0.1783,
      "grad_norm": 12.044028282165527,
      "learning_rate": 0.0001962974590644297,
      "loss": 5.9562,
      "step": 1783
    },
    {
      "epoch": 0.1784,
      "grad_norm": 2.863321304321289,
      "learning_rate": 0.00019628804265585877,
      "loss": 5.8238,
      "step": 1784
    },
    {
      "epoch": 0.1785,
      "grad_norm": 4.464944362640381,
      "learning_rate": 0.00019627861451488189,
      "loss": 6.0976,
      "step": 1785
    },
    {
      "epoch": 0.1786,
      "grad_norm": 2.5082104206085205,
      "learning_rate": 0.0001962691746426479,
      "loss": 5.6347,
      "step": 1786
    },
    {
      "epoch": 0.1787,
      "grad_norm": 2.531252145767212,
      "learning_rate": 0.00019625972304030697,
      "loss": 5.8885,
      "step": 1787
    },
    {
      "epoch": 0.1788,
      "grad_norm": 4.065119743347168,
      "learning_rate": 0.00019625025970901078,
      "loss": 5.9265,
      "step": 1788
    },
    {
      "epoch": 0.1789,
      "grad_norm": 2.437279224395752,
      "learning_rate": 0.0001962407846499124,
      "loss": 5.5341,
      "step": 1789
    },
    {
      "epoch": 0.179,
      "grad_norm": 2.8292124271392822,
      "learning_rate": 0.00019623129786416635,
      "loss": 6.3963,
      "step": 1790
    },
    {
      "epoch": 0.1791,
      "grad_norm": 4.567358016967773,
      "learning_rate": 0.00019622179935292855,
      "loss": 6.088,
      "step": 1791
    },
    {
      "epoch": 0.1792,
      "grad_norm": 2.1955432891845703,
      "learning_rate": 0.00019621228911735636,
      "loss": 5.6822,
      "step": 1792
    },
    {
      "epoch": 0.1793,
      "grad_norm": 14.695341110229492,
      "learning_rate": 0.0001962027671586086,
      "loss": 6.2635,
      "step": 1793
    },
    {
      "epoch": 0.1794,
      "grad_norm": 18.026525497436523,
      "learning_rate": 0.00019619323347784548,
      "loss": 6.4958,
      "step": 1794
    },
    {
      "epoch": 0.1795,
      "grad_norm": 23.532636642456055,
      "learning_rate": 0.00019618368807622862,
      "loss": 5.6758,
      "step": 1795
    },
    {
      "epoch": 0.1796,
      "grad_norm": 16.613767623901367,
      "learning_rate": 0.00019617413095492114,
      "loss": 5.408,
      "step": 1796
    },
    {
      "epoch": 0.1797,
      "grad_norm": 29.53449821472168,
      "learning_rate": 0.00019616456211508752,
      "loss": 6.0035,
      "step": 1797
    },
    {
      "epoch": 0.1798,
      "grad_norm": 4.6154069900512695,
      "learning_rate": 0.0001961549815578937,
      "loss": 6.075,
      "step": 1798
    },
    {
      "epoch": 0.1799,
      "grad_norm": 4.513603210449219,
      "learning_rate": 0.0001961453892845071,
      "loss": 5.9912,
      "step": 1799
    },
    {
      "epoch": 0.18,
      "grad_norm": 31.06067657470703,
      "learning_rate": 0.0001961357852960964,
      "loss": 6.3313,
      "step": 1800
    },
    {
      "epoch": 0.1801,
      "grad_norm": 8.058358192443848,
      "learning_rate": 0.0001961261695938319,
      "loss": 6.0069,
      "step": 1801
    },
    {
      "epoch": 0.1802,
      "grad_norm": 20.240432739257812,
      "learning_rate": 0.0001961165421788852,
      "loss": 5.9012,
      "step": 1802
    },
    {
      "epoch": 0.1803,
      "grad_norm": 3.2563891410827637,
      "learning_rate": 0.0001961069030524294,
      "loss": 5.7887,
      "step": 1803
    },
    {
      "epoch": 0.1804,
      "grad_norm": 3.9629814624786377,
      "learning_rate": 0.00019609725221563897,
      "loss": 6.5777,
      "step": 1804
    },
    {
      "epoch": 0.1805,
      "grad_norm": 8.04670238494873,
      "learning_rate": 0.00019608758966968988,
      "loss": 5.789,
      "step": 1805
    },
    {
      "epoch": 0.1806,
      "grad_norm": 6.525185585021973,
      "learning_rate": 0.00019607791541575943,
      "loss": 6.3451,
      "step": 1806
    },
    {
      "epoch": 0.1807,
      "grad_norm": 6.386545658111572,
      "learning_rate": 0.0001960682294550264,
      "loss": 6.0976,
      "step": 1807
    },
    {
      "epoch": 0.1808,
      "grad_norm": 3.8811933994293213,
      "learning_rate": 0.00019605853178867105,
      "loss": 5.6094,
      "step": 1808
    },
    {
      "epoch": 0.1809,
      "grad_norm": 2.493893623352051,
      "learning_rate": 0.00019604882241787498,
      "loss": 5.5087,
      "step": 1809
    },
    {
      "epoch": 0.181,
      "grad_norm": 4.572642803192139,
      "learning_rate": 0.00019603910134382123,
      "loss": 6.3121,
      "step": 1810
    },
    {
      "epoch": 0.1811,
      "grad_norm": 4.597893714904785,
      "learning_rate": 0.0001960293685676943,
      "loss": 6.3271,
      "step": 1811
    },
    {
      "epoch": 0.1812,
      "grad_norm": 2.8160006999969482,
      "learning_rate": 0.00019601962409068012,
      "loss": 6.0575,
      "step": 1812
    },
    {
      "epoch": 0.1813,
      "grad_norm": 16.58269500732422,
      "learning_rate": 0.000196009867913966,
      "loss": 6.4086,
      "step": 1813
    },
    {
      "epoch": 0.1814,
      "grad_norm": 4.625454902648926,
      "learning_rate": 0.00019600010003874069,
      "loss": 5.6763,
      "step": 1814
    },
    {
      "epoch": 0.1815,
      "grad_norm": 6.828894138336182,
      "learning_rate": 0.00019599032046619438,
      "loss": 5.6368,
      "step": 1815
    },
    {
      "epoch": 0.1816,
      "grad_norm": 2.510169744491577,
      "learning_rate": 0.0001959805291975187,
      "loss": 5.719,
      "step": 1816
    },
    {
      "epoch": 0.1817,
      "grad_norm": 6.222814559936523,
      "learning_rate": 0.00019597072623390668,
      "loss": 6.741,
      "step": 1817
    },
    {
      "epoch": 0.1818,
      "grad_norm": 2.379448890686035,
      "learning_rate": 0.00019596091157655278,
      "loss": 5.8019,
      "step": 1818
    },
    {
      "epoch": 0.1819,
      "grad_norm": 2.8843801021575928,
      "learning_rate": 0.0001959510852266529,
      "loss": 5.8781,
      "step": 1819
    },
    {
      "epoch": 0.182,
      "grad_norm": 2.287376880645752,
      "learning_rate": 0.0001959412471854043,
      "loss": 5.8525,
      "step": 1820
    },
    {
      "epoch": 0.1821,
      "grad_norm": 3.5780067443847656,
      "learning_rate": 0.00019593139745400576,
      "loss": 5.6994,
      "step": 1821
    },
    {
      "epoch": 0.1822,
      "grad_norm": 4.090850830078125,
      "learning_rate": 0.00019592153603365743,
      "loss": 5.7429,
      "step": 1822
    },
    {
      "epoch": 0.1823,
      "grad_norm": 12.353042602539062,
      "learning_rate": 0.0001959116629255609,
      "loss": 5.825,
      "step": 1823
    },
    {
      "epoch": 0.1824,
      "grad_norm": 4.439487457275391,
      "learning_rate": 0.0001959017781309192,
      "loss": 5.3857,
      "step": 1824
    },
    {
      "epoch": 0.1825,
      "grad_norm": 3.385716199874878,
      "learning_rate": 0.0001958918816509367,
      "loss": 6.1557,
      "step": 1825
    },
    {
      "epoch": 0.1826,
      "grad_norm": 4.956521034240723,
      "learning_rate": 0.0001958819734868193,
      "loss": 6.3424,
      "step": 1826
    },
    {
      "epoch": 0.1827,
      "grad_norm": 3.654254198074341,
      "learning_rate": 0.00019587205363977427,
      "loss": 5.5613,
      "step": 1827
    },
    {
      "epoch": 0.1828,
      "grad_norm": 5.679259777069092,
      "learning_rate": 0.00019586212211101037,
      "loss": 6.0734,
      "step": 1828
    },
    {
      "epoch": 0.1829,
      "grad_norm": 6.78184175491333,
      "learning_rate": 0.0001958521789017376,
      "loss": 4.9247,
      "step": 1829
    },
    {
      "epoch": 0.183,
      "grad_norm": 26.773420333862305,
      "learning_rate": 0.0001958422240131676,
      "loss": 7.6036,
      "step": 1830
    },
    {
      "epoch": 0.1831,
      "grad_norm": 5.919186592102051,
      "learning_rate": 0.00019583225744651333,
      "loss": 5.7175,
      "step": 1831
    },
    {
      "epoch": 0.1832,
      "grad_norm": 4.582914352416992,
      "learning_rate": 0.00019582227920298916,
      "loss": 6.1913,
      "step": 1832
    },
    {
      "epoch": 0.1833,
      "grad_norm": 5.624858379364014,
      "learning_rate": 0.00019581228928381098,
      "loss": 6.1049,
      "step": 1833
    },
    {
      "epoch": 0.1834,
      "grad_norm": 6.440640449523926,
      "learning_rate": 0.00019580228769019593,
      "loss": 5.6119,
      "step": 1834
    },
    {
      "epoch": 0.1835,
      "grad_norm": 5.511519908905029,
      "learning_rate": 0.00019579227442336278,
      "loss": 6.6863,
      "step": 1835
    },
    {
      "epoch": 0.1836,
      "grad_norm": 4.006153583526611,
      "learning_rate": 0.0001957822494845315,
      "loss": 5.8042,
      "step": 1836
    },
    {
      "epoch": 0.1837,
      "grad_norm": 4.034609317779541,
      "learning_rate": 0.00019577221287492367,
      "loss": 6.7431,
      "step": 1837
    },
    {
      "epoch": 0.1838,
      "grad_norm": 6.139560222625732,
      "learning_rate": 0.00019576216459576222,
      "loss": 6.6845,
      "step": 1838
    },
    {
      "epoch": 0.1839,
      "grad_norm": 3.8226125240325928,
      "learning_rate": 0.00019575210464827149,
      "loss": 6.0432,
      "step": 1839
    },
    {
      "epoch": 0.184,
      "grad_norm": 2.6244733333587646,
      "learning_rate": 0.00019574203303367727,
      "loss": 5.6519,
      "step": 1840
    },
    {
      "epoch": 0.1841,
      "grad_norm": 5.073551654815674,
      "learning_rate": 0.00019573194975320673,
      "loss": 6.6092,
      "step": 1841
    },
    {
      "epoch": 0.1842,
      "grad_norm": 4.095277786254883,
      "learning_rate": 0.00019572185480808848,
      "loss": 5.5561,
      "step": 1842
    },
    {
      "epoch": 0.1843,
      "grad_norm": 2.7274575233459473,
      "learning_rate": 0.00019571174819955263,
      "loss": 5.9821,
      "step": 1843
    },
    {
      "epoch": 0.1844,
      "grad_norm": 2.5123982429504395,
      "learning_rate": 0.00019570162992883054,
      "loss": 5.1522,
      "step": 1844
    },
    {
      "epoch": 0.1845,
      "grad_norm": 4.443167209625244,
      "learning_rate": 0.00019569149999715515,
      "loss": 5.4506,
      "step": 1845
    },
    {
      "epoch": 0.1846,
      "grad_norm": 2.7756824493408203,
      "learning_rate": 0.00019568135840576076,
      "loss": 5.7416,
      "step": 1846
    },
    {
      "epoch": 0.1847,
      "grad_norm": 8.501652717590332,
      "learning_rate": 0.00019567120515588308,
      "loss": 5.016,
      "step": 1847
    },
    {
      "epoch": 0.1848,
      "grad_norm": 3.3224923610687256,
      "learning_rate": 0.0001956610402487592,
      "loss": 6.5004,
      "step": 1848
    },
    {
      "epoch": 0.1849,
      "grad_norm": 4.069192886352539,
      "learning_rate": 0.0001956508636856278,
      "loss": 5.9475,
      "step": 1849
    },
    {
      "epoch": 0.185,
      "grad_norm": 3.2450039386749268,
      "learning_rate": 0.00019564067546772878,
      "loss": 5.8401,
      "step": 1850
    },
    {
      "epoch": 0.1851,
      "grad_norm": 5.402753829956055,
      "learning_rate": 0.00019563047559630357,
      "loss": 6.1254,
      "step": 1851
    },
    {
      "epoch": 0.1852,
      "grad_norm": 5.21713399887085,
      "learning_rate": 0.00019562026407259495,
      "loss": 7.2446,
      "step": 1852
    },
    {
      "epoch": 0.1853,
      "grad_norm": 4.467038631439209,
      "learning_rate": 0.00019561004089784723,
      "loss": 6.0019,
      "step": 1853
    },
    {
      "epoch": 0.1854,
      "grad_norm": 4.610062599182129,
      "learning_rate": 0.00019559980607330605,
      "loss": 5.7712,
      "step": 1854
    },
    {
      "epoch": 0.1855,
      "grad_norm": 16.567079544067383,
      "learning_rate": 0.00019558955960021849,
      "loss": 5.7504,
      "step": 1855
    },
    {
      "epoch": 0.1856,
      "grad_norm": 3.0925936698913574,
      "learning_rate": 0.00019557930147983302,
      "loss": 5.8998,
      "step": 1856
    },
    {
      "epoch": 0.1857,
      "grad_norm": 5.134648323059082,
      "learning_rate": 0.00019556903171339963,
      "loss": 5.5647,
      "step": 1857
    },
    {
      "epoch": 0.1858,
      "grad_norm": 2.820716142654419,
      "learning_rate": 0.00019555875030216954,
      "loss": 5.6861,
      "step": 1858
    },
    {
      "epoch": 0.1859,
      "grad_norm": 4.506610870361328,
      "learning_rate": 0.00019554845724739566,
      "loss": 6.4361,
      "step": 1859
    },
    {
      "epoch": 0.186,
      "grad_norm": 3.9133293628692627,
      "learning_rate": 0.00019553815255033205,
      "loss": 5.8522,
      "step": 1860
    },
    {
      "epoch": 0.1861,
      "grad_norm": 3.5506112575531006,
      "learning_rate": 0.00019552783621223436,
      "loss": 5.4634,
      "step": 1861
    },
    {
      "epoch": 0.1862,
      "grad_norm": 5.30897855758667,
      "learning_rate": 0.0001955175082343596,
      "loss": 5.3373,
      "step": 1862
    },
    {
      "epoch": 0.1863,
      "grad_norm": 4.70376443862915,
      "learning_rate": 0.00019550716861796623,
      "loss": 6.2058,
      "step": 1863
    },
    {
      "epoch": 0.1864,
      "grad_norm": 4.765412330627441,
      "learning_rate": 0.00019549681736431404,
      "loss": 5.6936,
      "step": 1864
    },
    {
      "epoch": 0.1865,
      "grad_norm": 4.844527721405029,
      "learning_rate": 0.00019548645447466431,
      "loss": 5.9927,
      "step": 1865
    },
    {
      "epoch": 0.1866,
      "grad_norm": 4.188244819641113,
      "learning_rate": 0.00019547607995027978,
      "loss": 6.0007,
      "step": 1866
    },
    {
      "epoch": 0.1867,
      "grad_norm": 23.331241607666016,
      "learning_rate": 0.00019546569379242444,
      "loss": 5.948,
      "step": 1867
    },
    {
      "epoch": 0.1868,
      "grad_norm": 4.135862350463867,
      "learning_rate": 0.00019545529600236398,
      "loss": 6.5972,
      "step": 1868
    },
    {
      "epoch": 0.1869,
      "grad_norm": 4.464096546173096,
      "learning_rate": 0.00019544488658136523,
      "loss": 6.0265,
      "step": 1869
    },
    {
      "epoch": 0.187,
      "grad_norm": 2.981785297393799,
      "learning_rate": 0.0001954344655306965,
      "loss": 5.9113,
      "step": 1870
    },
    {
      "epoch": 0.1871,
      "grad_norm": 42.67870330810547,
      "learning_rate": 0.0001954240328516277,
      "loss": 7.5355,
      "step": 1871
    },
    {
      "epoch": 0.1872,
      "grad_norm": 2.447568893432617,
      "learning_rate": 0.00019541358854542991,
      "loss": 5.9509,
      "step": 1872
    },
    {
      "epoch": 0.1873,
      "grad_norm": 4.090131759643555,
      "learning_rate": 0.0001954031326133758,
      "loss": 6.5809,
      "step": 1873
    },
    {
      "epoch": 0.1874,
      "grad_norm": 2.5728445053100586,
      "learning_rate": 0.00019539266505673938,
      "loss": 5.4782,
      "step": 1874
    },
    {
      "epoch": 0.1875,
      "grad_norm": 13.447710990905762,
      "learning_rate": 0.00019538218587679605,
      "loss": 6.8383,
      "step": 1875
    },
    {
      "epoch": 0.1876,
      "grad_norm": 4.688170433044434,
      "learning_rate": 0.0001953716950748227,
      "loss": 6.2034,
      "step": 1876
    },
    {
      "epoch": 0.1877,
      "grad_norm": 5.87640380859375,
      "learning_rate": 0.0001953611926520976,
      "loss": 6.2476,
      "step": 1877
    },
    {
      "epoch": 0.1878,
      "grad_norm": 3.790982484817505,
      "learning_rate": 0.00019535067860990046,
      "loss": 5.5247,
      "step": 1878
    },
    {
      "epoch": 0.1879,
      "grad_norm": 3.4364116191864014,
      "learning_rate": 0.00019534015294951233,
      "loss": 5.5599,
      "step": 1879
    },
    {
      "epoch": 0.188,
      "grad_norm": 7.848231315612793,
      "learning_rate": 0.00019532961567221576,
      "loss": 5.7862,
      "step": 1880
    },
    {
      "epoch": 0.1881,
      "grad_norm": 2.5485188961029053,
      "learning_rate": 0.0001953190667792947,
      "loss": 5.5694,
      "step": 1881
    },
    {
      "epoch": 0.1882,
      "grad_norm": 6.308555603027344,
      "learning_rate": 0.0001953085062720345,
      "loss": 5.358,
      "step": 1882
    },
    {
      "epoch": 0.1883,
      "grad_norm": 3.995948553085327,
      "learning_rate": 0.00019529793415172192,
      "loss": 6.2567,
      "step": 1883
    },
    {
      "epoch": 0.1884,
      "grad_norm": 4.8857293128967285,
      "learning_rate": 0.00019528735041964509,
      "loss": 5.8876,
      "step": 1884
    },
    {
      "epoch": 0.1885,
      "grad_norm": 5.78946590423584,
      "learning_rate": 0.00019527675507709366,
      "loss": 6.6056,
      "step": 1885
    },
    {
      "epoch": 0.1886,
      "grad_norm": 27.603425979614258,
      "learning_rate": 0.00019526614812535864,
      "loss": 6.4607,
      "step": 1886
    },
    {
      "epoch": 0.1887,
      "grad_norm": 3.068774700164795,
      "learning_rate": 0.00019525552956573244,
      "loss": 5.7437,
      "step": 1887
    },
    {
      "epoch": 0.1888,
      "grad_norm": 2.437843084335327,
      "learning_rate": 0.0001952448993995089,
      "loss": 6.0742,
      "step": 1888
    },
    {
      "epoch": 0.1889,
      "grad_norm": 4.153500080108643,
      "learning_rate": 0.00019523425762798329,
      "loss": 5.8377,
      "step": 1889
    },
    {
      "epoch": 0.189,
      "grad_norm": 4.070453643798828,
      "learning_rate": 0.00019522360425245226,
      "loss": 5.315,
      "step": 1890
    },
    {
      "epoch": 0.1891,
      "grad_norm": 5.602750301361084,
      "learning_rate": 0.00019521293927421388,
      "loss": 5.8345,
      "step": 1891
    },
    {
      "epoch": 0.1892,
      "grad_norm": 3.173646926879883,
      "learning_rate": 0.00019520226269456768,
      "loss": 6.4387,
      "step": 1892
    },
    {
      "epoch": 0.1893,
      "grad_norm": 2.4037668704986572,
      "learning_rate": 0.00019519157451481454,
      "loss": 5.7975,
      "step": 1893
    },
    {
      "epoch": 0.1894,
      "grad_norm": 7.328547954559326,
      "learning_rate": 0.0001951808747362568,
      "loss": 5.7026,
      "step": 1894
    },
    {
      "epoch": 0.1895,
      "grad_norm": 8.08935546875,
      "learning_rate": 0.0001951701633601982,
      "loss": 5.9683,
      "step": 1895
    },
    {
      "epoch": 0.1896,
      "grad_norm": 18.838279724121094,
      "learning_rate": 0.00019515944038794384,
      "loss": 6.1567,
      "step": 1896
    },
    {
      "epoch": 0.1897,
      "grad_norm": 5.898171901702881,
      "learning_rate": 0.00019514870582080032,
      "loss": 6.1603,
      "step": 1897
    },
    {
      "epoch": 0.1898,
      "grad_norm": 15.051395416259766,
      "learning_rate": 0.00019513795966007562,
      "loss": 6.0326,
      "step": 1898
    },
    {
      "epoch": 0.1899,
      "grad_norm": 2.8077144622802734,
      "learning_rate": 0.00019512720190707913,
      "loss": 5.9754,
      "step": 1899
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.029720306396484,
      "learning_rate": 0.00019511643256312164,
      "loss": 6.3828,
      "step": 1900
    },
    {
      "epoch": 0.1901,
      "grad_norm": 3.814983606338501,
      "learning_rate": 0.00019510565162951537,
      "loss": 6.2922,
      "step": 1901
    },
    {
      "epoch": 0.1902,
      "grad_norm": 16.990558624267578,
      "learning_rate": 0.0001950948591075739,
      "loss": 6.3066,
      "step": 1902
    },
    {
      "epoch": 0.1903,
      "grad_norm": 4.722686767578125,
      "learning_rate": 0.00019508405499861232,
      "loss": 5.9044,
      "step": 1903
    },
    {
      "epoch": 0.1904,
      "grad_norm": 2.682317018508911,
      "learning_rate": 0.00019507323930394708,
      "loss": 5.6031,
      "step": 1904
    },
    {
      "epoch": 0.1905,
      "grad_norm": 4.009721755981445,
      "learning_rate": 0.00019506241202489602,
      "loss": 6.1454,
      "step": 1905
    },
    {
      "epoch": 0.1906,
      "grad_norm": 9.497082710266113,
      "learning_rate": 0.00019505157316277837,
      "loss": 7.712,
      "step": 1906
    },
    {
      "epoch": 0.1907,
      "grad_norm": 3.8237345218658447,
      "learning_rate": 0.00019504072271891488,
      "loss": 5.6381,
      "step": 1907
    },
    {
      "epoch": 0.1908,
      "grad_norm": 5.120648384094238,
      "learning_rate": 0.00019502986069462762,
      "loss": 5.4901,
      "step": 1908
    },
    {
      "epoch": 0.1909,
      "grad_norm": 3.080514669418335,
      "learning_rate": 0.00019501898709124008,
      "loss": 5.643,
      "step": 1909
    },
    {
      "epoch": 0.191,
      "grad_norm": 2.9963667392730713,
      "learning_rate": 0.00019500810191007718,
      "loss": 5.7156,
      "step": 1910
    },
    {
      "epoch": 0.1911,
      "grad_norm": 3.646616220474243,
      "learning_rate": 0.00019499720515246525,
      "loss": 5.6862,
      "step": 1911
    },
    {
      "epoch": 0.1912,
      "grad_norm": 2.293877363204956,
      "learning_rate": 0.00019498629681973206,
      "loss": 5.5351,
      "step": 1912
    },
    {
      "epoch": 0.1913,
      "grad_norm": 4.163327217102051,
      "learning_rate": 0.00019497537691320668,
      "loss": 5.7016,
      "step": 1913
    },
    {
      "epoch": 0.1914,
      "grad_norm": 39.91741943359375,
      "learning_rate": 0.00019496444543421975,
      "loss": 7.064,
      "step": 1914
    },
    {
      "epoch": 0.1915,
      "grad_norm": 7.2125420570373535,
      "learning_rate": 0.0001949535023841032,
      "loss": 5.6521,
      "step": 1915
    },
    {
      "epoch": 0.1916,
      "grad_norm": 3.233781337738037,
      "learning_rate": 0.0001949425477641904,
      "loss": 5.5141,
      "step": 1916
    },
    {
      "epoch": 0.1917,
      "grad_norm": 4.0212812423706055,
      "learning_rate": 0.00019493158157581615,
      "loss": 5.6483,
      "step": 1917
    },
    {
      "epoch": 0.1918,
      "grad_norm": 5.491209030151367,
      "learning_rate": 0.00019492060382031662,
      "loss": 5.7022,
      "step": 1918
    },
    {
      "epoch": 0.1919,
      "grad_norm": 10.201300621032715,
      "learning_rate": 0.00019490961449902946,
      "loss": 5.6066,
      "step": 1919
    },
    {
      "epoch": 0.192,
      "grad_norm": 5.635716915130615,
      "learning_rate": 0.00019489861361329366,
      "loss": 5.7214,
      "step": 1920
    },
    {
      "epoch": 0.1921,
      "grad_norm": 9.307692527770996,
      "learning_rate": 0.00019488760116444966,
      "loss": 5.4784,
      "step": 1921
    },
    {
      "epoch": 0.1922,
      "grad_norm": 3.8241498470306396,
      "learning_rate": 0.00019487657715383926,
      "loss": 5.9666,
      "step": 1922
    },
    {
      "epoch": 0.1923,
      "grad_norm": 2.707608222961426,
      "learning_rate": 0.00019486554158280574,
      "loss": 5.8212,
      "step": 1923
    },
    {
      "epoch": 0.1924,
      "grad_norm": 2.2035973072052,
      "learning_rate": 0.00019485449445269377,
      "loss": 5.3495,
      "step": 1924
    },
    {
      "epoch": 0.1925,
      "grad_norm": 2.4506490230560303,
      "learning_rate": 0.00019484343576484933,
      "loss": 5.5865,
      "step": 1925
    },
    {
      "epoch": 0.1926,
      "grad_norm": 4.857569694519043,
      "learning_rate": 0.00019483236552061994,
      "loss": 6.2613,
      "step": 1926
    },
    {
      "epoch": 0.1927,
      "grad_norm": 5.590539455413818,
      "learning_rate": 0.00019482128372135446,
      "loss": 5.7543,
      "step": 1927
    },
    {
      "epoch": 0.1928,
      "grad_norm": 3.223494291305542,
      "learning_rate": 0.0001948101903684032,
      "loss": 5.5683,
      "step": 1928
    },
    {
      "epoch": 0.1929,
      "grad_norm": 3.364964008331299,
      "learning_rate": 0.00019479908546311781,
      "loss": 5.7917,
      "step": 1929
    },
    {
      "epoch": 0.193,
      "grad_norm": 2.3437507152557373,
      "learning_rate": 0.00019478796900685146,
      "loss": 5.8682,
      "step": 1930
    },
    {
      "epoch": 0.1931,
      "grad_norm": 4.232397556304932,
      "learning_rate": 0.0001947768410009586,
      "loss": 5.6284,
      "step": 1931
    },
    {
      "epoch": 0.1932,
      "grad_norm": 4.76163387298584,
      "learning_rate": 0.0001947657014467951,
      "loss": 6.0431,
      "step": 1932
    },
    {
      "epoch": 0.1933,
      "grad_norm": 2.1636550426483154,
      "learning_rate": 0.00019475455034571838,
      "loss": 5.2896,
      "step": 1933
    },
    {
      "epoch": 0.1934,
      "grad_norm": 3.4931023120880127,
      "learning_rate": 0.0001947433876990871,
      "loss": 5.7707,
      "step": 1934
    },
    {
      "epoch": 0.1935,
      "grad_norm": 5.771305561065674,
      "learning_rate": 0.00019473221350826142,
      "loss": 5.9918,
      "step": 1935
    },
    {
      "epoch": 0.1936,
      "grad_norm": 7.414418697357178,
      "learning_rate": 0.0001947210277746029,
      "loss": 5.8002,
      "step": 1936
    },
    {
      "epoch": 0.1937,
      "grad_norm": 9.216947555541992,
      "learning_rate": 0.00019470983049947444,
      "loss": 5.8497,
      "step": 1937
    },
    {
      "epoch": 0.1938,
      "grad_norm": 3.9808993339538574,
      "learning_rate": 0.00019469862168424042,
      "loss": 5.7329,
      "step": 1938
    },
    {
      "epoch": 0.1939,
      "grad_norm": 3.662963628768921,
      "learning_rate": 0.0001946874013302666,
      "loss": 6.0529,
      "step": 1939
    },
    {
      "epoch": 0.194,
      "grad_norm": 6.585302829742432,
      "learning_rate": 0.0001946761694389202,
      "loss": 6.97,
      "step": 1940
    },
    {
      "epoch": 0.1941,
      "grad_norm": 10.409296989440918,
      "learning_rate": 0.00019466492601156966,
      "loss": 6.1317,
      "step": 1941
    },
    {
      "epoch": 0.1942,
      "grad_norm": 2.2786757946014404,
      "learning_rate": 0.00019465367104958505,
      "loss": 5.8095,
      "step": 1942
    },
    {
      "epoch": 0.1943,
      "grad_norm": 5.986156463623047,
      "learning_rate": 0.00019464240455433775,
      "loss": 5.1654,
      "step": 1943
    },
    {
      "epoch": 0.1944,
      "grad_norm": 5.273672103881836,
      "learning_rate": 0.00019463112652720054,
      "loss": 5.7458,
      "step": 1944
    },
    {
      "epoch": 0.1945,
      "grad_norm": 4.741860389709473,
      "learning_rate": 0.00019461983696954758,
      "loss": 5.1916,
      "step": 1945
    },
    {
      "epoch": 0.1946,
      "grad_norm": 4.752035140991211,
      "learning_rate": 0.00019460853588275454,
      "loss": 5.7028,
      "step": 1946
    },
    {
      "epoch": 0.1947,
      "grad_norm": 3.633392810821533,
      "learning_rate": 0.00019459722326819838,
      "loss": 6.6811,
      "step": 1947
    },
    {
      "epoch": 0.1948,
      "grad_norm": 4.778956413269043,
      "learning_rate": 0.00019458589912725748,
      "loss": 6.2784,
      "step": 1948
    },
    {
      "epoch": 0.1949,
      "grad_norm": 4.505735397338867,
      "learning_rate": 0.0001945745634613117,
      "loss": 5.7685,
      "step": 1949
    },
    {
      "epoch": 0.195,
      "grad_norm": 4.009304523468018,
      "learning_rate": 0.00019456321627174221,
      "loss": 6.2551,
      "step": 1950
    },
    {
      "epoch": 0.1951,
      "grad_norm": 2.3057007789611816,
      "learning_rate": 0.0001945518575599317,
      "loss": 6.2959,
      "step": 1951
    },
    {
      "epoch": 0.1952,
      "grad_norm": 4.8827805519104,
      "learning_rate": 0.00019454048732726412,
      "loss": 6.4594,
      "step": 1952
    },
    {
      "epoch": 0.1953,
      "grad_norm": 2.527704954147339,
      "learning_rate": 0.00019452910557512496,
      "loss": 5.9485,
      "step": 1953
    },
    {
      "epoch": 0.1954,
      "grad_norm": 4.341801166534424,
      "learning_rate": 0.000194517712304901,
      "loss": 6.0791,
      "step": 1954
    },
    {
      "epoch": 0.1955,
      "grad_norm": 6.992699146270752,
      "learning_rate": 0.00019450630751798048,
      "loss": 7.398,
      "step": 1955
    },
    {
      "epoch": 0.1956,
      "grad_norm": 11.865696907043457,
      "learning_rate": 0.0001944948912157531,
      "loss": 6.2456,
      "step": 1956
    },
    {
      "epoch": 0.1957,
      "grad_norm": 2.6435842514038086,
      "learning_rate": 0.00019448346339960982,
      "loss": 5.6048,
      "step": 1957
    },
    {
      "epoch": 0.1958,
      "grad_norm": 4.29429292678833,
      "learning_rate": 0.00019447202407094316,
      "loss": 6.0133,
      "step": 1958
    },
    {
      "epoch": 0.1959,
      "grad_norm": 23.69799041748047,
      "learning_rate": 0.0001944605732311469,
      "loss": 6.3836,
      "step": 1959
    },
    {
      "epoch": 0.196,
      "grad_norm": 13.2357759475708,
      "learning_rate": 0.00019444911088161636,
      "loss": 6.2964,
      "step": 1960
    },
    {
      "epoch": 0.1961,
      "grad_norm": 14.659323692321777,
      "learning_rate": 0.00019443763702374812,
      "loss": 5.7108,
      "step": 1961
    },
    {
      "epoch": 0.1962,
      "grad_norm": 2.8994414806365967,
      "learning_rate": 0.00019442615165894027,
      "loss": 6.8705,
      "step": 1962
    },
    {
      "epoch": 0.1963,
      "grad_norm": 11.389138221740723,
      "learning_rate": 0.00019441465478859228,
      "loss": 5.6177,
      "step": 1963
    },
    {
      "epoch": 0.1964,
      "grad_norm": 10.383113861083984,
      "learning_rate": 0.000194403146414105,
      "loss": 6.0505,
      "step": 1964
    },
    {
      "epoch": 0.1965,
      "grad_norm": 4.112442970275879,
      "learning_rate": 0.00019439162653688065,
      "loss": 5.3673,
      "step": 1965
    },
    {
      "epoch": 0.1966,
      "grad_norm": 12.376358985900879,
      "learning_rate": 0.00019438009515832297,
      "loss": 5.7648,
      "step": 1966
    },
    {
      "epoch": 0.1967,
      "grad_norm": 3.697592258453369,
      "learning_rate": 0.00019436855227983695,
      "loss": 7.5166,
      "step": 1967
    },
    {
      "epoch": 0.1968,
      "grad_norm": 3.1094439029693604,
      "learning_rate": 0.00019435699790282908,
      "loss": 5.6056,
      "step": 1968
    },
    {
      "epoch": 0.1969,
      "grad_norm": 4.926400184631348,
      "learning_rate": 0.00019434543202870725,
      "loss": 5.984,
      "step": 1969
    },
    {
      "epoch": 0.197,
      "grad_norm": 2.982884645462036,
      "learning_rate": 0.0001943338546588807,
      "loss": 6.0451,
      "step": 1970
    },
    {
      "epoch": 0.1971,
      "grad_norm": 2.5942814350128174,
      "learning_rate": 0.0001943222657947601,
      "loss": 5.8076,
      "step": 1971
    },
    {
      "epoch": 0.1972,
      "grad_norm": 7.661454200744629,
      "learning_rate": 0.00019431066543775752,
      "loss": 7.5555,
      "step": 1972
    },
    {
      "epoch": 0.1973,
      "grad_norm": 4.893845558166504,
      "learning_rate": 0.00019429905358928646,
      "loss": 5.7234,
      "step": 1973
    },
    {
      "epoch": 0.1974,
      "grad_norm": 4.675678253173828,
      "learning_rate": 0.00019428743025076174,
      "loss": 5.8713,
      "step": 1974
    },
    {
      "epoch": 0.1975,
      "grad_norm": 5.9395527839660645,
      "learning_rate": 0.00019427579542359965,
      "loss": 6.0547,
      "step": 1975
    },
    {
      "epoch": 0.1976,
      "grad_norm": 8.666311264038086,
      "learning_rate": 0.00019426414910921787,
      "loss": 6.8869,
      "step": 1976
    },
    {
      "epoch": 0.1977,
      "grad_norm": 5.685393810272217,
      "learning_rate": 0.00019425249130903543,
      "loss": 5.3742,
      "step": 1977
    },
    {
      "epoch": 0.1978,
      "grad_norm": 3.18534779548645,
      "learning_rate": 0.00019424082202447283,
      "loss": 6.0504,
      "step": 1978
    },
    {
      "epoch": 0.1979,
      "grad_norm": 14.97274398803711,
      "learning_rate": 0.0001942291412569519,
      "loss": 5.7179,
      "step": 1979
    },
    {
      "epoch": 0.198,
      "grad_norm": 9.984042167663574,
      "learning_rate": 0.00019421744900789597,
      "loss": 6.1437,
      "step": 1980
    },
    {
      "epoch": 0.1981,
      "grad_norm": 5.600270748138428,
      "learning_rate": 0.00019420574527872968,
      "loss": 6.324,
      "step": 1981
    },
    {
      "epoch": 0.1982,
      "grad_norm": 15.80140209197998,
      "learning_rate": 0.00019419403007087907,
      "loss": 6.7844,
      "step": 1982
    },
    {
      "epoch": 0.1983,
      "grad_norm": 16.01495933532715,
      "learning_rate": 0.0001941823033857716,
      "loss": 5.91,
      "step": 1983
    },
    {
      "epoch": 0.1984,
      "grad_norm": 2.8791592121124268,
      "learning_rate": 0.0001941705652248362,
      "loss": 5.1753,
      "step": 1984
    },
    {
      "epoch": 0.1985,
      "grad_norm": 5.599843502044678,
      "learning_rate": 0.00019415881558950302,
      "loss": 6.0962,
      "step": 1985
    },
    {
      "epoch": 0.1986,
      "grad_norm": 12.584126472473145,
      "learning_rate": 0.0001941470544812038,
      "loss": 5.6619,
      "step": 1986
    },
    {
      "epoch": 0.1987,
      "grad_norm": 3.0123064517974854,
      "learning_rate": 0.0001941352819013716,
      "loss": 5.7794,
      "step": 1987
    },
    {
      "epoch": 0.1988,
      "grad_norm": 3.9567222595214844,
      "learning_rate": 0.00019412349785144078,
      "loss": 5.9423,
      "step": 1988
    },
    {
      "epoch": 0.1989,
      "grad_norm": 2.651810884475708,
      "learning_rate": 0.00019411170233284727,
      "loss": 6.5346,
      "step": 1989
    },
    {
      "epoch": 0.199,
      "grad_norm": 7.165063858032227,
      "learning_rate": 0.00019409989534702833,
      "loss": 6.1775,
      "step": 1990
    },
    {
      "epoch": 0.1991,
      "grad_norm": 9.063244819641113,
      "learning_rate": 0.00019408807689542257,
      "loss": 5.6077,
      "step": 1991
    },
    {
      "epoch": 0.1992,
      "grad_norm": 2.752800464630127,
      "learning_rate": 0.00019407624697947003,
      "loss": 5.8256,
      "step": 1992
    },
    {
      "epoch": 0.1993,
      "grad_norm": 3.586080551147461,
      "learning_rate": 0.00019406440560061216,
      "loss": 5.9013,
      "step": 1993
    },
    {
      "epoch": 0.1994,
      "grad_norm": 3.9430134296417236,
      "learning_rate": 0.0001940525527602918,
      "loss": 5.7021,
      "step": 1994
    },
    {
      "epoch": 0.1995,
      "grad_norm": 3.373464822769165,
      "learning_rate": 0.00019404068845995317,
      "loss": 5.5075,
      "step": 1995
    },
    {
      "epoch": 0.1996,
      "grad_norm": 2.643373966217041,
      "learning_rate": 0.0001940288127010419,
      "loss": 5.9614,
      "step": 1996
    },
    {
      "epoch": 0.1997,
      "grad_norm": 6.219358444213867,
      "learning_rate": 0.00019401692548500502,
      "loss": 5.4463,
      "step": 1997
    },
    {
      "epoch": 0.1998,
      "grad_norm": 5.91963005065918,
      "learning_rate": 0.00019400502681329098,
      "loss": 5.9663,
      "step": 1998
    },
    {
      "epoch": 0.1999,
      "grad_norm": 2.3587753772735596,
      "learning_rate": 0.00019399311668734956,
      "loss": 5.8224,
      "step": 1999
    },
    {
      "epoch": 0.2,
      "grad_norm": 6.445633411407471,
      "learning_rate": 0.00019398119510863197,
      "loss": 5.7444,
      "step": 2000
    },
    {
      "epoch": 0.2001,
      "grad_norm": 4.937867164611816,
      "learning_rate": 0.00019396926207859084,
      "loss": 5.7192,
      "step": 2001
    },
    {
      "epoch": 0.2002,
      "grad_norm": 4.282773494720459,
      "learning_rate": 0.00019395731759868018,
      "loss": 6.4046,
      "step": 2002
    },
    {
      "epoch": 0.2003,
      "grad_norm": 3.065312147140503,
      "learning_rate": 0.00019394536167035534,
      "loss": 5.0715,
      "step": 2003
    },
    {
      "epoch": 0.2004,
      "grad_norm": 2.829141616821289,
      "learning_rate": 0.00019393339429507318,
      "loss": 5.4535,
      "step": 2004
    },
    {
      "epoch": 0.2005,
      "grad_norm": 4.304609298706055,
      "learning_rate": 0.00019392141547429183,
      "loss": 5.5324,
      "step": 2005
    },
    {
      "epoch": 0.2006,
      "grad_norm": 3.161588430404663,
      "learning_rate": 0.0001939094252094709,
      "loss": 5.9349,
      "step": 2006
    },
    {
      "epoch": 0.2007,
      "grad_norm": 8.412795066833496,
      "learning_rate": 0.00019389742350207141,
      "loss": 5.8176,
      "step": 2007
    },
    {
      "epoch": 0.2008,
      "grad_norm": 3.431042194366455,
      "learning_rate": 0.00019388541035355564,
      "loss": 5.802,
      "step": 2008
    },
    {
      "epoch": 0.2009,
      "grad_norm": 3.4746124744415283,
      "learning_rate": 0.00019387338576538744,
      "loss": 5.6845,
      "step": 2009
    },
    {
      "epoch": 0.201,
      "grad_norm": 9.351471900939941,
      "learning_rate": 0.0001938613497390319,
      "loss": 6.4453,
      "step": 2010
    },
    {
      "epoch": 0.2011,
      "grad_norm": 3.103248119354248,
      "learning_rate": 0.0001938493022759556,
      "loss": 5.7872,
      "step": 2011
    },
    {
      "epoch": 0.2012,
      "grad_norm": 2.5525903701782227,
      "learning_rate": 0.0001938372433776265,
      "loss": 5.6941,
      "step": 2012
    },
    {
      "epoch": 0.2013,
      "grad_norm": 7.46622371673584,
      "learning_rate": 0.00019382517304551396,
      "loss": 7.1961,
      "step": 2013
    },
    {
      "epoch": 0.2014,
      "grad_norm": 14.475409507751465,
      "learning_rate": 0.00019381309128108865,
      "loss": 6.1422,
      "step": 2014
    },
    {
      "epoch": 0.2015,
      "grad_norm": 3.48580002784729,
      "learning_rate": 0.00019380099808582278,
      "loss": 5.4272,
      "step": 2015
    },
    {
      "epoch": 0.2016,
      "grad_norm": 2.618835926055908,
      "learning_rate": 0.0001937888934611898,
      "loss": 5.6583,
      "step": 2016
    },
    {
      "epoch": 0.2017,
      "grad_norm": 3.553460121154785,
      "learning_rate": 0.0001937767774086646,
      "loss": 5.6394,
      "step": 2017
    },
    {
      "epoch": 0.2018,
      "grad_norm": 3.4781758785247803,
      "learning_rate": 0.00019376464992972356,
      "loss": 6.8905,
      "step": 2018
    },
    {
      "epoch": 0.2019,
      "grad_norm": 6.479672908782959,
      "learning_rate": 0.0001937525110258444,
      "loss": 6.0319,
      "step": 2019
    },
    {
      "epoch": 0.202,
      "grad_norm": 17.533361434936523,
      "learning_rate": 0.00019374036069850608,
      "loss": 6.4751,
      "step": 2020
    },
    {
      "epoch": 0.2021,
      "grad_norm": 3.4021623134613037,
      "learning_rate": 0.00019372819894918915,
      "loss": 6.1972,
      "step": 2021
    },
    {
      "epoch": 0.2022,
      "grad_norm": 3.0344278812408447,
      "learning_rate": 0.00019371602577937554,
      "loss": 5.4331,
      "step": 2022
    },
    {
      "epoch": 0.2023,
      "grad_norm": 19.790483474731445,
      "learning_rate": 0.0001937038411905484,
      "loss": 5.8679,
      "step": 2023
    },
    {
      "epoch": 0.2024,
      "grad_norm": 3.0790517330169678,
      "learning_rate": 0.0001936916451841925,
      "loss": 6.6565,
      "step": 2024
    },
    {
      "epoch": 0.2025,
      "grad_norm": 7.568552494049072,
      "learning_rate": 0.0001936794377617938,
      "loss": 5.7699,
      "step": 2025
    },
    {
      "epoch": 0.2026,
      "grad_norm": 3.2531814575195312,
      "learning_rate": 0.00019366721892483978,
      "loss": 6.1017,
      "step": 2026
    },
    {
      "epoch": 0.2027,
      "grad_norm": 5.236907005310059,
      "learning_rate": 0.00019365498867481923,
      "loss": 6.0184,
      "step": 2027
    },
    {
      "epoch": 0.2028,
      "grad_norm": 4.766722202301025,
      "learning_rate": 0.00019364274701322244,
      "loss": 5.0843,
      "step": 2028
    },
    {
      "epoch": 0.2029,
      "grad_norm": 5.195804595947266,
      "learning_rate": 0.00019363049394154094,
      "loss": 5.3777,
      "step": 2029
    },
    {
      "epoch": 0.203,
      "grad_norm": 3.2699239253997803,
      "learning_rate": 0.0001936182294612678,
      "loss": 5.875,
      "step": 2030
    },
    {
      "epoch": 0.2031,
      "grad_norm": 4.644274711608887,
      "learning_rate": 0.00019360595357389735,
      "loss": 5.8212,
      "step": 2031
    },
    {
      "epoch": 0.2032,
      "grad_norm": 5.021501064300537,
      "learning_rate": 0.00019359366628092539,
      "loss": 6.6825,
      "step": 2032
    },
    {
      "epoch": 0.2033,
      "grad_norm": 3.4944634437561035,
      "learning_rate": 0.00019358136758384912,
      "loss": 5.3759,
      "step": 2033
    },
    {
      "epoch": 0.2034,
      "grad_norm": 5.51589298248291,
      "learning_rate": 0.00019356905748416702,
      "loss": 6.4519,
      "step": 2034
    },
    {
      "epoch": 0.2035,
      "grad_norm": 10.606785774230957,
      "learning_rate": 0.00019355673598337914,
      "loss": 6.1654,
      "step": 2035
    },
    {
      "epoch": 0.2036,
      "grad_norm": 4.336905479431152,
      "learning_rate": 0.00019354440308298675,
      "loss": 5.9924,
      "step": 2036
    },
    {
      "epoch": 0.2037,
      "grad_norm": 6.434743404388428,
      "learning_rate": 0.00019353205878449258,
      "loss": 5.5564,
      "step": 2037
    },
    {
      "epoch": 0.2038,
      "grad_norm": 2.414701223373413,
      "learning_rate": 0.0001935197030894008,
      "loss": 5.8702,
      "step": 2038
    },
    {
      "epoch": 0.2039,
      "grad_norm": 3.6249170303344727,
      "learning_rate": 0.00019350733599921683,
      "loss": 5.9198,
      "step": 2039
    },
    {
      "epoch": 0.204,
      "grad_norm": 3.4936513900756836,
      "learning_rate": 0.00019349495751544763,
      "loss": 5.2492,
      "step": 2040
    },
    {
      "epoch": 0.2041,
      "grad_norm": 12.854368209838867,
      "learning_rate": 0.00019348256763960145,
      "loss": 6.2945,
      "step": 2041
    },
    {
      "epoch": 0.2042,
      "grad_norm": 5.484251976013184,
      "learning_rate": 0.000193470166373188,
      "loss": 5.5801,
      "step": 2042
    },
    {
      "epoch": 0.2043,
      "grad_norm": 5.351827144622803,
      "learning_rate": 0.00019345775371771824,
      "loss": 5.6434,
      "step": 2043
    },
    {
      "epoch": 0.2044,
      "grad_norm": 3.943511724472046,
      "learning_rate": 0.0001934453296747047,
      "loss": 5.9512,
      "step": 2044
    },
    {
      "epoch": 0.2045,
      "grad_norm": 6.731863975524902,
      "learning_rate": 0.00019343289424566122,
      "loss": 6.0947,
      "step": 2045
    },
    {
      "epoch": 0.2046,
      "grad_norm": 4.906713485717773,
      "learning_rate": 0.00019342044743210295,
      "loss": 5.6276,
      "step": 2046
    },
    {
      "epoch": 0.2047,
      "grad_norm": 3.9746525287628174,
      "learning_rate": 0.00019340798923554657,
      "loss": 6.527,
      "step": 2047
    },
    {
      "epoch": 0.2048,
      "grad_norm": 3.197859048843384,
      "learning_rate": 0.00019339551965751002,
      "loss": 5.511,
      "step": 2048
    },
    {
      "epoch": 0.2049,
      "grad_norm": 15.1095552444458,
      "learning_rate": 0.00019338303869951269,
      "loss": 6.3372,
      "step": 2049
    },
    {
      "epoch": 0.205,
      "grad_norm": 6.79530668258667,
      "learning_rate": 0.00019337054636307536,
      "loss": 6.0078,
      "step": 2050
    },
    {
      "epoch": 0.2051,
      "grad_norm": 2.3271594047546387,
      "learning_rate": 0.00019335804264972018,
      "loss": 6.4747,
      "step": 2051
    },
    {
      "epoch": 0.2052,
      "grad_norm": 2.4604785442352295,
      "learning_rate": 0.0001933455275609707,
      "loss": 5.6952,
      "step": 2052
    },
    {
      "epoch": 0.2053,
      "grad_norm": 4.115812301635742,
      "learning_rate": 0.0001933330010983518,
      "loss": 5.3475,
      "step": 2053
    },
    {
      "epoch": 0.2054,
      "grad_norm": 3.2210466861724854,
      "learning_rate": 0.00019332046326338986,
      "loss": 5.7013,
      "step": 2054
    },
    {
      "epoch": 0.2055,
      "grad_norm": 3.8774118423461914,
      "learning_rate": 0.00019330791405761252,
      "loss": 5.6677,
      "step": 2055
    },
    {
      "epoch": 0.2056,
      "grad_norm": 3.701324462890625,
      "learning_rate": 0.00019329535348254893,
      "loss": 5.5703,
      "step": 2056
    },
    {
      "epoch": 0.2057,
      "grad_norm": 2.4709572792053223,
      "learning_rate": 0.00019328278153972947,
      "loss": 5.6916,
      "step": 2057
    },
    {
      "epoch": 0.2058,
      "grad_norm": 3.2217392921447754,
      "learning_rate": 0.00019327019823068604,
      "loss": 5.8553,
      "step": 2058
    },
    {
      "epoch": 0.2059,
      "grad_norm": 2.920689105987549,
      "learning_rate": 0.00019325760355695188,
      "loss": 5.6993,
      "step": 2059
    },
    {
      "epoch": 0.206,
      "grad_norm": 2.692683219909668,
      "learning_rate": 0.0001932449975200616,
      "loss": 5.3319,
      "step": 2060
    },
    {
      "epoch": 0.2061,
      "grad_norm": 2.259795665740967,
      "learning_rate": 0.00019323238012155123,
      "loss": 6.0578,
      "step": 2061
    },
    {
      "epoch": 0.2062,
      "grad_norm": 3.159073829650879,
      "learning_rate": 0.00019321975136295813,
      "loss": 5.5256,
      "step": 2062
    },
    {
      "epoch": 0.2063,
      "grad_norm": 2.9852237701416016,
      "learning_rate": 0.0001932071112458211,
      "loss": 6.058,
      "step": 2063
    },
    {
      "epoch": 0.2064,
      "grad_norm": 19.091585159301758,
      "learning_rate": 0.0001931944597716803,
      "loss": 5.8464,
      "step": 2064
    },
    {
      "epoch": 0.2065,
      "grad_norm": 10.71346664428711,
      "learning_rate": 0.00019318179694207725,
      "loss": 5.7699,
      "step": 2065
    },
    {
      "epoch": 0.2066,
      "grad_norm": 13.694106101989746,
      "learning_rate": 0.0001931691227585549,
      "loss": 5.4069,
      "step": 2066
    },
    {
      "epoch": 0.2067,
      "grad_norm": 12.709088325500488,
      "learning_rate": 0.00019315643722265757,
      "loss": 6.1767,
      "step": 2067
    },
    {
      "epoch": 0.2068,
      "grad_norm": 11.745285987854004,
      "learning_rate": 0.0001931437403359309,
      "loss": 5.7067,
      "step": 2068
    },
    {
      "epoch": 0.2069,
      "grad_norm": 9.39574146270752,
      "learning_rate": 0.00019313103209992204,
      "loss": 6.2576,
      "step": 2069
    },
    {
      "epoch": 0.207,
      "grad_norm": 6.333316326141357,
      "learning_rate": 0.0001931183125161794,
      "loss": 6.1401,
      "step": 2070
    },
    {
      "epoch": 0.2071,
      "grad_norm": 6.9987945556640625,
      "learning_rate": 0.00019310558158625285,
      "loss": 6.0612,
      "step": 2071
    },
    {
      "epoch": 0.2072,
      "grad_norm": 4.854059219360352,
      "learning_rate": 0.00019309283931169356,
      "loss": 5.8165,
      "step": 2072
    },
    {
      "epoch": 0.2073,
      "grad_norm": 3.451984167098999,
      "learning_rate": 0.00019308008569405422,
      "loss": 5.2868,
      "step": 2073
    },
    {
      "epoch": 0.2074,
      "grad_norm": 4.428050518035889,
      "learning_rate": 0.0001930673207348888,
      "loss": 5.3233,
      "step": 2074
    },
    {
      "epoch": 0.2075,
      "grad_norm": 3.29357647895813,
      "learning_rate": 0.00019305454443575262,
      "loss": 5.7449,
      "step": 2075
    },
    {
      "epoch": 0.2076,
      "grad_norm": 3.336698055267334,
      "learning_rate": 0.00019304175679820247,
      "loss": 5.0819,
      "step": 2076
    },
    {
      "epoch": 0.2077,
      "grad_norm": 3.7022831439971924,
      "learning_rate": 0.0001930289578237965,
      "loss": 6.1283,
      "step": 2077
    },
    {
      "epoch": 0.2078,
      "grad_norm": 3.0215237140655518,
      "learning_rate": 0.00019301614751409416,
      "loss": 5.8703,
      "step": 2078
    },
    {
      "epoch": 0.2079,
      "grad_norm": 4.4551873207092285,
      "learning_rate": 0.0001930033258706564,
      "loss": 5.8294,
      "step": 2079
    },
    {
      "epoch": 0.208,
      "grad_norm": 3.6612942218780518,
      "learning_rate": 0.00019299049289504553,
      "loss": 6.1698,
      "step": 2080
    },
    {
      "epoch": 0.2081,
      "grad_norm": 6.004714488983154,
      "learning_rate": 0.00019297764858882514,
      "loss": 6.1256,
      "step": 2081
    },
    {
      "epoch": 0.2082,
      "grad_norm": 3.1612677574157715,
      "learning_rate": 0.00019296479295356035,
      "loss": 6.3,
      "step": 2082
    },
    {
      "epoch": 0.2083,
      "grad_norm": 9.134725570678711,
      "learning_rate": 0.00019295192599081746,
      "loss": 6.0025,
      "step": 2083
    },
    {
      "epoch": 0.2084,
      "grad_norm": 3.163954973220825,
      "learning_rate": 0.00019293904770216437,
      "loss": 6.3144,
      "step": 2084
    },
    {
      "epoch": 0.2085,
      "grad_norm": 4.394771575927734,
      "learning_rate": 0.00019292615808917026,
      "loss": 6.1798,
      "step": 2085
    },
    {
      "epoch": 0.2086,
      "grad_norm": 3.111865520477295,
      "learning_rate": 0.00019291325715340563,
      "loss": 5.0936,
      "step": 2086
    },
    {
      "epoch": 0.2087,
      "grad_norm": 3.3181588649749756,
      "learning_rate": 0.00019290034489644246,
      "loss": 5.3264,
      "step": 2087
    },
    {
      "epoch": 0.2088,
      "grad_norm": 5.403226852416992,
      "learning_rate": 0.00019288742131985407,
      "loss": 5.8246,
      "step": 2088
    },
    {
      "epoch": 0.2089,
      "grad_norm": 11.759743690490723,
      "learning_rate": 0.00019287448642521513,
      "loss": 6.186,
      "step": 2089
    },
    {
      "epoch": 0.209,
      "grad_norm": 9.72570514678955,
      "learning_rate": 0.00019286154021410173,
      "loss": 5.4525,
      "step": 2090
    },
    {
      "epoch": 0.2091,
      "grad_norm": 4.480266094207764,
      "learning_rate": 0.00019284858268809137,
      "loss": 6.2803,
      "step": 2091
    },
    {
      "epoch": 0.2092,
      "grad_norm": 4.779378890991211,
      "learning_rate": 0.00019283561384876284,
      "loss": 5.4832,
      "step": 2092
    },
    {
      "epoch": 0.2093,
      "grad_norm": 15.37045669555664,
      "learning_rate": 0.00019282263369769633,
      "loss": 7.0848,
      "step": 2093
    },
    {
      "epoch": 0.2094,
      "grad_norm": 6.689272880554199,
      "learning_rate": 0.00019280964223647348,
      "loss": 5.8475,
      "step": 2094
    },
    {
      "epoch": 0.2095,
      "grad_norm": 3.620460033416748,
      "learning_rate": 0.00019279663946667727,
      "loss": 5.0791,
      "step": 2095
    },
    {
      "epoch": 0.2096,
      "grad_norm": 7.157323360443115,
      "learning_rate": 0.000192783625389892,
      "loss": 5.9469,
      "step": 2096
    },
    {
      "epoch": 0.2097,
      "grad_norm": 3.393159866333008,
      "learning_rate": 0.00019277060000770342,
      "loss": 5.8557,
      "step": 2097
    },
    {
      "epoch": 0.2098,
      "grad_norm": 5.951493740081787,
      "learning_rate": 0.00019275756332169867,
      "loss": 5.5799,
      "step": 2098
    },
    {
      "epoch": 0.2099,
      "grad_norm": 4.10254430770874,
      "learning_rate": 0.00019274451533346615,
      "loss": 5.8243,
      "step": 2099
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.0869510173797607,
      "learning_rate": 0.00019273145604459577,
      "loss": 6.0757,
      "step": 2100
    },
    {
      "epoch": 0.2101,
      "grad_norm": 3.323883295059204,
      "learning_rate": 0.00019271838545667876,
      "loss": 5.94,
      "step": 2101
    },
    {
      "epoch": 0.2102,
      "grad_norm": 7.589050769805908,
      "learning_rate": 0.0001927053035713077,
      "loss": 5.5628,
      "step": 2102
    },
    {
      "epoch": 0.2103,
      "grad_norm": 3.4237802028656006,
      "learning_rate": 0.00019269221039007665,
      "loss": 5.6893,
      "step": 2103
    },
    {
      "epoch": 0.2104,
      "grad_norm": 3.4828317165374756,
      "learning_rate": 0.0001926791059145809,
      "loss": 5.4888,
      "step": 2104
    },
    {
      "epoch": 0.2105,
      "grad_norm": 4.6789093017578125,
      "learning_rate": 0.0001926659901464172,
      "loss": 5.6719,
      "step": 2105
    },
    {
      "epoch": 0.2106,
      "grad_norm": 4.390766620635986,
      "learning_rate": 0.00019265286308718372,
      "loss": 6.2962,
      "step": 2106
    },
    {
      "epoch": 0.2107,
      "grad_norm": 3.8219170570373535,
      "learning_rate": 0.00019263972473847993,
      "loss": 5.8159,
      "step": 2107
    },
    {
      "epoch": 0.2108,
      "grad_norm": 2.5497140884399414,
      "learning_rate": 0.00019262657510190666,
      "loss": 5.6569,
      "step": 2108
    },
    {
      "epoch": 0.2109,
      "grad_norm": 4.787859916687012,
      "learning_rate": 0.00019261341417906621,
      "loss": 5.8112,
      "step": 2109
    },
    {
      "epoch": 0.211,
      "grad_norm": 5.434683799743652,
      "learning_rate": 0.00019260024197156214,
      "loss": 6.2694,
      "step": 2110
    },
    {
      "epoch": 0.2111,
      "grad_norm": 3.2133798599243164,
      "learning_rate": 0.0001925870584809995,
      "loss": 5.4542,
      "step": 2111
    },
    {
      "epoch": 0.2112,
      "grad_norm": 3.6866493225097656,
      "learning_rate": 0.00019257386370898457,
      "loss": 6.1638,
      "step": 2112
    },
    {
      "epoch": 0.2113,
      "grad_norm": 12.540678977966309,
      "learning_rate": 0.00019256065765712522,
      "loss": 6.3662,
      "step": 2113
    },
    {
      "epoch": 0.2114,
      "grad_norm": 7.877955436706543,
      "learning_rate": 0.00019254744032703048,
      "loss": 6.3113,
      "step": 2114
    },
    {
      "epoch": 0.2115,
      "grad_norm": 8.230093955993652,
      "learning_rate": 0.00019253421172031086,
      "loss": 5.6136,
      "step": 2115
    },
    {
      "epoch": 0.2116,
      "grad_norm": 5.537298202514648,
      "learning_rate": 0.00019252097183857823,
      "loss": 5.3792,
      "step": 2116
    },
    {
      "epoch": 0.2117,
      "grad_norm": 3.1806836128234863,
      "learning_rate": 0.0001925077206834458,
      "loss": 5.8726,
      "step": 2117
    },
    {
      "epoch": 0.2118,
      "grad_norm": 6.388739109039307,
      "learning_rate": 0.00019249445825652824,
      "loss": 5.5294,
      "step": 2118
    },
    {
      "epoch": 0.2119,
      "grad_norm": 2.908450126647949,
      "learning_rate": 0.0001924811845594415,
      "loss": 6.0063,
      "step": 2119
    },
    {
      "epoch": 0.212,
      "grad_norm": 2.5008115768432617,
      "learning_rate": 0.00019246789959380295,
      "loss": 5.6743,
      "step": 2120
    },
    {
      "epoch": 0.2121,
      "grad_norm": 3.0681252479553223,
      "learning_rate": 0.00019245460336123134,
      "loss": 5.731,
      "step": 2121
    },
    {
      "epoch": 0.2122,
      "grad_norm": 3.2271623611450195,
      "learning_rate": 0.00019244129586334672,
      "loss": 5.0468,
      "step": 2122
    },
    {
      "epoch": 0.2123,
      "grad_norm": 2.80643630027771,
      "learning_rate": 0.0001924279771017706,
      "loss": 5.7682,
      "step": 2123
    },
    {
      "epoch": 0.2124,
      "grad_norm": 3.9504053592681885,
      "learning_rate": 0.00019241464707812585,
      "loss": 5.4246,
      "step": 2124
    },
    {
      "epoch": 0.2125,
      "grad_norm": 14.5697021484375,
      "learning_rate": 0.0001924013057940367,
      "loss": 5.4122,
      "step": 2125
    },
    {
      "epoch": 0.2126,
      "grad_norm": 2.743694543838501,
      "learning_rate": 0.0001923879532511287,
      "loss": 5.4838,
      "step": 2126
    },
    {
      "epoch": 0.2127,
      "grad_norm": 3.9502689838409424,
      "learning_rate": 0.00019237458945102882,
      "loss": 6.1177,
      "step": 2127
    },
    {
      "epoch": 0.2128,
      "grad_norm": 14.16513729095459,
      "learning_rate": 0.00019236121439536542,
      "loss": 5.8305,
      "step": 2128
    },
    {
      "epoch": 0.2129,
      "grad_norm": 5.112298965454102,
      "learning_rate": 0.00019234782808576824,
      "loss": 5.568,
      "step": 2129
    },
    {
      "epoch": 0.213,
      "grad_norm": 2.6745054721832275,
      "learning_rate": 0.0001923344305238683,
      "loss": 5.7166,
      "step": 2130
    },
    {
      "epoch": 0.2131,
      "grad_norm": 4.820267200469971,
      "learning_rate": 0.00019232102171129811,
      "loss": 6.0692,
      "step": 2131
    },
    {
      "epoch": 0.2132,
      "grad_norm": 1.9019205570220947,
      "learning_rate": 0.00019230760164969143,
      "loss": 5.5984,
      "step": 2132
    },
    {
      "epoch": 0.2133,
      "grad_norm": 2.125511646270752,
      "learning_rate": 0.0001922941703406835,
      "loss": 5.9778,
      "step": 2133
    },
    {
      "epoch": 0.2134,
      "grad_norm": 9.9030179977417,
      "learning_rate": 0.00019228072778591088,
      "loss": 6.3523,
      "step": 2134
    },
    {
      "epoch": 0.2135,
      "grad_norm": 6.529533863067627,
      "learning_rate": 0.0001922672739870115,
      "loss": 5.5992,
      "step": 2135
    },
    {
      "epoch": 0.2136,
      "grad_norm": 4.30414342880249,
      "learning_rate": 0.00019225380894562463,
      "loss": 5.6049,
      "step": 2136
    },
    {
      "epoch": 0.2137,
      "grad_norm": 6.594366550445557,
      "learning_rate": 0.00019224033266339102,
      "loss": 5.5576,
      "step": 2137
    },
    {
      "epoch": 0.2138,
      "grad_norm": 7.760684490203857,
      "learning_rate": 0.00019222684514195264,
      "loss": 5.9806,
      "step": 2138
    },
    {
      "epoch": 0.2139,
      "grad_norm": 4.13138484954834,
      "learning_rate": 0.00019221334638295294,
      "loss": 5.9581,
      "step": 2139
    },
    {
      "epoch": 0.214,
      "grad_norm": 4.600522994995117,
      "learning_rate": 0.0001921998363880367,
      "loss": 6.4339,
      "step": 2140
    },
    {
      "epoch": 0.2141,
      "grad_norm": 13.102234840393066,
      "learning_rate": 0.00019218631515885006,
      "loss": 5.8178,
      "step": 2141
    },
    {
      "epoch": 0.2142,
      "grad_norm": 4.273313045501709,
      "learning_rate": 0.00019217278269704056,
      "loss": 6.0644,
      "step": 2142
    },
    {
      "epoch": 0.2143,
      "grad_norm": 3.7853288650512695,
      "learning_rate": 0.00019215923900425707,
      "loss": 5.5601,
      "step": 2143
    },
    {
      "epoch": 0.2144,
      "grad_norm": 3.671359062194824,
      "learning_rate": 0.00019214568408214985,
      "loss": 6.0931,
      "step": 2144
    },
    {
      "epoch": 0.2145,
      "grad_norm": 5.302005290985107,
      "learning_rate": 0.00019213211793237057,
      "loss": 6.4614,
      "step": 2145
    },
    {
      "epoch": 0.2146,
      "grad_norm": 4.087045192718506,
      "learning_rate": 0.00019211854055657215,
      "loss": 5.5828,
      "step": 2146
    },
    {
      "epoch": 0.2147,
      "grad_norm": 4.363244533538818,
      "learning_rate": 0.00019210495195640895,
      "loss": 5.9006,
      "step": 2147
    },
    {
      "epoch": 0.2148,
      "grad_norm": 2.915283203125,
      "learning_rate": 0.0001920913521335368,
      "loss": 4.9705,
      "step": 2148
    },
    {
      "epoch": 0.2149,
      "grad_norm": 4.941037178039551,
      "learning_rate": 0.00019207774108961272,
      "loss": 5.4923,
      "step": 2149
    },
    {
      "epoch": 0.215,
      "grad_norm": 3.1265132427215576,
      "learning_rate": 0.00019206411882629517,
      "loss": 5.6774,
      "step": 2150
    },
    {
      "epoch": 0.2151,
      "grad_norm": 3.597045421600342,
      "learning_rate": 0.00019205048534524406,
      "loss": 5.4771,
      "step": 2151
    },
    {
      "epoch": 0.2152,
      "grad_norm": 3.4995737075805664,
      "learning_rate": 0.00019203684064812045,
      "loss": 5.6781,
      "step": 2152
    },
    {
      "epoch": 0.2153,
      "grad_norm": 4.950976371765137,
      "learning_rate": 0.00019202318473658705,
      "loss": 5.9476,
      "step": 2153
    },
    {
      "epoch": 0.2154,
      "grad_norm": 4.743804454803467,
      "learning_rate": 0.0001920095176123077,
      "loss": 5.1477,
      "step": 2154
    },
    {
      "epoch": 0.2155,
      "grad_norm": 7.3057756423950195,
      "learning_rate": 0.00019199583927694772,
      "loss": 5.728,
      "step": 2155
    },
    {
      "epoch": 0.2156,
      "grad_norm": 2.7037906646728516,
      "learning_rate": 0.00019198214973217378,
      "loss": 5.2093,
      "step": 2156
    },
    {
      "epoch": 0.2157,
      "grad_norm": 2.492769956588745,
      "learning_rate": 0.00019196844897965393,
      "loss": 5.5484,
      "step": 2157
    },
    {
      "epoch": 0.2158,
      "grad_norm": 3.022624969482422,
      "learning_rate": 0.00019195473702105748,
      "loss": 5.7334,
      "step": 2158
    },
    {
      "epoch": 0.2159,
      "grad_norm": 3.5726890563964844,
      "learning_rate": 0.0001919410138580553,
      "loss": 5.3262,
      "step": 2159
    },
    {
      "epoch": 0.216,
      "grad_norm": 2.35561203956604,
      "learning_rate": 0.00019192727949231945,
      "loss": 5.8306,
      "step": 2160
    },
    {
      "epoch": 0.2161,
      "grad_norm": 3.9531328678131104,
      "learning_rate": 0.00019191353392552344,
      "loss": 5.8173,
      "step": 2161
    },
    {
      "epoch": 0.2162,
      "grad_norm": 3.4514222145080566,
      "learning_rate": 0.00019189977715934213,
      "loss": 5.8869,
      "step": 2162
    },
    {
      "epoch": 0.2163,
      "grad_norm": 4.84230899810791,
      "learning_rate": 0.00019188600919545174,
      "loss": 5.6978,
      "step": 2163
    },
    {
      "epoch": 0.2164,
      "grad_norm": 4.879822254180908,
      "learning_rate": 0.00019187223003552985,
      "loss": 6.1665,
      "step": 2164
    },
    {
      "epoch": 0.2165,
      "grad_norm": 2.1924238204956055,
      "learning_rate": 0.0001918584396812554,
      "loss": 5.772,
      "step": 2165
    },
    {
      "epoch": 0.2166,
      "grad_norm": 2.3923070430755615,
      "learning_rate": 0.00019184463813430873,
      "loss": 5.3744,
      "step": 2166
    },
    {
      "epoch": 0.2167,
      "grad_norm": 2.8480517864227295,
      "learning_rate": 0.00019183082539637146,
      "loss": 5.6735,
      "step": 2167
    },
    {
      "epoch": 0.2168,
      "grad_norm": 4.15550422668457,
      "learning_rate": 0.0001918170014691267,
      "loss": 5.0216,
      "step": 2168
    },
    {
      "epoch": 0.2169,
      "grad_norm": 16.65300941467285,
      "learning_rate": 0.0001918031663542588,
      "loss": 7.0252,
      "step": 2169
    },
    {
      "epoch": 0.217,
      "grad_norm": 14.37529468536377,
      "learning_rate": 0.0001917893200534536,
      "loss": 5.5529,
      "step": 2170
    },
    {
      "epoch": 0.2171,
      "grad_norm": 4.853829383850098,
      "learning_rate": 0.00019177546256839812,
      "loss": 5.8856,
      "step": 2171
    },
    {
      "epoch": 0.2172,
      "grad_norm": 4.806471824645996,
      "learning_rate": 0.00019176159390078094,
      "loss": 5.8681,
      "step": 2172
    },
    {
      "epoch": 0.2173,
      "grad_norm": 3.7039225101470947,
      "learning_rate": 0.00019174771405229186,
      "loss": 6.0032,
      "step": 2173
    },
    {
      "epoch": 0.2174,
      "grad_norm": 3.453080892562866,
      "learning_rate": 0.00019173382302462214,
      "loss": 6.3913,
      "step": 2174
    },
    {
      "epoch": 0.2175,
      "grad_norm": 2.381265640258789,
      "learning_rate": 0.00019171992081946435,
      "loss": 5.6366,
      "step": 2175
    },
    {
      "epoch": 0.2176,
      "grad_norm": 2.9639194011688232,
      "learning_rate": 0.0001917060074385124,
      "loss": 5.6302,
      "step": 2176
    },
    {
      "epoch": 0.2177,
      "grad_norm": 2.523000478744507,
      "learning_rate": 0.00019169208288346166,
      "loss": 5.6835,
      "step": 2177
    },
    {
      "epoch": 0.2178,
      "grad_norm": 8.545862197875977,
      "learning_rate": 0.0001916781471560087,
      "loss": 6.1976,
      "step": 2178
    },
    {
      "epoch": 0.2179,
      "grad_norm": 2.9540352821350098,
      "learning_rate": 0.00019166420025785164,
      "loss": 5.3306,
      "step": 2179
    },
    {
      "epoch": 0.218,
      "grad_norm": 8.07132339477539,
      "learning_rate": 0.0001916502421906898,
      "loss": 6.1404,
      "step": 2180
    },
    {
      "epoch": 0.2181,
      "grad_norm": 3.015045404434204,
      "learning_rate": 0.00019163627295622397,
      "loss": 5.433,
      "step": 2181
    },
    {
      "epoch": 0.2182,
      "grad_norm": 9.134276390075684,
      "learning_rate": 0.00019162229255615624,
      "loss": 5.3898,
      "step": 2182
    },
    {
      "epoch": 0.2183,
      "grad_norm": 3.679075241088867,
      "learning_rate": 0.00019160830099219006,
      "loss": 5.7726,
      "step": 2183
    },
    {
      "epoch": 0.2184,
      "grad_norm": 3.886181354522705,
      "learning_rate": 0.00019159429826603032,
      "loss": 6.4634,
      "step": 2184
    },
    {
      "epoch": 0.2185,
      "grad_norm": 5.669384479522705,
      "learning_rate": 0.00019158028437938317,
      "loss": 5.8504,
      "step": 2185
    },
    {
      "epoch": 0.2186,
      "grad_norm": 6.104648113250732,
      "learning_rate": 0.00019156625933395614,
      "loss": 6.6217,
      "step": 2186
    },
    {
      "epoch": 0.2187,
      "grad_norm": 3.6047110557556152,
      "learning_rate": 0.00019155222313145816,
      "loss": 6.1733,
      "step": 2187
    },
    {
      "epoch": 0.2188,
      "grad_norm": 8.449871063232422,
      "learning_rate": 0.0001915381757735995,
      "loss": 5.3933,
      "step": 2188
    },
    {
      "epoch": 0.2189,
      "grad_norm": 3.024399995803833,
      "learning_rate": 0.00019152411726209176,
      "loss": 5.7491,
      "step": 2189
    },
    {
      "epoch": 0.219,
      "grad_norm": 7.288807392120361,
      "learning_rate": 0.000191510047598648,
      "loss": 5.6984,
      "step": 2190
    },
    {
      "epoch": 0.2191,
      "grad_norm": 2.427769660949707,
      "learning_rate": 0.0001914959667849825,
      "loss": 5.3401,
      "step": 2191
    },
    {
      "epoch": 0.2192,
      "grad_norm": 3.2481372356414795,
      "learning_rate": 0.00019148187482281097,
      "loss": 5.1659,
      "step": 2192
    },
    {
      "epoch": 0.2193,
      "grad_norm": 3.8280303478240967,
      "learning_rate": 0.0001914677717138505,
      "loss": 5.909,
      "step": 2193
    },
    {
      "epoch": 0.2194,
      "grad_norm": 5.139510154724121,
      "learning_rate": 0.00019145365745981948,
      "loss": 5.7935,
      "step": 2194
    },
    {
      "epoch": 0.2195,
      "grad_norm": 10.25840950012207,
      "learning_rate": 0.00019143953206243776,
      "loss": 5.9583,
      "step": 2195
    },
    {
      "epoch": 0.2196,
      "grad_norm": 13.26622486114502,
      "learning_rate": 0.00019142539552342638,
      "loss": 5.5867,
      "step": 2196
    },
    {
      "epoch": 0.2197,
      "grad_norm": 3.9417147636413574,
      "learning_rate": 0.0001914112478445079,
      "loss": 5.7048,
      "step": 2197
    },
    {
      "epoch": 0.2198,
      "grad_norm": 6.5214457511901855,
      "learning_rate": 0.00019139708902740613,
      "loss": 6.1586,
      "step": 2198
    },
    {
      "epoch": 0.2199,
      "grad_norm": 9.684910774230957,
      "learning_rate": 0.0001913829190738463,
      "loss": 5.7299,
      "step": 2199
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.6560118198394775,
      "learning_rate": 0.000191368737985555,
      "loss": 5.3286,
      "step": 2200
    },
    {
      "epoch": 0.2201,
      "grad_norm": 3.177021026611328,
      "learning_rate": 0.0001913545457642601,
      "loss": 5.4489,
      "step": 2201
    },
    {
      "epoch": 0.2202,
      "grad_norm": 6.625150203704834,
      "learning_rate": 0.0001913403424116909,
      "loss": 6.6857,
      "step": 2202
    },
    {
      "epoch": 0.2203,
      "grad_norm": 4.815753936767578,
      "learning_rate": 0.00019132612792957808,
      "loss": 6.4508,
      "step": 2203
    },
    {
      "epoch": 0.2204,
      "grad_norm": 2.842029571533203,
      "learning_rate": 0.00019131190231965356,
      "loss": 5.4747,
      "step": 2204
    },
    {
      "epoch": 0.2205,
      "grad_norm": 3.8514561653137207,
      "learning_rate": 0.00019129766558365074,
      "loss": 5.7218,
      "step": 2205
    },
    {
      "epoch": 0.2206,
      "grad_norm": 2.8640623092651367,
      "learning_rate": 0.0001912834177233043,
      "loss": 4.8595,
      "step": 2206
    },
    {
      "epoch": 0.2207,
      "grad_norm": 7.247208595275879,
      "learning_rate": 0.0001912691587403503,
      "loss": 7.0171,
      "step": 2207
    },
    {
      "epoch": 0.2208,
      "grad_norm": 3.9029417037963867,
      "learning_rate": 0.00019125488863652615,
      "loss": 6.0545,
      "step": 2208
    },
    {
      "epoch": 0.2209,
      "grad_norm": 3.48335862159729,
      "learning_rate": 0.00019124060741357063,
      "loss": 5.5009,
      "step": 2209
    },
    {
      "epoch": 0.221,
      "grad_norm": 6.066159725189209,
      "learning_rate": 0.00019122631507322387,
      "loss": 5.2549,
      "step": 2210
    },
    {
      "epoch": 0.2211,
      "grad_norm": 4.984029293060303,
      "learning_rate": 0.0001912120116172273,
      "loss": 6.0636,
      "step": 2211
    },
    {
      "epoch": 0.2212,
      "grad_norm": 5.726056098937988,
      "learning_rate": 0.00019119769704732382,
      "loss": 5.9135,
      "step": 2212
    },
    {
      "epoch": 0.2213,
      "grad_norm": 15.089937210083008,
      "learning_rate": 0.0001911833713652576,
      "loss": 6.2769,
      "step": 2213
    },
    {
      "epoch": 0.2214,
      "grad_norm": 4.387356281280518,
      "learning_rate": 0.00019116903457277413,
      "loss": 6.4116,
      "step": 2214
    },
    {
      "epoch": 0.2215,
      "grad_norm": 5.75082540512085,
      "learning_rate": 0.00019115468667162038,
      "loss": 5.9129,
      "step": 2215
    },
    {
      "epoch": 0.2216,
      "grad_norm": 2.2348873615264893,
      "learning_rate": 0.00019114032766354453,
      "loss": 5.7272,
      "step": 2216
    },
    {
      "epoch": 0.2217,
      "grad_norm": 5.551529407501221,
      "learning_rate": 0.00019112595755029624,
      "loss": 5.2213,
      "step": 2217
    },
    {
      "epoch": 0.2218,
      "grad_norm": 7.943684101104736,
      "learning_rate": 0.0001911115763336264,
      "loss": 5.5976,
      "step": 2218
    },
    {
      "epoch": 0.2219,
      "grad_norm": 4.21608829498291,
      "learning_rate": 0.0001910971840152874,
      "loss": 5.5089,
      "step": 2219
    },
    {
      "epoch": 0.222,
      "grad_norm": 2.5206539630889893,
      "learning_rate": 0.0001910827805970328,
      "loss": 5.7276,
      "step": 2220
    },
    {
      "epoch": 0.2221,
      "grad_norm": 5.6060709953308105,
      "learning_rate": 0.00019106836608061772,
      "loss": 6.7429,
      "step": 2221
    },
    {
      "epoch": 0.2222,
      "grad_norm": 2.675924062728882,
      "learning_rate": 0.00019105394046779845,
      "loss": 5.2683,
      "step": 2222
    },
    {
      "epoch": 0.2223,
      "grad_norm": 4.024430274963379,
      "learning_rate": 0.00019103950376033276,
      "loss": 6.2632,
      "step": 2223
    },
    {
      "epoch": 0.2224,
      "grad_norm": 6.899057865142822,
      "learning_rate": 0.00019102505595997965,
      "loss": 6.2228,
      "step": 2224
    },
    {
      "epoch": 0.2225,
      "grad_norm": 2.451638698577881,
      "learning_rate": 0.00019101059706849957,
      "loss": 5.7444,
      "step": 2225
    },
    {
      "epoch": 0.2226,
      "grad_norm": 4.191677093505859,
      "learning_rate": 0.00019099612708765434,
      "loss": 6.0255,
      "step": 2226
    },
    {
      "epoch": 0.2227,
      "grad_norm": 9.128498077392578,
      "learning_rate": 0.000190981646019207,
      "loss": 5.5742,
      "step": 2227
    },
    {
      "epoch": 0.2228,
      "grad_norm": 6.205311298370361,
      "learning_rate": 0.0001909671538649221,
      "loss": 5.8614,
      "step": 2228
    },
    {
      "epoch": 0.2229,
      "grad_norm": 5.901429653167725,
      "learning_rate": 0.00019095265062656544,
      "loss": 5.2267,
      "step": 2229
    },
    {
      "epoch": 0.223,
      "grad_norm": 12.29652214050293,
      "learning_rate": 0.00019093813630590418,
      "loss": 6.7576,
      "step": 2230
    },
    {
      "epoch": 0.2231,
      "grad_norm": 15.51838207244873,
      "learning_rate": 0.00019092361090470688,
      "loss": 7.8935,
      "step": 2231
    },
    {
      "epoch": 0.2232,
      "grad_norm": 5.777015686035156,
      "learning_rate": 0.00019090907442474334,
      "loss": 5.7842,
      "step": 2232
    },
    {
      "epoch": 0.2233,
      "grad_norm": 3.347810745239258,
      "learning_rate": 0.00019089452686778488,
      "loss": 5.467,
      "step": 2233
    },
    {
      "epoch": 0.2234,
      "grad_norm": 3.249552011489868,
      "learning_rate": 0.00019087996823560402,
      "loss": 6.1703,
      "step": 2234
    },
    {
      "epoch": 0.2235,
      "grad_norm": 4.4496893882751465,
      "learning_rate": 0.0001908653985299747,
      "loss": 5.7231,
      "step": 2235
    },
    {
      "epoch": 0.2236,
      "grad_norm": 2.9430654048919678,
      "learning_rate": 0.0001908508177526722,
      "loss": 5.7638,
      "step": 2236
    },
    {
      "epoch": 0.2237,
      "grad_norm": 2.4572930335998535,
      "learning_rate": 0.00019083622590547312,
      "loss": 5.7257,
      "step": 2237
    },
    {
      "epoch": 0.2238,
      "grad_norm": 3.033629894256592,
      "learning_rate": 0.00019082162299015546,
      "loss": 5.8998,
      "step": 2238
    },
    {
      "epoch": 0.2239,
      "grad_norm": 3.0344114303588867,
      "learning_rate": 0.00019080700900849851,
      "loss": 5.4691,
      "step": 2239
    },
    {
      "epoch": 0.224,
      "grad_norm": 2.6494901180267334,
      "learning_rate": 0.000190792383962283,
      "loss": 6.0083,
      "step": 2240
    },
    {
      "epoch": 0.2241,
      "grad_norm": 4.027078151702881,
      "learning_rate": 0.00019077774785329087,
      "loss": 5.3559,
      "step": 2241
    },
    {
      "epoch": 0.2242,
      "grad_norm": 4.438051223754883,
      "learning_rate": 0.00019076310068330554,
      "loss": 6.0308,
      "step": 2242
    },
    {
      "epoch": 0.2243,
      "grad_norm": 7.780203342437744,
      "learning_rate": 0.0001907484424541117,
      "loss": 5.6859,
      "step": 2243
    },
    {
      "epoch": 0.2244,
      "grad_norm": 3.225978136062622,
      "learning_rate": 0.00019073377316749542,
      "loss": 5.3394,
      "step": 2244
    },
    {
      "epoch": 0.2245,
      "grad_norm": 7.066761493682861,
      "learning_rate": 0.00019071909282524413,
      "loss": 5.4833,
      "step": 2245
    },
    {
      "epoch": 0.2246,
      "grad_norm": 2.4651498794555664,
      "learning_rate": 0.0001907044014291465,
      "loss": 5.6771,
      "step": 2246
    },
    {
      "epoch": 0.2247,
      "grad_norm": 3.0828115940093994,
      "learning_rate": 0.0001906896989809927,
      "loss": 5.7039,
      "step": 2247
    },
    {
      "epoch": 0.2248,
      "grad_norm": 14.563668251037598,
      "learning_rate": 0.00019067498548257423,
      "loss": 5.9561,
      "step": 2248
    },
    {
      "epoch": 0.2249,
      "grad_norm": 3.646981954574585,
      "learning_rate": 0.00019066026093568378,
      "loss": 6.0664,
      "step": 2249
    },
    {
      "epoch": 0.225,
      "grad_norm": 3.2434334754943848,
      "learning_rate": 0.00019064552534211554,
      "loss": 5.1326,
      "step": 2250
    },
    {
      "epoch": 0.2251,
      "grad_norm": 2.9518275260925293,
      "learning_rate": 0.000190630778703665,
      "loss": 5.7527,
      "step": 2251
    },
    {
      "epoch": 0.2252,
      "grad_norm": 5.791699409484863,
      "learning_rate": 0.00019061602102212898,
      "loss": 5.2404,
      "step": 2252
    },
    {
      "epoch": 0.2253,
      "grad_norm": 7.1183881759643555,
      "learning_rate": 0.0001906012522993057,
      "loss": 6.1563,
      "step": 2253
    },
    {
      "epoch": 0.2254,
      "grad_norm": 4.055431365966797,
      "learning_rate": 0.0001905864725369946,
      "loss": 5.597,
      "step": 2254
    },
    {
      "epoch": 0.2255,
      "grad_norm": 5.878694534301758,
      "learning_rate": 0.00019057168173699664,
      "loss": 5.3335,
      "step": 2255
    },
    {
      "epoch": 0.2256,
      "grad_norm": 2.6647214889526367,
      "learning_rate": 0.00019055687990111398,
      "loss": 5.3412,
      "step": 2256
    },
    {
      "epoch": 0.2257,
      "grad_norm": 4.603827476501465,
      "learning_rate": 0.0001905420670311502,
      "loss": 5.6859,
      "step": 2257
    },
    {
      "epoch": 0.2258,
      "grad_norm": 2.9110491275787354,
      "learning_rate": 0.00019052724312891014,
      "loss": 5.8978,
      "step": 2258
    },
    {
      "epoch": 0.2259,
      "grad_norm": 2.690671920776367,
      "learning_rate": 0.00019051240819620014,
      "loss": 5.2453,
      "step": 2259
    },
    {
      "epoch": 0.226,
      "grad_norm": 8.688986778259277,
      "learning_rate": 0.0001904975622348278,
      "loss": 5.9656,
      "step": 2260
    },
    {
      "epoch": 0.2261,
      "grad_norm": 2.859498977661133,
      "learning_rate": 0.00019048270524660196,
      "loss": 5.9586,
      "step": 2261
    },
    {
      "epoch": 0.2262,
      "grad_norm": 4.05758810043335,
      "learning_rate": 0.00019046783723333297,
      "loss": 5.4656,
      "step": 2262
    },
    {
      "epoch": 0.2263,
      "grad_norm": 20.37089729309082,
      "learning_rate": 0.00019045295819683242,
      "loss": 6.6164,
      "step": 2263
    },
    {
      "epoch": 0.2264,
      "grad_norm": 4.500487327575684,
      "learning_rate": 0.0001904380681389133,
      "loss": 5.3435,
      "step": 2264
    },
    {
      "epoch": 0.2265,
      "grad_norm": 6.450089454650879,
      "learning_rate": 0.00019042316706138987,
      "loss": 5.7532,
      "step": 2265
    },
    {
      "epoch": 0.2266,
      "grad_norm": 8.885701179504395,
      "learning_rate": 0.00019040825496607786,
      "loss": 5.7184,
      "step": 2266
    },
    {
      "epoch": 0.2267,
      "grad_norm": 6.815227508544922,
      "learning_rate": 0.00019039333185479418,
      "loss": 5.6092,
      "step": 2267
    },
    {
      "epoch": 0.2268,
      "grad_norm": 5.487820625305176,
      "learning_rate": 0.0001903783977293572,
      "loss": 6.4299,
      "step": 2268
    },
    {
      "epoch": 0.2269,
      "grad_norm": 3.7698793411254883,
      "learning_rate": 0.00019036345259158667,
      "loss": 5.5314,
      "step": 2269
    },
    {
      "epoch": 0.227,
      "grad_norm": 2.1771044731140137,
      "learning_rate": 0.0001903484964433035,
      "loss": 5.5775,
      "step": 2270
    },
    {
      "epoch": 0.2271,
      "grad_norm": 4.266638278961182,
      "learning_rate": 0.0001903335292863301,
      "loss": 5.6345,
      "step": 2271
    },
    {
      "epoch": 0.2272,
      "grad_norm": 3.394643545150757,
      "learning_rate": 0.00019031855112249015,
      "loss": 5.7228,
      "step": 2272
    },
    {
      "epoch": 0.2273,
      "grad_norm": 15.891701698303223,
      "learning_rate": 0.00019030356195360874,
      "loss": 6.2976,
      "step": 2273
    },
    {
      "epoch": 0.2274,
      "grad_norm": 9.375031471252441,
      "learning_rate": 0.0001902885617815122,
      "loss": 5.2705,
      "step": 2274
    },
    {
      "epoch": 0.2275,
      "grad_norm": 3.1531882286071777,
      "learning_rate": 0.0001902735506080283,
      "loss": 5.8525,
      "step": 2275
    },
    {
      "epoch": 0.2276,
      "grad_norm": 3.5101072788238525,
      "learning_rate": 0.00019025852843498607,
      "loss": 5.2543,
      "step": 2276
    },
    {
      "epoch": 0.2277,
      "grad_norm": 3.972175121307373,
      "learning_rate": 0.00019024349526421594,
      "loss": 5.3958,
      "step": 2277
    },
    {
      "epoch": 0.2278,
      "grad_norm": 13.136713981628418,
      "learning_rate": 0.00019022845109754966,
      "loss": 6.59,
      "step": 2278
    },
    {
      "epoch": 0.2279,
      "grad_norm": 11.691570281982422,
      "learning_rate": 0.00019021339593682028,
      "loss": 5.4908,
      "step": 2279
    },
    {
      "epoch": 0.228,
      "grad_norm": 5.638294219970703,
      "learning_rate": 0.00019019832978386228,
      "loss": 5.6697,
      "step": 2280
    },
    {
      "epoch": 0.2281,
      "grad_norm": 10.637971878051758,
      "learning_rate": 0.0001901832526405114,
      "loss": 5.8554,
      "step": 2281
    },
    {
      "epoch": 0.2282,
      "grad_norm": 3.2279505729675293,
      "learning_rate": 0.00019016816450860474,
      "loss": 6.4002,
      "step": 2282
    },
    {
      "epoch": 0.2283,
      "grad_norm": 2.3993775844573975,
      "learning_rate": 0.0001901530653899807,
      "loss": 5.9481,
      "step": 2283
    },
    {
      "epoch": 0.2284,
      "grad_norm": 3.3658413887023926,
      "learning_rate": 0.00019013795528647912,
      "loss": 5.3375,
      "step": 2284
    },
    {
      "epoch": 0.2285,
      "grad_norm": 4.589962959289551,
      "learning_rate": 0.00019012283419994115,
      "loss": 5.8546,
      "step": 2285
    },
    {
      "epoch": 0.2286,
      "grad_norm": 3.3468031883239746,
      "learning_rate": 0.00019010770213220916,
      "loss": 5.6927,
      "step": 2286
    },
    {
      "epoch": 0.2287,
      "grad_norm": 2.6899735927581787,
      "learning_rate": 0.000190092559085127,
      "loss": 5.4375,
      "step": 2287
    },
    {
      "epoch": 0.2288,
      "grad_norm": 2.238710641860962,
      "learning_rate": 0.00019007740506053983,
      "loss": 5.1091,
      "step": 2288
    },
    {
      "epoch": 0.2289,
      "grad_norm": 2.4347550868988037,
      "learning_rate": 0.00019006224006029406,
      "loss": 6.2981,
      "step": 2289
    },
    {
      "epoch": 0.229,
      "grad_norm": 2.3276100158691406,
      "learning_rate": 0.0001900470640862375,
      "loss": 6.1326,
      "step": 2290
    },
    {
      "epoch": 0.2291,
      "grad_norm": 5.960839748382568,
      "learning_rate": 0.00019003187714021938,
      "loss": 7.5013,
      "step": 2291
    },
    {
      "epoch": 0.2292,
      "grad_norm": 5.032075881958008,
      "learning_rate": 0.00019001667922409008,
      "loss": 5.6902,
      "step": 2292
    },
    {
      "epoch": 0.2293,
      "grad_norm": 17.05318832397461,
      "learning_rate": 0.00019000147033970148,
      "loss": 6.0741,
      "step": 2293
    },
    {
      "epoch": 0.2294,
      "grad_norm": 2.8550822734832764,
      "learning_rate": 0.00018998625048890672,
      "loss": 5.6054,
      "step": 2294
    },
    {
      "epoch": 0.2295,
      "grad_norm": 16.890901565551758,
      "learning_rate": 0.0001899710196735603,
      "loss": 5.0672,
      "step": 2295
    },
    {
      "epoch": 0.2296,
      "grad_norm": 10.670011520385742,
      "learning_rate": 0.00018995577789551803,
      "loss": 5.8986,
      "step": 2296
    },
    {
      "epoch": 0.2297,
      "grad_norm": 5.057382583618164,
      "learning_rate": 0.0001899405251566371,
      "loss": 6.7953,
      "step": 2297
    },
    {
      "epoch": 0.2298,
      "grad_norm": 12.276496887207031,
      "learning_rate": 0.000189925261458776,
      "loss": 6.1987,
      "step": 2298
    },
    {
      "epoch": 0.2299,
      "grad_norm": 14.621031761169434,
      "learning_rate": 0.00018990998680379456,
      "loss": 6.3887,
      "step": 2299
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.61982250213623,
      "learning_rate": 0.00018989470119355398,
      "loss": 5.1001,
      "step": 2300
    },
    {
      "epoch": 0.2301,
      "grad_norm": 5.378514766693115,
      "learning_rate": 0.0001898794046299167,
      "loss": 6.0509,
      "step": 2301
    },
    {
      "epoch": 0.2302,
      "grad_norm": 3.539013147354126,
      "learning_rate": 0.00018986409711474665,
      "loss": 6.897,
      "step": 2302
    },
    {
      "epoch": 0.2303,
      "grad_norm": 4.065858364105225,
      "learning_rate": 0.00018984877864990888,
      "loss": 6.0672,
      "step": 2303
    },
    {
      "epoch": 0.2304,
      "grad_norm": 2.7219936847686768,
      "learning_rate": 0.00018983344923727003,
      "loss": 5.55,
      "step": 2304
    },
    {
      "epoch": 0.2305,
      "grad_norm": 2.7769041061401367,
      "learning_rate": 0.00018981810887869785,
      "loss": 5.6825,
      "step": 2305
    },
    {
      "epoch": 0.2306,
      "grad_norm": 3.417823553085327,
      "learning_rate": 0.00018980275757606157,
      "loss": 5.4881,
      "step": 2306
    },
    {
      "epoch": 0.2307,
      "grad_norm": 4.212622165679932,
      "learning_rate": 0.0001897873953312317,
      "loss": 4.9454,
      "step": 2307
    },
    {
      "epoch": 0.2308,
      "grad_norm": 4.082922458648682,
      "learning_rate": 0.00018977202214608,
      "loss": 5.204,
      "step": 2308
    },
    {
      "epoch": 0.2309,
      "grad_norm": 2.316669225692749,
      "learning_rate": 0.00018975663802247976,
      "loss": 5.384,
      "step": 2309
    },
    {
      "epoch": 0.231,
      "grad_norm": 3.167301654815674,
      "learning_rate": 0.0001897412429623054,
      "loss": 6.0629,
      "step": 2310
    },
    {
      "epoch": 0.2311,
      "grad_norm": 3.9662537574768066,
      "learning_rate": 0.00018972583696743285,
      "loss": 5.5338,
      "step": 2311
    },
    {
      "epoch": 0.2312,
      "grad_norm": 2.5631279945373535,
      "learning_rate": 0.00018971042003973924,
      "loss": 5.4802,
      "step": 2312
    },
    {
      "epoch": 0.2313,
      "grad_norm": 7.085266590118408,
      "learning_rate": 0.000189694992181103,
      "loss": 5.1911,
      "step": 2313
    },
    {
      "epoch": 0.2314,
      "grad_norm": 4.693726539611816,
      "learning_rate": 0.00018967955339340407,
      "loss": 6.6567,
      "step": 2314
    },
    {
      "epoch": 0.2315,
      "grad_norm": 4.318178176879883,
      "learning_rate": 0.00018966410367852362,
      "loss": 5.8746,
      "step": 2315
    },
    {
      "epoch": 0.2316,
      "grad_norm": 4.749048709869385,
      "learning_rate": 0.00018964864303834406,
      "loss": 5.9406,
      "step": 2316
    },
    {
      "epoch": 0.2317,
      "grad_norm": 3.7709028720855713,
      "learning_rate": 0.0001896331714747493,
      "loss": 5.5242,
      "step": 2317
    },
    {
      "epoch": 0.2318,
      "grad_norm": 18.639596939086914,
      "learning_rate": 0.0001896176889896245,
      "loss": 6.2839,
      "step": 2318
    },
    {
      "epoch": 0.2319,
      "grad_norm": 4.087204933166504,
      "learning_rate": 0.0001896021955848561,
      "loss": 5.545,
      "step": 2319
    },
    {
      "epoch": 0.232,
      "grad_norm": 2.2883448600769043,
      "learning_rate": 0.00018958669126233199,
      "loss": 5.7491,
      "step": 2320
    },
    {
      "epoch": 0.2321,
      "grad_norm": 2.916611433029175,
      "learning_rate": 0.0001895711760239413,
      "loss": 6.2366,
      "step": 2321
    },
    {
      "epoch": 0.2322,
      "grad_norm": 6.51270055770874,
      "learning_rate": 0.0001895556498715745,
      "loss": 5.7553,
      "step": 2322
    },
    {
      "epoch": 0.2323,
      "grad_norm": 5.309991836547852,
      "learning_rate": 0.0001895401128071234,
      "loss": 7.0502,
      "step": 2323
    },
    {
      "epoch": 0.2324,
      "grad_norm": 10.327960014343262,
      "learning_rate": 0.00018952456483248119,
      "loss": 5.7532,
      "step": 2324
    },
    {
      "epoch": 0.2325,
      "grad_norm": 2.12770938873291,
      "learning_rate": 0.00018950900594954227,
      "loss": 5.5172,
      "step": 2325
    },
    {
      "epoch": 0.2326,
      "grad_norm": 2.479365348815918,
      "learning_rate": 0.00018949343616020252,
      "loss": 5.6682,
      "step": 2326
    },
    {
      "epoch": 0.2327,
      "grad_norm": 3.0665230751037598,
      "learning_rate": 0.00018947785546635904,
      "loss": 5.167,
      "step": 2327
    },
    {
      "epoch": 0.2328,
      "grad_norm": 2.8029167652130127,
      "learning_rate": 0.00018946226386991027,
      "loss": 5.343,
      "step": 2328
    },
    {
      "epoch": 0.2329,
      "grad_norm": 4.176616191864014,
      "learning_rate": 0.000189446661372756,
      "loss": 5.496,
      "step": 2329
    },
    {
      "epoch": 0.233,
      "grad_norm": 4.510972023010254,
      "learning_rate": 0.0001894310479767974,
      "loss": 5.8933,
      "step": 2330
    },
    {
      "epoch": 0.2331,
      "grad_norm": 5.6890764236450195,
      "learning_rate": 0.0001894154236839368,
      "loss": 5.3204,
      "step": 2331
    },
    {
      "epoch": 0.2332,
      "grad_norm": 2.2775251865386963,
      "learning_rate": 0.00018939978849607814,
      "loss": 5.1284,
      "step": 2332
    },
    {
      "epoch": 0.2333,
      "grad_norm": 3.180778741836548,
      "learning_rate": 0.0001893841424151264,
      "loss": 5.4455,
      "step": 2333
    },
    {
      "epoch": 0.2334,
      "grad_norm": 7.31488561630249,
      "learning_rate": 0.000189368485442988,
      "loss": 7.9562,
      "step": 2334
    },
    {
      "epoch": 0.2335,
      "grad_norm": 3.004244565963745,
      "learning_rate": 0.00018935281758157078,
      "loss": 5.3956,
      "step": 2335
    },
    {
      "epoch": 0.2336,
      "grad_norm": 6.462453842163086,
      "learning_rate": 0.00018933713883278376,
      "loss": 5.5919,
      "step": 2336
    },
    {
      "epoch": 0.2337,
      "grad_norm": 3.4786221981048584,
      "learning_rate": 0.0001893214491985374,
      "loss": 5.557,
      "step": 2337
    },
    {
      "epoch": 0.2338,
      "grad_norm": 3.1216537952423096,
      "learning_rate": 0.00018930574868074334,
      "loss": 5.7071,
      "step": 2338
    },
    {
      "epoch": 0.2339,
      "grad_norm": 3.492985963821411,
      "learning_rate": 0.0001892900372813147,
      "loss": 5.7214,
      "step": 2339
    },
    {
      "epoch": 0.234,
      "grad_norm": 5.337524890899658,
      "learning_rate": 0.00018927431500216586,
      "loss": 5.9209,
      "step": 2340
    },
    {
      "epoch": 0.2341,
      "grad_norm": 5.134214878082275,
      "learning_rate": 0.00018925858184521256,
      "loss": 5.8694,
      "step": 2341
    },
    {
      "epoch": 0.2342,
      "grad_norm": 2.7308712005615234,
      "learning_rate": 0.0001892428378123718,
      "loss": 5.375,
      "step": 2342
    },
    {
      "epoch": 0.2343,
      "grad_norm": 2.081486701965332,
      "learning_rate": 0.00018922708290556198,
      "loss": 5.18,
      "step": 2343
    },
    {
      "epoch": 0.2344,
      "grad_norm": 4.461536884307861,
      "learning_rate": 0.0001892113171267027,
      "loss": 5.5759,
      "step": 2344
    },
    {
      "epoch": 0.2345,
      "grad_norm": 2.615919589996338,
      "learning_rate": 0.0001891955404777151,
      "loss": 5.1462,
      "step": 2345
    },
    {
      "epoch": 0.2346,
      "grad_norm": 3.4781577587127686,
      "learning_rate": 0.00018917975296052142,
      "loss": 5.8474,
      "step": 2346
    },
    {
      "epoch": 0.2347,
      "grad_norm": 3.2197489738464355,
      "learning_rate": 0.00018916395457704534,
      "loss": 6.0576,
      "step": 2347
    },
    {
      "epoch": 0.2348,
      "grad_norm": 3.66081166267395,
      "learning_rate": 0.00018914814532921187,
      "loss": 5.2507,
      "step": 2348
    },
    {
      "epoch": 0.2349,
      "grad_norm": 5.189724922180176,
      "learning_rate": 0.00018913232521894732,
      "loss": 6.2196,
      "step": 2349
    },
    {
      "epoch": 0.235,
      "grad_norm": 4.897153377532959,
      "learning_rate": 0.00018911649424817933,
      "loss": 5.2196,
      "step": 2350
    },
    {
      "epoch": 0.2351,
      "grad_norm": 3.251350164413452,
      "learning_rate": 0.0001891006524188368,
      "loss": 6.0121,
      "step": 2351
    },
    {
      "epoch": 0.2352,
      "grad_norm": 2.6665666103363037,
      "learning_rate": 0.00018908479973285005,
      "loss": 5.7878,
      "step": 2352
    },
    {
      "epoch": 0.2353,
      "grad_norm": 4.858261585235596,
      "learning_rate": 0.00018906893619215066,
      "loss": 5.3429,
      "step": 2353
    },
    {
      "epoch": 0.2354,
      "grad_norm": 3.4894914627075195,
      "learning_rate": 0.0001890530617986716,
      "loss": 6.2551,
      "step": 2354
    },
    {
      "epoch": 0.2355,
      "grad_norm": 6.1850128173828125,
      "learning_rate": 0.00018903717655434707,
      "loss": 6.2583,
      "step": 2355
    },
    {
      "epoch": 0.2356,
      "grad_norm": 3.000059127807617,
      "learning_rate": 0.00018902128046111266,
      "loss": 6.1749,
      "step": 2356
    },
    {
      "epoch": 0.2357,
      "grad_norm": 16.284339904785156,
      "learning_rate": 0.00018900537352090524,
      "loss": 6.2144,
      "step": 2357
    },
    {
      "epoch": 0.2358,
      "grad_norm": 17.62369155883789,
      "learning_rate": 0.00018898945573566308,
      "loss": 5.2047,
      "step": 2358
    },
    {
      "epoch": 0.2359,
      "grad_norm": 6.461252689361572,
      "learning_rate": 0.00018897352710732564,
      "loss": 5.4265,
      "step": 2359
    },
    {
      "epoch": 0.236,
      "grad_norm": 10.57401180267334,
      "learning_rate": 0.00018895758763783383,
      "loss": 5.9469,
      "step": 2360
    },
    {
      "epoch": 0.2361,
      "grad_norm": 5.654270172119141,
      "learning_rate": 0.00018894163732912977,
      "loss": 5.8133,
      "step": 2361
    },
    {
      "epoch": 0.2362,
      "grad_norm": 2.7907872200012207,
      "learning_rate": 0.000188925676183157,
      "loss": 6.7062,
      "step": 2362
    },
    {
      "epoch": 0.2363,
      "grad_norm": 8.737727165222168,
      "learning_rate": 0.00018890970420186033,
      "loss": 5.3925,
      "step": 2363
    },
    {
      "epoch": 0.2364,
      "grad_norm": 4.051711559295654,
      "learning_rate": 0.0001888937213871859,
      "loss": 5.9801,
      "step": 2364
    },
    {
      "epoch": 0.2365,
      "grad_norm": 2.6220455169677734,
      "learning_rate": 0.00018887772774108116,
      "loss": 5.7695,
      "step": 2365
    },
    {
      "epoch": 0.2366,
      "grad_norm": 2.190054416656494,
      "learning_rate": 0.0001888617232654949,
      "loss": 5.1094,
      "step": 2366
    },
    {
      "epoch": 0.2367,
      "grad_norm": 2.141339063644409,
      "learning_rate": 0.00018884570796237718,
      "loss": 5.1179,
      "step": 2367
    },
    {
      "epoch": 0.2368,
      "grad_norm": 2.5767292976379395,
      "learning_rate": 0.00018882968183367947,
      "loss": 5.7104,
      "step": 2368
    },
    {
      "epoch": 0.2369,
      "grad_norm": 3.3019495010375977,
      "learning_rate": 0.00018881364488135448,
      "loss": 5.0353,
      "step": 2369
    },
    {
      "epoch": 0.237,
      "grad_norm": 5.974609851837158,
      "learning_rate": 0.00018879759710735622,
      "loss": 5.9977,
      "step": 2370
    },
    {
      "epoch": 0.2371,
      "grad_norm": 8.442816734313965,
      "learning_rate": 0.00018878153851364013,
      "loss": 6.8498,
      "step": 2371
    },
    {
      "epoch": 0.2372,
      "grad_norm": 8.090008735656738,
      "learning_rate": 0.00018876546910216288,
      "loss": 5.7576,
      "step": 2372
    },
    {
      "epoch": 0.2373,
      "grad_norm": 3.603039264678955,
      "learning_rate": 0.00018874938887488248,
      "loss": 5.0364,
      "step": 2373
    },
    {
      "epoch": 0.2374,
      "grad_norm": 4.915457725524902,
      "learning_rate": 0.00018873329783375824,
      "loss": 5.5739,
      "step": 2374
    },
    {
      "epoch": 0.2375,
      "grad_norm": 5.040452003479004,
      "learning_rate": 0.0001887171959807508,
      "loss": 5.9088,
      "step": 2375
    },
    {
      "epoch": 0.2376,
      "grad_norm": 6.3807053565979,
      "learning_rate": 0.00018870108331782217,
      "loss": 5.4949,
      "step": 2376
    },
    {
      "epoch": 0.2377,
      "grad_norm": 3.994567394256592,
      "learning_rate": 0.0001886849598469356,
      "loss": 5.2345,
      "step": 2377
    },
    {
      "epoch": 0.2378,
      "grad_norm": 2.167919158935547,
      "learning_rate": 0.00018866882557005567,
      "loss": 5.8603,
      "step": 2378
    },
    {
      "epoch": 0.2379,
      "grad_norm": 2.9735541343688965,
      "learning_rate": 0.00018865268048914828,
      "loss": 6.0026,
      "step": 2379
    },
    {
      "epoch": 0.238,
      "grad_norm": 19.93646812438965,
      "learning_rate": 0.0001886365246061807,
      "loss": 6.3471,
      "step": 2380
    },
    {
      "epoch": 0.2381,
      "grad_norm": 3.006359577178955,
      "learning_rate": 0.00018862035792312147,
      "loss": 5.1467,
      "step": 2381
    },
    {
      "epoch": 0.2382,
      "grad_norm": 7.364262580871582,
      "learning_rate": 0.00018860418044194045,
      "loss": 5.794,
      "step": 2382
    },
    {
      "epoch": 0.2383,
      "grad_norm": 14.928060531616211,
      "learning_rate": 0.00018858799216460881,
      "loss": 6.2815,
      "step": 2383
    },
    {
      "epoch": 0.2384,
      "grad_norm": 2.56388521194458,
      "learning_rate": 0.00018857179309309901,
      "loss": 5.6208,
      "step": 2384
    },
    {
      "epoch": 0.2385,
      "grad_norm": 3.4603898525238037,
      "learning_rate": 0.00018855558322938493,
      "loss": 5.6877,
      "step": 2385
    },
    {
      "epoch": 0.2386,
      "grad_norm": 3.8272056579589844,
      "learning_rate": 0.0001885393625754416,
      "loss": 5.9863,
      "step": 2386
    },
    {
      "epoch": 0.2387,
      "grad_norm": 2.627397298812866,
      "learning_rate": 0.00018852313113324552,
      "loss": 5.3216,
      "step": 2387
    },
    {
      "epoch": 0.2388,
      "grad_norm": 4.629575729370117,
      "learning_rate": 0.00018850688890477445,
      "loss": 5.8164,
      "step": 2388
    },
    {
      "epoch": 0.2389,
      "grad_norm": 5.885373115539551,
      "learning_rate": 0.00018849063589200743,
      "loss": 6.4996,
      "step": 2389
    },
    {
      "epoch": 0.239,
      "grad_norm": 2.85870099067688,
      "learning_rate": 0.00018847437209692486,
      "loss": 5.7626,
      "step": 2390
    },
    {
      "epoch": 0.2391,
      "grad_norm": 3.380288600921631,
      "learning_rate": 0.0001884580975215084,
      "loss": 5.7413,
      "step": 2391
    },
    {
      "epoch": 0.2392,
      "grad_norm": 2.6702635288238525,
      "learning_rate": 0.0001884418121677411,
      "loss": 5.8197,
      "step": 2392
    },
    {
      "epoch": 0.2393,
      "grad_norm": 2.1600654125213623,
      "learning_rate": 0.00018842551603760724,
      "loss": 5.5021,
      "step": 2393
    },
    {
      "epoch": 0.2394,
      "grad_norm": 9.882226943969727,
      "learning_rate": 0.0001884092091330925,
      "loss": 5.5153,
      "step": 2394
    },
    {
      "epoch": 0.2395,
      "grad_norm": 10.801408767700195,
      "learning_rate": 0.00018839289145618378,
      "loss": 6.5872,
      "step": 2395
    },
    {
      "epoch": 0.2396,
      "grad_norm": 8.927903175354004,
      "learning_rate": 0.00018837656300886937,
      "loss": 5.6278,
      "step": 2396
    },
    {
      "epoch": 0.2397,
      "grad_norm": 4.197044849395752,
      "learning_rate": 0.00018836022379313883,
      "loss": 6.1347,
      "step": 2397
    },
    {
      "epoch": 0.2398,
      "grad_norm": 10.900369644165039,
      "learning_rate": 0.000188343873810983,
      "loss": 6.1165,
      "step": 2398
    },
    {
      "epoch": 0.2399,
      "grad_norm": 7.059217929840088,
      "learning_rate": 0.00018832751306439418,
      "loss": 5.2456,
      "step": 2399
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.819919109344482,
      "learning_rate": 0.0001883111415553658,
      "loss": 6.3366,
      "step": 2400
    },
    {
      "epoch": 0.2401,
      "grad_norm": 2.9006688594818115,
      "learning_rate": 0.00018829475928589271,
      "loss": 5.3954,
      "step": 2401
    },
    {
      "epoch": 0.2402,
      "grad_norm": 2.19280743598938,
      "learning_rate": 0.00018827836625797103,
      "loss": 5.6597,
      "step": 2402
    },
    {
      "epoch": 0.2403,
      "grad_norm": 6.053617477416992,
      "learning_rate": 0.00018826196247359817,
      "loss": 6.2763,
      "step": 2403
    },
    {
      "epoch": 0.2404,
      "grad_norm": 8.020133018493652,
      "learning_rate": 0.00018824554793477294,
      "loss": 6.649,
      "step": 2404
    },
    {
      "epoch": 0.2405,
      "grad_norm": 2.84061336517334,
      "learning_rate": 0.00018822912264349534,
      "loss": 5.8988,
      "step": 2405
    },
    {
      "epoch": 0.2406,
      "grad_norm": 6.603244781494141,
      "learning_rate": 0.00018821268660176678,
      "loss": 6.1664,
      "step": 2406
    },
    {
      "epoch": 0.2407,
      "grad_norm": 4.431100845336914,
      "learning_rate": 0.00018819623981158995,
      "loss": 5.7492,
      "step": 2407
    },
    {
      "epoch": 0.2408,
      "grad_norm": 4.728091239929199,
      "learning_rate": 0.00018817978227496883,
      "loss": 5.7052,
      "step": 2408
    },
    {
      "epoch": 0.2409,
      "grad_norm": 4.123027801513672,
      "learning_rate": 0.0001881633139939087,
      "loss": 6.3313,
      "step": 2409
    },
    {
      "epoch": 0.241,
      "grad_norm": 3.6149940490722656,
      "learning_rate": 0.0001881468349704162,
      "loss": 5.0161,
      "step": 2410
    },
    {
      "epoch": 0.2411,
      "grad_norm": 4.931407451629639,
      "learning_rate": 0.0001881303452064992,
      "loss": 5.4964,
      "step": 2411
    },
    {
      "epoch": 0.2412,
      "grad_norm": 10.269768714904785,
      "learning_rate": 0.00018811384470416705,
      "loss": 5.9891,
      "step": 2412
    },
    {
      "epoch": 0.2413,
      "grad_norm": 4.332967758178711,
      "learning_rate": 0.00018809733346543013,
      "loss": 5.6607,
      "step": 2413
    },
    {
      "epoch": 0.2414,
      "grad_norm": 3.3143961429595947,
      "learning_rate": 0.00018808081149230036,
      "loss": 5.2521,
      "step": 2414
    },
    {
      "epoch": 0.2415,
      "grad_norm": 2.8529300689697266,
      "learning_rate": 0.00018806427878679093,
      "loss": 6.0135,
      "step": 2415
    },
    {
      "epoch": 0.2416,
      "grad_norm": 4.238088130950928,
      "learning_rate": 0.0001880477353509162,
      "loss": 5.7292,
      "step": 2416
    },
    {
      "epoch": 0.2417,
      "grad_norm": 2.844189167022705,
      "learning_rate": 0.00018803118118669202,
      "loss": 5.4735,
      "step": 2417
    },
    {
      "epoch": 0.2418,
      "grad_norm": 3.3887200355529785,
      "learning_rate": 0.00018801461629613546,
      "loss": 5.3326,
      "step": 2418
    },
    {
      "epoch": 0.2419,
      "grad_norm": 3.5358822345733643,
      "learning_rate": 0.00018799804068126485,
      "loss": 5.6025,
      "step": 2419
    },
    {
      "epoch": 0.242,
      "grad_norm": 1.9635536670684814,
      "learning_rate": 0.0001879814543440999,
      "loss": 5.392,
      "step": 2420
    },
    {
      "epoch": 0.2421,
      "grad_norm": 6.321374893188477,
      "learning_rate": 0.00018796485728666165,
      "loss": 5.0243,
      "step": 2421
    },
    {
      "epoch": 0.2422,
      "grad_norm": 10.934194564819336,
      "learning_rate": 0.00018794824951097236,
      "loss": 5.8263,
      "step": 2422
    },
    {
      "epoch": 0.2423,
      "grad_norm": 24.707468032836914,
      "learning_rate": 0.00018793163101905563,
      "loss": 6.0105,
      "step": 2423
    },
    {
      "epoch": 0.2424,
      "grad_norm": 4.601532459259033,
      "learning_rate": 0.0001879150018129364,
      "loss": 5.7785,
      "step": 2424
    },
    {
      "epoch": 0.2425,
      "grad_norm": 6.792133808135986,
      "learning_rate": 0.00018789836189464086,
      "loss": 5.3047,
      "step": 2425
    },
    {
      "epoch": 0.2426,
      "grad_norm": 2.5471675395965576,
      "learning_rate": 0.00018788171126619653,
      "loss": 5.5365,
      "step": 2426
    },
    {
      "epoch": 0.2427,
      "grad_norm": 5.174701690673828,
      "learning_rate": 0.0001878650499296323,
      "loss": 5.562,
      "step": 2427
    },
    {
      "epoch": 0.2428,
      "grad_norm": 2.4794697761535645,
      "learning_rate": 0.00018784837788697823,
      "loss": 5.4086,
      "step": 2428
    },
    {
      "epoch": 0.2429,
      "grad_norm": 4.125171661376953,
      "learning_rate": 0.00018783169514026578,
      "loss": 5.2932,
      "step": 2429
    },
    {
      "epoch": 0.243,
      "grad_norm": 7.842416763305664,
      "learning_rate": 0.00018781500169152773,
      "loss": 5.3624,
      "step": 2430
    },
    {
      "epoch": 0.2431,
      "grad_norm": 3.1223740577697754,
      "learning_rate": 0.00018779829754279805,
      "loss": 6.0807,
      "step": 2431
    },
    {
      "epoch": 0.2432,
      "grad_norm": 2.3767433166503906,
      "learning_rate": 0.00018778158269611218,
      "loss": 5.4952,
      "step": 2432
    },
    {
      "epoch": 0.2433,
      "grad_norm": 8.125616073608398,
      "learning_rate": 0.00018776485715350671,
      "loss": 6.0719,
      "step": 2433
    },
    {
      "epoch": 0.2434,
      "grad_norm": 3.2539377212524414,
      "learning_rate": 0.00018774812091701962,
      "loss": 5.4488,
      "step": 2434
    },
    {
      "epoch": 0.2435,
      "grad_norm": 5.542023658752441,
      "learning_rate": 0.00018773137398869015,
      "loss": 5.2682,
      "step": 2435
    },
    {
      "epoch": 0.2436,
      "grad_norm": 5.097930908203125,
      "learning_rate": 0.00018771461637055888,
      "loss": 5.6298,
      "step": 2436
    },
    {
      "epoch": 0.2437,
      "grad_norm": 5.071180820465088,
      "learning_rate": 0.0001876978480646677,
      "loss": 5.7238,
      "step": 2437
    },
    {
      "epoch": 0.2438,
      "grad_norm": 4.066540241241455,
      "learning_rate": 0.00018768106907305973,
      "loss": 5.2561,
      "step": 2438
    },
    {
      "epoch": 0.2439,
      "grad_norm": 2.5828068256378174,
      "learning_rate": 0.00018766427939777945,
      "loss": 5.6412,
      "step": 2439
    },
    {
      "epoch": 0.244,
      "grad_norm": 6.807405948638916,
      "learning_rate": 0.00018764747904087263,
      "loss": 6.3713,
      "step": 2440
    },
    {
      "epoch": 0.2441,
      "grad_norm": 4.601139068603516,
      "learning_rate": 0.00018763066800438636,
      "loss": 4.9128,
      "step": 2441
    },
    {
      "epoch": 0.2442,
      "grad_norm": 2.8080010414123535,
      "learning_rate": 0.00018761384629036902,
      "loss": 6.2108,
      "step": 2442
    },
    {
      "epoch": 0.2443,
      "grad_norm": 3.1066057682037354,
      "learning_rate": 0.00018759701390087027,
      "loss": 5.8796,
      "step": 2443
    },
    {
      "epoch": 0.2444,
      "grad_norm": 2.532430410385132,
      "learning_rate": 0.0001875801708379411,
      "loss": 5.3086,
      "step": 2444
    },
    {
      "epoch": 0.2445,
      "grad_norm": 9.422342300415039,
      "learning_rate": 0.00018756331710363374,
      "loss": 5.0919,
      "step": 2445
    },
    {
      "epoch": 0.2446,
      "grad_norm": 2.1982991695404053,
      "learning_rate": 0.0001875464527000018,
      "loss": 5.361,
      "step": 2446
    },
    {
      "epoch": 0.2447,
      "grad_norm": 2.9513635635375977,
      "learning_rate": 0.00018752957762910018,
      "loss": 5.6511,
      "step": 2447
    },
    {
      "epoch": 0.2448,
      "grad_norm": 2.80918550491333,
      "learning_rate": 0.000187512691892985,
      "loss": 5.2384,
      "step": 2448
    },
    {
      "epoch": 0.2449,
      "grad_norm": 6.0558695793151855,
      "learning_rate": 0.0001874957954937138,
      "loss": 4.8308,
      "step": 2449
    },
    {
      "epoch": 0.245,
      "grad_norm": 8.041608810424805,
      "learning_rate": 0.0001874788884333453,
      "loss": 5.1404,
      "step": 2450
    },
    {
      "epoch": 0.2451,
      "grad_norm": 4.443957328796387,
      "learning_rate": 0.00018746197071393958,
      "loss": 4.9867,
      "step": 2451
    },
    {
      "epoch": 0.2452,
      "grad_norm": 3.4068312644958496,
      "learning_rate": 0.00018744504233755805,
      "loss": 5.8283,
      "step": 2452
    },
    {
      "epoch": 0.2453,
      "grad_norm": 4.132494926452637,
      "learning_rate": 0.00018742810330626337,
      "loss": 5.6177,
      "step": 2453
    },
    {
      "epoch": 0.2454,
      "grad_norm": 4.041040420532227,
      "learning_rate": 0.00018741115362211949,
      "loss": 6.2274,
      "step": 2454
    },
    {
      "epoch": 0.2455,
      "grad_norm": 2.8454625606536865,
      "learning_rate": 0.0001873941932871917,
      "loss": 6.5582,
      "step": 2455
    },
    {
      "epoch": 0.2456,
      "grad_norm": 4.370352268218994,
      "learning_rate": 0.00018737722230354655,
      "loss": 6.1064,
      "step": 2456
    },
    {
      "epoch": 0.2457,
      "grad_norm": 7.156678676605225,
      "learning_rate": 0.00018736024067325188,
      "loss": 6.2988,
      "step": 2457
    },
    {
      "epoch": 0.2458,
      "grad_norm": 13.558053970336914,
      "learning_rate": 0.0001873432483983769,
      "loss": 5.0686,
      "step": 2458
    },
    {
      "epoch": 0.2459,
      "grad_norm": 2.6889851093292236,
      "learning_rate": 0.00018732624548099204,
      "loss": 5.397,
      "step": 2459
    },
    {
      "epoch": 0.246,
      "grad_norm": 4.053489685058594,
      "learning_rate": 0.00018730923192316902,
      "loss": 6.1378,
      "step": 2460
    },
    {
      "epoch": 0.2461,
      "grad_norm": 2.524843215942383,
      "learning_rate": 0.00018729220772698097,
      "loss": 5.8616,
      "step": 2461
    },
    {
      "epoch": 0.2462,
      "grad_norm": 5.123845100402832,
      "learning_rate": 0.0001872751728945022,
      "loss": 5.1781,
      "step": 2462
    },
    {
      "epoch": 0.2463,
      "grad_norm": 2.6229970455169678,
      "learning_rate": 0.00018725812742780834,
      "loss": 5.0554,
      "step": 2463
    },
    {
      "epoch": 0.2464,
      "grad_norm": 4.987990856170654,
      "learning_rate": 0.0001872410713289763,
      "loss": 6.9928,
      "step": 2464
    },
    {
      "epoch": 0.2465,
      "grad_norm": 4.3924126625061035,
      "learning_rate": 0.0001872240046000844,
      "loss": 4.9293,
      "step": 2465
    },
    {
      "epoch": 0.2466,
      "grad_norm": 4.512756824493408,
      "learning_rate": 0.00018720692724321207,
      "loss": 5.4041,
      "step": 2466
    },
    {
      "epoch": 0.2467,
      "grad_norm": 10.651782989501953,
      "learning_rate": 0.0001871898392604402,
      "loss": 6.7558,
      "step": 2467
    },
    {
      "epoch": 0.2468,
      "grad_norm": 3.7964203357696533,
      "learning_rate": 0.0001871727406538509,
      "loss": 5.3313,
      "step": 2468
    },
    {
      "epoch": 0.2469,
      "grad_norm": 2.8312389850616455,
      "learning_rate": 0.00018715563142552758,
      "loss": 4.7923,
      "step": 2469
    },
    {
      "epoch": 0.247,
      "grad_norm": 2.6606974601745605,
      "learning_rate": 0.00018713851157755492,
      "loss": 5.6296,
      "step": 2470
    },
    {
      "epoch": 0.2471,
      "grad_norm": 2.2761611938476562,
      "learning_rate": 0.00018712138111201895,
      "loss": 5.4459,
      "step": 2471
    },
    {
      "epoch": 0.2472,
      "grad_norm": 2.357085704803467,
      "learning_rate": 0.00018710424003100698,
      "loss": 5.2574,
      "step": 2472
    },
    {
      "epoch": 0.2473,
      "grad_norm": 4.39935827255249,
      "learning_rate": 0.00018708708833660754,
      "loss": 5.2895,
      "step": 2473
    },
    {
      "epoch": 0.2474,
      "grad_norm": 4.325582981109619,
      "learning_rate": 0.00018706992603091058,
      "loss": 6.0787,
      "step": 2474
    },
    {
      "epoch": 0.2475,
      "grad_norm": 3.38547420501709,
      "learning_rate": 0.00018705275311600722,
      "loss": 5.7754,
      "step": 2475
    },
    {
      "epoch": 0.2476,
      "grad_norm": 4.236578464508057,
      "learning_rate": 0.00018703556959398998,
      "loss": 6.6324,
      "step": 2476
    },
    {
      "epoch": 0.2477,
      "grad_norm": 2.1968834400177,
      "learning_rate": 0.0001870183754669526,
      "loss": 5.8281,
      "step": 2477
    },
    {
      "epoch": 0.2478,
      "grad_norm": 7.898112773895264,
      "learning_rate": 0.0001870011707369901,
      "loss": 5.0406,
      "step": 2478
    },
    {
      "epoch": 0.2479,
      "grad_norm": 4.623898506164551,
      "learning_rate": 0.0001869839554061988,
      "loss": 5.0828,
      "step": 2479
    },
    {
      "epoch": 0.248,
      "grad_norm": 13.552009582519531,
      "learning_rate": 0.00018696672947667646,
      "loss": 5.4484,
      "step": 2480
    },
    {
      "epoch": 0.2481,
      "grad_norm": 4.791248798370361,
      "learning_rate": 0.0001869494929505219,
      "loss": 6.7181,
      "step": 2481
    },
    {
      "epoch": 0.2482,
      "grad_norm": 6.66261625289917,
      "learning_rate": 0.0001869322458298354,
      "loss": 5.3116,
      "step": 2482
    },
    {
      "epoch": 0.2483,
      "grad_norm": 3.9017176628112793,
      "learning_rate": 0.0001869149881167184,
      "loss": 5.9225,
      "step": 2483
    },
    {
      "epoch": 0.2484,
      "grad_norm": 4.188482761383057,
      "learning_rate": 0.00018689771981327376,
      "loss": 5.5412,
      "step": 2484
    },
    {
      "epoch": 0.2485,
      "grad_norm": 3.458143949508667,
      "learning_rate": 0.00018688044092160551,
      "loss": 5.6632,
      "step": 2485
    },
    {
      "epoch": 0.2486,
      "grad_norm": 3.2037441730499268,
      "learning_rate": 0.00018686315144381913,
      "loss": 6.3305,
      "step": 2486
    },
    {
      "epoch": 0.2487,
      "grad_norm": 2.46392560005188,
      "learning_rate": 0.00018684585138202122,
      "loss": 5.9219,
      "step": 2487
    },
    {
      "epoch": 0.2488,
      "grad_norm": 3.208308696746826,
      "learning_rate": 0.00018682854073831973,
      "loss": 5.0804,
      "step": 2488
    },
    {
      "epoch": 0.2489,
      "grad_norm": 5.973125457763672,
      "learning_rate": 0.00018681121951482393,
      "loss": 5.3067,
      "step": 2489
    },
    {
      "epoch": 0.249,
      "grad_norm": 4.727505207061768,
      "learning_rate": 0.00018679388771364436,
      "loss": 6.0103,
      "step": 2490
    },
    {
      "epoch": 0.2491,
      "grad_norm": 3.0309271812438965,
      "learning_rate": 0.00018677654533689287,
      "loss": 5.8993,
      "step": 2491
    },
    {
      "epoch": 0.2492,
      "grad_norm": 4.547187328338623,
      "learning_rate": 0.0001867591923866825,
      "loss": 5.369,
      "step": 2492
    },
    {
      "epoch": 0.2493,
      "grad_norm": 3.058143377304077,
      "learning_rate": 0.00018674182886512774,
      "loss": 5.8865,
      "step": 2493
    },
    {
      "epoch": 0.2494,
      "grad_norm": 7.824762344360352,
      "learning_rate": 0.00018672445477434425,
      "loss": 6.6092,
      "step": 2494
    },
    {
      "epoch": 0.2495,
      "grad_norm": 5.345363616943359,
      "learning_rate": 0.000186707070116449,
      "loss": 6.0512,
      "step": 2495
    },
    {
      "epoch": 0.2496,
      "grad_norm": 8.334959983825684,
      "learning_rate": 0.00018668967489356028,
      "loss": 5.6882,
      "step": 2496
    },
    {
      "epoch": 0.2497,
      "grad_norm": 8.053354263305664,
      "learning_rate": 0.00018667226910779765,
      "loss": 5.471,
      "step": 2497
    },
    {
      "epoch": 0.2498,
      "grad_norm": 3.3663837909698486,
      "learning_rate": 0.00018665485276128188,
      "loss": 5.5765,
      "step": 2498
    },
    {
      "epoch": 0.2499,
      "grad_norm": 6.1116719245910645,
      "learning_rate": 0.00018663742585613518,
      "loss": 5.5462,
      "step": 2499
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.764549493789673,
      "learning_rate": 0.00018661998839448094,
      "loss": 5.3694,
      "step": 2500
    },
    {
      "epoch": 0.2501,
      "grad_norm": 2.4880173206329346,
      "learning_rate": 0.00018660254037844388,
      "loss": 5.4466,
      "step": 2501
    },
    {
      "epoch": 0.2502,
      "grad_norm": 4.258087158203125,
      "learning_rate": 0.00018658508181014995,
      "loss": 5.9898,
      "step": 2502
    },
    {
      "epoch": 0.2503,
      "grad_norm": 5.018625736236572,
      "learning_rate": 0.00018656761269172643,
      "loss": 6.1438,
      "step": 2503
    },
    {
      "epoch": 0.2504,
      "grad_norm": 4.588842868804932,
      "learning_rate": 0.0001865501330253019,
      "loss": 4.8839,
      "step": 2504
    },
    {
      "epoch": 0.2505,
      "grad_norm": 3.443013906478882,
      "learning_rate": 0.00018653264281300622,
      "loss": 5.7663,
      "step": 2505
    },
    {
      "epoch": 0.2506,
      "grad_norm": 7.011101722717285,
      "learning_rate": 0.00018651514205697046,
      "loss": 5.6061,
      "step": 2506
    },
    {
      "epoch": 0.2507,
      "grad_norm": 2.5615999698638916,
      "learning_rate": 0.00018649763075932708,
      "loss": 5.7897,
      "step": 2507
    },
    {
      "epoch": 0.2508,
      "grad_norm": 3.365100622177124,
      "learning_rate": 0.00018648010892220978,
      "loss": 5.2486,
      "step": 2508
    },
    {
      "epoch": 0.2509,
      "grad_norm": 4.808720588684082,
      "learning_rate": 0.0001864625765477535,
      "loss": 5.2119,
      "step": 2509
    },
    {
      "epoch": 0.251,
      "grad_norm": 2.338986873626709,
      "learning_rate": 0.00018644503363809457,
      "loss": 5.7119,
      "step": 2510
    },
    {
      "epoch": 0.2511,
      "grad_norm": 6.834888935089111,
      "learning_rate": 0.0001864274801953705,
      "loss": 5.8115,
      "step": 2511
    },
    {
      "epoch": 0.2512,
      "grad_norm": 7.826887130737305,
      "learning_rate": 0.0001864099162217201,
      "loss": 6.5346,
      "step": 2512
    },
    {
      "epoch": 0.2513,
      "grad_norm": 10.37363052368164,
      "learning_rate": 0.00018639234171928353,
      "loss": 6.3583,
      "step": 2513
    },
    {
      "epoch": 0.2514,
      "grad_norm": 8.087769508361816,
      "learning_rate": 0.0001863747566902022,
      "loss": 5.6405,
      "step": 2514
    },
    {
      "epoch": 0.2515,
      "grad_norm": 5.503287315368652,
      "learning_rate": 0.00018635716113661873,
      "loss": 5.684,
      "step": 2515
    },
    {
      "epoch": 0.2516,
      "grad_norm": 3.6790056228637695,
      "learning_rate": 0.00018633955506067718,
      "loss": 5.2554,
      "step": 2516
    },
    {
      "epoch": 0.2517,
      "grad_norm": 2.3835065364837646,
      "learning_rate": 0.0001863219384645227,
      "loss": 5.5575,
      "step": 2517
    },
    {
      "epoch": 0.2518,
      "grad_norm": 2.4194817543029785,
      "learning_rate": 0.0001863043113503019,
      "loss": 5.1315,
      "step": 2518
    },
    {
      "epoch": 0.2519,
      "grad_norm": 4.114435195922852,
      "learning_rate": 0.0001862866737201625,
      "loss": 6.0932,
      "step": 2519
    },
    {
      "epoch": 0.252,
      "grad_norm": 3.1679413318634033,
      "learning_rate": 0.00018626902557625368,
      "loss": 6.0475,
      "step": 2520
    },
    {
      "epoch": 0.2521,
      "grad_norm": 4.702274799346924,
      "learning_rate": 0.00018625136692072575,
      "loss": 5.4704,
      "step": 2521
    },
    {
      "epoch": 0.2522,
      "grad_norm": 8.07321834564209,
      "learning_rate": 0.0001862336977557304,
      "loss": 5.4625,
      "step": 2522
    },
    {
      "epoch": 0.2523,
      "grad_norm": 6.732860565185547,
      "learning_rate": 0.00018621601808342056,
      "loss": 6.3288,
      "step": 2523
    },
    {
      "epoch": 0.2524,
      "grad_norm": 11.509590148925781,
      "learning_rate": 0.00018619832790595043,
      "loss": 6.4143,
      "step": 2524
    },
    {
      "epoch": 0.2525,
      "grad_norm": 7.59700345993042,
      "learning_rate": 0.0001861806272254755,
      "loss": 5.8888,
      "step": 2525
    },
    {
      "epoch": 0.2526,
      "grad_norm": 2.1069469451904297,
      "learning_rate": 0.00018616291604415258,
      "loss": 5.3036,
      "step": 2526
    },
    {
      "epoch": 0.2527,
      "grad_norm": 4.711550235748291,
      "learning_rate": 0.0001861451943641397,
      "loss": 5.2657,
      "step": 2527
    },
    {
      "epoch": 0.2528,
      "grad_norm": 4.213138103485107,
      "learning_rate": 0.00018612746218759618,
      "loss": 5.6699,
      "step": 2528
    },
    {
      "epoch": 0.2529,
      "grad_norm": 2.721090078353882,
      "learning_rate": 0.00018610971951668265,
      "loss": 5.8323,
      "step": 2529
    },
    {
      "epoch": 0.253,
      "grad_norm": 4.477640628814697,
      "learning_rate": 0.000186091966353561,
      "loss": 7.0121,
      "step": 2530
    },
    {
      "epoch": 0.2531,
      "grad_norm": 2.239332914352417,
      "learning_rate": 0.0001860742027003944,
      "loss": 5.1172,
      "step": 2531
    },
    {
      "epoch": 0.2532,
      "grad_norm": 5.9601731300354,
      "learning_rate": 0.00018605642855934725,
      "loss": 7.0726,
      "step": 2532
    },
    {
      "epoch": 0.2533,
      "grad_norm": 4.454159259796143,
      "learning_rate": 0.00018603864393258534,
      "loss": 5.4414,
      "step": 2533
    },
    {
      "epoch": 0.2534,
      "grad_norm": 7.39664363861084,
      "learning_rate": 0.00018602084882227566,
      "loss": 5.8623,
      "step": 2534
    },
    {
      "epoch": 0.2535,
      "grad_norm": 2.185986280441284,
      "learning_rate": 0.00018600304323058647,
      "loss": 5.7581,
      "step": 2535
    },
    {
      "epoch": 0.2536,
      "grad_norm": 6.665598392486572,
      "learning_rate": 0.00018598522715968736,
      "loss": 5.3918,
      "step": 2536
    },
    {
      "epoch": 0.2537,
      "grad_norm": 2.8032872676849365,
      "learning_rate": 0.0001859674006117491,
      "loss": 5.5527,
      "step": 2537
    },
    {
      "epoch": 0.2538,
      "grad_norm": 2.8675594329833984,
      "learning_rate": 0.00018594956358894388,
      "loss": 5.1718,
      "step": 2538
    },
    {
      "epoch": 0.2539,
      "grad_norm": 3.722062826156616,
      "learning_rate": 0.00018593171609344503,
      "loss": 5.0669,
      "step": 2539
    },
    {
      "epoch": 0.254,
      "grad_norm": 21.28429412841797,
      "learning_rate": 0.00018591385812742725,
      "loss": 6.3771,
      "step": 2540
    },
    {
      "epoch": 0.2541,
      "grad_norm": 6.057986259460449,
      "learning_rate": 0.00018589598969306645,
      "loss": 5.1031,
      "step": 2541
    },
    {
      "epoch": 0.2542,
      "grad_norm": 9.914308547973633,
      "learning_rate": 0.00018587811079253985,
      "loss": 5.5893,
      "step": 2542
    },
    {
      "epoch": 0.2543,
      "grad_norm": 6.026546478271484,
      "learning_rate": 0.00018586022142802597,
      "loss": 6.4519,
      "step": 2543
    },
    {
      "epoch": 0.2544,
      "grad_norm": 6.1983771324157715,
      "learning_rate": 0.00018584232160170452,
      "loss": 5.9598,
      "step": 2544
    },
    {
      "epoch": 0.2545,
      "grad_norm": 4.022098541259766,
      "learning_rate": 0.0001858244113157566,
      "loss": 6.045,
      "step": 2545
    },
    {
      "epoch": 0.2546,
      "grad_norm": 3.786834478378296,
      "learning_rate": 0.00018580649057236447,
      "loss": 6.3803,
      "step": 2546
    },
    {
      "epoch": 0.2547,
      "grad_norm": 3.136411666870117,
      "learning_rate": 0.00018578855937371173,
      "loss": 5.3334,
      "step": 2547
    },
    {
      "epoch": 0.2548,
      "grad_norm": 4.6000871658325195,
      "learning_rate": 0.0001857706177219833,
      "loss": 6.238,
      "step": 2548
    },
    {
      "epoch": 0.2549,
      "grad_norm": 8.394966125488281,
      "learning_rate": 0.00018575266561936523,
      "loss": 5.6555,
      "step": 2549
    },
    {
      "epoch": 0.255,
      "grad_norm": 2.749495267868042,
      "learning_rate": 0.00018573470306804498,
      "loss": 6.1092,
      "step": 2550
    },
    {
      "epoch": 0.2551,
      "grad_norm": 3.4485971927642822,
      "learning_rate": 0.00018571673007021123,
      "loss": 5.3122,
      "step": 2551
    },
    {
      "epoch": 0.2552,
      "grad_norm": 3.1270954608917236,
      "learning_rate": 0.00018569874662805393,
      "loss": 5.1729,
      "step": 2552
    },
    {
      "epoch": 0.2553,
      "grad_norm": 6.0764360427856445,
      "learning_rate": 0.0001856807527437643,
      "loss": 6.3156,
      "step": 2553
    },
    {
      "epoch": 0.2554,
      "grad_norm": 4.892404079437256,
      "learning_rate": 0.00018566274841953483,
      "loss": 5.7685,
      "step": 2554
    },
    {
      "epoch": 0.2555,
      "grad_norm": 3.6355860233306885,
      "learning_rate": 0.00018564473365755935,
      "loss": 5.3019,
      "step": 2555
    },
    {
      "epoch": 0.2556,
      "grad_norm": 4.09676456451416,
      "learning_rate": 0.00018562670846003284,
      "loss": 5.4746,
      "step": 2556
    },
    {
      "epoch": 0.2557,
      "grad_norm": 8.832869529724121,
      "learning_rate": 0.0001856086728291516,
      "loss": 5.6849,
      "step": 2557
    },
    {
      "epoch": 0.2558,
      "grad_norm": 6.609644889831543,
      "learning_rate": 0.00018559062676711332,
      "loss": 5.564,
      "step": 2558
    },
    {
      "epoch": 0.2559,
      "grad_norm": 2.077838182449341,
      "learning_rate": 0.00018557257027611675,
      "loss": 5.5785,
      "step": 2559
    },
    {
      "epoch": 0.256,
      "grad_norm": 11.995332717895508,
      "learning_rate": 0.00018555450335836206,
      "loss": 5.9999,
      "step": 2560
    },
    {
      "epoch": 0.2561,
      "grad_norm": 12.528913497924805,
      "learning_rate": 0.00018553642601605068,
      "loss": 5.54,
      "step": 2561
    },
    {
      "epoch": 0.2562,
      "grad_norm": 3.2932586669921875,
      "learning_rate": 0.0001855183382513852,
      "loss": 5.3176,
      "step": 2562
    },
    {
      "epoch": 0.2563,
      "grad_norm": 3.3880975246429443,
      "learning_rate": 0.00018550024006656966,
      "loss": 5.7652,
      "step": 2563
    },
    {
      "epoch": 0.2564,
      "grad_norm": 3.3539857864379883,
      "learning_rate": 0.00018548213146380918,
      "loss": 5.2552,
      "step": 2564
    },
    {
      "epoch": 0.2565,
      "grad_norm": 8.336670875549316,
      "learning_rate": 0.0001854640124453103,
      "loss": 6.5658,
      "step": 2565
    },
    {
      "epoch": 0.2566,
      "grad_norm": 3.8305068016052246,
      "learning_rate": 0.00018544588301328075,
      "loss": 5.7009,
      "step": 2566
    },
    {
      "epoch": 0.2567,
      "grad_norm": 1.7178230285644531,
      "learning_rate": 0.0001854277431699295,
      "loss": 4.9341,
      "step": 2567
    },
    {
      "epoch": 0.2568,
      "grad_norm": 40.13162612915039,
      "learning_rate": 0.00018540959291746693,
      "loss": 6.9112,
      "step": 2568
    },
    {
      "epoch": 0.2569,
      "grad_norm": 3.2152273654937744,
      "learning_rate": 0.0001853914322581045,
      "loss": 5.2528,
      "step": 2569
    },
    {
      "epoch": 0.257,
      "grad_norm": 3.6834306716918945,
      "learning_rate": 0.00018537326119405506,
      "loss": 5.817,
      "step": 2570
    },
    {
      "epoch": 0.2571,
      "grad_norm": 3.6225240230560303,
      "learning_rate": 0.00018535507972753274,
      "loss": 5.5701,
      "step": 2571
    },
    {
      "epoch": 0.2572,
      "grad_norm": 2.213998794555664,
      "learning_rate": 0.00018533688786075288,
      "loss": 5.6347,
      "step": 2572
    },
    {
      "epoch": 0.2573,
      "grad_norm": 21.060609817504883,
      "learning_rate": 0.00018531868559593204,
      "loss": 6.121,
      "step": 2573
    },
    {
      "epoch": 0.2574,
      "grad_norm": 2.451289176940918,
      "learning_rate": 0.00018530047293528819,
      "loss": 5.3065,
      "step": 2574
    },
    {
      "epoch": 0.2575,
      "grad_norm": 3.1589009761810303,
      "learning_rate": 0.00018528224988104044,
      "loss": 5.5572,
      "step": 2575
    },
    {
      "epoch": 0.2576,
      "grad_norm": 4.8654680252075195,
      "learning_rate": 0.00018526401643540922,
      "loss": 5.3106,
      "step": 2576
    },
    {
      "epoch": 0.2577,
      "grad_norm": 4.898959159851074,
      "learning_rate": 0.00018524577260061627,
      "loss": 5.9984,
      "step": 2577
    },
    {
      "epoch": 0.2578,
      "grad_norm": 2.6340692043304443,
      "learning_rate": 0.0001852275183788845,
      "loss": 5.6163,
      "step": 2578
    },
    {
      "epoch": 0.2579,
      "grad_norm": 3.8535895347595215,
      "learning_rate": 0.0001852092537724381,
      "loss": 5.8651,
      "step": 2579
    },
    {
      "epoch": 0.258,
      "grad_norm": 3.4571619033813477,
      "learning_rate": 0.00018519097878350263,
      "loss": 6.1218,
      "step": 2580
    },
    {
      "epoch": 0.2581,
      "grad_norm": 6.171290397644043,
      "learning_rate": 0.00018517269341430476,
      "loss": 5.7069,
      "step": 2581
    },
    {
      "epoch": 0.2582,
      "grad_norm": 5.905836582183838,
      "learning_rate": 0.00018515439766707262,
      "loss": 5.128,
      "step": 2582
    },
    {
      "epoch": 0.2583,
      "grad_norm": 12.426368713378906,
      "learning_rate": 0.00018513609154403534,
      "loss": 5.6085,
      "step": 2583
    },
    {
      "epoch": 0.2584,
      "grad_norm": 11.644067764282227,
      "learning_rate": 0.00018511777504742362,
      "loss": 6.2377,
      "step": 2584
    },
    {
      "epoch": 0.2585,
      "grad_norm": 2.8699469566345215,
      "learning_rate": 0.00018509944817946922,
      "loss": 5.7568,
      "step": 2585
    },
    {
      "epoch": 0.2586,
      "grad_norm": 8.647931098937988,
      "learning_rate": 0.00018508111094240514,
      "loss": 5.2284,
      "step": 2586
    },
    {
      "epoch": 0.2587,
      "grad_norm": 4.244480133056641,
      "learning_rate": 0.00018506276333846579,
      "loss": 5.1907,
      "step": 2587
    },
    {
      "epoch": 0.2588,
      "grad_norm": 5.719789981842041,
      "learning_rate": 0.00018504440536988673,
      "loss": 5.9523,
      "step": 2588
    },
    {
      "epoch": 0.2589,
      "grad_norm": 4.158243179321289,
      "learning_rate": 0.00018502603703890488,
      "loss": 5.1762,
      "step": 2589
    },
    {
      "epoch": 0.259,
      "grad_norm": 4.319550037384033,
      "learning_rate": 0.00018500765834775828,
      "loss": 5.5568,
      "step": 2590
    },
    {
      "epoch": 0.2591,
      "grad_norm": 2.3161025047302246,
      "learning_rate": 0.00018498926929868642,
      "loss": 5.4227,
      "step": 2591
    },
    {
      "epoch": 0.2592,
      "grad_norm": 4.001145362854004,
      "learning_rate": 0.00018497086989392988,
      "loss": 5.4092,
      "step": 2592
    },
    {
      "epoch": 0.2593,
      "grad_norm": 3.0049009323120117,
      "learning_rate": 0.00018495246013573054,
      "loss": 5.3147,
      "step": 2593
    },
    {
      "epoch": 0.2594,
      "grad_norm": 2.5222854614257812,
      "learning_rate": 0.00018493404002633166,
      "loss": 5.4853,
      "step": 2594
    },
    {
      "epoch": 0.2595,
      "grad_norm": 3.9237234592437744,
      "learning_rate": 0.00018491560956797765,
      "loss": 5.8963,
      "step": 2595
    },
    {
      "epoch": 0.2596,
      "grad_norm": 2.8145248889923096,
      "learning_rate": 0.00018489716876291415,
      "loss": 5.9711,
      "step": 2596
    },
    {
      "epoch": 0.2597,
      "grad_norm": 3.227881908416748,
      "learning_rate": 0.0001848787176133882,
      "loss": 5.5975,
      "step": 2597
    },
    {
      "epoch": 0.2598,
      "grad_norm": 6.716454029083252,
      "learning_rate": 0.00018486025612164794,
      "loss": 5.5011,
      "step": 2598
    },
    {
      "epoch": 0.2599,
      "grad_norm": 3.5992908477783203,
      "learning_rate": 0.0001848417842899429,
      "loss": 6.0528,
      "step": 2599
    },
    {
      "epoch": 0.26,
      "grad_norm": 7.552899360656738,
      "learning_rate": 0.00018482330212052378,
      "loss": 5.1314,
      "step": 2600
    },
    {
      "epoch": 0.2601,
      "grad_norm": 4.046154975891113,
      "learning_rate": 0.0001848048096156426,
      "loss": 5.1226,
      "step": 2601
    },
    {
      "epoch": 0.2602,
      "grad_norm": 7.185793876647949,
      "learning_rate": 0.00018478630677755262,
      "loss": 5.5664,
      "step": 2602
    },
    {
      "epoch": 0.2603,
      "grad_norm": 2.6133992671966553,
      "learning_rate": 0.00018476779360850832,
      "loss": 5.9431,
      "step": 2603
    },
    {
      "epoch": 0.2604,
      "grad_norm": 11.973648071289062,
      "learning_rate": 0.00018474927011076552,
      "loss": 6.4312,
      "step": 2604
    },
    {
      "epoch": 0.2605,
      "grad_norm": 5.6708502769470215,
      "learning_rate": 0.0001847307362865812,
      "loss": 5.3419,
      "step": 2605
    },
    {
      "epoch": 0.2606,
      "grad_norm": 2.8330488204956055,
      "learning_rate": 0.00018471219213821375,
      "loss": 5.1243,
      "step": 2606
    },
    {
      "epoch": 0.2607,
      "grad_norm": 3.438183546066284,
      "learning_rate": 0.00018469363766792255,
      "loss": 5.5287,
      "step": 2607
    },
    {
      "epoch": 0.2608,
      "grad_norm": 3.1127970218658447,
      "learning_rate": 0.00018467507287796856,
      "loss": 5.9986,
      "step": 2608
    },
    {
      "epoch": 0.2609,
      "grad_norm": 6.0376763343811035,
      "learning_rate": 0.0001846564977706138,
      "loss": 5.2566,
      "step": 2609
    },
    {
      "epoch": 0.261,
      "grad_norm": 4.100964069366455,
      "learning_rate": 0.00018463791234812153,
      "loss": 5.845,
      "step": 2610
    },
    {
      "epoch": 0.2611,
      "grad_norm": 2.4736416339874268,
      "learning_rate": 0.00018461931661275643,
      "loss": 6.0032,
      "step": 2611
    },
    {
      "epoch": 0.2612,
      "grad_norm": 8.588294982910156,
      "learning_rate": 0.00018460071056678422,
      "loss": 5.5942,
      "step": 2612
    },
    {
      "epoch": 0.2613,
      "grad_norm": 2.380091428756714,
      "learning_rate": 0.00018458209421247208,
      "loss": 5.776,
      "step": 2613
    },
    {
      "epoch": 0.2614,
      "grad_norm": 4.228795051574707,
      "learning_rate": 0.00018456346755208833,
      "loss": 5.2117,
      "step": 2614
    },
    {
      "epoch": 0.2615,
      "grad_norm": 2.9307963848114014,
      "learning_rate": 0.00018454483058790255,
      "loss": 5.4617,
      "step": 2615
    },
    {
      "epoch": 0.2616,
      "grad_norm": 4.0945963859558105,
      "learning_rate": 0.00018452618332218563,
      "loss": 5.1814,
      "step": 2616
    },
    {
      "epoch": 0.2617,
      "grad_norm": 2.968869686126709,
      "learning_rate": 0.00018450752575720967,
      "loss": 4.9547,
      "step": 2617
    },
    {
      "epoch": 0.2618,
      "grad_norm": 2.364424705505371,
      "learning_rate": 0.00018448885789524802,
      "loss": 5.0694,
      "step": 2618
    },
    {
      "epoch": 0.2619,
      "grad_norm": 6.269031047821045,
      "learning_rate": 0.00018447017973857532,
      "loss": 5.3244,
      "step": 2619
    },
    {
      "epoch": 0.262,
      "grad_norm": 3.7848007678985596,
      "learning_rate": 0.00018445149128946744,
      "loss": 5.9283,
      "step": 2620
    },
    {
      "epoch": 0.2621,
      "grad_norm": 17.90859031677246,
      "learning_rate": 0.00018443279255020152,
      "loss": 5.2742,
      "step": 2621
    },
    {
      "epoch": 0.2622,
      "grad_norm": 2.2180042266845703,
      "learning_rate": 0.00018441408352305594,
      "loss": 5.1077,
      "step": 2622
    },
    {
      "epoch": 0.2623,
      "grad_norm": 13.380126953125,
      "learning_rate": 0.00018439536421031033,
      "loss": 5.5602,
      "step": 2623
    },
    {
      "epoch": 0.2624,
      "grad_norm": 8.422517776489258,
      "learning_rate": 0.0001843766346142456,
      "loss": 5.6965,
      "step": 2624
    },
    {
      "epoch": 0.2625,
      "grad_norm": 2.5908589363098145,
      "learning_rate": 0.0001843578947371439,
      "loss": 5.5716,
      "step": 2625
    },
    {
      "epoch": 0.2626,
      "grad_norm": 2.512476682662964,
      "learning_rate": 0.0001843391445812886,
      "loss": 5.3541,
      "step": 2626
    },
    {
      "epoch": 0.2627,
      "grad_norm": 3.5167009830474854,
      "learning_rate": 0.00018432038414896434,
      "loss": 5.6228,
      "step": 2627
    },
    {
      "epoch": 0.2628,
      "grad_norm": 8.678203582763672,
      "learning_rate": 0.00018430161344245707,
      "loss": 5.4044,
      "step": 2628
    },
    {
      "epoch": 0.2629,
      "grad_norm": 3.529611349105835,
      "learning_rate": 0.0001842828324640539,
      "loss": 5.9192,
      "step": 2629
    },
    {
      "epoch": 0.263,
      "grad_norm": 3.194260358810425,
      "learning_rate": 0.00018426404121604323,
      "loss": 5.276,
      "step": 2630
    },
    {
      "epoch": 0.2631,
      "grad_norm": 2.499948024749756,
      "learning_rate": 0.00018424523970071477,
      "loss": 4.6157,
      "step": 2631
    },
    {
      "epoch": 0.2632,
      "grad_norm": 4.680108547210693,
      "learning_rate": 0.0001842264279203594,
      "loss": 5.677,
      "step": 2632
    },
    {
      "epoch": 0.2633,
      "grad_norm": 2.1938517093658447,
      "learning_rate": 0.00018420760587726923,
      "loss": 5.5001,
      "step": 2633
    },
    {
      "epoch": 0.2634,
      "grad_norm": 1.8108241558074951,
      "learning_rate": 0.00018418877357373776,
      "loss": 5.2255,
      "step": 2634
    },
    {
      "epoch": 0.2635,
      "grad_norm": 2.7569363117218018,
      "learning_rate": 0.00018416993101205958,
      "loss": 5.8428,
      "step": 2635
    },
    {
      "epoch": 0.2636,
      "grad_norm": 28.254745483398438,
      "learning_rate": 0.00018415107819453062,
      "loss": 5.9798,
      "step": 2636
    },
    {
      "epoch": 0.2637,
      "grad_norm": 2.3888022899627686,
      "learning_rate": 0.00018413221512344805,
      "loss": 5.3477,
      "step": 2637
    },
    {
      "epoch": 0.2638,
      "grad_norm": 13.949984550476074,
      "learning_rate": 0.00018411334180111027,
      "loss": 6.2691,
      "step": 2638
    },
    {
      "epoch": 0.2639,
      "grad_norm": 3.908165454864502,
      "learning_rate": 0.00018409445822981693,
      "loss": 5.8524,
      "step": 2639
    },
    {
      "epoch": 0.264,
      "grad_norm": 6.431516647338867,
      "learning_rate": 0.00018407556441186893,
      "loss": 5.1228,
      "step": 2640
    },
    {
      "epoch": 0.2641,
      "grad_norm": 15.055744171142578,
      "learning_rate": 0.00018405666034956844,
      "loss": 5.4808,
      "step": 2641
    },
    {
      "epoch": 0.2642,
      "grad_norm": 5.991715908050537,
      "learning_rate": 0.00018403774604521886,
      "loss": 5.6603,
      "step": 2642
    },
    {
      "epoch": 0.2643,
      "grad_norm": 3.742725372314453,
      "learning_rate": 0.00018401882150112484,
      "loss": 5.5176,
      "step": 2643
    },
    {
      "epoch": 0.2644,
      "grad_norm": 10.337864875793457,
      "learning_rate": 0.00018399988671959227,
      "loss": 5.9615,
      "step": 2644
    },
    {
      "epoch": 0.2645,
      "grad_norm": 7.746242046356201,
      "learning_rate": 0.0001839809417029283,
      "loss": 5.7287,
      "step": 2645
    },
    {
      "epoch": 0.2646,
      "grad_norm": 2.650125026702881,
      "learning_rate": 0.00018396198645344135,
      "loss": 5.6501,
      "step": 2646
    },
    {
      "epoch": 0.2647,
      "grad_norm": 2.8384523391723633,
      "learning_rate": 0.000183943020973441,
      "loss": 5.361,
      "step": 2647
    },
    {
      "epoch": 0.2648,
      "grad_norm": 4.667333126068115,
      "learning_rate": 0.00018392404526523817,
      "loss": 5.4321,
      "step": 2648
    },
    {
      "epoch": 0.2649,
      "grad_norm": 7.073169708251953,
      "learning_rate": 0.000183905059331145,
      "loss": 6.2665,
      "step": 2649
    },
    {
      "epoch": 0.265,
      "grad_norm": 6.180387020111084,
      "learning_rate": 0.0001838860631734749,
      "loss": 6.5055,
      "step": 2650
    },
    {
      "epoch": 0.2651,
      "grad_norm": 2.283207654953003,
      "learning_rate": 0.00018386705679454242,
      "loss": 5.4029,
      "step": 2651
    },
    {
      "epoch": 0.2652,
      "grad_norm": 2.586333990097046,
      "learning_rate": 0.00018384804019666345,
      "loss": 5.408,
      "step": 2652
    },
    {
      "epoch": 0.2653,
      "grad_norm": 2.163193702697754,
      "learning_rate": 0.00018382901338215516,
      "loss": 5.5492,
      "step": 2653
    },
    {
      "epoch": 0.2654,
      "grad_norm": 2.142031669616699,
      "learning_rate": 0.00018380997635333585,
      "loss": 6.0023,
      "step": 2654
    },
    {
      "epoch": 0.2655,
      "grad_norm": 3.731003522872925,
      "learning_rate": 0.00018379092911252514,
      "loss": 5.8309,
      "step": 2655
    },
    {
      "epoch": 0.2656,
      "grad_norm": 2.829035520553589,
      "learning_rate": 0.0001837718716620439,
      "loss": 5.4606,
      "step": 2656
    },
    {
      "epoch": 0.2657,
      "grad_norm": 4.11724853515625,
      "learning_rate": 0.0001837528040042142,
      "loss": 5.4547,
      "step": 2657
    },
    {
      "epoch": 0.2658,
      "grad_norm": 3.1522057056427,
      "learning_rate": 0.00018373372614135936,
      "loss": 4.789,
      "step": 2658
    },
    {
      "epoch": 0.2659,
      "grad_norm": 4.830076217651367,
      "learning_rate": 0.000183714638075804,
      "loss": 5.7921,
      "step": 2659
    },
    {
      "epoch": 0.266,
      "grad_norm": 3.823012590408325,
      "learning_rate": 0.0001836955398098739,
      "loss": 5.6216,
      "step": 2660
    },
    {
      "epoch": 0.2661,
      "grad_norm": 7.912096977233887,
      "learning_rate": 0.00018367643134589617,
      "loss": 5.1543,
      "step": 2661
    },
    {
      "epoch": 0.2662,
      "grad_norm": 2.915070056915283,
      "learning_rate": 0.0001836573126861991,
      "loss": 5.5563,
      "step": 2662
    },
    {
      "epoch": 0.2663,
      "grad_norm": 7.320728778839111,
      "learning_rate": 0.00018363818383311225,
      "loss": 5.7481,
      "step": 2663
    },
    {
      "epoch": 0.2664,
      "grad_norm": 2.978285312652588,
      "learning_rate": 0.0001836190447889664,
      "loss": 5.49,
      "step": 2664
    },
    {
      "epoch": 0.2665,
      "grad_norm": 4.4814772605896,
      "learning_rate": 0.00018359989555609353,
      "loss": 5.4455,
      "step": 2665
    },
    {
      "epoch": 0.2666,
      "grad_norm": 3.862027168273926,
      "learning_rate": 0.00018358073613682706,
      "loss": 6.6422,
      "step": 2666
    },
    {
      "epoch": 0.2667,
      "grad_norm": 12.811606407165527,
      "learning_rate": 0.00018356156653350137,
      "loss": 5.7073,
      "step": 2667
    },
    {
      "epoch": 0.2668,
      "grad_norm": 2.995245933532715,
      "learning_rate": 0.00018354238674845225,
      "loss": 5.0965,
      "step": 2668
    },
    {
      "epoch": 0.2669,
      "grad_norm": 3.5966250896453857,
      "learning_rate": 0.00018352319678401676,
      "loss": 5.4949,
      "step": 2669
    },
    {
      "epoch": 0.267,
      "grad_norm": 3.3904943466186523,
      "learning_rate": 0.00018350399664253305,
      "loss": 5.9689,
      "step": 2670
    },
    {
      "epoch": 0.2671,
      "grad_norm": 2.796010732650757,
      "learning_rate": 0.00018348478632634066,
      "loss": 5.4405,
      "step": 2671
    },
    {
      "epoch": 0.2672,
      "grad_norm": 2.492778778076172,
      "learning_rate": 0.0001834655658377803,
      "loss": 4.97,
      "step": 2672
    },
    {
      "epoch": 0.2673,
      "grad_norm": 4.343893527984619,
      "learning_rate": 0.00018344633517919392,
      "loss": 5.5103,
      "step": 2673
    },
    {
      "epoch": 0.2674,
      "grad_norm": 13.74405574798584,
      "learning_rate": 0.00018342709435292473,
      "loss": 5.9393,
      "step": 2674
    },
    {
      "epoch": 0.2675,
      "grad_norm": 6.357167720794678,
      "learning_rate": 0.00018340784336131713,
      "loss": 5.1595,
      "step": 2675
    },
    {
      "epoch": 0.2676,
      "grad_norm": 3.984754800796509,
      "learning_rate": 0.00018338858220671682,
      "loss": 5.6399,
      "step": 2676
    },
    {
      "epoch": 0.2677,
      "grad_norm": 2.5768063068389893,
      "learning_rate": 0.00018336931089147073,
      "loss": 5.5079,
      "step": 2677
    },
    {
      "epoch": 0.2678,
      "grad_norm": 7.349531650543213,
      "learning_rate": 0.00018335002941792698,
      "loss": 6.0846,
      "step": 2678
    },
    {
      "epoch": 0.2679,
      "grad_norm": 10.5476713180542,
      "learning_rate": 0.000183330737788435,
      "loss": 5.4456,
      "step": 2679
    },
    {
      "epoch": 0.268,
      "grad_norm": 5.483478546142578,
      "learning_rate": 0.00018331143600534535,
      "loss": 5.5782,
      "step": 2680
    },
    {
      "epoch": 0.2681,
      "grad_norm": 2.2464253902435303,
      "learning_rate": 0.00018329212407100994,
      "loss": 4.8707,
      "step": 2681
    },
    {
      "epoch": 0.2682,
      "grad_norm": 2.483177661895752,
      "learning_rate": 0.0001832728019877819,
      "loss": 5.3681,
      "step": 2682
    },
    {
      "epoch": 0.2683,
      "grad_norm": 4.6341118812561035,
      "learning_rate": 0.0001832534697580155,
      "loss": 5.4972,
      "step": 2683
    },
    {
      "epoch": 0.2684,
      "grad_norm": 3.5452892780303955,
      "learning_rate": 0.00018323412738406635,
      "loss": 5.3997,
      "step": 2684
    },
    {
      "epoch": 0.2685,
      "grad_norm": 3.9443447589874268,
      "learning_rate": 0.00018321477486829126,
      "loss": 5.2837,
      "step": 2685
    },
    {
      "epoch": 0.2686,
      "grad_norm": 3.5671815872192383,
      "learning_rate": 0.00018319541221304827,
      "loss": 5.1859,
      "step": 2686
    },
    {
      "epoch": 0.2687,
      "grad_norm": 3.3009984493255615,
      "learning_rate": 0.00018317603942069664,
      "loss": 5.205,
      "step": 2687
    },
    {
      "epoch": 0.2688,
      "grad_norm": 2.333454132080078,
      "learning_rate": 0.00018315665649359692,
      "loss": 5.0367,
      "step": 2688
    },
    {
      "epoch": 0.2689,
      "grad_norm": 4.274196147918701,
      "learning_rate": 0.00018313726343411086,
      "loss": 5.9548,
      "step": 2689
    },
    {
      "epoch": 0.269,
      "grad_norm": 2.767056465148926,
      "learning_rate": 0.0001831178602446014,
      "loss": 5.2126,
      "step": 2690
    },
    {
      "epoch": 0.2691,
      "grad_norm": 3.8288733959198,
      "learning_rate": 0.00018309844692743283,
      "loss": 5.4066,
      "step": 2691
    },
    {
      "epoch": 0.2692,
      "grad_norm": 3.4508607387542725,
      "learning_rate": 0.00018307902348497056,
      "loss": 4.9927,
      "step": 2692
    },
    {
      "epoch": 0.2693,
      "grad_norm": 3.575495958328247,
      "learning_rate": 0.00018305958991958127,
      "loss": 5.7412,
      "step": 2693
    },
    {
      "epoch": 0.2694,
      "grad_norm": 3.0832393169403076,
      "learning_rate": 0.0001830401462336329,
      "loss": 5.2568,
      "step": 2694
    },
    {
      "epoch": 0.2695,
      "grad_norm": 4.005279064178467,
      "learning_rate": 0.0001830206924294946,
      "loss": 5.1529,
      "step": 2695
    },
    {
      "epoch": 0.2696,
      "grad_norm": 5.005867958068848,
      "learning_rate": 0.00018300122850953675,
      "loss": 5.5597,
      "step": 2696
    },
    {
      "epoch": 0.2697,
      "grad_norm": 4.550533771514893,
      "learning_rate": 0.00018298175447613096,
      "loss": 5.8754,
      "step": 2697
    },
    {
      "epoch": 0.2698,
      "grad_norm": 4.067420482635498,
      "learning_rate": 0.00018296227033165013,
      "loss": 5.3733,
      "step": 2698
    },
    {
      "epoch": 0.2699,
      "grad_norm": 5.316220283508301,
      "learning_rate": 0.00018294277607846832,
      "loss": 5.5255,
      "step": 2699
    },
    {
      "epoch": 0.27,
      "grad_norm": 6.171837329864502,
      "learning_rate": 0.0001829232717189608,
      "loss": 5.8569,
      "step": 2700
    },
    {
      "epoch": 0.2701,
      "grad_norm": 5.763611793518066,
      "learning_rate": 0.00018290375725550417,
      "loss": 5.5423,
      "step": 2701
    },
    {
      "epoch": 0.2702,
      "grad_norm": 4.6137237548828125,
      "learning_rate": 0.0001828842326904762,
      "loss": 5.5666,
      "step": 2702
    },
    {
      "epoch": 0.2703,
      "grad_norm": 3.1591217517852783,
      "learning_rate": 0.00018286469802625589,
      "loss": 5.6246,
      "step": 2703
    },
    {
      "epoch": 0.2704,
      "grad_norm": 3.79034423828125,
      "learning_rate": 0.00018284515326522346,
      "loss": 5.2872,
      "step": 2704
    },
    {
      "epoch": 0.2705,
      "grad_norm": 6.253812313079834,
      "learning_rate": 0.00018282559840976042,
      "loss": 5.0673,
      "step": 2705
    },
    {
      "epoch": 0.2706,
      "grad_norm": 5.042981147766113,
      "learning_rate": 0.00018280603346224945,
      "loss": 6.1307,
      "step": 2706
    },
    {
      "epoch": 0.2707,
      "grad_norm": 3.34134840965271,
      "learning_rate": 0.00018278645842507448,
      "loss": 5.6421,
      "step": 2707
    },
    {
      "epoch": 0.2708,
      "grad_norm": 2.638489246368408,
      "learning_rate": 0.00018276687330062065,
      "loss": 5.13,
      "step": 2708
    },
    {
      "epoch": 0.2709,
      "grad_norm": 7.0063042640686035,
      "learning_rate": 0.00018274727809127438,
      "loss": 6.4869,
      "step": 2709
    },
    {
      "epoch": 0.271,
      "grad_norm": 4.501521587371826,
      "learning_rate": 0.00018272767279942328,
      "loss": 6.0969,
      "step": 2710
    },
    {
      "epoch": 0.2711,
      "grad_norm": 3.859851598739624,
      "learning_rate": 0.00018270805742745617,
      "loss": 5.9042,
      "step": 2711
    },
    {
      "epoch": 0.2712,
      "grad_norm": 2.363582134246826,
      "learning_rate": 0.00018268843197776318,
      "loss": 5.5945,
      "step": 2712
    },
    {
      "epoch": 0.2713,
      "grad_norm": 2.4202029705047607,
      "learning_rate": 0.00018266879645273556,
      "loss": 5.1675,
      "step": 2713
    },
    {
      "epoch": 0.2714,
      "grad_norm": 3.551102638244629,
      "learning_rate": 0.00018264915085476583,
      "loss": 5.2862,
      "step": 2714
    },
    {
      "epoch": 0.2715,
      "grad_norm": 3.346820592880249,
      "learning_rate": 0.0001826294951862478,
      "loss": 5.2287,
      "step": 2715
    },
    {
      "epoch": 0.2716,
      "grad_norm": 2.961416721343994,
      "learning_rate": 0.00018260982944957638,
      "loss": 5.8616,
      "step": 2716
    },
    {
      "epoch": 0.2717,
      "grad_norm": 2.4196159839630127,
      "learning_rate": 0.00018259015364714787,
      "loss": 5.0172,
      "step": 2717
    },
    {
      "epoch": 0.2718,
      "grad_norm": 6.818941593170166,
      "learning_rate": 0.00018257046778135964,
      "loss": 6.1112,
      "step": 2718
    },
    {
      "epoch": 0.2719,
      "grad_norm": 6.030742645263672,
      "learning_rate": 0.00018255077185461038,
      "loss": 5.1445,
      "step": 2719
    },
    {
      "epoch": 0.272,
      "grad_norm": 2.5305256843566895,
      "learning_rate": 0.00018253106586929997,
      "loss": 5.0809,
      "step": 2720
    },
    {
      "epoch": 0.2721,
      "grad_norm": 3.3271701335906982,
      "learning_rate": 0.00018251134982782952,
      "loss": 5.746,
      "step": 2721
    },
    {
      "epoch": 0.2722,
      "grad_norm": 2.476104259490967,
      "learning_rate": 0.00018249162373260141,
      "loss": 5.2196,
      "step": 2722
    },
    {
      "epoch": 0.2723,
      "grad_norm": 3.612180233001709,
      "learning_rate": 0.0001824718875860191,
      "loss": 5.4058,
      "step": 2723
    },
    {
      "epoch": 0.2724,
      "grad_norm": 3.190746545791626,
      "learning_rate": 0.00018245214139048753,
      "loss": 5.5422,
      "step": 2724
    },
    {
      "epoch": 0.2725,
      "grad_norm": 3.1574249267578125,
      "learning_rate": 0.0001824323851484126,
      "loss": 4.8338,
      "step": 2725
    },
    {
      "epoch": 0.2726,
      "grad_norm": 3.284080982208252,
      "learning_rate": 0.00018241261886220154,
      "loss": 4.9684,
      "step": 2726
    },
    {
      "epoch": 0.2727,
      "grad_norm": 3.319180965423584,
      "learning_rate": 0.00018239284253426295,
      "loss": 5.4762,
      "step": 2727
    },
    {
      "epoch": 0.2728,
      "grad_norm": 3.5093843936920166,
      "learning_rate": 0.00018237305616700637,
      "loss": 5.9748,
      "step": 2728
    },
    {
      "epoch": 0.2729,
      "grad_norm": 4.502373218536377,
      "learning_rate": 0.00018235325976284275,
      "loss": 6.217,
      "step": 2729
    },
    {
      "epoch": 0.273,
      "grad_norm": 6.9914774894714355,
      "learning_rate": 0.00018233345332418423,
      "loss": 5.3372,
      "step": 2730
    },
    {
      "epoch": 0.2731,
      "grad_norm": 6.297598838806152,
      "learning_rate": 0.0001823136368534442,
      "loss": 5.3372,
      "step": 2731
    },
    {
      "epoch": 0.2732,
      "grad_norm": 3.3544564247131348,
      "learning_rate": 0.00018229381035303718,
      "loss": 5.2332,
      "step": 2732
    },
    {
      "epoch": 0.2733,
      "grad_norm": 4.757925033569336,
      "learning_rate": 0.000182273973825379,
      "loss": 4.9106,
      "step": 2733
    },
    {
      "epoch": 0.2734,
      "grad_norm": 3.6217055320739746,
      "learning_rate": 0.00018225412727288667,
      "loss": 5.5134,
      "step": 2734
    },
    {
      "epoch": 0.2735,
      "grad_norm": 4.200356483459473,
      "learning_rate": 0.00018223427069797844,
      "loss": 5.1881,
      "step": 2735
    },
    {
      "epoch": 0.2736,
      "grad_norm": 7.515629768371582,
      "learning_rate": 0.00018221440410307374,
      "loss": 5.6365,
      "step": 2736
    },
    {
      "epoch": 0.2737,
      "grad_norm": 2.938361883163452,
      "learning_rate": 0.0001821945274905933,
      "loss": 5.5716,
      "step": 2737
    },
    {
      "epoch": 0.2738,
      "grad_norm": 2.784604549407959,
      "learning_rate": 0.00018217464086295904,
      "loss": 5.0564,
      "step": 2738
    },
    {
      "epoch": 0.2739,
      "grad_norm": 1.9898983240127563,
      "learning_rate": 0.00018215474422259402,
      "loss": 5.6819,
      "step": 2739
    },
    {
      "epoch": 0.274,
      "grad_norm": 6.820799827575684,
      "learning_rate": 0.00018213483757192263,
      "loss": 5.1545,
      "step": 2740
    },
    {
      "epoch": 0.2741,
      "grad_norm": 5.608919620513916,
      "learning_rate": 0.00018211492091337042,
      "loss": 5.2786,
      "step": 2741
    },
    {
      "epoch": 0.2742,
      "grad_norm": 10.240433692932129,
      "learning_rate": 0.00018209499424936415,
      "loss": 5.7025,
      "step": 2742
    },
    {
      "epoch": 0.2743,
      "grad_norm": 2.9937682151794434,
      "learning_rate": 0.0001820750575823319,
      "loss": 5.3658,
      "step": 2743
    },
    {
      "epoch": 0.2744,
      "grad_norm": 4.136473178863525,
      "learning_rate": 0.00018205511091470283,
      "loss": 7.2519,
      "step": 2744
    },
    {
      "epoch": 0.2745,
      "grad_norm": 4.157846927642822,
      "learning_rate": 0.0001820351542489074,
      "loss": 5.231,
      "step": 2745
    },
    {
      "epoch": 0.2746,
      "grad_norm": 6.541555881500244,
      "learning_rate": 0.00018201518758737724,
      "loss": 6.0881,
      "step": 2746
    },
    {
      "epoch": 0.2747,
      "grad_norm": 3.225083827972412,
      "learning_rate": 0.00018199521093254523,
      "loss": 5.5165,
      "step": 2747
    },
    {
      "epoch": 0.2748,
      "grad_norm": 6.086544036865234,
      "learning_rate": 0.00018197522428684552,
      "loss": 5.4677,
      "step": 2748
    },
    {
      "epoch": 0.2749,
      "grad_norm": 5.616893291473389,
      "learning_rate": 0.0001819552276527134,
      "loss": 6.2557,
      "step": 2749
    },
    {
      "epoch": 0.275,
      "grad_norm": 2.893369197845459,
      "learning_rate": 0.00018193522103258537,
      "loss": 6.4249,
      "step": 2750
    },
    {
      "epoch": 0.2751,
      "grad_norm": 5.213644504547119,
      "learning_rate": 0.0001819152044288992,
      "loss": 5.5454,
      "step": 2751
    },
    {
      "epoch": 0.2752,
      "grad_norm": 5.832185745239258,
      "learning_rate": 0.00018189517784409381,
      "loss": 5.4427,
      "step": 2752
    },
    {
      "epoch": 0.2753,
      "grad_norm": 2.412778377532959,
      "learning_rate": 0.00018187514128060946,
      "loss": 5.5832,
      "step": 2753
    },
    {
      "epoch": 0.2754,
      "grad_norm": 6.627700328826904,
      "learning_rate": 0.0001818550947408875,
      "loss": 6.043,
      "step": 2754
    },
    {
      "epoch": 0.2755,
      "grad_norm": 3.246513843536377,
      "learning_rate": 0.0001818350382273705,
      "loss": 6.364,
      "step": 2755
    },
    {
      "epoch": 0.2756,
      "grad_norm": 3.1165008544921875,
      "learning_rate": 0.00018181497174250236,
      "loss": 5.4903,
      "step": 2756
    },
    {
      "epoch": 0.2757,
      "grad_norm": 3.0534744262695312,
      "learning_rate": 0.00018179489528872807,
      "loss": 5.9773,
      "step": 2757
    },
    {
      "epoch": 0.2758,
      "grad_norm": 7.608018398284912,
      "learning_rate": 0.00018177480886849388,
      "loss": 5.8138,
      "step": 2758
    },
    {
      "epoch": 0.2759,
      "grad_norm": 11.174581527709961,
      "learning_rate": 0.0001817547124842473,
      "loss": 5.4096,
      "step": 2759
    },
    {
      "epoch": 0.276,
      "grad_norm": 3.3928987979888916,
      "learning_rate": 0.00018173460613843701,
      "loss": 5.337,
      "step": 2760
    },
    {
      "epoch": 0.2761,
      "grad_norm": 1.9970844984054565,
      "learning_rate": 0.00018171448983351284,
      "loss": 5.3426,
      "step": 2761
    },
    {
      "epoch": 0.2762,
      "grad_norm": 2.5097193717956543,
      "learning_rate": 0.00018169436357192602,
      "loss": 4.8624,
      "step": 2762
    },
    {
      "epoch": 0.2763,
      "grad_norm": 3.941918134689331,
      "learning_rate": 0.00018167422735612877,
      "loss": 5.6973,
      "step": 2763
    },
    {
      "epoch": 0.2764,
      "grad_norm": 3.628981113433838,
      "learning_rate": 0.00018165408118857464,
      "loss": 5.7916,
      "step": 2764
    },
    {
      "epoch": 0.2765,
      "grad_norm": 3.022580146789551,
      "learning_rate": 0.00018163392507171842,
      "loss": 5.3205,
      "step": 2765
    },
    {
      "epoch": 0.2766,
      "grad_norm": 2.911552667617798,
      "learning_rate": 0.00018161375900801604,
      "loss": 5.7155,
      "step": 2766
    },
    {
      "epoch": 0.2767,
      "grad_norm": 4.6182475090026855,
      "learning_rate": 0.00018159358299992467,
      "loss": 5.8237,
      "step": 2767
    },
    {
      "epoch": 0.2768,
      "grad_norm": 2.475855827331543,
      "learning_rate": 0.00018157339704990275,
      "loss": 5.2008,
      "step": 2768
    },
    {
      "epoch": 0.2769,
      "grad_norm": 3.689385175704956,
      "learning_rate": 0.00018155320116040982,
      "loss": 5.0413,
      "step": 2769
    },
    {
      "epoch": 0.277,
      "grad_norm": 7.803164005279541,
      "learning_rate": 0.00018153299533390672,
      "loss": 5.5262,
      "step": 2770
    },
    {
      "epoch": 0.2771,
      "grad_norm": 15.046921730041504,
      "learning_rate": 0.00018151277957285543,
      "loss": 5.669,
      "step": 2771
    },
    {
      "epoch": 0.2772,
      "grad_norm": 2.8676512241363525,
      "learning_rate": 0.00018149255387971922,
      "loss": 6.2988,
      "step": 2772
    },
    {
      "epoch": 0.2773,
      "grad_norm": 6.153730392456055,
      "learning_rate": 0.00018147231825696252,
      "loss": 5.5622,
      "step": 2773
    },
    {
      "epoch": 0.2774,
      "grad_norm": 6.399447917938232,
      "learning_rate": 0.00018145207270705096,
      "loss": 5.6671,
      "step": 2774
    },
    {
      "epoch": 0.2775,
      "grad_norm": 2.7839276790618896,
      "learning_rate": 0.0001814318172324514,
      "loss": 5.5305,
      "step": 2775
    },
    {
      "epoch": 0.2776,
      "grad_norm": 3.7380454540252686,
      "learning_rate": 0.00018141155183563193,
      "loss": 5.6946,
      "step": 2776
    },
    {
      "epoch": 0.2777,
      "grad_norm": 5.17465353012085,
      "learning_rate": 0.00018139127651906184,
      "loss": 5.2929,
      "step": 2777
    },
    {
      "epoch": 0.2778,
      "grad_norm": 2.4537322521209717,
      "learning_rate": 0.00018137099128521156,
      "loss": 5.4073,
      "step": 2778
    },
    {
      "epoch": 0.2779,
      "grad_norm": 7.0417633056640625,
      "learning_rate": 0.0001813506961365528,
      "loss": 5.8073,
      "step": 2779
    },
    {
      "epoch": 0.278,
      "grad_norm": 6.230091094970703,
      "learning_rate": 0.00018133039107555852,
      "loss": 5.3603,
      "step": 2780
    },
    {
      "epoch": 0.2781,
      "grad_norm": 2.9987635612487793,
      "learning_rate": 0.00018131007610470276,
      "loss": 5.8846,
      "step": 2781
    },
    {
      "epoch": 0.2782,
      "grad_norm": 4.443158149719238,
      "learning_rate": 0.0001812897512264609,
      "loss": 6.4967,
      "step": 2782
    },
    {
      "epoch": 0.2783,
      "grad_norm": 41.00222396850586,
      "learning_rate": 0.0001812694164433094,
      "loss": 7.26,
      "step": 2783
    },
    {
      "epoch": 0.2784,
      "grad_norm": 5.938882350921631,
      "learning_rate": 0.00018124907175772604,
      "loss": 5.3937,
      "step": 2784
    },
    {
      "epoch": 0.2785,
      "grad_norm": 2.378345489501953,
      "learning_rate": 0.0001812287171721897,
      "loss": 4.8758,
      "step": 2785
    },
    {
      "epoch": 0.2786,
      "grad_norm": 9.647140502929688,
      "learning_rate": 0.00018120835268918063,
      "loss": 5.6118,
      "step": 2786
    },
    {
      "epoch": 0.2787,
      "grad_norm": 3.054445266723633,
      "learning_rate": 0.0001811879783111801,
      "loss": 5.3103,
      "step": 2787
    },
    {
      "epoch": 0.2788,
      "grad_norm": 2.734070301055908,
      "learning_rate": 0.00018116759404067064,
      "loss": 5.1587,
      "step": 2788
    },
    {
      "epoch": 0.2789,
      "grad_norm": 4.585324287414551,
      "learning_rate": 0.00018114719988013612,
      "loss": 5.1977,
      "step": 2789
    },
    {
      "epoch": 0.279,
      "grad_norm": 2.7257347106933594,
      "learning_rate": 0.00018112679583206137,
      "loss": 5.0487,
      "step": 2790
    },
    {
      "epoch": 0.2791,
      "grad_norm": 4.771512985229492,
      "learning_rate": 0.00018110638189893267,
      "loss": 5.9181,
      "step": 2791
    },
    {
      "epoch": 0.2792,
      "grad_norm": 6.526870250701904,
      "learning_rate": 0.00018108595808323736,
      "loss": 5.4965,
      "step": 2792
    },
    {
      "epoch": 0.2793,
      "grad_norm": 5.317654609680176,
      "learning_rate": 0.000181065524387464,
      "loss": 5.6697,
      "step": 2793
    },
    {
      "epoch": 0.2794,
      "grad_norm": 6.799176216125488,
      "learning_rate": 0.0001810450808141024,
      "loss": 4.9602,
      "step": 2794
    },
    {
      "epoch": 0.2795,
      "grad_norm": 2.289069175720215,
      "learning_rate": 0.00018102462736564355,
      "loss": 5.2898,
      "step": 2795
    },
    {
      "epoch": 0.2796,
      "grad_norm": 2.556957483291626,
      "learning_rate": 0.00018100416404457961,
      "loss": 4.7256,
      "step": 2796
    },
    {
      "epoch": 0.2797,
      "grad_norm": 3.73872447013855,
      "learning_rate": 0.00018098369085340398,
      "loss": 5.9347,
      "step": 2797
    },
    {
      "epoch": 0.2798,
      "grad_norm": 3.0197083950042725,
      "learning_rate": 0.00018096320779461132,
      "loss": 5.625,
      "step": 2798
    },
    {
      "epoch": 0.2799,
      "grad_norm": 3.7838425636291504,
      "learning_rate": 0.00018094271487069735,
      "loss": 5.0735,
      "step": 2799
    },
    {
      "epoch": 0.28,
      "grad_norm": 2.0645253658294678,
      "learning_rate": 0.00018092221208415907,
      "loss": 5.3689,
      "step": 2800
    },
    {
      "epoch": 0.2801,
      "grad_norm": 3.7970077991485596,
      "learning_rate": 0.00018090169943749476,
      "loss": 5.6701,
      "step": 2801
    },
    {
      "epoch": 0.2802,
      "grad_norm": 5.667590618133545,
      "learning_rate": 0.00018088117693320374,
      "loss": 5.1737,
      "step": 2802
    },
    {
      "epoch": 0.2803,
      "grad_norm": 3.688007354736328,
      "learning_rate": 0.00018086064457378665,
      "loss": 6.0455,
      "step": 2803
    },
    {
      "epoch": 0.2804,
      "grad_norm": 12.64063549041748,
      "learning_rate": 0.00018084010236174534,
      "loss": 5.4183,
      "step": 2804
    },
    {
      "epoch": 0.2805,
      "grad_norm": 9.209149360656738,
      "learning_rate": 0.00018081955029958274,
      "loss": 6.2066,
      "step": 2805
    },
    {
      "epoch": 0.2806,
      "grad_norm": 6.702190399169922,
      "learning_rate": 0.00018079898838980305,
      "loss": 6.1267,
      "step": 2806
    },
    {
      "epoch": 0.2807,
      "grad_norm": 10.481171607971191,
      "learning_rate": 0.00018077841663491175,
      "loss": 5.6892,
      "step": 2807
    },
    {
      "epoch": 0.2808,
      "grad_norm": 7.181934833526611,
      "learning_rate": 0.0001807578350374154,
      "loss": 5.5487,
      "step": 2808
    },
    {
      "epoch": 0.2809,
      "grad_norm": 4.580722808837891,
      "learning_rate": 0.00018073724359982186,
      "loss": 6.3691,
      "step": 2809
    },
    {
      "epoch": 0.281,
      "grad_norm": 3.090749740600586,
      "learning_rate": 0.00018071664232464002,
      "loss": 5.1505,
      "step": 2810
    },
    {
      "epoch": 0.2811,
      "grad_norm": 2.799746036529541,
      "learning_rate": 0.00018069603121438022,
      "loss": 4.6698,
      "step": 2811
    },
    {
      "epoch": 0.2812,
      "grad_norm": 5.325982093811035,
      "learning_rate": 0.00018067541027155375,
      "loss": 6.3855,
      "step": 2812
    },
    {
      "epoch": 0.2813,
      "grad_norm": 3.7008719444274902,
      "learning_rate": 0.00018065477949867327,
      "loss": 5.0409,
      "step": 2813
    },
    {
      "epoch": 0.2814,
      "grad_norm": 4.537755489349365,
      "learning_rate": 0.00018063413889825254,
      "loss": 5.6009,
      "step": 2814
    },
    {
      "epoch": 0.2815,
      "grad_norm": 3.4157462120056152,
      "learning_rate": 0.0001806134884728066,
      "loss": 6.7134,
      "step": 2815
    },
    {
      "epoch": 0.2816,
      "grad_norm": 2.247720718383789,
      "learning_rate": 0.00018059282822485158,
      "loss": 5.2295,
      "step": 2816
    },
    {
      "epoch": 0.2817,
      "grad_norm": 2.690124988555908,
      "learning_rate": 0.00018057215815690494,
      "loss": 5.977,
      "step": 2817
    },
    {
      "epoch": 0.2818,
      "grad_norm": 7.871548175811768,
      "learning_rate": 0.00018055147827148523,
      "loss": 5.4574,
      "step": 2818
    },
    {
      "epoch": 0.2819,
      "grad_norm": 9.54426383972168,
      "learning_rate": 0.0001805307885711122,
      "loss": 5.3856,
      "step": 2819
    },
    {
      "epoch": 0.282,
      "grad_norm": 3.4690473079681396,
      "learning_rate": 0.0001805100890583069,
      "loss": 6.0659,
      "step": 2820
    },
    {
      "epoch": 0.2821,
      "grad_norm": 3.3459010124206543,
      "learning_rate": 0.0001804893797355914,
      "loss": 5.6815,
      "step": 2821
    },
    {
      "epoch": 0.2822,
      "grad_norm": 11.740500450134277,
      "learning_rate": 0.00018046866060548918,
      "loss": 6.8382,
      "step": 2822
    },
    {
      "epoch": 0.2823,
      "grad_norm": 1.9609824419021606,
      "learning_rate": 0.00018044793167052477,
      "loss": 5.6518,
      "step": 2823
    },
    {
      "epoch": 0.2824,
      "grad_norm": 5.130919456481934,
      "learning_rate": 0.00018042719293322388,
      "loss": 5.1137,
      "step": 2824
    },
    {
      "epoch": 0.2825,
      "grad_norm": 3.3440327644348145,
      "learning_rate": 0.00018040644439611348,
      "loss": 5.6382,
      "step": 2825
    },
    {
      "epoch": 0.2826,
      "grad_norm": 4.962672710418701,
      "learning_rate": 0.00018038568606172173,
      "loss": 5.8734,
      "step": 2826
    },
    {
      "epoch": 0.2827,
      "grad_norm": 4.468346118927002,
      "learning_rate": 0.00018036491793257798,
      "loss": 5.2997,
      "step": 2827
    },
    {
      "epoch": 0.2828,
      "grad_norm": 4.065969944000244,
      "learning_rate": 0.00018034414001121278,
      "loss": 5.4603,
      "step": 2828
    },
    {
      "epoch": 0.2829,
      "grad_norm": 2.6256966590881348,
      "learning_rate": 0.0001803233523001578,
      "loss": 4.9203,
      "step": 2829
    },
    {
      "epoch": 0.283,
      "grad_norm": 2.5168275833129883,
      "learning_rate": 0.000180302554801946,
      "loss": 5.7859,
      "step": 2830
    },
    {
      "epoch": 0.2831,
      "grad_norm": 3.1766552925109863,
      "learning_rate": 0.00018028174751911146,
      "loss": 5.4639,
      "step": 2831
    },
    {
      "epoch": 0.2832,
      "grad_norm": 2.085548162460327,
      "learning_rate": 0.00018026093045418954,
      "loss": 5.1106,
      "step": 2832
    },
    {
      "epoch": 0.2833,
      "grad_norm": 2.6526408195495605,
      "learning_rate": 0.0001802401036097167,
      "loss": 5.2026,
      "step": 2833
    },
    {
      "epoch": 0.2834,
      "grad_norm": 2.8944482803344727,
      "learning_rate": 0.00018021926698823059,
      "loss": 5.4379,
      "step": 2834
    },
    {
      "epoch": 0.2835,
      "grad_norm": 3.760880708694458,
      "learning_rate": 0.00018019842059227012,
      "loss": 5.2445,
      "step": 2835
    },
    {
      "epoch": 0.2836,
      "grad_norm": 4.943646430969238,
      "learning_rate": 0.0001801775644243754,
      "loss": 5.7092,
      "step": 2836
    },
    {
      "epoch": 0.2837,
      "grad_norm": 3.1980888843536377,
      "learning_rate": 0.00018015669848708767,
      "loss": 5.5544,
      "step": 2837
    },
    {
      "epoch": 0.2838,
      "grad_norm": 3.1224524974823,
      "learning_rate": 0.00018013582278294935,
      "loss": 4.7157,
      "step": 2838
    },
    {
      "epoch": 0.2839,
      "grad_norm": 11.246402740478516,
      "learning_rate": 0.00018011493731450413,
      "loss": 6.1933,
      "step": 2839
    },
    {
      "epoch": 0.284,
      "grad_norm": 2.9872562885284424,
      "learning_rate": 0.0001800940420842968,
      "loss": 5.5366,
      "step": 2840
    },
    {
      "epoch": 0.2841,
      "grad_norm": 3.289379596710205,
      "learning_rate": 0.00018007313709487334,
      "loss": 5.6033,
      "step": 2841
    },
    {
      "epoch": 0.2842,
      "grad_norm": 3.3337604999542236,
      "learning_rate": 0.0001800522223487811,
      "loss": 5.4007,
      "step": 2842
    },
    {
      "epoch": 0.2843,
      "grad_norm": 4.990576267242432,
      "learning_rate": 0.0001800312978485683,
      "loss": 6.237,
      "step": 2843
    },
    {
      "epoch": 0.2844,
      "grad_norm": 4.153465747833252,
      "learning_rate": 0.00018001036359678469,
      "loss": 5.4984,
      "step": 2844
    },
    {
      "epoch": 0.2845,
      "grad_norm": 4.846620082855225,
      "learning_rate": 0.00017998941959598095,
      "loss": 5.4403,
      "step": 2845
    },
    {
      "epoch": 0.2846,
      "grad_norm": 12.024687767028809,
      "learning_rate": 0.00017996846584870908,
      "loss": 5.2143,
      "step": 2846
    },
    {
      "epoch": 0.2847,
      "grad_norm": 3.990976095199585,
      "learning_rate": 0.0001799475023575222,
      "loss": 5.6391,
      "step": 2847
    },
    {
      "epoch": 0.2848,
      "grad_norm": 3.328070878982544,
      "learning_rate": 0.00017992652912497464,
      "loss": 5.9551,
      "step": 2848
    },
    {
      "epoch": 0.2849,
      "grad_norm": 3.1685400009155273,
      "learning_rate": 0.00017990554615362198,
      "loss": 5.6576,
      "step": 2849
    },
    {
      "epoch": 0.285,
      "grad_norm": 15.693193435668945,
      "learning_rate": 0.00017988455344602092,
      "loss": 6.8175,
      "step": 2850
    },
    {
      "epoch": 0.2851,
      "grad_norm": 3.5548970699310303,
      "learning_rate": 0.00017986355100472928,
      "loss": 6.6805,
      "step": 2851
    },
    {
      "epoch": 0.2852,
      "grad_norm": 2.206057071685791,
      "learning_rate": 0.00017984253883230627,
      "loss": 5.2307,
      "step": 2852
    },
    {
      "epoch": 0.2853,
      "grad_norm": 10.458955764770508,
      "learning_rate": 0.00017982151693131203,
      "loss": 4.8862,
      "step": 2853
    },
    {
      "epoch": 0.2854,
      "grad_norm": 2.9186697006225586,
      "learning_rate": 0.0001798004853043081,
      "loss": 5.101,
      "step": 2854
    },
    {
      "epoch": 0.2855,
      "grad_norm": 10.774382591247559,
      "learning_rate": 0.0001797794439538571,
      "loss": 5.7654,
      "step": 2855
    },
    {
      "epoch": 0.2856,
      "grad_norm": 3.365683078765869,
      "learning_rate": 0.00017975839288252287,
      "loss": 5.698,
      "step": 2856
    },
    {
      "epoch": 0.2857,
      "grad_norm": 2.262589931488037,
      "learning_rate": 0.00017973733209287036,
      "loss": 5.6264,
      "step": 2857
    },
    {
      "epoch": 0.2858,
      "grad_norm": 5.9254536628723145,
      "learning_rate": 0.00017971626158746584,
      "loss": 5.8869,
      "step": 2858
    },
    {
      "epoch": 0.2859,
      "grad_norm": 11.102778434753418,
      "learning_rate": 0.00017969518136887663,
      "loss": 5.3207,
      "step": 2859
    },
    {
      "epoch": 0.286,
      "grad_norm": 4.959065914154053,
      "learning_rate": 0.00017967409143967132,
      "loss": 5.5163,
      "step": 2860
    },
    {
      "epoch": 0.2861,
      "grad_norm": 3.6302363872528076,
      "learning_rate": 0.00017965299180241963,
      "loss": 5.1398,
      "step": 2861
    },
    {
      "epoch": 0.2862,
      "grad_norm": 3.1261258125305176,
      "learning_rate": 0.00017963188245969253,
      "loss": 5.0208,
      "step": 2862
    },
    {
      "epoch": 0.2863,
      "grad_norm": 3.359898805618286,
      "learning_rate": 0.00017961076341406208,
      "loss": 5.3537,
      "step": 2863
    },
    {
      "epoch": 0.2864,
      "grad_norm": 3.1516056060791016,
      "learning_rate": 0.0001795896346681016,
      "loss": 4.9565,
      "step": 2864
    },
    {
      "epoch": 0.2865,
      "grad_norm": 3.8142476081848145,
      "learning_rate": 0.00017956849622438554,
      "loss": 5.2036,
      "step": 2865
    },
    {
      "epoch": 0.2866,
      "grad_norm": 4.426968574523926,
      "learning_rate": 0.00017954734808548958,
      "loss": 5.5728,
      "step": 2866
    },
    {
      "epoch": 0.2867,
      "grad_norm": 2.9751875400543213,
      "learning_rate": 0.00017952619025399057,
      "loss": 5.2883,
      "step": 2867
    },
    {
      "epoch": 0.2868,
      "grad_norm": 5.667436122894287,
      "learning_rate": 0.00017950502273246649,
      "loss": 5.3284,
      "step": 2868
    },
    {
      "epoch": 0.2869,
      "grad_norm": 7.096568584442139,
      "learning_rate": 0.00017948384552349657,
      "loss": 5.4936,
      "step": 2869
    },
    {
      "epoch": 0.287,
      "grad_norm": 4.420072555541992,
      "learning_rate": 0.00017946265862966114,
      "loss": 6.7795,
      "step": 2870
    },
    {
      "epoch": 0.2871,
      "grad_norm": 7.4643096923828125,
      "learning_rate": 0.00017944146205354182,
      "loss": 4.9397,
      "step": 2871
    },
    {
      "epoch": 0.2872,
      "grad_norm": 3.6722323894500732,
      "learning_rate": 0.00017942025579772132,
      "loss": 5.2582,
      "step": 2872
    },
    {
      "epoch": 0.2873,
      "grad_norm": 3.356196880340576,
      "learning_rate": 0.00017939903986478355,
      "loss": 5.4875,
      "step": 2873
    },
    {
      "epoch": 0.2874,
      "grad_norm": 3.5476198196411133,
      "learning_rate": 0.0001793778142573136,
      "loss": 4.816,
      "step": 2874
    },
    {
      "epoch": 0.2875,
      "grad_norm": 2.7464828491210938,
      "learning_rate": 0.0001793565789778978,
      "loss": 5.1564,
      "step": 2875
    },
    {
      "epoch": 0.2876,
      "grad_norm": 5.307430267333984,
      "learning_rate": 0.00017933533402912354,
      "loss": 4.7253,
      "step": 2876
    },
    {
      "epoch": 0.2877,
      "grad_norm": 4.037013530731201,
      "learning_rate": 0.00017931407941357947,
      "loss": 5.2029,
      "step": 2877
    },
    {
      "epoch": 0.2878,
      "grad_norm": 2.2222795486450195,
      "learning_rate": 0.0001792928151338554,
      "loss": 5.2517,
      "step": 2878
    },
    {
      "epoch": 0.2879,
      "grad_norm": 4.640401363372803,
      "learning_rate": 0.00017927154119254236,
      "loss": 5.2368,
      "step": 2879
    },
    {
      "epoch": 0.288,
      "grad_norm": 3.289156436920166,
      "learning_rate": 0.00017925025759223245,
      "loss": 5.3552,
      "step": 2880
    },
    {
      "epoch": 0.2881,
      "grad_norm": 5.4246296882629395,
      "learning_rate": 0.00017922896433551907,
      "loss": 5.7847,
      "step": 2881
    },
    {
      "epoch": 0.2882,
      "grad_norm": 3.2969672679901123,
      "learning_rate": 0.00017920766142499672,
      "loss": 5.3127,
      "step": 2882
    },
    {
      "epoch": 0.2883,
      "grad_norm": 2.6906726360321045,
      "learning_rate": 0.00017918634886326108,
      "loss": 5.1507,
      "step": 2883
    },
    {
      "epoch": 0.2884,
      "grad_norm": 4.110598087310791,
      "learning_rate": 0.00017916502665290903,
      "loss": 5.0911,
      "step": 2884
    },
    {
      "epoch": 0.2885,
      "grad_norm": 7.478288650512695,
      "learning_rate": 0.0001791436947965386,
      "loss": 6.0333,
      "step": 2885
    },
    {
      "epoch": 0.2886,
      "grad_norm": 7.874764442443848,
      "learning_rate": 0.00017912235329674902,
      "loss": 5.6793,
      "step": 2886
    },
    {
      "epoch": 0.2887,
      "grad_norm": 8.81761646270752,
      "learning_rate": 0.0001791010021561407,
      "loss": 5.3839,
      "step": 2887
    },
    {
      "epoch": 0.2888,
      "grad_norm": 2.8637006282806396,
      "learning_rate": 0.0001790796413773152,
      "loss": 5.2877,
      "step": 2888
    },
    {
      "epoch": 0.2889,
      "grad_norm": 4.028423309326172,
      "learning_rate": 0.0001790582709628753,
      "loss": 5.249,
      "step": 2889
    },
    {
      "epoch": 0.289,
      "grad_norm": 8.272892951965332,
      "learning_rate": 0.0001790368909154249,
      "loss": 5.6436,
      "step": 2890
    },
    {
      "epoch": 0.2891,
      "grad_norm": 3.7571890354156494,
      "learning_rate": 0.00017901550123756906,
      "loss": 5.4408,
      "step": 2891
    },
    {
      "epoch": 0.2892,
      "grad_norm": 9.79236888885498,
      "learning_rate": 0.00017899410193191406,
      "loss": 6.3816,
      "step": 2892
    },
    {
      "epoch": 0.2893,
      "grad_norm": 2.906367778778076,
      "learning_rate": 0.00017897269300106737,
      "loss": 5.2976,
      "step": 2893
    },
    {
      "epoch": 0.2894,
      "grad_norm": 6.8006744384765625,
      "learning_rate": 0.0001789512744476376,
      "loss": 6.2739,
      "step": 2894
    },
    {
      "epoch": 0.2895,
      "grad_norm": 4.5579447746276855,
      "learning_rate": 0.0001789298462742345,
      "loss": 5.4984,
      "step": 2895
    },
    {
      "epoch": 0.2896,
      "grad_norm": 5.839450359344482,
      "learning_rate": 0.00017890840848346908,
      "loss": 6.0575,
      "step": 2896
    },
    {
      "epoch": 0.2897,
      "grad_norm": 5.616231441497803,
      "learning_rate": 0.00017888696107795342,
      "loss": 5.4824,
      "step": 2897
    },
    {
      "epoch": 0.2898,
      "grad_norm": 3.0371551513671875,
      "learning_rate": 0.00017886550406030085,
      "loss": 5.2469,
      "step": 2898
    },
    {
      "epoch": 0.2899,
      "grad_norm": 2.661224842071533,
      "learning_rate": 0.00017884403743312582,
      "loss": 5.1537,
      "step": 2899
    },
    {
      "epoch": 0.29,
      "grad_norm": 2.6646981239318848,
      "learning_rate": 0.00017882256119904403,
      "loss": 5.4438,
      "step": 2900
    },
    {
      "epoch": 0.2901,
      "grad_norm": 2.8743245601654053,
      "learning_rate": 0.00017880107536067218,
      "loss": 4.8566,
      "step": 2901
    },
    {
      "epoch": 0.2902,
      "grad_norm": 5.41933536529541,
      "learning_rate": 0.0001787795799206284,
      "loss": 5.1994,
      "step": 2902
    },
    {
      "epoch": 0.2903,
      "grad_norm": 2.491196393966675,
      "learning_rate": 0.00017875807488153175,
      "loss": 5.6271,
      "step": 2903
    },
    {
      "epoch": 0.2904,
      "grad_norm": 13.461886405944824,
      "learning_rate": 0.00017873656024600254,
      "loss": 5.4182,
      "step": 2904
    },
    {
      "epoch": 0.2905,
      "grad_norm": 4.286015033721924,
      "learning_rate": 0.00017871503601666233,
      "loss": 5.2218,
      "step": 2905
    },
    {
      "epoch": 0.2906,
      "grad_norm": 3.50974178314209,
      "learning_rate": 0.00017869350219613375,
      "loss": 6.2943,
      "step": 2906
    },
    {
      "epoch": 0.2907,
      "grad_norm": 3.1876442432403564,
      "learning_rate": 0.0001786719587870406,
      "loss": 5.405,
      "step": 2907
    },
    {
      "epoch": 0.2908,
      "grad_norm": 3.364816665649414,
      "learning_rate": 0.00017865040579200794,
      "loss": 5.303,
      "step": 2908
    },
    {
      "epoch": 0.2909,
      "grad_norm": 3.3081719875335693,
      "learning_rate": 0.00017862884321366188,
      "loss": 6.5548,
      "step": 2909
    },
    {
      "epoch": 0.291,
      "grad_norm": 3.2954554557800293,
      "learning_rate": 0.0001786072710546298,
      "loss": 5.3964,
      "step": 2910
    },
    {
      "epoch": 0.2911,
      "grad_norm": 3.742473602294922,
      "learning_rate": 0.0001785856893175402,
      "loss": 5.1328,
      "step": 2911
    },
    {
      "epoch": 0.2912,
      "grad_norm": 7.857283115386963,
      "learning_rate": 0.00017856409800502272,
      "loss": 5.5677,
      "step": 2912
    },
    {
      "epoch": 0.2913,
      "grad_norm": 3.4974312782287598,
      "learning_rate": 0.00017854249711970818,
      "loss": 5.404,
      "step": 2913
    },
    {
      "epoch": 0.2914,
      "grad_norm": 2.547431468963623,
      "learning_rate": 0.00017852088666422863,
      "loss": 5.4282,
      "step": 2914
    },
    {
      "epoch": 0.2915,
      "grad_norm": 3.9616334438323975,
      "learning_rate": 0.00017849926664121726,
      "loss": 5.5093,
      "step": 2915
    },
    {
      "epoch": 0.2916,
      "grad_norm": 6.170868396759033,
      "learning_rate": 0.0001784776370533083,
      "loss": 6.082,
      "step": 2916
    },
    {
      "epoch": 0.2917,
      "grad_norm": 12.67625617980957,
      "learning_rate": 0.00017845599790313735,
      "loss": 5.7176,
      "step": 2917
    },
    {
      "epoch": 0.2918,
      "grad_norm": 4.610630989074707,
      "learning_rate": 0.000178434349193341,
      "loss": 5.3085,
      "step": 2918
    },
    {
      "epoch": 0.2919,
      "grad_norm": 2.473715305328369,
      "learning_rate": 0.00017841269092655715,
      "loss": 5.761,
      "step": 2919
    },
    {
      "epoch": 0.292,
      "grad_norm": 3.4445669651031494,
      "learning_rate": 0.00017839102310542477,
      "loss": 5.4743,
      "step": 2920
    },
    {
      "epoch": 0.2921,
      "grad_norm": 5.087316989898682,
      "learning_rate": 0.000178369345732584,
      "loss": 5.329,
      "step": 2921
    },
    {
      "epoch": 0.2922,
      "grad_norm": 3.291902780532837,
      "learning_rate": 0.00017834765881067616,
      "loss": 5.4893,
      "step": 2922
    },
    {
      "epoch": 0.2923,
      "grad_norm": 2.269481658935547,
      "learning_rate": 0.00017832596234234376,
      "loss": 5.0421,
      "step": 2923
    },
    {
      "epoch": 0.2924,
      "grad_norm": 3.322335958480835,
      "learning_rate": 0.00017830425633023043,
      "loss": 5.2171,
      "step": 2924
    },
    {
      "epoch": 0.2925,
      "grad_norm": 2.7467939853668213,
      "learning_rate": 0.000178282540776981,
      "loss": 5.1491,
      "step": 2925
    },
    {
      "epoch": 0.2926,
      "grad_norm": 2.9509663581848145,
      "learning_rate": 0.0001782608156852414,
      "loss": 5.2156,
      "step": 2926
    },
    {
      "epoch": 0.2927,
      "grad_norm": 2.876457929611206,
      "learning_rate": 0.0001782390810576588,
      "loss": 5.5444,
      "step": 2927
    },
    {
      "epoch": 0.2928,
      "grad_norm": 6.478153228759766,
      "learning_rate": 0.00017821733689688153,
      "loss": 5.6824,
      "step": 2928
    },
    {
      "epoch": 0.2929,
      "grad_norm": 4.674193859100342,
      "learning_rate": 0.000178195583205559,
      "loss": 5.6978,
      "step": 2929
    },
    {
      "epoch": 0.293,
      "grad_norm": 4.993865489959717,
      "learning_rate": 0.00017817381998634185,
      "loss": 5.3576,
      "step": 2930
    },
    {
      "epoch": 0.2931,
      "grad_norm": 4.969654560089111,
      "learning_rate": 0.00017815204724188187,
      "loss": 4.8886,
      "step": 2931
    },
    {
      "epoch": 0.2932,
      "grad_norm": 5.1199421882629395,
      "learning_rate": 0.000178130264974832,
      "loss": 5.166,
      "step": 2932
    },
    {
      "epoch": 0.2933,
      "grad_norm": 2.4679224491119385,
      "learning_rate": 0.0001781084731878463,
      "loss": 5.459,
      "step": 2933
    },
    {
      "epoch": 0.2934,
      "grad_norm": 6.447225093841553,
      "learning_rate": 0.00017808667188358012,
      "loss": 5.1885,
      "step": 2934
    },
    {
      "epoch": 0.2935,
      "grad_norm": 19.935745239257812,
      "learning_rate": 0.00017806486106468981,
      "loss": 6.6985,
      "step": 2935
    },
    {
      "epoch": 0.2936,
      "grad_norm": 12.060554504394531,
      "learning_rate": 0.000178043040733833,
      "loss": 7.6359,
      "step": 2936
    },
    {
      "epoch": 0.2937,
      "grad_norm": 3.61661434173584,
      "learning_rate": 0.00017802121089366836,
      "loss": 6.3092,
      "step": 2937
    },
    {
      "epoch": 0.2938,
      "grad_norm": 2.2989940643310547,
      "learning_rate": 0.00017799937154685586,
      "loss": 5.072,
      "step": 2938
    },
    {
      "epoch": 0.2939,
      "grad_norm": 7.515536785125732,
      "learning_rate": 0.00017797752269605653,
      "loss": 5.2324,
      "step": 2939
    },
    {
      "epoch": 0.294,
      "grad_norm": 3.2353055477142334,
      "learning_rate": 0.00017795566434393258,
      "loss": 5.2879,
      "step": 2940
    },
    {
      "epoch": 0.2941,
      "grad_norm": 3.135542631149292,
      "learning_rate": 0.00017793379649314744,
      "loss": 5.3209,
      "step": 2941
    },
    {
      "epoch": 0.2942,
      "grad_norm": 2.287956953048706,
      "learning_rate": 0.00017791191914636554,
      "loss": 4.9845,
      "step": 2942
    },
    {
      "epoch": 0.2943,
      "grad_norm": 3.7022006511688232,
      "learning_rate": 0.00017789003230625266,
      "loss": 4.9997,
      "step": 2943
    },
    {
      "epoch": 0.2944,
      "grad_norm": 2.878105878829956,
      "learning_rate": 0.00017786813597547562,
      "loss": 5.3267,
      "step": 2944
    },
    {
      "epoch": 0.2945,
      "grad_norm": 3.3089182376861572,
      "learning_rate": 0.00017784623015670238,
      "loss": 5.8347,
      "step": 2945
    },
    {
      "epoch": 0.2946,
      "grad_norm": 4.0127363204956055,
      "learning_rate": 0.00017782431485260212,
      "loss": 4.9159,
      "step": 2946
    },
    {
      "epoch": 0.2947,
      "grad_norm": 4.297247409820557,
      "learning_rate": 0.00017780239006584515,
      "loss": 5.2197,
      "step": 2947
    },
    {
      "epoch": 0.2948,
      "grad_norm": 5.973016738891602,
      "learning_rate": 0.00017778045579910302,
      "loss": 4.9967,
      "step": 2948
    },
    {
      "epoch": 0.2949,
      "grad_norm": 2.9656598567962646,
      "learning_rate": 0.0001777585120550482,
      "loss": 5.2712,
      "step": 2949
    },
    {
      "epoch": 0.295,
      "grad_norm": 12.384346961975098,
      "learning_rate": 0.0001777365588363546,
      "loss": 5.2793,
      "step": 2950
    },
    {
      "epoch": 0.2951,
      "grad_norm": 3.7754080295562744,
      "learning_rate": 0.0001777145961456971,
      "loss": 5.7468,
      "step": 2951
    },
    {
      "epoch": 0.2952,
      "grad_norm": 5.101579189300537,
      "learning_rate": 0.0001776926239857518,
      "loss": 5.5439,
      "step": 2952
    },
    {
      "epoch": 0.2953,
      "grad_norm": 3.874807596206665,
      "learning_rate": 0.00017767064235919592,
      "loss": 5.4933,
      "step": 2953
    },
    {
      "epoch": 0.2954,
      "grad_norm": 3.0687201023101807,
      "learning_rate": 0.00017764865126870786,
      "loss": 5.925,
      "step": 2954
    },
    {
      "epoch": 0.2955,
      "grad_norm": 2.866825580596924,
      "learning_rate": 0.0001776266507169672,
      "loss": 5.3873,
      "step": 2955
    },
    {
      "epoch": 0.2956,
      "grad_norm": 18.13474464416504,
      "learning_rate": 0.0001776046407066546,
      "loss": 5.9507,
      "step": 2956
    },
    {
      "epoch": 0.2957,
      "grad_norm": 2.639171600341797,
      "learning_rate": 0.00017758262124045195,
      "loss": 5.2107,
      "step": 2957
    },
    {
      "epoch": 0.2958,
      "grad_norm": 9.714716911315918,
      "learning_rate": 0.0001775605923210422,
      "loss": 5.9014,
      "step": 2958
    },
    {
      "epoch": 0.2959,
      "grad_norm": 4.841752052307129,
      "learning_rate": 0.0001775385539511096,
      "loss": 5.3844,
      "step": 2959
    },
    {
      "epoch": 0.296,
      "grad_norm": 3.414358377456665,
      "learning_rate": 0.00017751650613333935,
      "loss": 5.6373,
      "step": 2960
    },
    {
      "epoch": 0.2961,
      "grad_norm": 5.950947284698486,
      "learning_rate": 0.00017749444887041799,
      "loss": 5.6032,
      "step": 2961
    },
    {
      "epoch": 0.2962,
      "grad_norm": 2.3767881393432617,
      "learning_rate": 0.00017747238216503307,
      "loss": 4.9565,
      "step": 2962
    },
    {
      "epoch": 0.2963,
      "grad_norm": 4.545330047607422,
      "learning_rate": 0.00017745030601987337,
      "loss": 6.1467,
      "step": 2963
    },
    {
      "epoch": 0.2964,
      "grad_norm": 5.585211753845215,
      "learning_rate": 0.00017742822043762888,
      "loss": 6.1433,
      "step": 2964
    },
    {
      "epoch": 0.2965,
      "grad_norm": 2.688643455505371,
      "learning_rate": 0.00017740612542099053,
      "loss": 4.7363,
      "step": 2965
    },
    {
      "epoch": 0.2966,
      "grad_norm": 5.30292272567749,
      "learning_rate": 0.00017738402097265064,
      "loss": 5.3222,
      "step": 2966
    },
    {
      "epoch": 0.2967,
      "grad_norm": 3.9252190589904785,
      "learning_rate": 0.0001773619070953025,
      "loss": 5.3465,
      "step": 2967
    },
    {
      "epoch": 0.2968,
      "grad_norm": 3.62306809425354,
      "learning_rate": 0.00017733978379164066,
      "loss": 5.3904,
      "step": 2968
    },
    {
      "epoch": 0.2969,
      "grad_norm": 2.2480628490448,
      "learning_rate": 0.00017731765106436073,
      "loss": 5.383,
      "step": 2969
    },
    {
      "epoch": 0.297,
      "grad_norm": 1.9993873834609985,
      "learning_rate": 0.00017729550891615957,
      "loss": 5.3007,
      "step": 2970
    },
    {
      "epoch": 0.2971,
      "grad_norm": 9.65993595123291,
      "learning_rate": 0.00017727335734973512,
      "loss": 5.8174,
      "step": 2971
    },
    {
      "epoch": 0.2972,
      "grad_norm": 2.5003113746643066,
      "learning_rate": 0.00017725119636778644,
      "loss": 5.2888,
      "step": 2972
    },
    {
      "epoch": 0.2973,
      "grad_norm": 5.6741156578063965,
      "learning_rate": 0.00017722902597301383,
      "loss": 5.2918,
      "step": 2973
    },
    {
      "epoch": 0.2974,
      "grad_norm": 5.502175331115723,
      "learning_rate": 0.00017720684616811866,
      "loss": 5.2374,
      "step": 2974
    },
    {
      "epoch": 0.2975,
      "grad_norm": 2.508213758468628,
      "learning_rate": 0.0001771846569558035,
      "loss": 5.0489,
      "step": 2975
    },
    {
      "epoch": 0.2976,
      "grad_norm": 3.056805372238159,
      "learning_rate": 0.00017716245833877201,
      "loss": 5.1542,
      "step": 2976
    },
    {
      "epoch": 0.2977,
      "grad_norm": 4.8914408683776855,
      "learning_rate": 0.00017714025031972903,
      "loss": 6.5425,
      "step": 2977
    },
    {
      "epoch": 0.2978,
      "grad_norm": 3.158006429672241,
      "learning_rate": 0.00017711803290138052,
      "loss": 5.0671,
      "step": 2978
    },
    {
      "epoch": 0.2979,
      "grad_norm": 3.180361270904541,
      "learning_rate": 0.00017709580608643363,
      "loss": 5.4413,
      "step": 2979
    },
    {
      "epoch": 0.298,
      "grad_norm": 21.110065460205078,
      "learning_rate": 0.0001770735698775966,
      "loss": 5.7505,
      "step": 2980
    },
    {
      "epoch": 0.2981,
      "grad_norm": 2.6073648929595947,
      "learning_rate": 0.00017705132427757895,
      "loss": 5.4167,
      "step": 2981
    },
    {
      "epoch": 0.2982,
      "grad_norm": 2.6281142234802246,
      "learning_rate": 0.00017702906928909108,
      "loss": 5.2432,
      "step": 2982
    },
    {
      "epoch": 0.2983,
      "grad_norm": 9.108809471130371,
      "learning_rate": 0.0001770068049148448,
      "loss": 5.516,
      "step": 2983
    },
    {
      "epoch": 0.2984,
      "grad_norm": 3.3123810291290283,
      "learning_rate": 0.00017698453115755293,
      "loss": 5.5438,
      "step": 2984
    },
    {
      "epoch": 0.2985,
      "grad_norm": 2.9472744464874268,
      "learning_rate": 0.00017696224801992945,
      "loss": 5.5926,
      "step": 2985
    },
    {
      "epoch": 0.2986,
      "grad_norm": 4.949767589569092,
      "learning_rate": 0.0001769399555046895,
      "loss": 5.1794,
      "step": 2986
    },
    {
      "epoch": 0.2987,
      "grad_norm": 3.155287742614746,
      "learning_rate": 0.00017691765361454938,
      "loss": 5.1974,
      "step": 2987
    },
    {
      "epoch": 0.2988,
      "grad_norm": 2.608281135559082,
      "learning_rate": 0.00017689534235222648,
      "loss": 5.5081,
      "step": 2988
    },
    {
      "epoch": 0.2989,
      "grad_norm": 2.0810630321502686,
      "learning_rate": 0.00017687302172043933,
      "loss": 5.1063,
      "step": 2989
    },
    {
      "epoch": 0.299,
      "grad_norm": 3.457662343978882,
      "learning_rate": 0.00017685069172190766,
      "loss": 5.6771,
      "step": 2990
    },
    {
      "epoch": 0.2991,
      "grad_norm": 6.537090301513672,
      "learning_rate": 0.00017682835235935236,
      "loss": 5.8992,
      "step": 2991
    },
    {
      "epoch": 0.2992,
      "grad_norm": 2.893259286880493,
      "learning_rate": 0.00017680600363549533,
      "loss": 5.4647,
      "step": 2992
    },
    {
      "epoch": 0.2993,
      "grad_norm": 10.797114372253418,
      "learning_rate": 0.00017678364555305978,
      "loss": 5.6084,
      "step": 2993
    },
    {
      "epoch": 0.2994,
      "grad_norm": 2.6748926639556885,
      "learning_rate": 0.00017676127811476987,
      "loss": 5.6644,
      "step": 2994
    },
    {
      "epoch": 0.2995,
      "grad_norm": 4.859403610229492,
      "learning_rate": 0.0001767389013233511,
      "loss": 5.9921,
      "step": 2995
    },
    {
      "epoch": 0.2996,
      "grad_norm": 4.4628214836120605,
      "learning_rate": 0.00017671651518153,
      "loss": 4.9238,
      "step": 2996
    },
    {
      "epoch": 0.2997,
      "grad_norm": 4.397302627563477,
      "learning_rate": 0.00017669411969203417,
      "loss": 5.1216,
      "step": 2997
    },
    {
      "epoch": 0.2998,
      "grad_norm": 6.258809566497803,
      "learning_rate": 0.00017667171485759252,
      "loss": 5.5305,
      "step": 2998
    },
    {
      "epoch": 0.2999,
      "grad_norm": 9.838446617126465,
      "learning_rate": 0.00017664930068093498,
      "loss": 5.2726,
      "step": 2999
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.133929967880249,
      "learning_rate": 0.00017662687716479266,
      "loss": 5.315,
      "step": 3000
    },
    {
      "epoch": 0.3001,
      "grad_norm": 4.117702484130859,
      "learning_rate": 0.0001766044443118978,
      "loss": 6.5707,
      "step": 3001
    },
    {
      "epoch": 0.3002,
      "grad_norm": 4.325991153717041,
      "learning_rate": 0.00017658200212498378,
      "loss": 5.3846,
      "step": 3002
    },
    {
      "epoch": 0.3003,
      "grad_norm": 3.3091461658477783,
      "learning_rate": 0.00017655955060678506,
      "loss": 5.3872,
      "step": 3003
    },
    {
      "epoch": 0.3004,
      "grad_norm": 2.3936235904693604,
      "learning_rate": 0.00017653708976003737,
      "loss": 4.6361,
      "step": 3004
    },
    {
      "epoch": 0.3005,
      "grad_norm": 3.7092695236206055,
      "learning_rate": 0.00017651461958747745,
      "loss": 5.4503,
      "step": 3005
    },
    {
      "epoch": 0.3006,
      "grad_norm": 2.4021217823028564,
      "learning_rate": 0.0001764921400918432,
      "loss": 5.1215,
      "step": 3006
    },
    {
      "epoch": 0.3007,
      "grad_norm": 4.169247627258301,
      "learning_rate": 0.0001764696512758737,
      "loss": 6.283,
      "step": 3007
    },
    {
      "epoch": 0.3008,
      "grad_norm": 3.0752859115600586,
      "learning_rate": 0.00017644715314230918,
      "loss": 5.1507,
      "step": 3008
    },
    {
      "epoch": 0.3009,
      "grad_norm": 7.541386604309082,
      "learning_rate": 0.0001764246456938909,
      "loss": 6.5645,
      "step": 3009
    },
    {
      "epoch": 0.301,
      "grad_norm": 5.582626819610596,
      "learning_rate": 0.00017640212893336142,
      "loss": 5.8837,
      "step": 3010
    },
    {
      "epoch": 0.3011,
      "grad_norm": 2.3699772357940674,
      "learning_rate": 0.00017637960286346425,
      "loss": 5.4257,
      "step": 3011
    },
    {
      "epoch": 0.3012,
      "grad_norm": 2.8395907878875732,
      "learning_rate": 0.00017635706748694413,
      "loss": 5.2647,
      "step": 3012
    },
    {
      "epoch": 0.3013,
      "grad_norm": 2.7124698162078857,
      "learning_rate": 0.000176334522806547,
      "loss": 5.8223,
      "step": 3013
    },
    {
      "epoch": 0.3014,
      "grad_norm": 4.806459903717041,
      "learning_rate": 0.00017631196882501973,
      "loss": 5.2858,
      "step": 3014
    },
    {
      "epoch": 0.3015,
      "grad_norm": 7.1489338874816895,
      "learning_rate": 0.00017628940554511061,
      "loss": 5.827,
      "step": 3015
    },
    {
      "epoch": 0.3016,
      "grad_norm": 7.1557464599609375,
      "learning_rate": 0.00017626683296956882,
      "loss": 6.124,
      "step": 3016
    },
    {
      "epoch": 0.3017,
      "grad_norm": 2.530379056930542,
      "learning_rate": 0.0001762442511011448,
      "loss": 5.606,
      "step": 3017
    },
    {
      "epoch": 0.3018,
      "grad_norm": 1.9679242372512817,
      "learning_rate": 0.00017622165994259,
      "loss": 5.1241,
      "step": 3018
    },
    {
      "epoch": 0.3019,
      "grad_norm": 7.519789218902588,
      "learning_rate": 0.0001761990594966572,
      "loss": 5.7567,
      "step": 3019
    },
    {
      "epoch": 0.302,
      "grad_norm": 3.131887197494507,
      "learning_rate": 0.0001761764497661001,
      "loss": 4.8211,
      "step": 3020
    },
    {
      "epoch": 0.3021,
      "grad_norm": 6.2940850257873535,
      "learning_rate": 0.0001761538307536737,
      "loss": 6.2539,
      "step": 3021
    },
    {
      "epoch": 0.3022,
      "grad_norm": 3.333186149597168,
      "learning_rate": 0.000176131202462134,
      "loss": 5.1959,
      "step": 3022
    },
    {
      "epoch": 0.3023,
      "grad_norm": 2.287292718887329,
      "learning_rate": 0.0001761085648942382,
      "loss": 4.7274,
      "step": 3023
    },
    {
      "epoch": 0.3024,
      "grad_norm": 7.241552352905273,
      "learning_rate": 0.00017608591805274464,
      "loss": 5.5324,
      "step": 3024
    },
    {
      "epoch": 0.3025,
      "grad_norm": 2.5897669792175293,
      "learning_rate": 0.00017606326194041273,
      "loss": 5.1836,
      "step": 3025
    },
    {
      "epoch": 0.3026,
      "grad_norm": 3.6045989990234375,
      "learning_rate": 0.0001760405965600031,
      "loss": 5.0002,
      "step": 3026
    },
    {
      "epoch": 0.3027,
      "grad_norm": 3.64072847366333,
      "learning_rate": 0.00017601792191427741,
      "loss": 4.9699,
      "step": 3027
    },
    {
      "epoch": 0.3028,
      "grad_norm": 2.4311769008636475,
      "learning_rate": 0.0001759952380059986,
      "loss": 5.5888,
      "step": 3028
    },
    {
      "epoch": 0.3029,
      "grad_norm": 3.5634374618530273,
      "learning_rate": 0.00017597254483793048,
      "loss": 5.0126,
      "step": 3029
    },
    {
      "epoch": 0.303,
      "grad_norm": 5.083425045013428,
      "learning_rate": 0.00017594984241283825,
      "loss": 5.6395,
      "step": 3030
    },
    {
      "epoch": 0.3031,
      "grad_norm": 2.7009594440460205,
      "learning_rate": 0.00017592713073348807,
      "loss": 5.2248,
      "step": 3031
    },
    {
      "epoch": 0.3032,
      "grad_norm": 2.9348304271698,
      "learning_rate": 0.00017590440980264738,
      "loss": 5.4918,
      "step": 3032
    },
    {
      "epoch": 0.3033,
      "grad_norm": 6.655057430267334,
      "learning_rate": 0.00017588167962308458,
      "loss": 6.1774,
      "step": 3033
    },
    {
      "epoch": 0.3034,
      "grad_norm": 2.3461921215057373,
      "learning_rate": 0.00017585894019756925,
      "loss": 5.1958,
      "step": 3034
    },
    {
      "epoch": 0.3035,
      "grad_norm": 5.2488555908203125,
      "learning_rate": 0.0001758361915288722,
      "loss": 5.9996,
      "step": 3035
    },
    {
      "epoch": 0.3036,
      "grad_norm": 3.3803205490112305,
      "learning_rate": 0.00017581343361976524,
      "loss": 5.1569,
      "step": 3036
    },
    {
      "epoch": 0.3037,
      "grad_norm": 2.861999750137329,
      "learning_rate": 0.00017579066647302133,
      "loss": 5.5448,
      "step": 3037
    },
    {
      "epoch": 0.3038,
      "grad_norm": 2.9587366580963135,
      "learning_rate": 0.00017576789009141465,
      "loss": 4.9745,
      "step": 3038
    },
    {
      "epoch": 0.3039,
      "grad_norm": 3.4788830280303955,
      "learning_rate": 0.00017574510447772039,
      "loss": 4.9755,
      "step": 3039
    },
    {
      "epoch": 0.304,
      "grad_norm": 2.8470394611358643,
      "learning_rate": 0.00017572230963471488,
      "loss": 5.0721,
      "step": 3040
    },
    {
      "epoch": 0.3041,
      "grad_norm": 2.7615091800689697,
      "learning_rate": 0.00017569950556517566,
      "loss": 5.5607,
      "step": 3041
    },
    {
      "epoch": 0.3042,
      "grad_norm": 8.337615966796875,
      "learning_rate": 0.00017567669227188128,
      "loss": 5.2337,
      "step": 3042
    },
    {
      "epoch": 0.3043,
      "grad_norm": 8.323890686035156,
      "learning_rate": 0.0001756538697576115,
      "loss": 5.7061,
      "step": 3043
    },
    {
      "epoch": 0.3044,
      "grad_norm": 3.0420217514038086,
      "learning_rate": 0.0001756310380251472,
      "loss": 5.8091,
      "step": 3044
    },
    {
      "epoch": 0.3045,
      "grad_norm": 2.3198506832122803,
      "learning_rate": 0.00017560819707727033,
      "loss": 5.65,
      "step": 3045
    },
    {
      "epoch": 0.3046,
      "grad_norm": 2.5132312774658203,
      "learning_rate": 0.00017558534691676397,
      "loss": 5.6746,
      "step": 3046
    },
    {
      "epoch": 0.3047,
      "grad_norm": 2.712845802307129,
      "learning_rate": 0.00017556248754641235,
      "loss": 5.1829,
      "step": 3047
    },
    {
      "epoch": 0.3048,
      "grad_norm": 2.4591305255889893,
      "learning_rate": 0.00017553961896900087,
      "loss": 5.2728,
      "step": 3048
    },
    {
      "epoch": 0.3049,
      "grad_norm": 2.4983370304107666,
      "learning_rate": 0.00017551674118731591,
      "loss": 4.7738,
      "step": 3049
    },
    {
      "epoch": 0.305,
      "grad_norm": 4.969968318939209,
      "learning_rate": 0.00017549385420414514,
      "loss": 5.4612,
      "step": 3050
    },
    {
      "epoch": 0.3051,
      "grad_norm": 2.1692333221435547,
      "learning_rate": 0.00017547095802227723,
      "loss": 5.0766,
      "step": 3051
    },
    {
      "epoch": 0.3052,
      "grad_norm": 2.694648027420044,
      "learning_rate": 0.00017544805264450196,
      "loss": 4.8123,
      "step": 3052
    },
    {
      "epoch": 0.3053,
      "grad_norm": 7.127232551574707,
      "learning_rate": 0.00017542513807361037,
      "loss": 5.3003,
      "step": 3053
    },
    {
      "epoch": 0.3054,
      "grad_norm": 2.5721209049224854,
      "learning_rate": 0.00017540221431239453,
      "loss": 5.4955,
      "step": 3054
    },
    {
      "epoch": 0.3055,
      "grad_norm": 4.742898464202881,
      "learning_rate": 0.00017537928136364755,
      "loss": 4.9652,
      "step": 3055
    },
    {
      "epoch": 0.3056,
      "grad_norm": 3.9665019512176514,
      "learning_rate": 0.0001753563392301638,
      "loss": 5.6238,
      "step": 3056
    },
    {
      "epoch": 0.3057,
      "grad_norm": 2.2188310623168945,
      "learning_rate": 0.0001753333879147387,
      "loss": 5.0246,
      "step": 3057
    },
    {
      "epoch": 0.3058,
      "grad_norm": 7.546825408935547,
      "learning_rate": 0.00017531042742016876,
      "loss": 5.3423,
      "step": 3058
    },
    {
      "epoch": 0.3059,
      "grad_norm": 2.4726755619049072,
      "learning_rate": 0.00017528745774925172,
      "loss": 5.7807,
      "step": 3059
    },
    {
      "epoch": 0.306,
      "grad_norm": 2.5764834880828857,
      "learning_rate": 0.00017526447890478633,
      "loss": 5.5797,
      "step": 3060
    },
    {
      "epoch": 0.3061,
      "grad_norm": 9.400018692016602,
      "learning_rate": 0.00017524149088957245,
      "loss": 5.4573,
      "step": 3061
    },
    {
      "epoch": 0.3062,
      "grad_norm": 3.30637526512146,
      "learning_rate": 0.00017521849370641114,
      "loss": 5.6242,
      "step": 3062
    },
    {
      "epoch": 0.3063,
      "grad_norm": 2.7618515491485596,
      "learning_rate": 0.00017519548735810456,
      "loss": 5.4666,
      "step": 3063
    },
    {
      "epoch": 0.3064,
      "grad_norm": 3.3794243335723877,
      "learning_rate": 0.00017517247184745593,
      "loss": 5.0681,
      "step": 3064
    },
    {
      "epoch": 0.3065,
      "grad_norm": 5.106950759887695,
      "learning_rate": 0.00017514944717726962,
      "loss": 5.072,
      "step": 3065
    },
    {
      "epoch": 0.3066,
      "grad_norm": 3.4529645442962646,
      "learning_rate": 0.00017512641335035113,
      "loss": 5.3247,
      "step": 3066
    },
    {
      "epoch": 0.3067,
      "grad_norm": 3.1515772342681885,
      "learning_rate": 0.00017510337036950703,
      "loss": 5.2103,
      "step": 3067
    },
    {
      "epoch": 0.3068,
      "grad_norm": 3.25189471244812,
      "learning_rate": 0.0001750803182375451,
      "loss": 5.441,
      "step": 3068
    },
    {
      "epoch": 0.3069,
      "grad_norm": 6.001576900482178,
      "learning_rate": 0.00017505725695727412,
      "loss": 5.2051,
      "step": 3069
    },
    {
      "epoch": 0.307,
      "grad_norm": 4.74181604385376,
      "learning_rate": 0.00017503418653150405,
      "loss": 5.3935,
      "step": 3070
    },
    {
      "epoch": 0.3071,
      "grad_norm": 3.8020458221435547,
      "learning_rate": 0.00017501110696304596,
      "loss": 5.7057,
      "step": 3071
    },
    {
      "epoch": 0.3072,
      "grad_norm": 2.9943621158599854,
      "learning_rate": 0.00017498801825471203,
      "loss": 5.3248,
      "step": 3072
    },
    {
      "epoch": 0.3073,
      "grad_norm": 2.2665488719940186,
      "learning_rate": 0.00017496492040931552,
      "loss": 4.9496,
      "step": 3073
    },
    {
      "epoch": 0.3074,
      "grad_norm": 2.070892810821533,
      "learning_rate": 0.00017494181342967083,
      "loss": 5.229,
      "step": 3074
    },
    {
      "epoch": 0.3075,
      "grad_norm": 3.4556691646575928,
      "learning_rate": 0.00017491869731859353,
      "loss": 5.2307,
      "step": 3075
    },
    {
      "epoch": 0.3076,
      "grad_norm": 2.3402087688446045,
      "learning_rate": 0.00017489557207890023,
      "loss": 5.2246,
      "step": 3076
    },
    {
      "epoch": 0.3077,
      "grad_norm": 3.275888442993164,
      "learning_rate": 0.0001748724377134086,
      "loss": 5.7389,
      "step": 3077
    },
    {
      "epoch": 0.3078,
      "grad_norm": 4.774658679962158,
      "learning_rate": 0.0001748492942249376,
      "loss": 5.0302,
      "step": 3078
    },
    {
      "epoch": 0.3079,
      "grad_norm": 3.9042298793792725,
      "learning_rate": 0.00017482614161630714,
      "loss": 5.1409,
      "step": 3079
    },
    {
      "epoch": 0.308,
      "grad_norm": 3.0405309200286865,
      "learning_rate": 0.00017480297989033825,
      "loss": 5.4318,
      "step": 3080
    },
    {
      "epoch": 0.3081,
      "grad_norm": 5.712641716003418,
      "learning_rate": 0.0001747798090498532,
      "loss": 5.4549,
      "step": 3081
    },
    {
      "epoch": 0.3082,
      "grad_norm": 15.735087394714355,
      "learning_rate": 0.00017475662909767522,
      "loss": 6.1529,
      "step": 3082
    },
    {
      "epoch": 0.3083,
      "grad_norm": 5.33223295211792,
      "learning_rate": 0.00017473344003662877,
      "loss": 5.1461,
      "step": 3083
    },
    {
      "epoch": 0.3084,
      "grad_norm": 3.723891019821167,
      "learning_rate": 0.00017471024186953936,
      "loss": 6.3016,
      "step": 3084
    },
    {
      "epoch": 0.3085,
      "grad_norm": 4.441618919372559,
      "learning_rate": 0.00017468703459923356,
      "loss": 5.4658,
      "step": 3085
    },
    {
      "epoch": 0.3086,
      "grad_norm": 4.102380275726318,
      "learning_rate": 0.00017466381822853915,
      "loss": 5.6162,
      "step": 3086
    },
    {
      "epoch": 0.3087,
      "grad_norm": 2.4513392448425293,
      "learning_rate": 0.00017464059276028497,
      "loss": 5.4729,
      "step": 3087
    },
    {
      "epoch": 0.3088,
      "grad_norm": 2.0119082927703857,
      "learning_rate": 0.00017461735819730096,
      "loss": 5.5101,
      "step": 3088
    },
    {
      "epoch": 0.3089,
      "grad_norm": 2.4720962047576904,
      "learning_rate": 0.00017459411454241822,
      "loss": 5.2043,
      "step": 3089
    },
    {
      "epoch": 0.309,
      "grad_norm": 2.5007200241088867,
      "learning_rate": 0.00017457086179846888,
      "loss": 5.0845,
      "step": 3090
    },
    {
      "epoch": 0.3091,
      "grad_norm": 2.6318445205688477,
      "learning_rate": 0.00017454759996828623,
      "loss": 5.556,
      "step": 3091
    },
    {
      "epoch": 0.3092,
      "grad_norm": 4.926275730133057,
      "learning_rate": 0.00017452432905470464,
      "loss": 5.3514,
      "step": 3092
    },
    {
      "epoch": 0.3093,
      "grad_norm": 8.174728393554688,
      "learning_rate": 0.00017450104906055963,
      "loss": 5.4552,
      "step": 3093
    },
    {
      "epoch": 0.3094,
      "grad_norm": 3.2565159797668457,
      "learning_rate": 0.00017447775998868776,
      "loss": 5.0472,
      "step": 3094
    },
    {
      "epoch": 0.3095,
      "grad_norm": 4.403658390045166,
      "learning_rate": 0.00017445446184192676,
      "loss": 5.3175,
      "step": 3095
    },
    {
      "epoch": 0.3096,
      "grad_norm": 5.45024299621582,
      "learning_rate": 0.0001744311546231154,
      "loss": 5.0339,
      "step": 3096
    },
    {
      "epoch": 0.3097,
      "grad_norm": 6.144473552703857,
      "learning_rate": 0.00017440783833509366,
      "loss": 5.1911,
      "step": 3097
    },
    {
      "epoch": 0.3098,
      "grad_norm": 6.157329559326172,
      "learning_rate": 0.00017438451298070252,
      "loss": 5.1041,
      "step": 3098
    },
    {
      "epoch": 0.3099,
      "grad_norm": 5.018340587615967,
      "learning_rate": 0.0001743611785627841,
      "loss": 5.1179,
      "step": 3099
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.332580089569092,
      "learning_rate": 0.00017433783508418162,
      "loss": 5.5016,
      "step": 3100
    },
    {
      "epoch": 0.3101,
      "grad_norm": 4.944673538208008,
      "learning_rate": 0.00017431448254773944,
      "loss": 5.0895,
      "step": 3101
    },
    {
      "epoch": 0.3102,
      "grad_norm": 3.521611213684082,
      "learning_rate": 0.00017429112095630295,
      "loss": 5.6276,
      "step": 3102
    },
    {
      "epoch": 0.3103,
      "grad_norm": 3.561206102371216,
      "learning_rate": 0.00017426775031271875,
      "loss": 5.2586,
      "step": 3103
    },
    {
      "epoch": 0.3104,
      "grad_norm": 4.273228168487549,
      "learning_rate": 0.00017424437061983446,
      "loss": 5.4649,
      "step": 3104
    },
    {
      "epoch": 0.3105,
      "grad_norm": 5.163867950439453,
      "learning_rate": 0.00017422098188049883,
      "loss": 5.617,
      "step": 3105
    },
    {
      "epoch": 0.3106,
      "grad_norm": 2.8935253620147705,
      "learning_rate": 0.00017419758409756164,
      "loss": 5.6352,
      "step": 3106
    },
    {
      "epoch": 0.3107,
      "grad_norm": 2.8743062019348145,
      "learning_rate": 0.00017417417727387394,
      "loss": 5.2588,
      "step": 3107
    },
    {
      "epoch": 0.3108,
      "grad_norm": 4.583456516265869,
      "learning_rate": 0.0001741507614122877,
      "loss": 5.9997,
      "step": 3108
    },
    {
      "epoch": 0.3109,
      "grad_norm": 6.666719913482666,
      "learning_rate": 0.0001741273365156561,
      "loss": 5.252,
      "step": 3109
    },
    {
      "epoch": 0.311,
      "grad_norm": 4.7575788497924805,
      "learning_rate": 0.00017410390258683345,
      "loss": 5.2972,
      "step": 3110
    },
    {
      "epoch": 0.3111,
      "grad_norm": 2.7515268325805664,
      "learning_rate": 0.000174080459628675,
      "loss": 5.6063,
      "step": 3111
    },
    {
      "epoch": 0.3112,
      "grad_norm": 4.338598728179932,
      "learning_rate": 0.00017405700764403726,
      "loss": 5.0147,
      "step": 3112
    },
    {
      "epoch": 0.3113,
      "grad_norm": 5.917651176452637,
      "learning_rate": 0.00017403354663577783,
      "loss": 5.6982,
      "step": 3113
    },
    {
      "epoch": 0.3114,
      "grad_norm": 4.1908111572265625,
      "learning_rate": 0.00017401007660675525,
      "loss": 5.518,
      "step": 3114
    },
    {
      "epoch": 0.3115,
      "grad_norm": 2.381941080093384,
      "learning_rate": 0.00017398659755982936,
      "loss": 5.2085,
      "step": 3115
    },
    {
      "epoch": 0.3116,
      "grad_norm": 2.187342882156372,
      "learning_rate": 0.000173963109497861,
      "loss": 5.1549,
      "step": 3116
    },
    {
      "epoch": 0.3117,
      "grad_norm": 4.3472514152526855,
      "learning_rate": 0.00017393961242371205,
      "loss": 6.409,
      "step": 3117
    },
    {
      "epoch": 0.3118,
      "grad_norm": 6.667789459228516,
      "learning_rate": 0.00017391610634024564,
      "loss": 4.9212,
      "step": 3118
    },
    {
      "epoch": 0.3119,
      "grad_norm": 5.662012577056885,
      "learning_rate": 0.0001738925912503259,
      "loss": 5.1876,
      "step": 3119
    },
    {
      "epoch": 0.312,
      "grad_norm": 2.8152987957000732,
      "learning_rate": 0.000173869067156818,
      "loss": 4.9729,
      "step": 3120
    },
    {
      "epoch": 0.3121,
      "grad_norm": 4.086686134338379,
      "learning_rate": 0.00017384553406258842,
      "loss": 5.0359,
      "step": 3121
    },
    {
      "epoch": 0.3122,
      "grad_norm": 3.179260015487671,
      "learning_rate": 0.0001738219919705044,
      "loss": 5.0295,
      "step": 3122
    },
    {
      "epoch": 0.3123,
      "grad_norm": 3.876429557800293,
      "learning_rate": 0.00017379844088343468,
      "loss": 5.3497,
      "step": 3123
    },
    {
      "epoch": 0.3124,
      "grad_norm": 2.4223217964172363,
      "learning_rate": 0.00017377488080424876,
      "loss": 4.6978,
      "step": 3124
    },
    {
      "epoch": 0.3125,
      "grad_norm": 3.409271717071533,
      "learning_rate": 0.0001737513117358174,
      "loss": 5.3898,
      "step": 3125
    },
    {
      "epoch": 0.3126,
      "grad_norm": 7.10009765625,
      "learning_rate": 0.0001737277336810124,
      "loss": 6.0292,
      "step": 3126
    },
    {
      "epoch": 0.3127,
      "grad_norm": 3.039905548095703,
      "learning_rate": 0.00017370414664270674,
      "loss": 4.92,
      "step": 3127
    },
    {
      "epoch": 0.3128,
      "grad_norm": 2.170276165008545,
      "learning_rate": 0.00017368055062377434,
      "loss": 5.7438,
      "step": 3128
    },
    {
      "epoch": 0.3129,
      "grad_norm": 6.458451747894287,
      "learning_rate": 0.00017365694562709034,
      "loss": 5.5298,
      "step": 3129
    },
    {
      "epoch": 0.313,
      "grad_norm": 4.152144908905029,
      "learning_rate": 0.00017363333165553092,
      "loss": 4.8402,
      "step": 3130
    },
    {
      "epoch": 0.3131,
      "grad_norm": 13.92390251159668,
      "learning_rate": 0.00017360970871197346,
      "loss": 5.1908,
      "step": 3131
    },
    {
      "epoch": 0.3132,
      "grad_norm": 7.013578414916992,
      "learning_rate": 0.0001735860767992962,
      "loss": 5.2886,
      "step": 3132
    },
    {
      "epoch": 0.3133,
      "grad_norm": 3.5786914825439453,
      "learning_rate": 0.00017356243592037875,
      "loss": 5.543,
      "step": 3133
    },
    {
      "epoch": 0.3134,
      "grad_norm": 3.094593048095703,
      "learning_rate": 0.0001735387860781016,
      "loss": 5.0742,
      "step": 3134
    },
    {
      "epoch": 0.3135,
      "grad_norm": 2.4729130268096924,
      "learning_rate": 0.00017351512727534644,
      "loss": 5.319,
      "step": 3135
    },
    {
      "epoch": 0.3136,
      "grad_norm": 2.444427013397217,
      "learning_rate": 0.000173491459514996,
      "loss": 5.3145,
      "step": 3136
    },
    {
      "epoch": 0.3137,
      "grad_norm": 2.0394814014434814,
      "learning_rate": 0.00017346778279993415,
      "loss": 4.9339,
      "step": 3137
    },
    {
      "epoch": 0.3138,
      "grad_norm": 7.02643346786499,
      "learning_rate": 0.0001734440971330458,
      "loss": 5.2406,
      "step": 3138
    },
    {
      "epoch": 0.3139,
      "grad_norm": 3.202885627746582,
      "learning_rate": 0.00017342040251721702,
      "loss": 5.5759,
      "step": 3139
    },
    {
      "epoch": 0.314,
      "grad_norm": 3.2579824924468994,
      "learning_rate": 0.0001733966989553349,
      "loss": 5.7769,
      "step": 3140
    },
    {
      "epoch": 0.3141,
      "grad_norm": 5.299618721008301,
      "learning_rate": 0.00017337298645028764,
      "loss": 5.318,
      "step": 3141
    },
    {
      "epoch": 0.3142,
      "grad_norm": 11.644442558288574,
      "learning_rate": 0.00017334926500496456,
      "loss": 6.1685,
      "step": 3142
    },
    {
      "epoch": 0.3143,
      "grad_norm": 5.2927327156066895,
      "learning_rate": 0.00017332553462225602,
      "loss": 6.0056,
      "step": 3143
    },
    {
      "epoch": 0.3144,
      "grad_norm": 2.0699658393859863,
      "learning_rate": 0.00017330179530505348,
      "loss": 5.0725,
      "step": 3144
    },
    {
      "epoch": 0.3145,
      "grad_norm": 3.0071828365325928,
      "learning_rate": 0.00017327804705624957,
      "loss": 5.3226,
      "step": 3145
    },
    {
      "epoch": 0.3146,
      "grad_norm": 2.7746033668518066,
      "learning_rate": 0.0001732542898787379,
      "loss": 5.0632,
      "step": 3146
    },
    {
      "epoch": 0.3147,
      "grad_norm": 2.8752238750457764,
      "learning_rate": 0.00017323052377541317,
      "loss": 5.4125,
      "step": 3147
    },
    {
      "epoch": 0.3148,
      "grad_norm": 5.67249059677124,
      "learning_rate": 0.0001732067487491713,
      "loss": 4.8403,
      "step": 3148
    },
    {
      "epoch": 0.3149,
      "grad_norm": 12.484209060668945,
      "learning_rate": 0.0001731829648029091,
      "loss": 5.0263,
      "step": 3149
    },
    {
      "epoch": 0.315,
      "grad_norm": 2.8567798137664795,
      "learning_rate": 0.0001731591719395247,
      "loss": 5.4522,
      "step": 3150
    },
    {
      "epoch": 0.3151,
      "grad_norm": 18.37646484375,
      "learning_rate": 0.00017313537016191706,
      "loss": 5.5891,
      "step": 3151
    },
    {
      "epoch": 0.3152,
      "grad_norm": 3.26320219039917,
      "learning_rate": 0.00017311155947298643,
      "loss": 5.913,
      "step": 3152
    },
    {
      "epoch": 0.3153,
      "grad_norm": 3.047415018081665,
      "learning_rate": 0.00017308773987563406,
      "loss": 5.2281,
      "step": 3153
    },
    {
      "epoch": 0.3154,
      "grad_norm": 3.5584611892700195,
      "learning_rate": 0.00017306391137276224,
      "loss": 4.7911,
      "step": 3154
    },
    {
      "epoch": 0.3155,
      "grad_norm": 2.6473376750946045,
      "learning_rate": 0.00017304007396727448,
      "loss": 4.9523,
      "step": 3155
    },
    {
      "epoch": 0.3156,
      "grad_norm": 9.688016891479492,
      "learning_rate": 0.00017301622766207527,
      "loss": 6.2694,
      "step": 3156
    },
    {
      "epoch": 0.3157,
      "grad_norm": 2.427833080291748,
      "learning_rate": 0.00017299237246007015,
      "loss": 5.2535,
      "step": 3157
    },
    {
      "epoch": 0.3158,
      "grad_norm": 3.5214149951934814,
      "learning_rate": 0.00017296850836416588,
      "loss": 4.929,
      "step": 3158
    },
    {
      "epoch": 0.3159,
      "grad_norm": 3.1279993057250977,
      "learning_rate": 0.00017294463537727024,
      "loss": 5.3189,
      "step": 3159
    },
    {
      "epoch": 0.316,
      "grad_norm": 2.191178798675537,
      "learning_rate": 0.000172920753502292,
      "loss": 4.9586,
      "step": 3160
    },
    {
      "epoch": 0.3161,
      "grad_norm": 2.6992945671081543,
      "learning_rate": 0.00017289686274214118,
      "loss": 4.815,
      "step": 3161
    },
    {
      "epoch": 0.3162,
      "grad_norm": 2.9771728515625,
      "learning_rate": 0.0001728729630997287,
      "loss": 5.0677,
      "step": 3162
    },
    {
      "epoch": 0.3163,
      "grad_norm": 2.292558193206787,
      "learning_rate": 0.00017284905457796675,
      "loss": 5.1618,
      "step": 3163
    },
    {
      "epoch": 0.3164,
      "grad_norm": 4.6396636962890625,
      "learning_rate": 0.00017282513717976848,
      "loss": 5.3715,
      "step": 3164
    },
    {
      "epoch": 0.3165,
      "grad_norm": 3.9005634784698486,
      "learning_rate": 0.00017280121090804812,
      "loss": 4.8579,
      "step": 3165
    },
    {
      "epoch": 0.3166,
      "grad_norm": 2.7158100605010986,
      "learning_rate": 0.00017277727576572107,
      "loss": 5.6585,
      "step": 3166
    },
    {
      "epoch": 0.3167,
      "grad_norm": 4.419057369232178,
      "learning_rate": 0.00017275333175570368,
      "loss": 6.1224,
      "step": 3167
    },
    {
      "epoch": 0.3168,
      "grad_norm": 11.620479583740234,
      "learning_rate": 0.00017272937888091353,
      "loss": 5.6264,
      "step": 3168
    },
    {
      "epoch": 0.3169,
      "grad_norm": 2.277998924255371,
      "learning_rate": 0.0001727054171442692,
      "loss": 4.9881,
      "step": 3169
    },
    {
      "epoch": 0.317,
      "grad_norm": 13.574300765991211,
      "learning_rate": 0.0001726814465486903,
      "loss": 5.0141,
      "step": 3170
    },
    {
      "epoch": 0.3171,
      "grad_norm": 5.0140156745910645,
      "learning_rate": 0.0001726574670970976,
      "loss": 4.9119,
      "step": 3171
    },
    {
      "epoch": 0.3172,
      "grad_norm": 3.6077866554260254,
      "learning_rate": 0.00017263347879241291,
      "loss": 5.2038,
      "step": 3172
    },
    {
      "epoch": 0.3173,
      "grad_norm": 8.11462116241455,
      "learning_rate": 0.00017260948163755918,
      "loss": 5.7954,
      "step": 3173
    },
    {
      "epoch": 0.3174,
      "grad_norm": 2.65044903755188,
      "learning_rate": 0.00017258547563546038,
      "loss": 5.1981,
      "step": 3174
    },
    {
      "epoch": 0.3175,
      "grad_norm": 5.488699436187744,
      "learning_rate": 0.00017256146078904153,
      "loss": 5.0973,
      "step": 3175
    },
    {
      "epoch": 0.3176,
      "grad_norm": 3.782500982284546,
      "learning_rate": 0.00017253743710122875,
      "loss": 5.5941,
      "step": 3176
    },
    {
      "epoch": 0.3177,
      "grad_norm": 3.2561116218566895,
      "learning_rate": 0.00017251340457494934,
      "loss": 5.4014,
      "step": 3177
    },
    {
      "epoch": 0.3178,
      "grad_norm": 2.1682517528533936,
      "learning_rate": 0.0001724893632131315,
      "loss": 5.4376,
      "step": 3178
    },
    {
      "epoch": 0.3179,
      "grad_norm": 2.556673049926758,
      "learning_rate": 0.0001724653130187047,
      "loss": 5.2623,
      "step": 3179
    },
    {
      "epoch": 0.318,
      "grad_norm": 3.6540346145629883,
      "learning_rate": 0.00017244125399459926,
      "loss": 5.3661,
      "step": 3180
    },
    {
      "epoch": 0.3181,
      "grad_norm": 4.453672409057617,
      "learning_rate": 0.00017241718614374678,
      "loss": 5.3901,
      "step": 3181
    },
    {
      "epoch": 0.3182,
      "grad_norm": 2.8771891593933105,
      "learning_rate": 0.0001723931094690798,
      "loss": 5.2528,
      "step": 3182
    },
    {
      "epoch": 0.3183,
      "grad_norm": 2.754912853240967,
      "learning_rate": 0.00017236902397353205,
      "loss": 5.1247,
      "step": 3183
    },
    {
      "epoch": 0.3184,
      "grad_norm": 20.81771469116211,
      "learning_rate": 0.00017234492966003824,
      "loss": 7.089,
      "step": 3184
    },
    {
      "epoch": 0.3185,
      "grad_norm": 4.229422569274902,
      "learning_rate": 0.00017232082653153422,
      "loss": 5.4616,
      "step": 3185
    },
    {
      "epoch": 0.3186,
      "grad_norm": 4.3446526527404785,
      "learning_rate": 0.00017229671459095683,
      "loss": 5.1295,
      "step": 3186
    },
    {
      "epoch": 0.3187,
      "grad_norm": 3.947965621948242,
      "learning_rate": 0.0001722725938412441,
      "loss": 5.1686,
      "step": 3187
    },
    {
      "epoch": 0.3188,
      "grad_norm": 2.933114767074585,
      "learning_rate": 0.00017224846428533499,
      "loss": 5.2229,
      "step": 3188
    },
    {
      "epoch": 0.3189,
      "grad_norm": 3.668651819229126,
      "learning_rate": 0.0001722243259261697,
      "loss": 5.6342,
      "step": 3189
    },
    {
      "epoch": 0.319,
      "grad_norm": 4.244904041290283,
      "learning_rate": 0.00017220017876668934,
      "loss": 5.6038,
      "step": 3190
    },
    {
      "epoch": 0.3191,
      "grad_norm": 11.214171409606934,
      "learning_rate": 0.00017217602280983623,
      "loss": 5.6802,
      "step": 3191
    },
    {
      "epoch": 0.3192,
      "grad_norm": 3.180821180343628,
      "learning_rate": 0.00017215185805855368,
      "loss": 5.8681,
      "step": 3192
    },
    {
      "epoch": 0.3193,
      "grad_norm": 4.6527018547058105,
      "learning_rate": 0.00017212768451578607,
      "loss": 6.2531,
      "step": 3193
    },
    {
      "epoch": 0.3194,
      "grad_norm": 6.312183856964111,
      "learning_rate": 0.00017210350218447887,
      "loss": 6.6687,
      "step": 3194
    },
    {
      "epoch": 0.3195,
      "grad_norm": 12.429105758666992,
      "learning_rate": 0.00017207931106757868,
      "loss": 4.9958,
      "step": 3195
    },
    {
      "epoch": 0.3196,
      "grad_norm": 11.17943000793457,
      "learning_rate": 0.00017205511116803306,
      "loss": 5.0156,
      "step": 3196
    },
    {
      "epoch": 0.3197,
      "grad_norm": 3.6429126262664795,
      "learning_rate": 0.0001720309024887907,
      "loss": 4.9595,
      "step": 3197
    },
    {
      "epoch": 0.3198,
      "grad_norm": 3.131535768508911,
      "learning_rate": 0.00017200668503280136,
      "loss": 5.2162,
      "step": 3198
    },
    {
      "epoch": 0.3199,
      "grad_norm": 3.084974765777588,
      "learning_rate": 0.0001719824588030159,
      "loss": 5.3949,
      "step": 3199
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.7358598709106445,
      "learning_rate": 0.00017195822380238615,
      "loss": 5.2619,
      "step": 3200
    },
    {
      "epoch": 0.3201,
      "grad_norm": 6.330494403839111,
      "learning_rate": 0.0001719339800338651,
      "loss": 5.5133,
      "step": 3201
    },
    {
      "epoch": 0.3202,
      "grad_norm": 13.365460395812988,
      "learning_rate": 0.00017190972750040682,
      "loss": 6.2221,
      "step": 3202
    },
    {
      "epoch": 0.3203,
      "grad_norm": 2.8058087825775146,
      "learning_rate": 0.00017188546620496635,
      "loss": 5.4304,
      "step": 3203
    },
    {
      "epoch": 0.3204,
      "grad_norm": 1.962013840675354,
      "learning_rate": 0.00017186119615049988,
      "loss": 5.8251,
      "step": 3204
    },
    {
      "epoch": 0.3205,
      "grad_norm": 2.488086462020874,
      "learning_rate": 0.00017183691733996462,
      "loss": 5.3433,
      "step": 3205
    },
    {
      "epoch": 0.3206,
      "grad_norm": 7.533576965332031,
      "learning_rate": 0.00017181262977631888,
      "loss": 5.2169,
      "step": 3206
    },
    {
      "epoch": 0.3207,
      "grad_norm": 3.128328800201416,
      "learning_rate": 0.00017178833346252206,
      "loss": 4.7747,
      "step": 3207
    },
    {
      "epoch": 0.3208,
      "grad_norm": 3.137483835220337,
      "learning_rate": 0.00017176402840153455,
      "loss": 4.8566,
      "step": 3208
    },
    {
      "epoch": 0.3209,
      "grad_norm": 2.972959518432617,
      "learning_rate": 0.00017173971459631787,
      "loss": 5.3663,
      "step": 3209
    },
    {
      "epoch": 0.321,
      "grad_norm": 4.5377583503723145,
      "learning_rate": 0.0001717153920498346,
      "loss": 5.391,
      "step": 3210
    },
    {
      "epoch": 0.3211,
      "grad_norm": 2.5597643852233887,
      "learning_rate": 0.0001716910607650483,
      "loss": 5.3341,
      "step": 3211
    },
    {
      "epoch": 0.3212,
      "grad_norm": 17.73623275756836,
      "learning_rate": 0.0001716667207449237,
      "loss": 6.4085,
      "step": 3212
    },
    {
      "epoch": 0.3213,
      "grad_norm": 2.495384693145752,
      "learning_rate": 0.0001716423719924266,
      "loss": 5.0087,
      "step": 3213
    },
    {
      "epoch": 0.3214,
      "grad_norm": 2.4588098526000977,
      "learning_rate": 0.00017161801451052376,
      "loss": 5.0943,
      "step": 3214
    },
    {
      "epoch": 0.3215,
      "grad_norm": 2.464412212371826,
      "learning_rate": 0.00017159364830218312,
      "loss": 5.5068,
      "step": 3215
    },
    {
      "epoch": 0.3216,
      "grad_norm": 6.472148895263672,
      "learning_rate": 0.0001715692733703736,
      "loss": 6.555,
      "step": 3216
    },
    {
      "epoch": 0.3217,
      "grad_norm": 4.277796745300293,
      "learning_rate": 0.00017154488971806518,
      "loss": 5.2108,
      "step": 3217
    },
    {
      "epoch": 0.3218,
      "grad_norm": 4.576182842254639,
      "learning_rate": 0.00017152049734822902,
      "loss": 5.8168,
      "step": 3218
    },
    {
      "epoch": 0.3219,
      "grad_norm": 2.959415912628174,
      "learning_rate": 0.00017149609626383717,
      "loss": 5.2909,
      "step": 3219
    },
    {
      "epoch": 0.322,
      "grad_norm": 2.584214448928833,
      "learning_rate": 0.00017147168646786286,
      "loss": 4.8515,
      "step": 3220
    },
    {
      "epoch": 0.3221,
      "grad_norm": 3.3081135749816895,
      "learning_rate": 0.00017144726796328034,
      "loss": 5.3113,
      "step": 3221
    },
    {
      "epoch": 0.3222,
      "grad_norm": 2.3322067260742188,
      "learning_rate": 0.00017142284075306497,
      "loss": 5.0106,
      "step": 3222
    },
    {
      "epoch": 0.3223,
      "grad_norm": 5.651757717132568,
      "learning_rate": 0.0001713984048401931,
      "loss": 4.9274,
      "step": 3223
    },
    {
      "epoch": 0.3224,
      "grad_norm": 3.276585102081299,
      "learning_rate": 0.00017137396022764214,
      "loss": 5.9264,
      "step": 3224
    },
    {
      "epoch": 0.3225,
      "grad_norm": 3.9236321449279785,
      "learning_rate": 0.00017134950691839065,
      "loss": 4.8477,
      "step": 3225
    },
    {
      "epoch": 0.3226,
      "grad_norm": 21.09899139404297,
      "learning_rate": 0.00017132504491541818,
      "loss": 6.0797,
      "step": 3226
    },
    {
      "epoch": 0.3227,
      "grad_norm": 2.937116861343384,
      "learning_rate": 0.0001713005742217053,
      "loss": 5.1404,
      "step": 3227
    },
    {
      "epoch": 0.3228,
      "grad_norm": 31.75925636291504,
      "learning_rate": 0.00017127609484023377,
      "loss": 7.5625,
      "step": 3228
    },
    {
      "epoch": 0.3229,
      "grad_norm": 7.382253646850586,
      "learning_rate": 0.00017125160677398626,
      "loss": 5.2405,
      "step": 3229
    },
    {
      "epoch": 0.323,
      "grad_norm": 5.61705207824707,
      "learning_rate": 0.0001712271100259466,
      "loss": 5.2086,
      "step": 3230
    },
    {
      "epoch": 0.3231,
      "grad_norm": 16.06414031982422,
      "learning_rate": 0.00017120260459909967,
      "loss": 5.5527,
      "step": 3231
    },
    {
      "epoch": 0.3232,
      "grad_norm": 5.963083267211914,
      "learning_rate": 0.0001711780904964313,
      "loss": 5.3886,
      "step": 3232
    },
    {
      "epoch": 0.3233,
      "grad_norm": 2.5832347869873047,
      "learning_rate": 0.00017115356772092857,
      "loss": 5.5218,
      "step": 3233
    },
    {
      "epoch": 0.3234,
      "grad_norm": 4.6452202796936035,
      "learning_rate": 0.0001711290362755794,
      "loss": 5.4137,
      "step": 3234
    },
    {
      "epoch": 0.3235,
      "grad_norm": 3.3304858207702637,
      "learning_rate": 0.00017110449616337289,
      "loss": 4.8547,
      "step": 3235
    },
    {
      "epoch": 0.3236,
      "grad_norm": 2.333235263824463,
      "learning_rate": 0.00017107994738729926,
      "loss": 5.2382,
      "step": 3236
    },
    {
      "epoch": 0.3237,
      "grad_norm": 3.6059842109680176,
      "learning_rate": 0.00017105538995034963,
      "loss": 5.4735,
      "step": 3237
    },
    {
      "epoch": 0.3238,
      "grad_norm": 3.1608409881591797,
      "learning_rate": 0.00017103082385551627,
      "loss": 5.0713,
      "step": 3238
    },
    {
      "epoch": 0.3239,
      "grad_norm": 3.285428762435913,
      "learning_rate": 0.0001710062491057925,
      "loss": 5.8493,
      "step": 3239
    },
    {
      "epoch": 0.324,
      "grad_norm": 3.3664419651031494,
      "learning_rate": 0.00017098166570417262,
      "loss": 5.4967,
      "step": 3240
    },
    {
      "epoch": 0.3241,
      "grad_norm": 3.6035053730010986,
      "learning_rate": 0.0001709570736536521,
      "loss": 5.5656,
      "step": 3241
    },
    {
      "epoch": 0.3242,
      "grad_norm": 4.100790977478027,
      "learning_rate": 0.0001709324729572274,
      "loss": 5.4822,
      "step": 3242
    },
    {
      "epoch": 0.3243,
      "grad_norm": 2.2193121910095215,
      "learning_rate": 0.000170907863617896,
      "loss": 5.0159,
      "step": 3243
    },
    {
      "epoch": 0.3244,
      "grad_norm": 2.394641876220703,
      "learning_rate": 0.00017088324563865656,
      "loss": 5.1293,
      "step": 3244
    },
    {
      "epoch": 0.3245,
      "grad_norm": 2.74396014213562,
      "learning_rate": 0.00017085861902250865,
      "loss": 4.8483,
      "step": 3245
    },
    {
      "epoch": 0.3246,
      "grad_norm": 7.358772277832031,
      "learning_rate": 0.0001708339837724529,
      "loss": 4.7376,
      "step": 3246
    },
    {
      "epoch": 0.3247,
      "grad_norm": 2.4594273567199707,
      "learning_rate": 0.0001708093398914911,
      "loss": 4.7458,
      "step": 3247
    },
    {
      "epoch": 0.3248,
      "grad_norm": 6.333447456359863,
      "learning_rate": 0.00017078468738262602,
      "loss": 5.0713,
      "step": 3248
    },
    {
      "epoch": 0.3249,
      "grad_norm": 5.419327259063721,
      "learning_rate": 0.00017076002624886154,
      "loss": 5.2146,
      "step": 3249
    },
    {
      "epoch": 0.325,
      "grad_norm": 2.662167549133301,
      "learning_rate": 0.00017073535649320248,
      "loss": 5.7891,
      "step": 3250
    },
    {
      "epoch": 0.3251,
      "grad_norm": 4.025168418884277,
      "learning_rate": 0.00017071067811865476,
      "loss": 5.7465,
      "step": 3251
    },
    {
      "epoch": 0.3252,
      "grad_norm": 3.9197885990142822,
      "learning_rate": 0.00017068599112822543,
      "loss": 5.469,
      "step": 3252
    },
    {
      "epoch": 0.3253,
      "grad_norm": 3.0949316024780273,
      "learning_rate": 0.0001706612955249225,
      "loss": 5.339,
      "step": 3253
    },
    {
      "epoch": 0.3254,
      "grad_norm": 2.764312267303467,
      "learning_rate": 0.000170636591311755,
      "loss": 4.7622,
      "step": 3254
    },
    {
      "epoch": 0.3255,
      "grad_norm": 2.0994575023651123,
      "learning_rate": 0.00017061187849173317,
      "loss": 5.2302,
      "step": 3255
    },
    {
      "epoch": 0.3256,
      "grad_norm": 2.485811710357666,
      "learning_rate": 0.0001705871570678681,
      "loss": 4.956,
      "step": 3256
    },
    {
      "epoch": 0.3257,
      "grad_norm": 9.151593208312988,
      "learning_rate": 0.0001705624270431721,
      "loss": 5.0299,
      "step": 3257
    },
    {
      "epoch": 0.3258,
      "grad_norm": 2.9094526767730713,
      "learning_rate": 0.00017053768842065833,
      "loss": 5.4288,
      "step": 3258
    },
    {
      "epoch": 0.3259,
      "grad_norm": 4.629036903381348,
      "learning_rate": 0.00017051294120334125,
      "loss": 6.042,
      "step": 3259
    },
    {
      "epoch": 0.326,
      "grad_norm": 2.4536209106445312,
      "learning_rate": 0.00017048818539423615,
      "loss": 5.168,
      "step": 3260
    },
    {
      "epoch": 0.3261,
      "grad_norm": 2.5623602867126465,
      "learning_rate": 0.00017046342099635948,
      "loss": 5.2582,
      "step": 3261
    },
    {
      "epoch": 0.3262,
      "grad_norm": 2.0933825969696045,
      "learning_rate": 0.00017043864801272868,
      "loss": 4.9821,
      "step": 3262
    },
    {
      "epoch": 0.3263,
      "grad_norm": 2.0703699588775635,
      "learning_rate": 0.0001704138664463623,
      "loss": 5.1706,
      "step": 3263
    },
    {
      "epoch": 0.3264,
      "grad_norm": 4.974003314971924,
      "learning_rate": 0.00017038907630027988,
      "loss": 5.654,
      "step": 3264
    },
    {
      "epoch": 0.3265,
      "grad_norm": 4.627954959869385,
      "learning_rate": 0.00017036427757750205,
      "loss": 5.3949,
      "step": 3265
    },
    {
      "epoch": 0.3266,
      "grad_norm": 3.823997974395752,
      "learning_rate": 0.00017033947028105039,
      "loss": 4.9121,
      "step": 3266
    },
    {
      "epoch": 0.3267,
      "grad_norm": 6.010074615478516,
      "learning_rate": 0.00017031465441394767,
      "loss": 5.6593,
      "step": 3267
    },
    {
      "epoch": 0.3268,
      "grad_norm": 10.757473945617676,
      "learning_rate": 0.00017028982997921758,
      "loss": 5.7668,
      "step": 3268
    },
    {
      "epoch": 0.3269,
      "grad_norm": 6.535094738006592,
      "learning_rate": 0.00017026499697988493,
      "loss": 7.0302,
      "step": 3269
    },
    {
      "epoch": 0.327,
      "grad_norm": 6.032433032989502,
      "learning_rate": 0.0001702401554189755,
      "loss": 5.328,
      "step": 3270
    },
    {
      "epoch": 0.3271,
      "grad_norm": 2.756190299987793,
      "learning_rate": 0.00017021530529951625,
      "loss": 5.5939,
      "step": 3271
    },
    {
      "epoch": 0.3272,
      "grad_norm": 14.89538288116455,
      "learning_rate": 0.000170190446624535,
      "loss": 6.4347,
      "step": 3272
    },
    {
      "epoch": 0.3273,
      "grad_norm": 16.796964645385742,
      "learning_rate": 0.00017016557939706075,
      "loss": 5.5888,
      "step": 3273
    },
    {
      "epoch": 0.3274,
      "grad_norm": 6.256404399871826,
      "learning_rate": 0.00017014070362012348,
      "loss": 5.3564,
      "step": 3274
    },
    {
      "epoch": 0.3275,
      "grad_norm": 2.5318973064422607,
      "learning_rate": 0.00017011581929675425,
      "loss": 5.2897,
      "step": 3275
    },
    {
      "epoch": 0.3276,
      "grad_norm": 2.5865423679351807,
      "learning_rate": 0.0001700909264299851,
      "loss": 5.3977,
      "step": 3276
    },
    {
      "epoch": 0.3277,
      "grad_norm": 8.715333938598633,
      "learning_rate": 0.00017006602502284918,
      "loss": 4.8273,
      "step": 3277
    },
    {
      "epoch": 0.3278,
      "grad_norm": 2.5304107666015625,
      "learning_rate": 0.00017004111507838064,
      "loss": 4.7178,
      "step": 3278
    },
    {
      "epoch": 0.3279,
      "grad_norm": 2.930807113647461,
      "learning_rate": 0.00017001619659961467,
      "loss": 5.8211,
      "step": 3279
    },
    {
      "epoch": 0.328,
      "grad_norm": 3.433414936065674,
      "learning_rate": 0.00016999126958958756,
      "loss": 5.2214,
      "step": 3280
    },
    {
      "epoch": 0.3281,
      "grad_norm": 10.320745468139648,
      "learning_rate": 0.00016996633405133655,
      "loss": 5.4616,
      "step": 3281
    },
    {
      "epoch": 0.3282,
      "grad_norm": 6.441842079162598,
      "learning_rate": 0.00016994138998789997,
      "loss": 5.5391,
      "step": 3282
    },
    {
      "epoch": 0.3283,
      "grad_norm": 2.7204599380493164,
      "learning_rate": 0.00016991643740231716,
      "loss": 5.2476,
      "step": 3283
    },
    {
      "epoch": 0.3284,
      "grad_norm": 2.445939064025879,
      "learning_rate": 0.0001698914762976285,
      "loss": 5.1516,
      "step": 3284
    },
    {
      "epoch": 0.3285,
      "grad_norm": 3.206277370452881,
      "learning_rate": 0.00016986650667687552,
      "loss": 5.3626,
      "step": 3285
    },
    {
      "epoch": 0.3286,
      "grad_norm": 8.057491302490234,
      "learning_rate": 0.0001698415285431006,
      "loss": 5.4415,
      "step": 3286
    },
    {
      "epoch": 0.3287,
      "grad_norm": 2.9746081829071045,
      "learning_rate": 0.00016981654189934727,
      "loss": 5.0167,
      "step": 3287
    },
    {
      "epoch": 0.3288,
      "grad_norm": 3.536653757095337,
      "learning_rate": 0.0001697915467486601,
      "loss": 6.0473,
      "step": 3288
    },
    {
      "epoch": 0.3289,
      "grad_norm": 3.9680540561676025,
      "learning_rate": 0.00016976654309408464,
      "loss": 4.8314,
      "step": 3289
    },
    {
      "epoch": 0.329,
      "grad_norm": 7.262027263641357,
      "learning_rate": 0.00016974153093866757,
      "loss": 5.7539,
      "step": 3290
    },
    {
      "epoch": 0.3291,
      "grad_norm": 3.368154287338257,
      "learning_rate": 0.00016971651028545648,
      "loss": 6.092,
      "step": 3291
    },
    {
      "epoch": 0.3292,
      "grad_norm": 7.323160648345947,
      "learning_rate": 0.00016969148113750007,
      "loss": 6.257,
      "step": 3292
    },
    {
      "epoch": 0.3293,
      "grad_norm": 2.5359251499176025,
      "learning_rate": 0.00016966644349784808,
      "loss": 4.4621,
      "step": 3293
    },
    {
      "epoch": 0.3294,
      "grad_norm": 2.776179075241089,
      "learning_rate": 0.0001696413973695513,
      "loss": 5.7666,
      "step": 3294
    },
    {
      "epoch": 0.3295,
      "grad_norm": 3.1852166652679443,
      "learning_rate": 0.00016961634275566146,
      "loss": 5.445,
      "step": 3295
    },
    {
      "epoch": 0.3296,
      "grad_norm": 4.2911248207092285,
      "learning_rate": 0.00016959127965923142,
      "loss": 5.2565,
      "step": 3296
    },
    {
      "epoch": 0.3297,
      "grad_norm": 6.543774127960205,
      "learning_rate": 0.0001695662080833151,
      "loss": 4.8363,
      "step": 3297
    },
    {
      "epoch": 0.3298,
      "grad_norm": 5.446486473083496,
      "learning_rate": 0.00016954112803096728,
      "loss": 4.7102,
      "step": 3298
    },
    {
      "epoch": 0.3299,
      "grad_norm": 1.802133560180664,
      "learning_rate": 0.000169516039505244,
      "loss": 4.9443,
      "step": 3299
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.247014045715332,
      "learning_rate": 0.00016949094250920217,
      "loss": 5.0057,
      "step": 3300
    },
    {
      "epoch": 0.3301,
      "grad_norm": 3.780998945236206,
      "learning_rate": 0.00016946583704589973,
      "loss": 5.2147,
      "step": 3301
    },
    {
      "epoch": 0.3302,
      "grad_norm": 2.5058279037475586,
      "learning_rate": 0.00016944072311839581,
      "loss": 5.025,
      "step": 3302
    },
    {
      "epoch": 0.3303,
      "grad_norm": 2.2334675788879395,
      "learning_rate": 0.0001694156007297504,
      "loss": 5.6974,
      "step": 3303
    },
    {
      "epoch": 0.3304,
      "grad_norm": 2.7022948265075684,
      "learning_rate": 0.00016939046988302458,
      "loss": 5.2496,
      "step": 3304
    },
    {
      "epoch": 0.3305,
      "grad_norm": 2.9496617317199707,
      "learning_rate": 0.0001693653305812805,
      "loss": 5.6618,
      "step": 3305
    },
    {
      "epoch": 0.3306,
      "grad_norm": 5.949965953826904,
      "learning_rate": 0.0001693401828275813,
      "loss": 5.081,
      "step": 3306
    },
    {
      "epoch": 0.3307,
      "grad_norm": 4.674259185791016,
      "learning_rate": 0.00016931502662499118,
      "loss": 5.1921,
      "step": 3307
    },
    {
      "epoch": 0.3308,
      "grad_norm": 2.791269063949585,
      "learning_rate": 0.00016928986197657525,
      "loss": 5.1537,
      "step": 3308
    },
    {
      "epoch": 0.3309,
      "grad_norm": 8.834040641784668,
      "learning_rate": 0.0001692646888853999,
      "loss": 5.0507,
      "step": 3309
    },
    {
      "epoch": 0.331,
      "grad_norm": 2.932232618331909,
      "learning_rate": 0.00016923950735453226,
      "loss": 5.4416,
      "step": 3310
    },
    {
      "epoch": 0.3311,
      "grad_norm": 2.1924796104431152,
      "learning_rate": 0.0001692143173870407,
      "loss": 5.5423,
      "step": 3311
    },
    {
      "epoch": 0.3312,
      "grad_norm": 3.5595571994781494,
      "learning_rate": 0.0001691891189859945,
      "loss": 5.1278,
      "step": 3312
    },
    {
      "epoch": 0.3313,
      "grad_norm": 3.016667127609253,
      "learning_rate": 0.000169163912154464,
      "loss": 4.8596,
      "step": 3313
    },
    {
      "epoch": 0.3314,
      "grad_norm": 2.427931070327759,
      "learning_rate": 0.00016913869689552064,
      "loss": 5.0829,
      "step": 3314
    },
    {
      "epoch": 0.3315,
      "grad_norm": 2.064145088195801,
      "learning_rate": 0.0001691134732122368,
      "loss": 5.2919,
      "step": 3315
    },
    {
      "epoch": 0.3316,
      "grad_norm": 4.277704238891602,
      "learning_rate": 0.00016908824110768584,
      "loss": 5.5451,
      "step": 3316
    },
    {
      "epoch": 0.3317,
      "grad_norm": 2.1330506801605225,
      "learning_rate": 0.00016906300058494228,
      "loss": 4.9209,
      "step": 3317
    },
    {
      "epoch": 0.3318,
      "grad_norm": 6.745793342590332,
      "learning_rate": 0.00016903775164708163,
      "loss": 5.1302,
      "step": 3318
    },
    {
      "epoch": 0.3319,
      "grad_norm": 2.7340176105499268,
      "learning_rate": 0.00016901249429718032,
      "loss": 5.1853,
      "step": 3319
    },
    {
      "epoch": 0.332,
      "grad_norm": 3.516946315765381,
      "learning_rate": 0.00016898722853831593,
      "loss": 5.2429,
      "step": 3320
    },
    {
      "epoch": 0.3321,
      "grad_norm": 3.5792481899261475,
      "learning_rate": 0.000168961954373567,
      "loss": 5.0714,
      "step": 3321
    },
    {
      "epoch": 0.3322,
      "grad_norm": 6.264875411987305,
      "learning_rate": 0.00016893667180601312,
      "loss": 6.3289,
      "step": 3322
    },
    {
      "epoch": 0.3323,
      "grad_norm": 3.366135835647583,
      "learning_rate": 0.00016891138083873487,
      "loss": 5.099,
      "step": 3323
    },
    {
      "epoch": 0.3324,
      "grad_norm": 1.8679414987564087,
      "learning_rate": 0.00016888608147481388,
      "loss": 5.1667,
      "step": 3324
    },
    {
      "epoch": 0.3325,
      "grad_norm": 3.9369096755981445,
      "learning_rate": 0.00016886077371733283,
      "loss": 5.0261,
      "step": 3325
    },
    {
      "epoch": 0.3326,
      "grad_norm": 4.1045756340026855,
      "learning_rate": 0.0001688354575693754,
      "loss": 5.4447,
      "step": 3326
    },
    {
      "epoch": 0.3327,
      "grad_norm": 3.4205939769744873,
      "learning_rate": 0.0001688101330340263,
      "loss": 5.5845,
      "step": 3327
    },
    {
      "epoch": 0.3328,
      "grad_norm": 10.047897338867188,
      "learning_rate": 0.0001687848001143711,
      "loss": 6.0321,
      "step": 3328
    },
    {
      "epoch": 0.3329,
      "grad_norm": 2.1112802028656006,
      "learning_rate": 0.00016875945881349676,
      "loss": 5.4722,
      "step": 3329
    },
    {
      "epoch": 0.333,
      "grad_norm": 2.7084801197052,
      "learning_rate": 0.00016873410913449091,
      "loss": 5.178,
      "step": 3330
    },
    {
      "epoch": 0.3331,
      "grad_norm": 2.3028690814971924,
      "learning_rate": 0.0001687087510804423,
      "loss": 4.8963,
      "step": 3331
    },
    {
      "epoch": 0.3332,
      "grad_norm": 2.206533193588257,
      "learning_rate": 0.00016868338465444085,
      "loss": 4.909,
      "step": 3332
    },
    {
      "epoch": 0.3333,
      "grad_norm": 3.0908520221710205,
      "learning_rate": 0.00016865800985957726,
      "loss": 5.265,
      "step": 3333
    },
    {
      "epoch": 0.3334,
      "grad_norm": 2.124842643737793,
      "learning_rate": 0.0001686326266989435,
      "loss": 4.9534,
      "step": 3334
    },
    {
      "epoch": 0.3335,
      "grad_norm": 3.0866103172302246,
      "learning_rate": 0.0001686072351756323,
      "loss": 5.251,
      "step": 3335
    },
    {
      "epoch": 0.3336,
      "grad_norm": 2.241741418838501,
      "learning_rate": 0.00016858183529273765,
      "loss": 5.2106,
      "step": 3336
    },
    {
      "epoch": 0.3337,
      "grad_norm": 2.282134532928467,
      "learning_rate": 0.00016855642705335437,
      "loss": 5.0002,
      "step": 3337
    },
    {
      "epoch": 0.3338,
      "grad_norm": 7.842081546783447,
      "learning_rate": 0.0001685310104605784,
      "loss": 6.1002,
      "step": 3338
    },
    {
      "epoch": 0.3339,
      "grad_norm": 13.475752830505371,
      "learning_rate": 0.0001685055855175067,
      "loss": 5.0664,
      "step": 3339
    },
    {
      "epoch": 0.334,
      "grad_norm": 5.1048197746276855,
      "learning_rate": 0.0001684801522272372,
      "loss": 5.3048,
      "step": 3340
    },
    {
      "epoch": 0.3341,
      "grad_norm": 3.116518974304199,
      "learning_rate": 0.00016845471059286887,
      "loss": 5.6126,
      "step": 3341
    },
    {
      "epoch": 0.3342,
      "grad_norm": 3.419135808944702,
      "learning_rate": 0.0001684292606175017,
      "loss": 4.9518,
      "step": 3342
    },
    {
      "epoch": 0.3343,
      "grad_norm": 9.127632141113281,
      "learning_rate": 0.00016840380230423668,
      "loss": 5.335,
      "step": 3343
    },
    {
      "epoch": 0.3344,
      "grad_norm": 6.858264446258545,
      "learning_rate": 0.0001683783356561759,
      "loss": 5.2545,
      "step": 3344
    },
    {
      "epoch": 0.3345,
      "grad_norm": 8.765423774719238,
      "learning_rate": 0.00016835286067642228,
      "loss": 5.3784,
      "step": 3345
    },
    {
      "epoch": 0.3346,
      "grad_norm": 2.448044776916504,
      "learning_rate": 0.00016832737736807994,
      "loss": 5.1293,
      "step": 3346
    },
    {
      "epoch": 0.3347,
      "grad_norm": 8.13670539855957,
      "learning_rate": 0.00016830188573425389,
      "loss": 5.7263,
      "step": 3347
    },
    {
      "epoch": 0.3348,
      "grad_norm": 3.0061352252960205,
      "learning_rate": 0.00016827638577805026,
      "loss": 6.0144,
      "step": 3348
    },
    {
      "epoch": 0.3349,
      "grad_norm": 2.9557759761810303,
      "learning_rate": 0.0001682508775025762,
      "loss": 4.9236,
      "step": 3349
    },
    {
      "epoch": 0.335,
      "grad_norm": 3.1828322410583496,
      "learning_rate": 0.00016822536091093965,
      "loss": 5.2973,
      "step": 3350
    },
    {
      "epoch": 0.3351,
      "grad_norm": 3.322023868560791,
      "learning_rate": 0.00016819983600624986,
      "loss": 5.4012,
      "step": 3351
    },
    {
      "epoch": 0.3352,
      "grad_norm": 3.333162546157837,
      "learning_rate": 0.0001681743027916169,
      "loss": 4.787,
      "step": 3352
    },
    {
      "epoch": 0.3353,
      "grad_norm": 2.888965368270874,
      "learning_rate": 0.000168148761270152,
      "loss": 6.0456,
      "step": 3353
    },
    {
      "epoch": 0.3354,
      "grad_norm": 5.56376838684082,
      "learning_rate": 0.0001681232114449672,
      "loss": 6.1161,
      "step": 3354
    },
    {
      "epoch": 0.3355,
      "grad_norm": 3.3224685192108154,
      "learning_rate": 0.00016809765331917575,
      "loss": 5.4194,
      "step": 3355
    },
    {
      "epoch": 0.3356,
      "grad_norm": 2.756094217300415,
      "learning_rate": 0.0001680720868958918,
      "loss": 4.9458,
      "step": 3356
    },
    {
      "epoch": 0.3357,
      "grad_norm": 6.172860145568848,
      "learning_rate": 0.00016804651217823053,
      "loss": 4.9631,
      "step": 3357
    },
    {
      "epoch": 0.3358,
      "grad_norm": 26.067869186401367,
      "learning_rate": 0.00016802092916930818,
      "loss": 5.9824,
      "step": 3358
    },
    {
      "epoch": 0.3359,
      "grad_norm": 2.561868667602539,
      "learning_rate": 0.00016799533787224192,
      "loss": 5.3244,
      "step": 3359
    },
    {
      "epoch": 0.336,
      "grad_norm": 5.0944132804870605,
      "learning_rate": 0.00016796973829015,
      "loss": 5.1006,
      "step": 3360
    },
    {
      "epoch": 0.3361,
      "grad_norm": 2.4434762001037598,
      "learning_rate": 0.00016794413042615168,
      "loss": 5.3597,
      "step": 3361
    },
    {
      "epoch": 0.3362,
      "grad_norm": 6.924977779388428,
      "learning_rate": 0.00016791851428336711,
      "loss": 5.6832,
      "step": 3362
    },
    {
      "epoch": 0.3363,
      "grad_norm": 3.5626790523529053,
      "learning_rate": 0.00016789288986491762,
      "loss": 5.5146,
      "step": 3363
    },
    {
      "epoch": 0.3364,
      "grad_norm": 2.251239776611328,
      "learning_rate": 0.00016786725717392545,
      "loss": 5.2847,
      "step": 3364
    },
    {
      "epoch": 0.3365,
      "grad_norm": 4.839903831481934,
      "learning_rate": 0.00016784161621351382,
      "loss": 5.3437,
      "step": 3365
    },
    {
      "epoch": 0.3366,
      "grad_norm": 2.5748798847198486,
      "learning_rate": 0.0001678159669868071,
      "loss": 5.2169,
      "step": 3366
    },
    {
      "epoch": 0.3367,
      "grad_norm": 3.915405750274658,
      "learning_rate": 0.00016779030949693044,
      "loss": 4.8651,
      "step": 3367
    },
    {
      "epoch": 0.3368,
      "grad_norm": 8.624011993408203,
      "learning_rate": 0.00016776464374701025,
      "loss": 5.6976,
      "step": 3368
    },
    {
      "epoch": 0.3369,
      "grad_norm": 3.2023723125457764,
      "learning_rate": 0.00016773896974017373,
      "loss": 4.7783,
      "step": 3369
    },
    {
      "epoch": 0.337,
      "grad_norm": 3.6145336627960205,
      "learning_rate": 0.00016771328747954925,
      "loss": 5.1046,
      "step": 3370
    },
    {
      "epoch": 0.3371,
      "grad_norm": 5.428323268890381,
      "learning_rate": 0.00016768759696826608,
      "loss": 5.7911,
      "step": 3371
    },
    {
      "epoch": 0.3372,
      "grad_norm": 4.395740509033203,
      "learning_rate": 0.00016766189820945456,
      "loss": 5.397,
      "step": 3372
    },
    {
      "epoch": 0.3373,
      "grad_norm": 3.195483922958374,
      "learning_rate": 0.00016763619120624594,
      "loss": 5.1859,
      "step": 3373
    },
    {
      "epoch": 0.3374,
      "grad_norm": 2.157116174697876,
      "learning_rate": 0.00016761047596177263,
      "loss": 5.1918,
      "step": 3374
    },
    {
      "epoch": 0.3375,
      "grad_norm": 8.136451721191406,
      "learning_rate": 0.00016758475247916787,
      "loss": 5.3308,
      "step": 3375
    },
    {
      "epoch": 0.3376,
      "grad_norm": 11.224508285522461,
      "learning_rate": 0.00016755902076156604,
      "loss": 6.2817,
      "step": 3376
    },
    {
      "epoch": 0.3377,
      "grad_norm": 2.0763635635375977,
      "learning_rate": 0.00016753328081210245,
      "loss": 5.138,
      "step": 3377
    },
    {
      "epoch": 0.3378,
      "grad_norm": 3.582146644592285,
      "learning_rate": 0.00016750753263391346,
      "loss": 4.8644,
      "step": 3378
    },
    {
      "epoch": 0.3379,
      "grad_norm": 6.3212785720825195,
      "learning_rate": 0.00016748177623013638,
      "loss": 5.7987,
      "step": 3379
    },
    {
      "epoch": 0.338,
      "grad_norm": 2.6711266040802,
      "learning_rate": 0.00016745601160390958,
      "loss": 5.5347,
      "step": 3380
    },
    {
      "epoch": 0.3381,
      "grad_norm": 4.893423557281494,
      "learning_rate": 0.00016743023875837233,
      "loss": 4.9375,
      "step": 3381
    },
    {
      "epoch": 0.3382,
      "grad_norm": 2.5749928951263428,
      "learning_rate": 0.00016740445769666509,
      "loss": 5.4807,
      "step": 3382
    },
    {
      "epoch": 0.3383,
      "grad_norm": 9.963428497314453,
      "learning_rate": 0.0001673786684219291,
      "loss": 5.8456,
      "step": 3383
    },
    {
      "epoch": 0.3384,
      "grad_norm": 3.129901647567749,
      "learning_rate": 0.00016735287093730676,
      "loss": 5.0729,
      "step": 3384
    },
    {
      "epoch": 0.3385,
      "grad_norm": 2.7076776027679443,
      "learning_rate": 0.00016732706524594137,
      "loss": 5.5638,
      "step": 3385
    },
    {
      "epoch": 0.3386,
      "grad_norm": 6.737016201019287,
      "learning_rate": 0.00016730125135097735,
      "loss": 7.6769,
      "step": 3386
    },
    {
      "epoch": 0.3387,
      "grad_norm": 4.489473342895508,
      "learning_rate": 0.00016727542925555996,
      "loss": 5.5038,
      "step": 3387
    },
    {
      "epoch": 0.3388,
      "grad_norm": 2.552579879760742,
      "learning_rate": 0.00016724959896283559,
      "loss": 4.9799,
      "step": 3388
    },
    {
      "epoch": 0.3389,
      "grad_norm": 8.628547668457031,
      "learning_rate": 0.00016722376047595164,
      "loss": 6.3955,
      "step": 3389
    },
    {
      "epoch": 0.339,
      "grad_norm": 6.13525390625,
      "learning_rate": 0.0001671979137980563,
      "loss": 5.3698,
      "step": 3390
    },
    {
      "epoch": 0.3391,
      "grad_norm": 6.061637878417969,
      "learning_rate": 0.00016717205893229903,
      "loss": 4.8153,
      "step": 3391
    },
    {
      "epoch": 0.3392,
      "grad_norm": 2.11928653717041,
      "learning_rate": 0.00016714619588183014,
      "loss": 4.7383,
      "step": 3392
    },
    {
      "epoch": 0.3393,
      "grad_norm": 2.538229465484619,
      "learning_rate": 0.00016712032464980095,
      "loss": 5.3587,
      "step": 3393
    },
    {
      "epoch": 0.3394,
      "grad_norm": 3.7506728172302246,
      "learning_rate": 0.0001670944452393638,
      "loss": 4.9711,
      "step": 3394
    },
    {
      "epoch": 0.3395,
      "grad_norm": 2.1316025257110596,
      "learning_rate": 0.000167068557653672,
      "loss": 5.1758,
      "step": 3395
    },
    {
      "epoch": 0.3396,
      "grad_norm": 2.5928282737731934,
      "learning_rate": 0.0001670426618958799,
      "loss": 4.7337,
      "step": 3396
    },
    {
      "epoch": 0.3397,
      "grad_norm": 2.135057210922241,
      "learning_rate": 0.00016701675796914286,
      "loss": 5.1618,
      "step": 3397
    },
    {
      "epoch": 0.3398,
      "grad_norm": 2.45090913772583,
      "learning_rate": 0.0001669908458766171,
      "loss": 5.4204,
      "step": 3398
    },
    {
      "epoch": 0.3399,
      "grad_norm": 2.5074622631073,
      "learning_rate": 0.00016696492562145996,
      "loss": 5.0515,
      "step": 3399
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.6317360401153564,
      "learning_rate": 0.00016693899720682977,
      "loss": 5.3906,
      "step": 3400
    },
    {
      "epoch": 0.3401,
      "grad_norm": 4.199395656585693,
      "learning_rate": 0.00016691306063588583,
      "loss": 5.9767,
      "step": 3401
    },
    {
      "epoch": 0.3402,
      "grad_norm": 1.9569315910339355,
      "learning_rate": 0.00016688711591178842,
      "loss": 4.9222,
      "step": 3402
    },
    {
      "epoch": 0.3403,
      "grad_norm": 8.491572380065918,
      "learning_rate": 0.00016686116303769882,
      "loss": 5.047,
      "step": 3403
    },
    {
      "epoch": 0.3404,
      "grad_norm": 3.9831604957580566,
      "learning_rate": 0.0001668352020167793,
      "loss": 4.8759,
      "step": 3404
    },
    {
      "epoch": 0.3405,
      "grad_norm": 3.0439300537109375,
      "learning_rate": 0.00016680923285219317,
      "loss": 5.27,
      "step": 3405
    },
    {
      "epoch": 0.3406,
      "grad_norm": 7.159900665283203,
      "learning_rate": 0.00016678325554710468,
      "loss": 5.205,
      "step": 3406
    },
    {
      "epoch": 0.3407,
      "grad_norm": 4.132561683654785,
      "learning_rate": 0.00016675727010467906,
      "loss": 5.2381,
      "step": 3407
    },
    {
      "epoch": 0.3408,
      "grad_norm": 3.2154595851898193,
      "learning_rate": 0.0001667312765280826,
      "loss": 5.5339,
      "step": 3408
    },
    {
      "epoch": 0.3409,
      "grad_norm": 2.956510305404663,
      "learning_rate": 0.00016670527482048246,
      "loss": 4.9795,
      "step": 3409
    },
    {
      "epoch": 0.341,
      "grad_norm": 4.068884372711182,
      "learning_rate": 0.00016667926498504696,
      "loss": 6.0589,
      "step": 3410
    },
    {
      "epoch": 0.3411,
      "grad_norm": 1.9445184469223022,
      "learning_rate": 0.00016665324702494524,
      "loss": 5.186,
      "step": 3411
    },
    {
      "epoch": 0.3412,
      "grad_norm": 2.889932155609131,
      "learning_rate": 0.0001666272209433476,
      "loss": 4.9099,
      "step": 3412
    },
    {
      "epoch": 0.3413,
      "grad_norm": 8.59227180480957,
      "learning_rate": 0.00016660118674342517,
      "loss": 5.3061,
      "step": 3413
    },
    {
      "epoch": 0.3414,
      "grad_norm": 5.396921157836914,
      "learning_rate": 0.00016657514442835014,
      "loss": 5.5358,
      "step": 3414
    },
    {
      "epoch": 0.3415,
      "grad_norm": 3.9787890911102295,
      "learning_rate": 0.00016654909400129575,
      "loss": 5.1495,
      "step": 3415
    },
    {
      "epoch": 0.3416,
      "grad_norm": 4.241540908813477,
      "learning_rate": 0.00016652303546543608,
      "loss": 5.0872,
      "step": 3416
    },
    {
      "epoch": 0.3417,
      "grad_norm": 2.185364007949829,
      "learning_rate": 0.00016649696882394633,
      "loss": 4.9592,
      "step": 3417
    },
    {
      "epoch": 0.3418,
      "grad_norm": 3.556605339050293,
      "learning_rate": 0.00016647089408000266,
      "loss": 5.4234,
      "step": 3418
    },
    {
      "epoch": 0.3419,
      "grad_norm": 2.20109224319458,
      "learning_rate": 0.00016644481123678217,
      "loss": 4.9271,
      "step": 3419
    },
    {
      "epoch": 0.342,
      "grad_norm": 2.7789838314056396,
      "learning_rate": 0.00016641872029746297,
      "loss": 5.4576,
      "step": 3420
    },
    {
      "epoch": 0.3421,
      "grad_norm": 3.2385177612304688,
      "learning_rate": 0.00016639262126522418,
      "loss": 5.2466,
      "step": 3421
    },
    {
      "epoch": 0.3422,
      "grad_norm": 4.17047643661499,
      "learning_rate": 0.00016636651414324587,
      "loss": 4.8673,
      "step": 3422
    },
    {
      "epoch": 0.3423,
      "grad_norm": 2.249077081680298,
      "learning_rate": 0.00016634039893470912,
      "loss": 4.8908,
      "step": 3423
    },
    {
      "epoch": 0.3424,
      "grad_norm": 3.4751293659210205,
      "learning_rate": 0.000166314275642796,
      "loss": 5.5368,
      "step": 3424
    },
    {
      "epoch": 0.3425,
      "grad_norm": 2.874444007873535,
      "learning_rate": 0.00016628814427068953,
      "loss": 5.0298,
      "step": 3425
    },
    {
      "epoch": 0.3426,
      "grad_norm": 4.839580535888672,
      "learning_rate": 0.00016626200482157378,
      "loss": 5.52,
      "step": 3426
    },
    {
      "epoch": 0.3427,
      "grad_norm": 3.987884521484375,
      "learning_rate": 0.00016623585729863368,
      "loss": 4.9604,
      "step": 3427
    },
    {
      "epoch": 0.3428,
      "grad_norm": 3.4349396228790283,
      "learning_rate": 0.00016620970170505534,
      "loss": 4.8966,
      "step": 3428
    },
    {
      "epoch": 0.3429,
      "grad_norm": 4.967717170715332,
      "learning_rate": 0.00016618353804402568,
      "loss": 5.0378,
      "step": 3429
    },
    {
      "epoch": 0.343,
      "grad_norm": 2.2895278930664062,
      "learning_rate": 0.00016615736631873262,
      "loss": 4.9064,
      "step": 3430
    },
    {
      "epoch": 0.3431,
      "grad_norm": 2.1759533882141113,
      "learning_rate": 0.00016613118653236518,
      "loss": 5.4635,
      "step": 3431
    },
    {
      "epoch": 0.3432,
      "grad_norm": 6.083454608917236,
      "learning_rate": 0.00016610499868811328,
      "loss": 5.1389,
      "step": 3432
    },
    {
      "epoch": 0.3433,
      "grad_norm": 7.009731292724609,
      "learning_rate": 0.00016607880278916777,
      "loss": 6.349,
      "step": 3433
    },
    {
      "epoch": 0.3434,
      "grad_norm": 4.157683372497559,
      "learning_rate": 0.0001660525988387206,
      "loss": 5.7717,
      "step": 3434
    },
    {
      "epoch": 0.3435,
      "grad_norm": 7.874264240264893,
      "learning_rate": 0.00016602638683996463,
      "loss": 5.2769,
      "step": 3435
    },
    {
      "epoch": 0.3436,
      "grad_norm": 3.4022209644317627,
      "learning_rate": 0.0001660001667960937,
      "loss": 4.645,
      "step": 3436
    },
    {
      "epoch": 0.3437,
      "grad_norm": 1.809404969215393,
      "learning_rate": 0.00016597393871030264,
      "loss": 5.4663,
      "step": 3437
    },
    {
      "epoch": 0.3438,
      "grad_norm": 7.811855792999268,
      "learning_rate": 0.0001659477025857872,
      "loss": 5.0985,
      "step": 3438
    },
    {
      "epoch": 0.3439,
      "grad_norm": 3.419106960296631,
      "learning_rate": 0.00016592145842574433,
      "loss": 5.4805,
      "step": 3439
    },
    {
      "epoch": 0.344,
      "grad_norm": 4.748898506164551,
      "learning_rate": 0.0001658952062333717,
      "loss": 5.059,
      "step": 3440
    },
    {
      "epoch": 0.3441,
      "grad_norm": 3.1931419372558594,
      "learning_rate": 0.00016586894601186805,
      "loss": 5.2607,
      "step": 3441
    },
    {
      "epoch": 0.3442,
      "grad_norm": 4.138723373413086,
      "learning_rate": 0.00016584267776443317,
      "loss": 5.6398,
      "step": 3442
    },
    {
      "epoch": 0.3443,
      "grad_norm": 4.547636032104492,
      "learning_rate": 0.00016581640149426768,
      "loss": 5.3872,
      "step": 3443
    },
    {
      "epoch": 0.3444,
      "grad_norm": 3.369361639022827,
      "learning_rate": 0.00016579011720457333,
      "loss": 4.9943,
      "step": 3444
    },
    {
      "epoch": 0.3445,
      "grad_norm": 2.8790290355682373,
      "learning_rate": 0.00016576382489855274,
      "loss": 5.8017,
      "step": 3445
    },
    {
      "epoch": 0.3446,
      "grad_norm": 4.911459922790527,
      "learning_rate": 0.0001657375245794096,
      "loss": 5.8029,
      "step": 3446
    },
    {
      "epoch": 0.3447,
      "grad_norm": 4.953139305114746,
      "learning_rate": 0.00016571121625034847,
      "loss": 5.7324,
      "step": 3447
    },
    {
      "epoch": 0.3448,
      "grad_norm": 4.911296367645264,
      "learning_rate": 0.00016568489991457497,
      "loss": 5.1751,
      "step": 3448
    },
    {
      "epoch": 0.3449,
      "grad_norm": 2.673262357711792,
      "learning_rate": 0.00016565857557529566,
      "loss": 5.3264,
      "step": 3449
    },
    {
      "epoch": 0.345,
      "grad_norm": 2.591764211654663,
      "learning_rate": 0.00016563224323571806,
      "loss": 5.1149,
      "step": 3450
    },
    {
      "epoch": 0.3451,
      "grad_norm": 3.7979652881622314,
      "learning_rate": 0.00016560590289905073,
      "loss": 5.7314,
      "step": 3451
    },
    {
      "epoch": 0.3452,
      "grad_norm": 3.611321449279785,
      "learning_rate": 0.00016557955456850313,
      "loss": 5.1044,
      "step": 3452
    },
    {
      "epoch": 0.3453,
      "grad_norm": 3.4010229110717773,
      "learning_rate": 0.00016555319824728575,
      "loss": 5.4664,
      "step": 3453
    },
    {
      "epoch": 0.3454,
      "grad_norm": 2.1699934005737305,
      "learning_rate": 0.00016552683393860997,
      "loss": 5.2436,
      "step": 3454
    },
    {
      "epoch": 0.3455,
      "grad_norm": 7.458680629730225,
      "learning_rate": 0.00016550046164568827,
      "loss": 4.8727,
      "step": 3455
    },
    {
      "epoch": 0.3456,
      "grad_norm": 5.607057094573975,
      "learning_rate": 0.00016547408137173396,
      "loss": 5.2206,
      "step": 3456
    },
    {
      "epoch": 0.3457,
      "grad_norm": 9.850472450256348,
      "learning_rate": 0.00016544769311996148,
      "loss": 5.8219,
      "step": 3457
    },
    {
      "epoch": 0.3458,
      "grad_norm": 4.693068504333496,
      "learning_rate": 0.00016542129689358612,
      "loss": 5.1004,
      "step": 3458
    },
    {
      "epoch": 0.3459,
      "grad_norm": 5.932913303375244,
      "learning_rate": 0.00016539489269582416,
      "loss": 5.3543,
      "step": 3459
    },
    {
      "epoch": 0.346,
      "grad_norm": 9.703887939453125,
      "learning_rate": 0.00016536848052989291,
      "loss": 5.4256,
      "step": 3460
    },
    {
      "epoch": 0.3461,
      "grad_norm": 4.034893035888672,
      "learning_rate": 0.00016534206039901057,
      "loss": 5.282,
      "step": 3461
    },
    {
      "epoch": 0.3462,
      "grad_norm": 4.665827751159668,
      "learning_rate": 0.00016531563230639637,
      "loss": 5.1544,
      "step": 3462
    },
    {
      "epoch": 0.3463,
      "grad_norm": 8.16281795501709,
      "learning_rate": 0.0001652891962552705,
      "loss": 5.2876,
      "step": 3463
    },
    {
      "epoch": 0.3464,
      "grad_norm": 2.947084903717041,
      "learning_rate": 0.00016526275224885411,
      "loss": 5.1298,
      "step": 3464
    },
    {
      "epoch": 0.3465,
      "grad_norm": 2.8019585609436035,
      "learning_rate": 0.00016523630029036931,
      "loss": 5.5469,
      "step": 3465
    },
    {
      "epoch": 0.3466,
      "grad_norm": 4.315515041351318,
      "learning_rate": 0.00016520984038303924,
      "loss": 5.4383,
      "step": 3466
    },
    {
      "epoch": 0.3467,
      "grad_norm": 4.592276096343994,
      "learning_rate": 0.0001651833725300879,
      "loss": 4.94,
      "step": 3467
    },
    {
      "epoch": 0.3468,
      "grad_norm": 2.9991660118103027,
      "learning_rate": 0.00016515689673474033,
      "loss": 5.7786,
      "step": 3468
    },
    {
      "epoch": 0.3469,
      "grad_norm": 2.353968381881714,
      "learning_rate": 0.00016513041300022255,
      "loss": 4.8485,
      "step": 3469
    },
    {
      "epoch": 0.347,
      "grad_norm": 3.910555601119995,
      "learning_rate": 0.0001651039213297615,
      "loss": 5.1936,
      "step": 3470
    },
    {
      "epoch": 0.3471,
      "grad_norm": 2.008342742919922,
      "learning_rate": 0.0001650774217265851,
      "loss": 5.7928,
      "step": 3471
    },
    {
      "epoch": 0.3472,
      "grad_norm": 2.3327345848083496,
      "learning_rate": 0.00016505091419392228,
      "loss": 5.4186,
      "step": 3472
    },
    {
      "epoch": 0.3473,
      "grad_norm": 3.962707757949829,
      "learning_rate": 0.00016502439873500289,
      "loss": 5.0306,
      "step": 3473
    },
    {
      "epoch": 0.3474,
      "grad_norm": 2.577174663543701,
      "learning_rate": 0.00016499787535305776,
      "loss": 5.6849,
      "step": 3474
    },
    {
      "epoch": 0.3475,
      "grad_norm": 2.151893377304077,
      "learning_rate": 0.00016497134405131866,
      "loss": 5.1407,
      "step": 3475
    },
    {
      "epoch": 0.3476,
      "grad_norm": 9.861145973205566,
      "learning_rate": 0.00016494480483301836,
      "loss": 5.5926,
      "step": 3476
    },
    {
      "epoch": 0.3477,
      "grad_norm": 10.419118881225586,
      "learning_rate": 0.00016491825770139062,
      "loss": 5.6146,
      "step": 3477
    },
    {
      "epoch": 0.3478,
      "grad_norm": 5.492389678955078,
      "learning_rate": 0.00016489170265967008,
      "loss": 5.8417,
      "step": 3478
    },
    {
      "epoch": 0.3479,
      "grad_norm": 7.173908710479736,
      "learning_rate": 0.00016486513971109243,
      "loss": 5.9209,
      "step": 3479
    },
    {
      "epoch": 0.348,
      "grad_norm": 19.323835372924805,
      "learning_rate": 0.0001648385688588942,
      "loss": 5.2811,
      "step": 3480
    },
    {
      "epoch": 0.3481,
      "grad_norm": 5.15094518661499,
      "learning_rate": 0.0001648119901063131,
      "loss": 4.9351,
      "step": 3481
    },
    {
      "epoch": 0.3482,
      "grad_norm": 5.493136882781982,
      "learning_rate": 0.00016478540345658759,
      "loss": 5.6018,
      "step": 3482
    },
    {
      "epoch": 0.3483,
      "grad_norm": 1.91474187374115,
      "learning_rate": 0.00016475880891295716,
      "loss": 5.1814,
      "step": 3483
    },
    {
      "epoch": 0.3484,
      "grad_norm": 5.633931636810303,
      "learning_rate": 0.0001647322064786623,
      "loss": 4.8635,
      "step": 3484
    },
    {
      "epoch": 0.3485,
      "grad_norm": 3.5346014499664307,
      "learning_rate": 0.00016470559615694446,
      "loss": 5.2416,
      "step": 3485
    },
    {
      "epoch": 0.3486,
      "grad_norm": 8.495972633361816,
      "learning_rate": 0.000164678977951046,
      "loss": 5.1461,
      "step": 3486
    },
    {
      "epoch": 0.3487,
      "grad_norm": 3.3549964427948,
      "learning_rate": 0.0001646523518642102,
      "loss": 5.6903,
      "step": 3487
    },
    {
      "epoch": 0.3488,
      "grad_norm": 2.9665400981903076,
      "learning_rate": 0.00016462571789968152,
      "loss": 5.5811,
      "step": 3488
    },
    {
      "epoch": 0.3489,
      "grad_norm": 3.952573299407959,
      "learning_rate": 0.0001645990760607051,
      "loss": 5.3768,
      "step": 3489
    },
    {
      "epoch": 0.349,
      "grad_norm": 2.391845464706421,
      "learning_rate": 0.00016457242635052724,
      "loss": 5.7188,
      "step": 3490
    },
    {
      "epoch": 0.3491,
      "grad_norm": 4.427984714508057,
      "learning_rate": 0.00016454576877239507,
      "loss": 5.2391,
      "step": 3491
    },
    {
      "epoch": 0.3492,
      "grad_norm": 2.4530675411224365,
      "learning_rate": 0.00016451910332955679,
      "loss": 5.2316,
      "step": 3492
    },
    {
      "epoch": 0.3493,
      "grad_norm": 3.184187650680542,
      "learning_rate": 0.00016449243002526144,
      "loss": 4.967,
      "step": 3493
    },
    {
      "epoch": 0.3494,
      "grad_norm": 3.248767852783203,
      "learning_rate": 0.00016446574886275913,
      "loss": 5.2219,
      "step": 3494
    },
    {
      "epoch": 0.3495,
      "grad_norm": 4.693120956420898,
      "learning_rate": 0.0001644390598453009,
      "loss": 4.9624,
      "step": 3495
    },
    {
      "epoch": 0.3496,
      "grad_norm": 2.8951072692871094,
      "learning_rate": 0.00016441236297613866,
      "loss": 4.9203,
      "step": 3496
    },
    {
      "epoch": 0.3497,
      "grad_norm": 4.7666401863098145,
      "learning_rate": 0.0001643856582585254,
      "loss": 4.6567,
      "step": 3497
    },
    {
      "epoch": 0.3498,
      "grad_norm": 3.7450122833251953,
      "learning_rate": 0.00016435894569571496,
      "loss": 5.354,
      "step": 3498
    },
    {
      "epoch": 0.3499,
      "grad_norm": 9.137309074401855,
      "learning_rate": 0.0001643322252909622,
      "loss": 5.4918,
      "step": 3499
    },
    {
      "epoch": 0.35,
      "grad_norm": 4.220043659210205,
      "learning_rate": 0.00016430549704752294,
      "loss": 5.6938,
      "step": 3500
    },
    {
      "epoch": 0.3501,
      "grad_norm": 1.9442999362945557,
      "learning_rate": 0.00016427876096865394,
      "loss": 5.3241,
      "step": 3501
    },
    {
      "epoch": 0.3502,
      "grad_norm": 2.6358721256256104,
      "learning_rate": 0.00016425201705761288,
      "loss": 4.6766,
      "step": 3502
    },
    {
      "epoch": 0.3503,
      "grad_norm": 6.496501445770264,
      "learning_rate": 0.00016422526531765846,
      "loss": 5.3963,
      "step": 3503
    },
    {
      "epoch": 0.3504,
      "grad_norm": 3.40842342376709,
      "learning_rate": 0.00016419850575205024,
      "loss": 5.0906,
      "step": 3504
    },
    {
      "epoch": 0.3505,
      "grad_norm": 2.387256622314453,
      "learning_rate": 0.00016417173836404887,
      "loss": 5.1863,
      "step": 3505
    },
    {
      "epoch": 0.3506,
      "grad_norm": 2.9794528484344482,
      "learning_rate": 0.00016414496315691581,
      "loss": 4.9502,
      "step": 3506
    },
    {
      "epoch": 0.3507,
      "grad_norm": 5.881515979766846,
      "learning_rate": 0.00016411818013391355,
      "loss": 5.3362,
      "step": 3507
    },
    {
      "epoch": 0.3508,
      "grad_norm": 3.0103914737701416,
      "learning_rate": 0.00016409138929830553,
      "loss": 5.0467,
      "step": 3508
    },
    {
      "epoch": 0.3509,
      "grad_norm": 4.563115119934082,
      "learning_rate": 0.00016406459065335615,
      "loss": 5.0614,
      "step": 3509
    },
    {
      "epoch": 0.351,
      "grad_norm": 3.612612009048462,
      "learning_rate": 0.00016403778420233075,
      "loss": 5.4393,
      "step": 3510
    },
    {
      "epoch": 0.3511,
      "grad_norm": 6.332382678985596,
      "learning_rate": 0.00016401096994849557,
      "loss": 5.8101,
      "step": 3511
    },
    {
      "epoch": 0.3512,
      "grad_norm": 4.318439960479736,
      "learning_rate": 0.00016398414789511786,
      "loss": 5.8783,
      "step": 3512
    },
    {
      "epoch": 0.3513,
      "grad_norm": 4.420598030090332,
      "learning_rate": 0.0001639573180454658,
      "loss": 5.9454,
      "step": 3513
    },
    {
      "epoch": 0.3514,
      "grad_norm": 5.245269775390625,
      "learning_rate": 0.00016393048040280855,
      "loss": 5.5095,
      "step": 3514
    },
    {
      "epoch": 0.3515,
      "grad_norm": 3.072925090789795,
      "learning_rate": 0.00016390363497041622,
      "loss": 5.5559,
      "step": 3515
    },
    {
      "epoch": 0.3516,
      "grad_norm": 2.147965669631958,
      "learning_rate": 0.00016387678175155978,
      "loss": 4.9596,
      "step": 3516
    },
    {
      "epoch": 0.3517,
      "grad_norm": 2.9986751079559326,
      "learning_rate": 0.00016384992074951123,
      "loss": 5.9872,
      "step": 3517
    },
    {
      "epoch": 0.3518,
      "grad_norm": 19.477148056030273,
      "learning_rate": 0.00016382305196754356,
      "loss": 5.0316,
      "step": 3518
    },
    {
      "epoch": 0.3519,
      "grad_norm": 19.221744537353516,
      "learning_rate": 0.00016379617540893056,
      "loss": 5.8117,
      "step": 3519
    },
    {
      "epoch": 0.352,
      "grad_norm": 23.774648666381836,
      "learning_rate": 0.0001637692910769471,
      "loss": 5.9261,
      "step": 3520
    },
    {
      "epoch": 0.3521,
      "grad_norm": 6.231064319610596,
      "learning_rate": 0.000163742398974869,
      "loss": 4.9546,
      "step": 3521
    },
    {
      "epoch": 0.3522,
      "grad_norm": 5.261959552764893,
      "learning_rate": 0.00016371549910597287,
      "loss": 4.962,
      "step": 3522
    },
    {
      "epoch": 0.3523,
      "grad_norm": 9.875401496887207,
      "learning_rate": 0.0001636885914735365,
      "loss": 5.5327,
      "step": 3523
    },
    {
      "epoch": 0.3524,
      "grad_norm": 3.4961555004119873,
      "learning_rate": 0.00016366167608083843,
      "loss": 4.9586,
      "step": 3524
    },
    {
      "epoch": 0.3525,
      "grad_norm": 7.797249794006348,
      "learning_rate": 0.00016363475293115824,
      "loss": 5.5859,
      "step": 3525
    },
    {
      "epoch": 0.3526,
      "grad_norm": 2.2327194213867188,
      "learning_rate": 0.0001636078220277764,
      "loss": 5.1343,
      "step": 3526
    },
    {
      "epoch": 0.3527,
      "grad_norm": 3.1418044567108154,
      "learning_rate": 0.00016358088337397442,
      "loss": 4.9768,
      "step": 3527
    },
    {
      "epoch": 0.3528,
      "grad_norm": 2.2829225063323975,
      "learning_rate": 0.00016355393697303465,
      "loss": 4.6111,
      "step": 3528
    },
    {
      "epoch": 0.3529,
      "grad_norm": 2.4933876991271973,
      "learning_rate": 0.00016352698282824044,
      "loss": 4.9214,
      "step": 3529
    },
    {
      "epoch": 0.353,
      "grad_norm": 6.2541680335998535,
      "learning_rate": 0.00016350002094287609,
      "loss": 6.5613,
      "step": 3530
    },
    {
      "epoch": 0.3531,
      "grad_norm": 7.23848819732666,
      "learning_rate": 0.00016347305132022677,
      "loss": 6.0623,
      "step": 3531
    },
    {
      "epoch": 0.3532,
      "grad_norm": 2.9305551052093506,
      "learning_rate": 0.0001634460739635787,
      "loss": 5.603,
      "step": 3532
    },
    {
      "epoch": 0.3533,
      "grad_norm": 3.0221405029296875,
      "learning_rate": 0.00016341908887621895,
      "loss": 5.4674,
      "step": 3533
    },
    {
      "epoch": 0.3534,
      "grad_norm": 2.691526174545288,
      "learning_rate": 0.00016339209606143563,
      "loss": 5.7177,
      "step": 3534
    },
    {
      "epoch": 0.3535,
      "grad_norm": 2.3410961627960205,
      "learning_rate": 0.00016336509552251766,
      "loss": 5.2776,
      "step": 3535
    },
    {
      "epoch": 0.3536,
      "grad_norm": 3.4845798015594482,
      "learning_rate": 0.000163338087262755,
      "loss": 5.7609,
      "step": 3536
    },
    {
      "epoch": 0.3537,
      "grad_norm": 6.136795997619629,
      "learning_rate": 0.00016331107128543857,
      "loss": 4.7507,
      "step": 3537
    },
    {
      "epoch": 0.3538,
      "grad_norm": 2.7591559886932373,
      "learning_rate": 0.00016328404759386014,
      "loss": 5.052,
      "step": 3538
    },
    {
      "epoch": 0.3539,
      "grad_norm": 3.4950714111328125,
      "learning_rate": 0.00016325701619131246,
      "loss": 5.4008,
      "step": 3539
    },
    {
      "epoch": 0.354,
      "grad_norm": 3.263007879257202,
      "learning_rate": 0.00016322997708108922,
      "loss": 5.0275,
      "step": 3540
    },
    {
      "epoch": 0.3541,
      "grad_norm": 3.389556407928467,
      "learning_rate": 0.0001632029302664851,
      "loss": 5.0478,
      "step": 3541
    },
    {
      "epoch": 0.3542,
      "grad_norm": 2.3046488761901855,
      "learning_rate": 0.00016317587575079563,
      "loss": 5.2176,
      "step": 3542
    },
    {
      "epoch": 0.3543,
      "grad_norm": 9.31805419921875,
      "learning_rate": 0.00016314881353731732,
      "loss": 5.9683,
      "step": 3543
    },
    {
      "epoch": 0.3544,
      "grad_norm": 3.7338569164276123,
      "learning_rate": 0.00016312174362934765,
      "loss": 5.5598,
      "step": 3544
    },
    {
      "epoch": 0.3545,
      "grad_norm": 2.0529839992523193,
      "learning_rate": 0.00016309466603018496,
      "loss": 5.5682,
      "step": 3545
    },
    {
      "epoch": 0.3546,
      "grad_norm": 2.91774845123291,
      "learning_rate": 0.00016306758074312864,
      "loss": 5.1391,
      "step": 3546
    },
    {
      "epoch": 0.3547,
      "grad_norm": 2.503770112991333,
      "learning_rate": 0.0001630404877714789,
      "loss": 4.8712,
      "step": 3547
    },
    {
      "epoch": 0.3548,
      "grad_norm": 2.395176649093628,
      "learning_rate": 0.00016301338711853693,
      "loss": 5.1138,
      "step": 3548
    },
    {
      "epoch": 0.3549,
      "grad_norm": 3.562518358230591,
      "learning_rate": 0.00016298627878760487,
      "loss": 5.4391,
      "step": 3549
    },
    {
      "epoch": 0.355,
      "grad_norm": 7.2828898429870605,
      "learning_rate": 0.00016295916278198584,
      "loss": 5.6752,
      "step": 3550
    },
    {
      "epoch": 0.3551,
      "grad_norm": 4.227139472961426,
      "learning_rate": 0.00016293203910498376,
      "loss": 5.4331,
      "step": 3551
    },
    {
      "epoch": 0.3552,
      "grad_norm": 2.5513627529144287,
      "learning_rate": 0.0001629049077599036,
      "loss": 5.0409,
      "step": 3552
    },
    {
      "epoch": 0.3553,
      "grad_norm": 2.0043561458587646,
      "learning_rate": 0.0001628777687500513,
      "loss": 5.1758,
      "step": 3553
    },
    {
      "epoch": 0.3554,
      "grad_norm": 1.9514544010162354,
      "learning_rate": 0.00016285062207873355,
      "loss": 5.0007,
      "step": 3554
    },
    {
      "epoch": 0.3555,
      "grad_norm": 1.8425183296203613,
      "learning_rate": 0.00016282346774925817,
      "loss": 4.7455,
      "step": 3555
    },
    {
      "epoch": 0.3556,
      "grad_norm": 5.026706218719482,
      "learning_rate": 0.00016279630576493382,
      "loss": 5.0703,
      "step": 3556
    },
    {
      "epoch": 0.3557,
      "grad_norm": 2.8265674114227295,
      "learning_rate": 0.00016276913612907007,
      "loss": 4.8969,
      "step": 3557
    },
    {
      "epoch": 0.3558,
      "grad_norm": 3.6387059688568115,
      "learning_rate": 0.0001627419588449775,
      "loss": 5.1173,
      "step": 3558
    },
    {
      "epoch": 0.3559,
      "grad_norm": 2.8666045665740967,
      "learning_rate": 0.00016271477391596753,
      "loss": 5.4863,
      "step": 3559
    },
    {
      "epoch": 0.356,
      "grad_norm": 8.801528930664062,
      "learning_rate": 0.0001626875813453526,
      "loss": 5.5105,
      "step": 3560
    },
    {
      "epoch": 0.3561,
      "grad_norm": 5.518276214599609,
      "learning_rate": 0.00016266038113644607,
      "loss": 4.7209,
      "step": 3561
    },
    {
      "epoch": 0.3562,
      "grad_norm": 3.271289825439453,
      "learning_rate": 0.00016263317329256213,
      "loss": 5.0708,
      "step": 3562
    },
    {
      "epoch": 0.3563,
      "grad_norm": 2.982121229171753,
      "learning_rate": 0.00016260595781701604,
      "loss": 5.2439,
      "step": 3563
    },
    {
      "epoch": 0.3564,
      "grad_norm": 5.097856044769287,
      "learning_rate": 0.0001625787347131239,
      "loss": 4.7085,
      "step": 3564
    },
    {
      "epoch": 0.3565,
      "grad_norm": 4.042386531829834,
      "learning_rate": 0.00016255150398420271,
      "loss": 5.0481,
      "step": 3565
    },
    {
      "epoch": 0.3566,
      "grad_norm": 5.35045051574707,
      "learning_rate": 0.00016252426563357055,
      "loss": 5.9806,
      "step": 3566
    },
    {
      "epoch": 0.3567,
      "grad_norm": 7.404446601867676,
      "learning_rate": 0.00016249701966454625,
      "loss": 7.1212,
      "step": 3567
    },
    {
      "epoch": 0.3568,
      "grad_norm": 6.857364177703857,
      "learning_rate": 0.0001624697660804497,
      "loss": 4.8933,
      "step": 3568
    },
    {
      "epoch": 0.3569,
      "grad_norm": 6.68425989151001,
      "learning_rate": 0.00016244250488460158,
      "loss": 5.1637,
      "step": 3569
    },
    {
      "epoch": 0.357,
      "grad_norm": 3.038789749145508,
      "learning_rate": 0.00016241523608032373,
      "loss": 5.6874,
      "step": 3570
    },
    {
      "epoch": 0.3571,
      "grad_norm": 4.373527526855469,
      "learning_rate": 0.00016238795967093864,
      "loss": 5.5281,
      "step": 3571
    },
    {
      "epoch": 0.3572,
      "grad_norm": 2.299834966659546,
      "learning_rate": 0.00016236067565976992,
      "loss": 5.2861,
      "step": 3572
    },
    {
      "epoch": 0.3573,
      "grad_norm": 3.3679044246673584,
      "learning_rate": 0.00016233338405014202,
      "loss": 5.3267,
      "step": 3573
    },
    {
      "epoch": 0.3574,
      "grad_norm": 2.455361843109131,
      "learning_rate": 0.00016230608484538034,
      "loss": 4.8154,
      "step": 3574
    },
    {
      "epoch": 0.3575,
      "grad_norm": 2.9881958961486816,
      "learning_rate": 0.00016227877804881127,
      "loss": 5.324,
      "step": 3575
    },
    {
      "epoch": 0.3576,
      "grad_norm": 2.264357328414917,
      "learning_rate": 0.00016225146366376198,
      "loss": 5.0734,
      "step": 3576
    },
    {
      "epoch": 0.3577,
      "grad_norm": 2.167836904525757,
      "learning_rate": 0.00016222414169356065,
      "loss": 4.7463,
      "step": 3577
    },
    {
      "epoch": 0.3578,
      "grad_norm": 2.2981977462768555,
      "learning_rate": 0.00016219681214153643,
      "loss": 5.2295,
      "step": 3578
    },
    {
      "epoch": 0.3579,
      "grad_norm": 2.015456438064575,
      "learning_rate": 0.0001621694750110193,
      "loss": 5.0069,
      "step": 3579
    },
    {
      "epoch": 0.358,
      "grad_norm": 2.952484130859375,
      "learning_rate": 0.0001621421303053402,
      "loss": 5.1195,
      "step": 3580
    },
    {
      "epoch": 0.3581,
      "grad_norm": 3.1708967685699463,
      "learning_rate": 0.00016211477802783103,
      "loss": 4.9081,
      "step": 3581
    },
    {
      "epoch": 0.3582,
      "grad_norm": 2.9304380416870117,
      "learning_rate": 0.0001620874181818246,
      "loss": 5.3903,
      "step": 3582
    },
    {
      "epoch": 0.3583,
      "grad_norm": 2.1909635066986084,
      "learning_rate": 0.00016206005077065458,
      "loss": 5.2091,
      "step": 3583
    },
    {
      "epoch": 0.3584,
      "grad_norm": 9.5924711227417,
      "learning_rate": 0.00016203267579765563,
      "loss": 5.7326,
      "step": 3584
    },
    {
      "epoch": 0.3585,
      "grad_norm": 7.063661575317383,
      "learning_rate": 0.00016200529326616328,
      "loss": 5.9366,
      "step": 3585
    },
    {
      "epoch": 0.3586,
      "grad_norm": 2.7811996936798096,
      "learning_rate": 0.00016197790317951403,
      "loss": 4.7836,
      "step": 3586
    },
    {
      "epoch": 0.3587,
      "grad_norm": 2.3904123306274414,
      "learning_rate": 0.00016195050554104528,
      "loss": 5.6714,
      "step": 3587
    },
    {
      "epoch": 0.3588,
      "grad_norm": 9.316290855407715,
      "learning_rate": 0.00016192310035409536,
      "loss": 5.3603,
      "step": 3588
    },
    {
      "epoch": 0.3589,
      "grad_norm": 5.986927509307861,
      "learning_rate": 0.00016189568762200348,
      "loss": 4.9936,
      "step": 3589
    },
    {
      "epoch": 0.359,
      "grad_norm": 4.152308464050293,
      "learning_rate": 0.00016186826734810979,
      "loss": 5.3593,
      "step": 3590
    },
    {
      "epoch": 0.3591,
      "grad_norm": 6.153663635253906,
      "learning_rate": 0.0001618408395357554,
      "loss": 4.9823,
      "step": 3591
    },
    {
      "epoch": 0.3592,
      "grad_norm": 4.461214542388916,
      "learning_rate": 0.00016181340418828233,
      "loss": 5.2158,
      "step": 3592
    },
    {
      "epoch": 0.3593,
      "grad_norm": 3.5639281272888184,
      "learning_rate": 0.00016178596130903344,
      "loss": 5.006,
      "step": 3593
    },
    {
      "epoch": 0.3594,
      "grad_norm": 3.2112529277801514,
      "learning_rate": 0.0001617585109013526,
      "loss": 5.3463,
      "step": 3594
    },
    {
      "epoch": 0.3595,
      "grad_norm": 5.7212090492248535,
      "learning_rate": 0.00016173105296858452,
      "loss": 5.3574,
      "step": 3595
    },
    {
      "epoch": 0.3596,
      "grad_norm": 2.400697946548462,
      "learning_rate": 0.00016170358751407487,
      "loss": 5.12,
      "step": 3596
    },
    {
      "epoch": 0.3597,
      "grad_norm": 3.009261131286621,
      "learning_rate": 0.00016167611454117025,
      "loss": 4.8147,
      "step": 3597
    },
    {
      "epoch": 0.3598,
      "grad_norm": 2.3669962882995605,
      "learning_rate": 0.0001616486340532182,
      "loss": 4.5938,
      "step": 3598
    },
    {
      "epoch": 0.3599,
      "grad_norm": 2.47689151763916,
      "learning_rate": 0.00016162114605356703,
      "loss": 5.0174,
      "step": 3599
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.23789644241333,
      "learning_rate": 0.0001615936505455662,
      "loss": 4.9796,
      "step": 3600
    },
    {
      "epoch": 0.3601,
      "grad_norm": 2.8438515663146973,
      "learning_rate": 0.0001615661475325658,
      "loss": 5.2009,
      "step": 3601
    },
    {
      "epoch": 0.3602,
      "grad_norm": 4.33348274230957,
      "learning_rate": 0.00016153863701791717,
      "loss": 4.9296,
      "step": 3602
    },
    {
      "epoch": 0.3603,
      "grad_norm": 2.089247226715088,
      "learning_rate": 0.00016151111900497225,
      "loss": 5.146,
      "step": 3603
    },
    {
      "epoch": 0.3604,
      "grad_norm": 2.0882651805877686,
      "learning_rate": 0.00016148359349708402,
      "loss": 5.0246,
      "step": 3604
    },
    {
      "epoch": 0.3605,
      "grad_norm": 6.604158401489258,
      "learning_rate": 0.00016145606049760644,
      "loss": 4.5566,
      "step": 3605
    },
    {
      "epoch": 0.3606,
      "grad_norm": 2.0595641136169434,
      "learning_rate": 0.00016142852000989433,
      "loss": 5.7383,
      "step": 3606
    },
    {
      "epoch": 0.3607,
      "grad_norm": 3.2369625568389893,
      "learning_rate": 0.00016140097203730337,
      "loss": 4.9206,
      "step": 3607
    },
    {
      "epoch": 0.3608,
      "grad_norm": 2.997403383255005,
      "learning_rate": 0.00016137341658319023,
      "loss": 4.6578,
      "step": 3608
    },
    {
      "epoch": 0.3609,
      "grad_norm": 2.1718788146972656,
      "learning_rate": 0.00016134585365091243,
      "loss": 4.8998,
      "step": 3609
    },
    {
      "epoch": 0.361,
      "grad_norm": 5.279642581939697,
      "learning_rate": 0.00016131828324382846,
      "loss": 5.5381,
      "step": 3610
    },
    {
      "epoch": 0.3611,
      "grad_norm": 2.3427953720092773,
      "learning_rate": 0.00016129070536529766,
      "loss": 5.1202,
      "step": 3611
    },
    {
      "epoch": 0.3612,
      "grad_norm": 4.4410881996154785,
      "learning_rate": 0.00016126312001868033,
      "loss": 5.7553,
      "step": 3612
    },
    {
      "epoch": 0.3613,
      "grad_norm": 2.5126290321350098,
      "learning_rate": 0.00016123552720733765,
      "loss": 5.4262,
      "step": 3613
    },
    {
      "epoch": 0.3614,
      "grad_norm": 2.7511088848114014,
      "learning_rate": 0.00016120792693463174,
      "loss": 5.4608,
      "step": 3614
    },
    {
      "epoch": 0.3615,
      "grad_norm": 7.563016414642334,
      "learning_rate": 0.00016118031920392558,
      "loss": 4.6838,
      "step": 3615
    },
    {
      "epoch": 0.3616,
      "grad_norm": 3.433018445968628,
      "learning_rate": 0.00016115270401858314,
      "loss": 5.545,
      "step": 3616
    },
    {
      "epoch": 0.3617,
      "grad_norm": 2.5811755657196045,
      "learning_rate": 0.00016112508138196917,
      "loss": 4.9552,
      "step": 3617
    },
    {
      "epoch": 0.3618,
      "grad_norm": 3.014063835144043,
      "learning_rate": 0.00016109745129744946,
      "loss": 4.9262,
      "step": 3618
    },
    {
      "epoch": 0.3619,
      "grad_norm": 2.3559787273406982,
      "learning_rate": 0.00016106981376839066,
      "loss": 5.7144,
      "step": 3619
    },
    {
      "epoch": 0.362,
      "grad_norm": 5.113919734954834,
      "learning_rate": 0.00016104216879816026,
      "loss": 5.1111,
      "step": 3620
    },
    {
      "epoch": 0.3621,
      "grad_norm": 3.4931135177612305,
      "learning_rate": 0.0001610145163901268,
      "loss": 5.2648,
      "step": 3621
    },
    {
      "epoch": 0.3622,
      "grad_norm": 3.981438636779785,
      "learning_rate": 0.00016098685654765955,
      "loss": 4.9781,
      "step": 3622
    },
    {
      "epoch": 0.3623,
      "grad_norm": 3.9873435497283936,
      "learning_rate": 0.0001609591892741288,
      "loss": 6.0583,
      "step": 3623
    },
    {
      "epoch": 0.3624,
      "grad_norm": 3.3586623668670654,
      "learning_rate": 0.0001609315145729058,
      "loss": 5.2381,
      "step": 3624
    },
    {
      "epoch": 0.3625,
      "grad_norm": 4.269761085510254,
      "learning_rate": 0.00016090383244736256,
      "loss": 5.8407,
      "step": 3625
    },
    {
      "epoch": 0.3626,
      "grad_norm": 5.87043571472168,
      "learning_rate": 0.00016087614290087208,
      "loss": 4.9305,
      "step": 3626
    },
    {
      "epoch": 0.3627,
      "grad_norm": 3.5850822925567627,
      "learning_rate": 0.0001608484459368082,
      "loss": 5.2166,
      "step": 3627
    },
    {
      "epoch": 0.3628,
      "grad_norm": 4.290431022644043,
      "learning_rate": 0.00016082074155854582,
      "loss": 4.8499,
      "step": 3628
    },
    {
      "epoch": 0.3629,
      "grad_norm": 2.4786152839660645,
      "learning_rate": 0.00016079302976946055,
      "loss": 5.0224,
      "step": 3629
    },
    {
      "epoch": 0.363,
      "grad_norm": 6.956276893615723,
      "learning_rate": 0.000160765310572929,
      "loss": 5.566,
      "step": 3630
    },
    {
      "epoch": 0.3631,
      "grad_norm": 3.232236623764038,
      "learning_rate": 0.00016073758397232868,
      "loss": 4.8835,
      "step": 3631
    },
    {
      "epoch": 0.3632,
      "grad_norm": 4.624008655548096,
      "learning_rate": 0.000160709849971038,
      "loss": 5.4244,
      "step": 3632
    },
    {
      "epoch": 0.3633,
      "grad_norm": 4.430325031280518,
      "learning_rate": 0.00016068210857243624,
      "loss": 5.1925,
      "step": 3633
    },
    {
      "epoch": 0.3634,
      "grad_norm": 9.412099838256836,
      "learning_rate": 0.0001606543597799036,
      "loss": 5.4219,
      "step": 3634
    },
    {
      "epoch": 0.3635,
      "grad_norm": 3.977483034133911,
      "learning_rate": 0.00016062660359682124,
      "loss": 6.0134,
      "step": 3635
    },
    {
      "epoch": 0.3636,
      "grad_norm": 6.261490821838379,
      "learning_rate": 0.0001605988400265711,
      "loss": 5.4813,
      "step": 3636
    },
    {
      "epoch": 0.3637,
      "grad_norm": 2.190645933151245,
      "learning_rate": 0.00016057106907253616,
      "loss": 5.5588,
      "step": 3637
    },
    {
      "epoch": 0.3638,
      "grad_norm": 3.5088460445404053,
      "learning_rate": 0.00016054329073810015,
      "loss": 5.9639,
      "step": 3638
    },
    {
      "epoch": 0.3639,
      "grad_norm": 5.262531757354736,
      "learning_rate": 0.0001605155050266478,
      "loss": 5.4472,
      "step": 3639
    },
    {
      "epoch": 0.364,
      "grad_norm": 7.4259514808654785,
      "learning_rate": 0.00016048771194156477,
      "loss": 5.2979,
      "step": 3640
    },
    {
      "epoch": 0.3641,
      "grad_norm": 2.4743971824645996,
      "learning_rate": 0.0001604599114862375,
      "loss": 5.3887,
      "step": 3641
    },
    {
      "epoch": 0.3642,
      "grad_norm": 2.5419881343841553,
      "learning_rate": 0.0001604321036640534,
      "loss": 5.0254,
      "step": 3642
    },
    {
      "epoch": 0.3643,
      "grad_norm": 3.360002279281616,
      "learning_rate": 0.00016040428847840079,
      "loss": 5.5487,
      "step": 3643
    },
    {
      "epoch": 0.3644,
      "grad_norm": 4.581014156341553,
      "learning_rate": 0.00016037646593266883,
      "loss": 5.1805,
      "step": 3644
    },
    {
      "epoch": 0.3645,
      "grad_norm": 2.5815391540527344,
      "learning_rate": 0.00016034863603024767,
      "loss": 4.8534,
      "step": 3645
    },
    {
      "epoch": 0.3646,
      "grad_norm": 3.8499138355255127,
      "learning_rate": 0.00016032079877452825,
      "loss": 5.1685,
      "step": 3646
    },
    {
      "epoch": 0.3647,
      "grad_norm": 3.38126277923584,
      "learning_rate": 0.00016029295416890248,
      "loss": 5.258,
      "step": 3647
    },
    {
      "epoch": 0.3648,
      "grad_norm": 1.9035248756408691,
      "learning_rate": 0.00016026510221676311,
      "loss": 4.9426,
      "step": 3648
    },
    {
      "epoch": 0.3649,
      "grad_norm": 2.3849239349365234,
      "learning_rate": 0.00016023724292150385,
      "loss": 5.3367,
      "step": 3649
    },
    {
      "epoch": 0.365,
      "grad_norm": 8.598565101623535,
      "learning_rate": 0.00016020937628651927,
      "loss": 5.3966,
      "step": 3650
    },
    {
      "epoch": 0.3651,
      "grad_norm": 9.501439094543457,
      "learning_rate": 0.00016018150231520486,
      "loss": 5.4415,
      "step": 3651
    },
    {
      "epoch": 0.3652,
      "grad_norm": 4.910306930541992,
      "learning_rate": 0.0001601536210109569,
      "loss": 5.6084,
      "step": 3652
    },
    {
      "epoch": 0.3653,
      "grad_norm": 2.8988966941833496,
      "learning_rate": 0.0001601257323771727,
      "loss": 4.6897,
      "step": 3653
    },
    {
      "epoch": 0.3654,
      "grad_norm": 2.4313228130340576,
      "learning_rate": 0.00016009783641725034,
      "loss": 5.0007,
      "step": 3654
    },
    {
      "epoch": 0.3655,
      "grad_norm": 2.8105950355529785,
      "learning_rate": 0.00016006993313458896,
      "loss": 4.4548,
      "step": 3655
    },
    {
      "epoch": 0.3656,
      "grad_norm": 5.740596294403076,
      "learning_rate": 0.00016004202253258842,
      "loss": 5.2809,
      "step": 3656
    },
    {
      "epoch": 0.3657,
      "grad_norm": 2.311398983001709,
      "learning_rate": 0.00016001410461464956,
      "loss": 5.0041,
      "step": 3657
    },
    {
      "epoch": 0.3658,
      "grad_norm": 2.212108850479126,
      "learning_rate": 0.0001599861793841741,
      "loss": 4.8928,
      "step": 3658
    },
    {
      "epoch": 0.3659,
      "grad_norm": 2.235654592514038,
      "learning_rate": 0.00015995824684456465,
      "loss": 5.3212,
      "step": 3659
    },
    {
      "epoch": 0.366,
      "grad_norm": 2.8953440189361572,
      "learning_rate": 0.00015993030699922468,
      "loss": 4.8847,
      "step": 3660
    },
    {
      "epoch": 0.3661,
      "grad_norm": 4.066795349121094,
      "learning_rate": 0.0001599023598515586,
      "loss": 4.4208,
      "step": 3661
    },
    {
      "epoch": 0.3662,
      "grad_norm": 2.568021535873413,
      "learning_rate": 0.00015987440540497167,
      "loss": 5.1919,
      "step": 3662
    },
    {
      "epoch": 0.3663,
      "grad_norm": 5.979730606079102,
      "learning_rate": 0.00015984644366287006,
      "loss": 4.924,
      "step": 3663
    },
    {
      "epoch": 0.3664,
      "grad_norm": 2.6440131664276123,
      "learning_rate": 0.00015981847462866083,
      "loss": 5.0508,
      "step": 3664
    },
    {
      "epoch": 0.3665,
      "grad_norm": 4.940854072570801,
      "learning_rate": 0.0001597904983057519,
      "loss": 5.3266,
      "step": 3665
    },
    {
      "epoch": 0.3666,
      "grad_norm": 1.9226044416427612,
      "learning_rate": 0.00015976251469755214,
      "loss": 4.7823,
      "step": 3666
    },
    {
      "epoch": 0.3667,
      "grad_norm": 4.114530563354492,
      "learning_rate": 0.00015973452380747122,
      "loss": 5.0803,
      "step": 3667
    },
    {
      "epoch": 0.3668,
      "grad_norm": 3.259831190109253,
      "learning_rate": 0.00015970652563891978,
      "loss": 5.3962,
      "step": 3668
    },
    {
      "epoch": 0.3669,
      "grad_norm": 2.182724952697754,
      "learning_rate": 0.00015967852019530926,
      "loss": 4.6848,
      "step": 3669
    },
    {
      "epoch": 0.367,
      "grad_norm": 2.4359211921691895,
      "learning_rate": 0.00015965050748005215,
      "loss": 4.7988,
      "step": 3670
    },
    {
      "epoch": 0.3671,
      "grad_norm": 2.6070878505706787,
      "learning_rate": 0.0001596224874965616,
      "loss": 5.1683,
      "step": 3671
    },
    {
      "epoch": 0.3672,
      "grad_norm": 3.3089616298675537,
      "learning_rate": 0.0001595944602482518,
      "loss": 4.7978,
      "step": 3672
    },
    {
      "epoch": 0.3673,
      "grad_norm": 2.62555193901062,
      "learning_rate": 0.0001595664257385378,
      "loss": 4.6284,
      "step": 3673
    },
    {
      "epoch": 0.3674,
      "grad_norm": 3.4884986877441406,
      "learning_rate": 0.00015953838397083552,
      "loss": 5.1728,
      "step": 3674
    },
    {
      "epoch": 0.3675,
      "grad_norm": 3.0456483364105225,
      "learning_rate": 0.00015951033494856175,
      "loss": 5.2078,
      "step": 3675
    },
    {
      "epoch": 0.3676,
      "grad_norm": 2.7429397106170654,
      "learning_rate": 0.00015948227867513415,
      "loss": 5.0464,
      "step": 3676
    },
    {
      "epoch": 0.3677,
      "grad_norm": 2.9096193313598633,
      "learning_rate": 0.00015945421515397133,
      "loss": 5.7049,
      "step": 3677
    },
    {
      "epoch": 0.3678,
      "grad_norm": 6.685516834259033,
      "learning_rate": 0.00015942614438849275,
      "loss": 6.2006,
      "step": 3678
    },
    {
      "epoch": 0.3679,
      "grad_norm": 4.453888416290283,
      "learning_rate": 0.0001593980663821187,
      "loss": 5.1182,
      "step": 3679
    },
    {
      "epoch": 0.368,
      "grad_norm": 6.840296745300293,
      "learning_rate": 0.00015936998113827048,
      "loss": 5.2108,
      "step": 3680
    },
    {
      "epoch": 0.3681,
      "grad_norm": 7.4318671226501465,
      "learning_rate": 0.00015934188866037016,
      "loss": 5.1224,
      "step": 3681
    },
    {
      "epoch": 0.3682,
      "grad_norm": 2.798494338989258,
      "learning_rate": 0.0001593137889518407,
      "loss": 5.5555,
      "step": 3682
    },
    {
      "epoch": 0.3683,
      "grad_norm": 1.843361258506775,
      "learning_rate": 0.00015928568201610595,
      "loss": 4.9233,
      "step": 3683
    },
    {
      "epoch": 0.3684,
      "grad_norm": 2.724675178527832,
      "learning_rate": 0.00015925756785659069,
      "loss": 5.2289,
      "step": 3684
    },
    {
      "epoch": 0.3685,
      "grad_norm": 2.5268213748931885,
      "learning_rate": 0.00015922944647672052,
      "loss": 5.3021,
      "step": 3685
    },
    {
      "epoch": 0.3686,
      "grad_norm": 3.6625473499298096,
      "learning_rate": 0.00015920131787992197,
      "loss": 4.6296,
      "step": 3686
    },
    {
      "epoch": 0.3687,
      "grad_norm": 2.4323832988739014,
      "learning_rate": 0.0001591731820696224,
      "loss": 4.8655,
      "step": 3687
    },
    {
      "epoch": 0.3688,
      "grad_norm": 3.0710155963897705,
      "learning_rate": 0.0001591450390492501,
      "loss": 5.3733,
      "step": 3688
    },
    {
      "epoch": 0.3689,
      "grad_norm": 3.676192045211792,
      "learning_rate": 0.0001591168888222342,
      "loss": 5.15,
      "step": 3689
    },
    {
      "epoch": 0.369,
      "grad_norm": 5.294004440307617,
      "learning_rate": 0.00015908873139200473,
      "loss": 5.3542,
      "step": 3690
    },
    {
      "epoch": 0.3691,
      "grad_norm": 2.5201871395111084,
      "learning_rate": 0.00015906056676199255,
      "loss": 4.9389,
      "step": 3691
    },
    {
      "epoch": 0.3692,
      "grad_norm": 11.237199783325195,
      "learning_rate": 0.00015903239493562948,
      "loss": 6.046,
      "step": 3692
    },
    {
      "epoch": 0.3693,
      "grad_norm": 11.23511028289795,
      "learning_rate": 0.00015900421591634814,
      "loss": 5.2012,
      "step": 3693
    },
    {
      "epoch": 0.3694,
      "grad_norm": 4.7550950050354,
      "learning_rate": 0.00015897602970758206,
      "loss": 5.1828,
      "step": 3694
    },
    {
      "epoch": 0.3695,
      "grad_norm": 5.544837951660156,
      "learning_rate": 0.00015894783631276567,
      "loss": 5.1079,
      "step": 3695
    },
    {
      "epoch": 0.3696,
      "grad_norm": 11.01254653930664,
      "learning_rate": 0.0001589196357353342,
      "loss": 5.2916,
      "step": 3696
    },
    {
      "epoch": 0.3697,
      "grad_norm": 2.7447803020477295,
      "learning_rate": 0.00015889142797872387,
      "loss": 4.9394,
      "step": 3697
    },
    {
      "epoch": 0.3698,
      "grad_norm": 2.541508197784424,
      "learning_rate": 0.0001588632130463717,
      "loss": 4.9599,
      "step": 3698
    },
    {
      "epoch": 0.3699,
      "grad_norm": 3.6154539585113525,
      "learning_rate": 0.00015883499094171554,
      "loss": 5.3113,
      "step": 3699
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.167060136795044,
      "learning_rate": 0.0001588067616681942,
      "loss": 5.2321,
      "step": 3700
    },
    {
      "epoch": 0.3701,
      "grad_norm": 7.4674530029296875,
      "learning_rate": 0.00015877852522924732,
      "loss": 4.9962,
      "step": 3701
    },
    {
      "epoch": 0.3702,
      "grad_norm": 3.815540313720703,
      "learning_rate": 0.00015875028162831546,
      "loss": 4.8706,
      "step": 3702
    },
    {
      "epoch": 0.3703,
      "grad_norm": 3.8099253177642822,
      "learning_rate": 0.00015872203086883996,
      "loss": 5.1789,
      "step": 3703
    },
    {
      "epoch": 0.3704,
      "grad_norm": 4.475642204284668,
      "learning_rate": 0.00015869377295426316,
      "loss": 4.8411,
      "step": 3704
    },
    {
      "epoch": 0.3705,
      "grad_norm": 3.801353693008423,
      "learning_rate": 0.00015866550788802813,
      "loss": 5.4727,
      "step": 3705
    },
    {
      "epoch": 0.3706,
      "grad_norm": 2.087735891342163,
      "learning_rate": 0.00015863723567357892,
      "loss": 5.3388,
      "step": 3706
    },
    {
      "epoch": 0.3707,
      "grad_norm": 4.215925693511963,
      "learning_rate": 0.00015860895631436043,
      "loss": 5.122,
      "step": 3707
    },
    {
      "epoch": 0.3708,
      "grad_norm": 3.082951784133911,
      "learning_rate": 0.0001585806698138184,
      "loss": 6.0007,
      "step": 3708
    },
    {
      "epoch": 0.3709,
      "grad_norm": 2.292431116104126,
      "learning_rate": 0.0001585523761753994,
      "loss": 5.2116,
      "step": 3709
    },
    {
      "epoch": 0.371,
      "grad_norm": 2.855778217315674,
      "learning_rate": 0.00015852407540255104,
      "loss": 5.4048,
      "step": 3710
    },
    {
      "epoch": 0.3711,
      "grad_norm": 3.6123569011688232,
      "learning_rate": 0.00015849576749872157,
      "loss": 5.0963,
      "step": 3711
    },
    {
      "epoch": 0.3712,
      "grad_norm": 5.270050525665283,
      "learning_rate": 0.00015846745246736026,
      "loss": 5.9991,
      "step": 3712
    },
    {
      "epoch": 0.3713,
      "grad_norm": 3.2015111446380615,
      "learning_rate": 0.00015843913031191723,
      "loss": 5.6971,
      "step": 3713
    },
    {
      "epoch": 0.3714,
      "grad_norm": 3.812079429626465,
      "learning_rate": 0.00015841080103584342,
      "loss": 5.4898,
      "step": 3714
    },
    {
      "epoch": 0.3715,
      "grad_norm": 2.6757030487060547,
      "learning_rate": 0.0001583824646425907,
      "loss": 5.3192,
      "step": 3715
    },
    {
      "epoch": 0.3716,
      "grad_norm": 12.51317024230957,
      "learning_rate": 0.00015835412113561175,
      "loss": 5.2933,
      "step": 3716
    },
    {
      "epoch": 0.3717,
      "grad_norm": 6.082845687866211,
      "learning_rate": 0.00015832577051836015,
      "loss": 5.139,
      "step": 3717
    },
    {
      "epoch": 0.3718,
      "grad_norm": 11.847944259643555,
      "learning_rate": 0.00015829741279429035,
      "loss": 5.7653,
      "step": 3718
    },
    {
      "epoch": 0.3719,
      "grad_norm": 11.289405822753906,
      "learning_rate": 0.00015826904796685762,
      "loss": 5.7411,
      "step": 3719
    },
    {
      "epoch": 0.372,
      "grad_norm": 7.888276100158691,
      "learning_rate": 0.00015824067603951812,
      "loss": 5.1654,
      "step": 3720
    },
    {
      "epoch": 0.3721,
      "grad_norm": 3.6864919662475586,
      "learning_rate": 0.00015821229701572896,
      "loss": 5.0722,
      "step": 3721
    },
    {
      "epoch": 0.3722,
      "grad_norm": 8.867057800292969,
      "learning_rate": 0.00015818391089894796,
      "loss": 5.082,
      "step": 3722
    },
    {
      "epoch": 0.3723,
      "grad_norm": 2.3327531814575195,
      "learning_rate": 0.00015815551769263387,
      "loss": 5.2277,
      "step": 3723
    },
    {
      "epoch": 0.3724,
      "grad_norm": 3.1431665420532227,
      "learning_rate": 0.0001581271174002464,
      "loss": 5.566,
      "step": 3724
    },
    {
      "epoch": 0.3725,
      "grad_norm": 11.730887413024902,
      "learning_rate": 0.000158098710025246,
      "loss": 5.5558,
      "step": 3725
    },
    {
      "epoch": 0.3726,
      "grad_norm": 2.7340598106384277,
      "learning_rate": 0.00015807029557109398,
      "loss": 4.8048,
      "step": 3726
    },
    {
      "epoch": 0.3727,
      "grad_norm": 7.254354476928711,
      "learning_rate": 0.0001580418740412526,
      "loss": 4.9975,
      "step": 3727
    },
    {
      "epoch": 0.3728,
      "grad_norm": 3.4266302585601807,
      "learning_rate": 0.00015801344543918495,
      "loss": 4.8946,
      "step": 3728
    },
    {
      "epoch": 0.3729,
      "grad_norm": 2.702054738998413,
      "learning_rate": 0.00015798500976835493,
      "loss": 5.5409,
      "step": 3729
    },
    {
      "epoch": 0.373,
      "grad_norm": 2.6578311920166016,
      "learning_rate": 0.00015795656703222736,
      "loss": 4.816,
      "step": 3730
    },
    {
      "epoch": 0.3731,
      "grad_norm": 4.648913860321045,
      "learning_rate": 0.0001579281172342679,
      "loss": 5.4448,
      "step": 3731
    },
    {
      "epoch": 0.3732,
      "grad_norm": 4.267664909362793,
      "learning_rate": 0.00015789966037794306,
      "loss": 4.5269,
      "step": 3732
    },
    {
      "epoch": 0.3733,
      "grad_norm": 4.669946670532227,
      "learning_rate": 0.00015787119646672025,
      "loss": 4.8967,
      "step": 3733
    },
    {
      "epoch": 0.3734,
      "grad_norm": 3.04695200920105,
      "learning_rate": 0.00015784272550406765,
      "loss": 5.2689,
      "step": 3734
    },
    {
      "epoch": 0.3735,
      "grad_norm": 3.3445448875427246,
      "learning_rate": 0.00015781424749345446,
      "loss": 5.0331,
      "step": 3735
    },
    {
      "epoch": 0.3736,
      "grad_norm": 2.2963924407958984,
      "learning_rate": 0.00015778576243835054,
      "loss": 4.7924,
      "step": 3736
    },
    {
      "epoch": 0.3737,
      "grad_norm": 2.470039129257202,
      "learning_rate": 0.00015775727034222675,
      "loss": 5.006,
      "step": 3737
    },
    {
      "epoch": 0.3738,
      "grad_norm": 1.8831760883331299,
      "learning_rate": 0.0001577287712085548,
      "loss": 4.8748,
      "step": 3738
    },
    {
      "epoch": 0.3739,
      "grad_norm": 2.5469839572906494,
      "learning_rate": 0.00015770026504080718,
      "loss": 5.0958,
      "step": 3739
    },
    {
      "epoch": 0.374,
      "grad_norm": 3.3680710792541504,
      "learning_rate": 0.00015767175184245726,
      "loss": 4.3058,
      "step": 3740
    },
    {
      "epoch": 0.3741,
      "grad_norm": 2.651278495788574,
      "learning_rate": 0.00015764323161697935,
      "loss": 5.1438,
      "step": 3741
    },
    {
      "epoch": 0.3742,
      "grad_norm": 3.0414600372314453,
      "learning_rate": 0.00015761470436784846,
      "loss": 4.5916,
      "step": 3742
    },
    {
      "epoch": 0.3743,
      "grad_norm": 3.7437195777893066,
      "learning_rate": 0.0001575861700985407,
      "loss": 4.9848,
      "step": 3743
    },
    {
      "epoch": 0.3744,
      "grad_norm": 3.284501552581787,
      "learning_rate": 0.0001575576288125327,
      "loss": 5.2502,
      "step": 3744
    },
    {
      "epoch": 0.3745,
      "grad_norm": 8.963765144348145,
      "learning_rate": 0.00015752908051330227,
      "loss": 5.2107,
      "step": 3745
    },
    {
      "epoch": 0.3746,
      "grad_norm": 4.275839805603027,
      "learning_rate": 0.00015750052520432787,
      "loss": 5.3563,
      "step": 3746
    },
    {
      "epoch": 0.3747,
      "grad_norm": 2.007120370864868,
      "learning_rate": 0.00015747196288908887,
      "loss": 5.0136,
      "step": 3747
    },
    {
      "epoch": 0.3748,
      "grad_norm": 2.4656784534454346,
      "learning_rate": 0.00015744339357106558,
      "loss": 4.889,
      "step": 3748
    },
    {
      "epoch": 0.3749,
      "grad_norm": 2.4532129764556885,
      "learning_rate": 0.000157414817253739,
      "loss": 5.3251,
      "step": 3749
    },
    {
      "epoch": 0.375,
      "grad_norm": 3.974574327468872,
      "learning_rate": 0.00015738623394059107,
      "loss": 4.8373,
      "step": 3750
    },
    {
      "epoch": 0.3751,
      "grad_norm": 8.104355812072754,
      "learning_rate": 0.0001573576436351046,
      "loss": 5.7754,
      "step": 3751
    },
    {
      "epoch": 0.3752,
      "grad_norm": 2.3106672763824463,
      "learning_rate": 0.00015732904634076328,
      "loss": 5.4344,
      "step": 3752
    },
    {
      "epoch": 0.3753,
      "grad_norm": 3.899399757385254,
      "learning_rate": 0.00015730044206105156,
      "loss": 5.1091,
      "step": 3753
    },
    {
      "epoch": 0.3754,
      "grad_norm": 3.658705234527588,
      "learning_rate": 0.00015727183079945476,
      "loss": 5.5223,
      "step": 3754
    },
    {
      "epoch": 0.3755,
      "grad_norm": 3.779356002807617,
      "learning_rate": 0.0001572432125594591,
      "loss": 4.3197,
      "step": 3755
    },
    {
      "epoch": 0.3756,
      "grad_norm": 1.8226667642593384,
      "learning_rate": 0.00015721458734455163,
      "loss": 4.9961,
      "step": 3756
    },
    {
      "epoch": 0.3757,
      "grad_norm": 2.9064929485321045,
      "learning_rate": 0.00015718595515822027,
      "loss": 6.1111,
      "step": 3757
    },
    {
      "epoch": 0.3758,
      "grad_norm": 3.499368906021118,
      "learning_rate": 0.0001571573160039537,
      "loss": 5.5374,
      "step": 3758
    },
    {
      "epoch": 0.3759,
      "grad_norm": 4.7730889320373535,
      "learning_rate": 0.00015712866988524155,
      "loss": 5.1414,
      "step": 3759
    },
    {
      "epoch": 0.376,
      "grad_norm": 3.038071870803833,
      "learning_rate": 0.0001571000168055743,
      "loss": 5.7849,
      "step": 3760
    },
    {
      "epoch": 0.3761,
      "grad_norm": 2.327759265899658,
      "learning_rate": 0.0001570713567684432,
      "loss": 5.0416,
      "step": 3761
    },
    {
      "epoch": 0.3762,
      "grad_norm": 3.7375190258026123,
      "learning_rate": 0.00015704268977734037,
      "loss": 5.0002,
      "step": 3762
    },
    {
      "epoch": 0.3763,
      "grad_norm": 4.122392654418945,
      "learning_rate": 0.00015701401583575884,
      "loss": 5.5389,
      "step": 3763
    },
    {
      "epoch": 0.3764,
      "grad_norm": 2.5517654418945312,
      "learning_rate": 0.00015698533494719238,
      "loss": 4.906,
      "step": 3764
    },
    {
      "epoch": 0.3765,
      "grad_norm": 4.555659770965576,
      "learning_rate": 0.00015695664711513576,
      "loss": 4.8673,
      "step": 3765
    },
    {
      "epoch": 0.3766,
      "grad_norm": 7.7634687423706055,
      "learning_rate": 0.00015692795234308445,
      "loss": 5.1197,
      "step": 3766
    },
    {
      "epoch": 0.3767,
      "grad_norm": 2.44838285446167,
      "learning_rate": 0.0001568992506345348,
      "loss": 5.2743,
      "step": 3767
    },
    {
      "epoch": 0.3768,
      "grad_norm": 3.722210645675659,
      "learning_rate": 0.00015687054199298408,
      "loss": 5.2153,
      "step": 3768
    },
    {
      "epoch": 0.3769,
      "grad_norm": 5.057201385498047,
      "learning_rate": 0.0001568418264219303,
      "loss": 5.1856,
      "step": 3769
    },
    {
      "epoch": 0.377,
      "grad_norm": 2.5678951740264893,
      "learning_rate": 0.0001568131039248724,
      "loss": 5.7753,
      "step": 3770
    },
    {
      "epoch": 0.3771,
      "grad_norm": 3.3541321754455566,
      "learning_rate": 0.00015678437450531013,
      "loss": 5.7539,
      "step": 3771
    },
    {
      "epoch": 0.3772,
      "grad_norm": 6.0282979011535645,
      "learning_rate": 0.00015675563816674407,
      "loss": 4.7407,
      "step": 3772
    },
    {
      "epoch": 0.3773,
      "grad_norm": 2.120327949523926,
      "learning_rate": 0.00015672689491267567,
      "loss": 5.4274,
      "step": 3773
    },
    {
      "epoch": 0.3774,
      "grad_norm": 3.674077033996582,
      "learning_rate": 0.00015669814474660718,
      "loss": 5.0303,
      "step": 3774
    },
    {
      "epoch": 0.3775,
      "grad_norm": 2.2021591663360596,
      "learning_rate": 0.0001566693876720417,
      "loss": 5.2482,
      "step": 3775
    },
    {
      "epoch": 0.3776,
      "grad_norm": 2.982666254043579,
      "learning_rate": 0.00015664062369248328,
      "loss": 4.7254,
      "step": 3776
    },
    {
      "epoch": 0.3777,
      "grad_norm": 16.885377883911133,
      "learning_rate": 0.00015661185281143667,
      "loss": 5.9493,
      "step": 3777
    },
    {
      "epoch": 0.3778,
      "grad_norm": 5.198866844177246,
      "learning_rate": 0.0001565830750324075,
      "loss": 6.2668,
      "step": 3778
    },
    {
      "epoch": 0.3779,
      "grad_norm": 2.7920007705688477,
      "learning_rate": 0.0001565542903589023,
      "loss": 5.2619,
      "step": 3779
    },
    {
      "epoch": 0.378,
      "grad_norm": 3.0053365230560303,
      "learning_rate": 0.00015652549879442834,
      "loss": 5.4437,
      "step": 3780
    },
    {
      "epoch": 0.3781,
      "grad_norm": 2.7807822227478027,
      "learning_rate": 0.0001564967003424938,
      "loss": 4.9223,
      "step": 3781
    },
    {
      "epoch": 0.3782,
      "grad_norm": 4.265754222869873,
      "learning_rate": 0.00015646789500660773,
      "loss": 5.6888,
      "step": 3782
    },
    {
      "epoch": 0.3783,
      "grad_norm": 4.275457382202148,
      "learning_rate": 0.00015643908279027992,
      "loss": 5.7355,
      "step": 3783
    },
    {
      "epoch": 0.3784,
      "grad_norm": 3.921895980834961,
      "learning_rate": 0.00015641026369702106,
      "loss": 4.4738,
      "step": 3784
    },
    {
      "epoch": 0.3785,
      "grad_norm": 3.896526336669922,
      "learning_rate": 0.00015638143773034267,
      "loss": 4.7919,
      "step": 3785
    },
    {
      "epoch": 0.3786,
      "grad_norm": 4.408327102661133,
      "learning_rate": 0.00015635260489375714,
      "loss": 5.3678,
      "step": 3786
    },
    {
      "epoch": 0.3787,
      "grad_norm": 4.3397979736328125,
      "learning_rate": 0.00015632376519077767,
      "loss": 4.7899,
      "step": 3787
    },
    {
      "epoch": 0.3788,
      "grad_norm": 2.13382887840271,
      "learning_rate": 0.0001562949186249182,
      "loss": 4.8866,
      "step": 3788
    },
    {
      "epoch": 0.3789,
      "grad_norm": 3.776658296585083,
      "learning_rate": 0.0001562660651996937,
      "loss": 5.6336,
      "step": 3789
    },
    {
      "epoch": 0.379,
      "grad_norm": 2.545053243637085,
      "learning_rate": 0.0001562372049186198,
      "loss": 5.2176,
      "step": 3790
    },
    {
      "epoch": 0.3791,
      "grad_norm": 2.366645097732544,
      "learning_rate": 0.00015620833778521307,
      "loss": 5.3674,
      "step": 3791
    },
    {
      "epoch": 0.3792,
      "grad_norm": 4.276769161224365,
      "learning_rate": 0.00015617946380299088,
      "loss": 5.153,
      "step": 3792
    },
    {
      "epoch": 0.3793,
      "grad_norm": 4.775831699371338,
      "learning_rate": 0.00015615058297547145,
      "loss": 5.2092,
      "step": 3793
    },
    {
      "epoch": 0.3794,
      "grad_norm": 5.1017746925354,
      "learning_rate": 0.0001561216953061738,
      "loss": 5.5667,
      "step": 3794
    },
    {
      "epoch": 0.3795,
      "grad_norm": 2.721372604370117,
      "learning_rate": 0.0001560928007986178,
      "loss": 4.3455,
      "step": 3795
    },
    {
      "epoch": 0.3796,
      "grad_norm": 3.366901159286499,
      "learning_rate": 0.0001560638994563242,
      "loss": 5.1615,
      "step": 3796
    },
    {
      "epoch": 0.3797,
      "grad_norm": 4.656524658203125,
      "learning_rate": 0.00015603499128281448,
      "loss": 5.5335,
      "step": 3797
    },
    {
      "epoch": 0.3798,
      "grad_norm": 2.4845221042633057,
      "learning_rate": 0.00015600607628161103,
      "loss": 4.6716,
      "step": 3798
    },
    {
      "epoch": 0.3799,
      "grad_norm": 3.8967854976654053,
      "learning_rate": 0.00015597715445623712,
      "loss": 6.6253,
      "step": 3799
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.95802903175354,
      "learning_rate": 0.0001559482258102167,
      "loss": 4.8608,
      "step": 3800
    },
    {
      "epoch": 0.3801,
      "grad_norm": 3.9017045497894287,
      "learning_rate": 0.0001559192903470747,
      "loss": 4.6928,
      "step": 3801
    },
    {
      "epoch": 0.3802,
      "grad_norm": 3.4243900775909424,
      "learning_rate": 0.00015589034807033677,
      "loss": 5.0477,
      "step": 3802
    },
    {
      "epoch": 0.3803,
      "grad_norm": 2.1956231594085693,
      "learning_rate": 0.00015586139898352946,
      "loss": 5.1478,
      "step": 3803
    },
    {
      "epoch": 0.3804,
      "grad_norm": 2.036834239959717,
      "learning_rate": 0.00015583244309018014,
      "loss": 5.0927,
      "step": 3804
    },
    {
      "epoch": 0.3805,
      "grad_norm": 4.418098449707031,
      "learning_rate": 0.000155803480393817,
      "loss": 7.2346,
      "step": 3805
    },
    {
      "epoch": 0.3806,
      "grad_norm": 2.396549940109253,
      "learning_rate": 0.00015577451089796905,
      "loss": 4.9914,
      "step": 3806
    },
    {
      "epoch": 0.3807,
      "grad_norm": 2.847869873046875,
      "learning_rate": 0.00015574553460616608,
      "loss": 4.9775,
      "step": 3807
    },
    {
      "epoch": 0.3808,
      "grad_norm": 2.3236513137817383,
      "learning_rate": 0.00015571655152193885,
      "loss": 4.9699,
      "step": 3808
    },
    {
      "epoch": 0.3809,
      "grad_norm": 4.825192928314209,
      "learning_rate": 0.00015568756164881882,
      "loss": 4.761,
      "step": 3809
    },
    {
      "epoch": 0.381,
      "grad_norm": 2.2276203632354736,
      "learning_rate": 0.00015565856499033832,
      "loss": 5.0272,
      "step": 3810
    },
    {
      "epoch": 0.3811,
      "grad_norm": 2.874829053878784,
      "learning_rate": 0.0001556295615500305,
      "loss": 6.2243,
      "step": 3811
    },
    {
      "epoch": 0.3812,
      "grad_norm": 5.846498966217041,
      "learning_rate": 0.0001556005513314293,
      "loss": 4.7994,
      "step": 3812
    },
    {
      "epoch": 0.3813,
      "grad_norm": 8.417187690734863,
      "learning_rate": 0.00015557153433806966,
      "loss": 5.4513,
      "step": 3813
    },
    {
      "epoch": 0.3814,
      "grad_norm": 3.701296329498291,
      "learning_rate": 0.00015554251057348713,
      "loss": 5.4759,
      "step": 3814
    },
    {
      "epoch": 0.3815,
      "grad_norm": 2.01570463180542,
      "learning_rate": 0.0001555134800412181,
      "loss": 4.9966,
      "step": 3815
    },
    {
      "epoch": 0.3816,
      "grad_norm": 4.555514335632324,
      "learning_rate": 0.00015548444274479995,
      "loss": 5.6655,
      "step": 3816
    },
    {
      "epoch": 0.3817,
      "grad_norm": 3.5984373092651367,
      "learning_rate": 0.00015545539868777074,
      "loss": 4.965,
      "step": 3817
    },
    {
      "epoch": 0.3818,
      "grad_norm": 3.245405435562134,
      "learning_rate": 0.00015542634787366942,
      "loss": 5.3994,
      "step": 3818
    },
    {
      "epoch": 0.3819,
      "grad_norm": 3.611888885498047,
      "learning_rate": 0.00015539729030603574,
      "loss": 5.5919,
      "step": 3819
    },
    {
      "epoch": 0.382,
      "grad_norm": 10.671337127685547,
      "learning_rate": 0.00015536822598841024,
      "loss": 5.4967,
      "step": 3820
    },
    {
      "epoch": 0.3821,
      "grad_norm": 2.5737757682800293,
      "learning_rate": 0.00015533915492433443,
      "loss": 4.8449,
      "step": 3821
    },
    {
      "epoch": 0.3822,
      "grad_norm": 2.3797805309295654,
      "learning_rate": 0.0001553100771173504,
      "loss": 4.5447,
      "step": 3822
    },
    {
      "epoch": 0.3823,
      "grad_norm": 5.670876979827881,
      "learning_rate": 0.00015528099257100127,
      "loss": 5.5044,
      "step": 3823
    },
    {
      "epoch": 0.3824,
      "grad_norm": 4.270078659057617,
      "learning_rate": 0.00015525190128883083,
      "loss": 5.2083,
      "step": 3824
    },
    {
      "epoch": 0.3825,
      "grad_norm": 2.4702718257904053,
      "learning_rate": 0.00015522280327438388,
      "loss": 4.7223,
      "step": 3825
    },
    {
      "epoch": 0.3826,
      "grad_norm": 13.328025817871094,
      "learning_rate": 0.0001551936985312058,
      "loss": 5.4872,
      "step": 3826
    },
    {
      "epoch": 0.3827,
      "grad_norm": 2.5737252235412598,
      "learning_rate": 0.00015516458706284303,
      "loss": 5.0423,
      "step": 3827
    },
    {
      "epoch": 0.3828,
      "grad_norm": 3.20526123046875,
      "learning_rate": 0.00015513546887284264,
      "loss": 5.1179,
      "step": 3828
    },
    {
      "epoch": 0.3829,
      "grad_norm": 3.8233463764190674,
      "learning_rate": 0.0001551063439647526,
      "loss": 5.5874,
      "step": 3829
    },
    {
      "epoch": 0.383,
      "grad_norm": 2.6672754287719727,
      "learning_rate": 0.00015507721234212172,
      "loss": 5.153,
      "step": 3830
    },
    {
      "epoch": 0.3831,
      "grad_norm": 3.500835418701172,
      "learning_rate": 0.00015504807400849958,
      "loss": 4.8636,
      "step": 3831
    },
    {
      "epoch": 0.3832,
      "grad_norm": 5.124509334564209,
      "learning_rate": 0.0001550189289674366,
      "loss": 4.8462,
      "step": 3832
    },
    {
      "epoch": 0.3833,
      "grad_norm": 12.170418739318848,
      "learning_rate": 0.000154989777222484,
      "loss": 5.8832,
      "step": 3833
    },
    {
      "epoch": 0.3834,
      "grad_norm": 10.904394149780273,
      "learning_rate": 0.00015496061877719384,
      "loss": 5.9908,
      "step": 3834
    },
    {
      "epoch": 0.3835,
      "grad_norm": 11.345632553100586,
      "learning_rate": 0.000154931453635119,
      "loss": 5.7716,
      "step": 3835
    },
    {
      "epoch": 0.3836,
      "grad_norm": 17.36412239074707,
      "learning_rate": 0.0001549022817998132,
      "loss": 5.3048,
      "step": 3836
    },
    {
      "epoch": 0.3837,
      "grad_norm": 12.305880546569824,
      "learning_rate": 0.00015487310327483086,
      "loss": 5.3279,
      "step": 3837
    },
    {
      "epoch": 0.3838,
      "grad_norm": 3.5326995849609375,
      "learning_rate": 0.00015484391806372733,
      "loss": 5.2178,
      "step": 3838
    },
    {
      "epoch": 0.3839,
      "grad_norm": 2.4750537872314453,
      "learning_rate": 0.00015481472617005876,
      "loss": 5.4803,
      "step": 3839
    },
    {
      "epoch": 0.384,
      "grad_norm": 3.644608736038208,
      "learning_rate": 0.00015478552759738207,
      "loss": 5.1735,
      "step": 3840
    },
    {
      "epoch": 0.3841,
      "grad_norm": 3.9475350379943848,
      "learning_rate": 0.00015475632234925504,
      "loss": 4.8659,
      "step": 3841
    },
    {
      "epoch": 0.3842,
      "grad_norm": 3.7509853839874268,
      "learning_rate": 0.00015472711042923621,
      "loss": 4.6245,
      "step": 3842
    },
    {
      "epoch": 0.3843,
      "grad_norm": 5.6492109298706055,
      "learning_rate": 0.00015469789184088497,
      "loss": 5.9603,
      "step": 3843
    },
    {
      "epoch": 0.3844,
      "grad_norm": 3.924812078475952,
      "learning_rate": 0.00015466866658776155,
      "loss": 5.3648,
      "step": 3844
    },
    {
      "epoch": 0.3845,
      "grad_norm": 3.3575003147125244,
      "learning_rate": 0.00015463943467342693,
      "loss": 5.7395,
      "step": 3845
    },
    {
      "epoch": 0.3846,
      "grad_norm": 4.435554027557373,
      "learning_rate": 0.0001546101961014429,
      "loss": 5.0751,
      "step": 3846
    },
    {
      "epoch": 0.3847,
      "grad_norm": 3.6977744102478027,
      "learning_rate": 0.00015458095087537218,
      "loss": 5.6692,
      "step": 3847
    },
    {
      "epoch": 0.3848,
      "grad_norm": 5.843740463256836,
      "learning_rate": 0.00015455169899877813,
      "loss": 5.2639,
      "step": 3848
    },
    {
      "epoch": 0.3849,
      "grad_norm": 2.0758132934570312,
      "learning_rate": 0.00015452244047522502,
      "loss": 4.9506,
      "step": 3849
    },
    {
      "epoch": 0.385,
      "grad_norm": 3.487727642059326,
      "learning_rate": 0.00015449317530827794,
      "loss": 4.9612,
      "step": 3850
    },
    {
      "epoch": 0.3851,
      "grad_norm": 3.7429425716400146,
      "learning_rate": 0.00015446390350150273,
      "loss": 5.3623,
      "step": 3851
    },
    {
      "epoch": 0.3852,
      "grad_norm": 1.995604395866394,
      "learning_rate": 0.0001544346250584661,
      "loss": 5.0756,
      "step": 3852
    },
    {
      "epoch": 0.3853,
      "grad_norm": 3.5590014457702637,
      "learning_rate": 0.00015440533998273547,
      "loss": 4.7961,
      "step": 3853
    },
    {
      "epoch": 0.3854,
      "grad_norm": 8.812822341918945,
      "learning_rate": 0.00015437604827787927,
      "loss": 6.7242,
      "step": 3854
    },
    {
      "epoch": 0.3855,
      "grad_norm": 4.280649185180664,
      "learning_rate": 0.0001543467499474665,
      "loss": 4.822,
      "step": 3855
    },
    {
      "epoch": 0.3856,
      "grad_norm": 5.948354721069336,
      "learning_rate": 0.00015431744499506706,
      "loss": 4.9136,
      "step": 3856
    },
    {
      "epoch": 0.3857,
      "grad_norm": 3.468738079071045,
      "learning_rate": 0.00015428813342425177,
      "loss": 5.4089,
      "step": 3857
    },
    {
      "epoch": 0.3858,
      "grad_norm": 11.167182922363281,
      "learning_rate": 0.00015425881523859207,
      "loss": 5.4338,
      "step": 3858
    },
    {
      "epoch": 0.3859,
      "grad_norm": 2.7885868549346924,
      "learning_rate": 0.0001542294904416603,
      "loss": 5.183,
      "step": 3859
    },
    {
      "epoch": 0.386,
      "grad_norm": 3.207935094833374,
      "learning_rate": 0.00015420015903702962,
      "loss": 5.3041,
      "step": 3860
    },
    {
      "epoch": 0.3861,
      "grad_norm": 2.9318065643310547,
      "learning_rate": 0.000154170821028274,
      "loss": 5.5127,
      "step": 3861
    },
    {
      "epoch": 0.3862,
      "grad_norm": 2.9471609592437744,
      "learning_rate": 0.00015414147641896813,
      "loss": 5.2994,
      "step": 3862
    },
    {
      "epoch": 0.3863,
      "grad_norm": 2.436941623687744,
      "learning_rate": 0.00015411212521268758,
      "loss": 5.0018,
      "step": 3863
    },
    {
      "epoch": 0.3864,
      "grad_norm": 6.20158576965332,
      "learning_rate": 0.00015408276741300873,
      "loss": 5.6833,
      "step": 3864
    },
    {
      "epoch": 0.3865,
      "grad_norm": 2.949723720550537,
      "learning_rate": 0.00015405340302350871,
      "loss": 5.3564,
      "step": 3865
    },
    {
      "epoch": 0.3866,
      "grad_norm": 2.7317516803741455,
      "learning_rate": 0.0001540240320477655,
      "loss": 5.1218,
      "step": 3866
    },
    {
      "epoch": 0.3867,
      "grad_norm": 2.2166998386383057,
      "learning_rate": 0.00015399465448935788,
      "loss": 4.8884,
      "step": 3867
    },
    {
      "epoch": 0.3868,
      "grad_norm": 4.835622787475586,
      "learning_rate": 0.00015396527035186537,
      "loss": 5.7758,
      "step": 3868
    },
    {
      "epoch": 0.3869,
      "grad_norm": 3.766592264175415,
      "learning_rate": 0.00015393587963886835,
      "loss": 5.443,
      "step": 3869
    },
    {
      "epoch": 0.387,
      "grad_norm": 5.572547435760498,
      "learning_rate": 0.00015390648235394803,
      "loss": 5.9517,
      "step": 3870
    },
    {
      "epoch": 0.3871,
      "grad_norm": 3.8381640911102295,
      "learning_rate": 0.0001538770785006863,
      "loss": 5.1452,
      "step": 3871
    },
    {
      "epoch": 0.3872,
      "grad_norm": 6.159384727478027,
      "learning_rate": 0.00015384766808266602,
      "loss": 5.3526,
      "step": 3872
    },
    {
      "epoch": 0.3873,
      "grad_norm": 2.2198216915130615,
      "learning_rate": 0.00015381825110347074,
      "loss": 4.5746,
      "step": 3873
    },
    {
      "epoch": 0.3874,
      "grad_norm": 2.147188186645508,
      "learning_rate": 0.00015378882756668478,
      "loss": 5.0542,
      "step": 3874
    },
    {
      "epoch": 0.3875,
      "grad_norm": 2.1216704845428467,
      "learning_rate": 0.00015375939747589335,
      "loss": 4.766,
      "step": 3875
    },
    {
      "epoch": 0.3876,
      "grad_norm": 4.371271133422852,
      "learning_rate": 0.0001537299608346824,
      "loss": 4.7572,
      "step": 3876
    },
    {
      "epoch": 0.3877,
      "grad_norm": 5.082224369049072,
      "learning_rate": 0.0001537005176466387,
      "loss": 5.5262,
      "step": 3877
    },
    {
      "epoch": 0.3878,
      "grad_norm": 4.847775459289551,
      "learning_rate": 0.00015367106791534983,
      "loss": 4.5975,
      "step": 3878
    },
    {
      "epoch": 0.3879,
      "grad_norm": 2.7723004817962646,
      "learning_rate": 0.0001536416116444041,
      "loss": 5.3509,
      "step": 3879
    },
    {
      "epoch": 0.388,
      "grad_norm": 5.596477031707764,
      "learning_rate": 0.00015361214883739076,
      "loss": 5.1851,
      "step": 3880
    },
    {
      "epoch": 0.3881,
      "grad_norm": 5.547904014587402,
      "learning_rate": 0.00015358267949789966,
      "loss": 5.55,
      "step": 3881
    },
    {
      "epoch": 0.3882,
      "grad_norm": 3.2523672580718994,
      "learning_rate": 0.00015355320362952162,
      "loss": 5.2478,
      "step": 3882
    },
    {
      "epoch": 0.3883,
      "grad_norm": 4.422754764556885,
      "learning_rate": 0.00015352372123584814,
      "loss": 5.1776,
      "step": 3883
    },
    {
      "epoch": 0.3884,
      "grad_norm": 5.615268707275391,
      "learning_rate": 0.00015349423232047162,
      "loss": 5.279,
      "step": 3884
    },
    {
      "epoch": 0.3885,
      "grad_norm": 3.1279122829437256,
      "learning_rate": 0.00015346473688698513,
      "loss": 4.4212,
      "step": 3885
    },
    {
      "epoch": 0.3886,
      "grad_norm": 3.9478583335876465,
      "learning_rate": 0.00015343523493898265,
      "loss": 5.6729,
      "step": 3886
    },
    {
      "epoch": 0.3887,
      "grad_norm": 3.337695837020874,
      "learning_rate": 0.00015340572648005888,
      "loss": 4.8306,
      "step": 3887
    },
    {
      "epoch": 0.3888,
      "grad_norm": 3.480701446533203,
      "learning_rate": 0.0001533762115138093,
      "loss": 5.297,
      "step": 3888
    },
    {
      "epoch": 0.3889,
      "grad_norm": 8.005728721618652,
      "learning_rate": 0.0001533466900438303,
      "loss": 5.2495,
      "step": 3889
    },
    {
      "epoch": 0.389,
      "grad_norm": 2.7173707485198975,
      "learning_rate": 0.00015331716207371888,
      "loss": 5.0583,
      "step": 3890
    },
    {
      "epoch": 0.3891,
      "grad_norm": 6.5037031173706055,
      "learning_rate": 0.000153287627607073,
      "loss": 5.5888,
      "step": 3891
    },
    {
      "epoch": 0.3892,
      "grad_norm": 2.9816951751708984,
      "learning_rate": 0.00015325808664749135,
      "loss": 4.9606,
      "step": 3892
    },
    {
      "epoch": 0.3893,
      "grad_norm": 4.045719623565674,
      "learning_rate": 0.0001532285391985734,
      "loss": 5.2043,
      "step": 3893
    },
    {
      "epoch": 0.3894,
      "grad_norm": 3.460203170776367,
      "learning_rate": 0.0001531989852639194,
      "loss": 6.0911,
      "step": 3894
    },
    {
      "epoch": 0.3895,
      "grad_norm": 2.636260747909546,
      "learning_rate": 0.0001531694248471304,
      "loss": 5.6194,
      "step": 3895
    },
    {
      "epoch": 0.3896,
      "grad_norm": 2.5187368392944336,
      "learning_rate": 0.00015313985795180828,
      "loss": 4.5315,
      "step": 3896
    },
    {
      "epoch": 0.3897,
      "grad_norm": 2.4343717098236084,
      "learning_rate": 0.00015311028458155567,
      "loss": 5.3504,
      "step": 3897
    },
    {
      "epoch": 0.3898,
      "grad_norm": 2.8764564990997314,
      "learning_rate": 0.00015308070473997598,
      "loss": 5.3381,
      "step": 3898
    },
    {
      "epoch": 0.3899,
      "grad_norm": 5.046579360961914,
      "learning_rate": 0.0001530511184306734,
      "loss": 5.1689,
      "step": 3899
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.262382507324219,
      "learning_rate": 0.00015302152565725298,
      "loss": 6.9492,
      "step": 3900
    },
    {
      "epoch": 0.3901,
      "grad_norm": 2.4003982543945312,
      "learning_rate": 0.0001529919264233205,
      "loss": 5.0224,
      "step": 3901
    },
    {
      "epoch": 0.3902,
      "grad_norm": 3.4969098567962646,
      "learning_rate": 0.00015296232073248251,
      "loss": 4.5659,
      "step": 3902
    },
    {
      "epoch": 0.3903,
      "grad_norm": 3.51720929145813,
      "learning_rate": 0.00015293270858834644,
      "loss": 4.9016,
      "step": 3903
    },
    {
      "epoch": 0.3904,
      "grad_norm": 2.2317092418670654,
      "learning_rate": 0.00015290308999452032,
      "loss": 5.4732,
      "step": 3904
    },
    {
      "epoch": 0.3905,
      "grad_norm": 3.141226291656494,
      "learning_rate": 0.00015287346495461315,
      "loss": 5.204,
      "step": 3905
    },
    {
      "epoch": 0.3906,
      "grad_norm": 3.3221018314361572,
      "learning_rate": 0.00015284383347223472,
      "loss": 4.94,
      "step": 3906
    },
    {
      "epoch": 0.3907,
      "grad_norm": 7.785981178283691,
      "learning_rate": 0.00015281419555099546,
      "loss": 5.5839,
      "step": 3907
    },
    {
      "epoch": 0.3908,
      "grad_norm": 4.2538251876831055,
      "learning_rate": 0.00015278455119450664,
      "loss": 5.8529,
      "step": 3908
    },
    {
      "epoch": 0.3909,
      "grad_norm": 8.332098007202148,
      "learning_rate": 0.00015275490040638038,
      "loss": 4.5901,
      "step": 3909
    },
    {
      "epoch": 0.391,
      "grad_norm": 3.654810905456543,
      "learning_rate": 0.00015272524319022955,
      "loss": 5.4014,
      "step": 3910
    },
    {
      "epoch": 0.3911,
      "grad_norm": 3.6206226348876953,
      "learning_rate": 0.00015269557954966778,
      "loss": 4.7886,
      "step": 3911
    },
    {
      "epoch": 0.3912,
      "grad_norm": 3.1170685291290283,
      "learning_rate": 0.00015266590948830947,
      "loss": 5.2018,
      "step": 3912
    },
    {
      "epoch": 0.3913,
      "grad_norm": 2.283214569091797,
      "learning_rate": 0.00015263623300976978,
      "loss": 4.883,
      "step": 3913
    },
    {
      "epoch": 0.3914,
      "grad_norm": 2.298583745956421,
      "learning_rate": 0.00015260655011766484,
      "loss": 5.0342,
      "step": 3914
    },
    {
      "epoch": 0.3915,
      "grad_norm": 3.4901013374328613,
      "learning_rate": 0.00015257686081561134,
      "loss": 4.896,
      "step": 3915
    },
    {
      "epoch": 0.3916,
      "grad_norm": 3.3885738849639893,
      "learning_rate": 0.00015254716510722677,
      "loss": 4.6782,
      "step": 3916
    },
    {
      "epoch": 0.3917,
      "grad_norm": 4.266900539398193,
      "learning_rate": 0.0001525174629961296,
      "loss": 4.6359,
      "step": 3917
    },
    {
      "epoch": 0.3918,
      "grad_norm": 5.82711124420166,
      "learning_rate": 0.00015248775448593882,
      "loss": 5.9562,
      "step": 3918
    },
    {
      "epoch": 0.3919,
      "grad_norm": 3.6622605323791504,
      "learning_rate": 0.00015245803958027434,
      "loss": 4.7955,
      "step": 3919
    },
    {
      "epoch": 0.392,
      "grad_norm": 2.926525115966797,
      "learning_rate": 0.00015242831828275692,
      "loss": 5.1174,
      "step": 3920
    },
    {
      "epoch": 0.3921,
      "grad_norm": 2.4777793884277344,
      "learning_rate": 0.00015239859059700794,
      "loss": 5.4309,
      "step": 3921
    },
    {
      "epoch": 0.3922,
      "grad_norm": 6.188035011291504,
      "learning_rate": 0.00015236885652664963,
      "loss": 4.5424,
      "step": 3922
    },
    {
      "epoch": 0.3923,
      "grad_norm": 3.967576026916504,
      "learning_rate": 0.00015233911607530497,
      "loss": 5.5976,
      "step": 3923
    },
    {
      "epoch": 0.3924,
      "grad_norm": 2.7322709560394287,
      "learning_rate": 0.0001523093692465978,
      "loss": 4.8908,
      "step": 3924
    },
    {
      "epoch": 0.3925,
      "grad_norm": 3.781933546066284,
      "learning_rate": 0.00015227961604415266,
      "loss": 4.9577,
      "step": 3925
    },
    {
      "epoch": 0.3926,
      "grad_norm": 4.8347625732421875,
      "learning_rate": 0.0001522498564715949,
      "loss": 5.3431,
      "step": 3926
    },
    {
      "epoch": 0.3927,
      "grad_norm": 3.420428514480591,
      "learning_rate": 0.0001522200905325506,
      "loss": 5.1149,
      "step": 3927
    },
    {
      "epoch": 0.3928,
      "grad_norm": 2.573046922683716,
      "learning_rate": 0.00015219031823064666,
      "loss": 5.0929,
      "step": 3928
    },
    {
      "epoch": 0.3929,
      "grad_norm": 2.7568867206573486,
      "learning_rate": 0.0001521605395695108,
      "loss": 4.9188,
      "step": 3929
    },
    {
      "epoch": 0.393,
      "grad_norm": 2.1395134925842285,
      "learning_rate": 0.0001521307545527714,
      "loss": 5.0839,
      "step": 3930
    },
    {
      "epoch": 0.3931,
      "grad_norm": 2.2568233013153076,
      "learning_rate": 0.00015210096318405767,
      "loss": 4.8097,
      "step": 3931
    },
    {
      "epoch": 0.3932,
      "grad_norm": 3.615093946456909,
      "learning_rate": 0.0001520711654669996,
      "loss": 5.6515,
      "step": 3932
    },
    {
      "epoch": 0.3933,
      "grad_norm": 2.5407817363739014,
      "learning_rate": 0.00015204136140522799,
      "loss": 5.0179,
      "step": 3933
    },
    {
      "epoch": 0.3934,
      "grad_norm": 2.242373466491699,
      "learning_rate": 0.0001520115510023743,
      "loss": 4.854,
      "step": 3934
    },
    {
      "epoch": 0.3935,
      "grad_norm": 2.217028856277466,
      "learning_rate": 0.00015198173426207094,
      "loss": 5.0131,
      "step": 3935
    },
    {
      "epoch": 0.3936,
      "grad_norm": 5.30352258682251,
      "learning_rate": 0.00015195191118795096,
      "loss": 4.8851,
      "step": 3936
    },
    {
      "epoch": 0.3937,
      "grad_norm": 2.572401285171509,
      "learning_rate": 0.00015192208178364816,
      "loss": 4.9075,
      "step": 3937
    },
    {
      "epoch": 0.3938,
      "grad_norm": 3.338486909866333,
      "learning_rate": 0.00015189224605279718,
      "loss": 5.5762,
      "step": 3938
    },
    {
      "epoch": 0.3939,
      "grad_norm": 4.082182884216309,
      "learning_rate": 0.00015186240399903342,
      "loss": 4.6685,
      "step": 3939
    },
    {
      "epoch": 0.394,
      "grad_norm": 7.155563831329346,
      "learning_rate": 0.00015183255562599307,
      "loss": 4.7342,
      "step": 3940
    },
    {
      "epoch": 0.3941,
      "grad_norm": 3.4429614543914795,
      "learning_rate": 0.00015180270093731303,
      "loss": 4.5325,
      "step": 3941
    },
    {
      "epoch": 0.3942,
      "grad_norm": 6.11724853515625,
      "learning_rate": 0.000151772839936631,
      "loss": 5.0179,
      "step": 3942
    },
    {
      "epoch": 0.3943,
      "grad_norm": 2.3202435970306396,
      "learning_rate": 0.0001517429726275855,
      "loss": 5.1378,
      "step": 3943
    },
    {
      "epoch": 0.3944,
      "grad_norm": 6.446569442749023,
      "learning_rate": 0.00015171309901381572,
      "loss": 5.5008,
      "step": 3944
    },
    {
      "epoch": 0.3945,
      "grad_norm": 2.4398417472839355,
      "learning_rate": 0.00015168321909896172,
      "loss": 4.8425,
      "step": 3945
    },
    {
      "epoch": 0.3946,
      "grad_norm": 3.265805721282959,
      "learning_rate": 0.0001516533328866642,
      "loss": 5.0628,
      "step": 3946
    },
    {
      "epoch": 0.3947,
      "grad_norm": 4.66032600402832,
      "learning_rate": 0.00015162344038056476,
      "loss": 5.1924,
      "step": 3947
    },
    {
      "epoch": 0.3948,
      "grad_norm": 5.176969051361084,
      "learning_rate": 0.00015159354158430572,
      "loss": 5.5024,
      "step": 3948
    },
    {
      "epoch": 0.3949,
      "grad_norm": 3.1394686698913574,
      "learning_rate": 0.00015156363650153012,
      "loss": 4.7507,
      "step": 3949
    },
    {
      "epoch": 0.395,
      "grad_norm": 2.794628620147705,
      "learning_rate": 0.00015153372513588182,
      "loss": 5.3356,
      "step": 3950
    },
    {
      "epoch": 0.3951,
      "grad_norm": 4.391885757446289,
      "learning_rate": 0.00015150380749100545,
      "loss": 5.4781,
      "step": 3951
    },
    {
      "epoch": 0.3952,
      "grad_norm": 2.2780745029449463,
      "learning_rate": 0.00015147388357054632,
      "loss": 4.9531,
      "step": 3952
    },
    {
      "epoch": 0.3953,
      "grad_norm": 1.9899675846099854,
      "learning_rate": 0.00015144395337815064,
      "loss": 5.1503,
      "step": 3953
    },
    {
      "epoch": 0.3954,
      "grad_norm": 2.858077049255371,
      "learning_rate": 0.0001514140169174653,
      "loss": 6.04,
      "step": 3954
    },
    {
      "epoch": 0.3955,
      "grad_norm": 4.448555946350098,
      "learning_rate": 0.00015138407419213796,
      "loss": 5.1358,
      "step": 3955
    },
    {
      "epoch": 0.3956,
      "grad_norm": 8.790107727050781,
      "learning_rate": 0.00015135412520581702,
      "loss": 4.9718,
      "step": 3956
    },
    {
      "epoch": 0.3957,
      "grad_norm": 11.7772855758667,
      "learning_rate": 0.0001513241699621517,
      "loss": 5.7215,
      "step": 3957
    },
    {
      "epoch": 0.3958,
      "grad_norm": 6.86590051651001,
      "learning_rate": 0.00015129420846479196,
      "loss": 5.0365,
      "step": 3958
    },
    {
      "epoch": 0.3959,
      "grad_norm": 5.800752639770508,
      "learning_rate": 0.0001512642407173885,
      "loss": 4.681,
      "step": 3959
    },
    {
      "epoch": 0.396,
      "grad_norm": 4.225568771362305,
      "learning_rate": 0.00015123426672359285,
      "loss": 5.0952,
      "step": 3960
    },
    {
      "epoch": 0.3961,
      "grad_norm": 2.366035223007202,
      "learning_rate": 0.00015120428648705717,
      "loss": 5.6262,
      "step": 3961
    },
    {
      "epoch": 0.3962,
      "grad_norm": 1.9905295372009277,
      "learning_rate": 0.00015117430001143452,
      "loss": 5.2287,
      "step": 3962
    },
    {
      "epoch": 0.3963,
      "grad_norm": 3.6094069480895996,
      "learning_rate": 0.0001511443073003786,
      "loss": 5.2694,
      "step": 3963
    },
    {
      "epoch": 0.3964,
      "grad_norm": 2.060070753097534,
      "learning_rate": 0.000151114308357544,
      "loss": 4.7127,
      "step": 3964
    },
    {
      "epoch": 0.3965,
      "grad_norm": 2.3527543544769287,
      "learning_rate": 0.000151084303186586,
      "loss": 4.4327,
      "step": 3965
    },
    {
      "epoch": 0.3966,
      "grad_norm": 3.089632272720337,
      "learning_rate": 0.0001510542917911606,
      "loss": 4.6169,
      "step": 3966
    },
    {
      "epoch": 0.3967,
      "grad_norm": 2.375277519226074,
      "learning_rate": 0.0001510242741749246,
      "loss": 5.1647,
      "step": 3967
    },
    {
      "epoch": 0.3968,
      "grad_norm": 4.956357479095459,
      "learning_rate": 0.00015099425034153553,
      "loss": 5.5505,
      "step": 3968
    },
    {
      "epoch": 0.3969,
      "grad_norm": 7.069380760192871,
      "learning_rate": 0.00015096422029465178,
      "loss": 5.8252,
      "step": 3969
    },
    {
      "epoch": 0.397,
      "grad_norm": 2.479325771331787,
      "learning_rate": 0.00015093418403793238,
      "loss": 4.9594,
      "step": 3970
    },
    {
      "epoch": 0.3971,
      "grad_norm": 3.1391687393188477,
      "learning_rate": 0.00015090414157503714,
      "loss": 5.8576,
      "step": 3971
    },
    {
      "epoch": 0.3972,
      "grad_norm": 5.944414138793945,
      "learning_rate": 0.00015087409290962667,
      "loss": 5.6413,
      "step": 3972
    },
    {
      "epoch": 0.3973,
      "grad_norm": 4.099064826965332,
      "learning_rate": 0.0001508440380453623,
      "loss": 4.8757,
      "step": 3973
    },
    {
      "epoch": 0.3974,
      "grad_norm": 2.647294044494629,
      "learning_rate": 0.0001508139769859061,
      "loss": 5.2223,
      "step": 3974
    },
    {
      "epoch": 0.3975,
      "grad_norm": 3.4899752140045166,
      "learning_rate": 0.00015078390973492092,
      "loss": 5.9833,
      "step": 3975
    },
    {
      "epoch": 0.3976,
      "grad_norm": 2.4594578742980957,
      "learning_rate": 0.00015075383629607042,
      "loss": 5.1173,
      "step": 3976
    },
    {
      "epoch": 0.3977,
      "grad_norm": 4.794008255004883,
      "learning_rate": 0.00015072375667301893,
      "loss": 5.7375,
      "step": 3977
    },
    {
      "epoch": 0.3978,
      "grad_norm": 17.586509704589844,
      "learning_rate": 0.00015069367086943154,
      "loss": 7.7832,
      "step": 3978
    },
    {
      "epoch": 0.3979,
      "grad_norm": 2.4333202838897705,
      "learning_rate": 0.0001506635788889741,
      "loss": 5.5791,
      "step": 3979
    },
    {
      "epoch": 0.398,
      "grad_norm": 18.597505569458008,
      "learning_rate": 0.00015063348073531324,
      "loss": 5.4469,
      "step": 3980
    },
    {
      "epoch": 0.3981,
      "grad_norm": 2.8898708820343018,
      "learning_rate": 0.00015060337641211637,
      "loss": 4.7987,
      "step": 3981
    },
    {
      "epoch": 0.3982,
      "grad_norm": 2.7373857498168945,
      "learning_rate": 0.0001505732659230516,
      "loss": 5.2436,
      "step": 3982
    },
    {
      "epoch": 0.3983,
      "grad_norm": 4.125609397888184,
      "learning_rate": 0.0001505431492717878,
      "loss": 4.6899,
      "step": 3983
    },
    {
      "epoch": 0.3984,
      "grad_norm": 2.5154364109039307,
      "learning_rate": 0.0001505130264619945,
      "loss": 4.9883,
      "step": 3984
    },
    {
      "epoch": 0.3985,
      "grad_norm": 4.226368427276611,
      "learning_rate": 0.0001504828974973422,
      "loss": 5.176,
      "step": 3985
    },
    {
      "epoch": 0.3986,
      "grad_norm": 4.192867755889893,
      "learning_rate": 0.00015045276238150192,
      "loss": 5.0127,
      "step": 3986
    },
    {
      "epoch": 0.3987,
      "grad_norm": 4.039165019989014,
      "learning_rate": 0.00015042262111814565,
      "loss": 5.1513,
      "step": 3987
    },
    {
      "epoch": 0.3988,
      "grad_norm": 2.528094530105591,
      "learning_rate": 0.00015039247371094588,
      "loss": 5.0199,
      "step": 3988
    },
    {
      "epoch": 0.3989,
      "grad_norm": 2.1852316856384277,
      "learning_rate": 0.0001503623201635761,
      "loss": 5.2898,
      "step": 3989
    },
    {
      "epoch": 0.399,
      "grad_norm": 4.158294677734375,
      "learning_rate": 0.00015033216047971031,
      "loss": 5.1464,
      "step": 3990
    },
    {
      "epoch": 0.3991,
      "grad_norm": 3.8487131595611572,
      "learning_rate": 0.00015030199466302353,
      "loss": 5.2246,
      "step": 3991
    },
    {
      "epoch": 0.3992,
      "grad_norm": 4.822324275970459,
      "learning_rate": 0.00015027182271719122,
      "loss": 5.1901,
      "step": 3992
    },
    {
      "epoch": 0.3993,
      "grad_norm": 2.3553764820098877,
      "learning_rate": 0.00015024164464588982,
      "loss": 4.964,
      "step": 3993
    },
    {
      "epoch": 0.3994,
      "grad_norm": 2.917569398880005,
      "learning_rate": 0.0001502114604527964,
      "loss": 5.3561,
      "step": 3994
    },
    {
      "epoch": 0.3995,
      "grad_norm": 4.179385662078857,
      "learning_rate": 0.00015018127014158886,
      "loss": 5.8446,
      "step": 3995
    },
    {
      "epoch": 0.3996,
      "grad_norm": 2.7633447647094727,
      "learning_rate": 0.00015015107371594573,
      "loss": 4.9353,
      "step": 3996
    },
    {
      "epoch": 0.3997,
      "grad_norm": 6.858094215393066,
      "learning_rate": 0.00015012087117954642,
      "loss": 5.801,
      "step": 3997
    },
    {
      "epoch": 0.3998,
      "grad_norm": 2.848284959793091,
      "learning_rate": 0.000150090662536071,
      "loss": 4.8831,
      "step": 3998
    },
    {
      "epoch": 0.3999,
      "grad_norm": 3.6484224796295166,
      "learning_rate": 0.0001500604477892003,
      "loss": 5.0184,
      "step": 3999
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.31323766708374,
      "learning_rate": 0.00015003022694261585,
      "loss": 5.3774,
      "step": 4000
    },
    {
      "epoch": 0.4001,
      "grad_norm": 4.904125690460205,
      "learning_rate": 0.00015000000000000001,
      "loss": 5.4044,
      "step": 4001
    },
    {
      "epoch": 0.4002,
      "grad_norm": 4.028195381164551,
      "learning_rate": 0.00014996976696503587,
      "loss": 5.3376,
      "step": 4002
    },
    {
      "epoch": 0.4003,
      "grad_norm": 3.6660680770874023,
      "learning_rate": 0.00014993952784140717,
      "loss": 4.8558,
      "step": 4003
    },
    {
      "epoch": 0.4004,
      "grad_norm": 7.097280979156494,
      "learning_rate": 0.00014990928263279848,
      "loss": 5.4902,
      "step": 4004
    },
    {
      "epoch": 0.4005,
      "grad_norm": 2.0762441158294678,
      "learning_rate": 0.00014987903134289508,
      "loss": 4.6011,
      "step": 4005
    },
    {
      "epoch": 0.4006,
      "grad_norm": 3.8706023693084717,
      "learning_rate": 0.00014984877397538303,
      "loss": 4.9978,
      "step": 4006
    },
    {
      "epoch": 0.4007,
      "grad_norm": 3.5693089962005615,
      "learning_rate": 0.0001498185105339491,
      "loss": 5.0934,
      "step": 4007
    },
    {
      "epoch": 0.4008,
      "grad_norm": 2.56050181388855,
      "learning_rate": 0.00014978824102228076,
      "loss": 5.1705,
      "step": 4008
    },
    {
      "epoch": 0.4009,
      "grad_norm": 3.156460762023926,
      "learning_rate": 0.00014975796544406625,
      "loss": 5.7306,
      "step": 4009
    },
    {
      "epoch": 0.401,
      "grad_norm": 2.2604618072509766,
      "learning_rate": 0.0001497276838029946,
      "loss": 5.1389,
      "step": 4010
    },
    {
      "epoch": 0.4011,
      "grad_norm": 2.3076655864715576,
      "learning_rate": 0.00014969739610275556,
      "loss": 5.3442,
      "step": 4011
    },
    {
      "epoch": 0.4012,
      "grad_norm": 2.1544189453125,
      "learning_rate": 0.0001496671023470395,
      "loss": 4.8779,
      "step": 4012
    },
    {
      "epoch": 0.4013,
      "grad_norm": 3.4567947387695312,
      "learning_rate": 0.0001496368025395377,
      "loss": 5.1231,
      "step": 4013
    },
    {
      "epoch": 0.4014,
      "grad_norm": 2.278545618057251,
      "learning_rate": 0.00014960649668394207,
      "loss": 4.8989,
      "step": 4014
    },
    {
      "epoch": 0.4015,
      "grad_norm": 3.423693895339966,
      "learning_rate": 0.00014957618478394529,
      "loss": 5.11,
      "step": 4015
    },
    {
      "epoch": 0.4016,
      "grad_norm": 4.8462090492248535,
      "learning_rate": 0.00014954586684324078,
      "loss": 4.8452,
      "step": 4016
    },
    {
      "epoch": 0.4017,
      "grad_norm": 3.463782787322998,
      "learning_rate": 0.00014951554286552266,
      "loss": 4.6496,
      "step": 4017
    },
    {
      "epoch": 0.4018,
      "grad_norm": 3.7920944690704346,
      "learning_rate": 0.00014948521285448586,
      "loss": 5.4836,
      "step": 4018
    },
    {
      "epoch": 0.4019,
      "grad_norm": 2.3810582160949707,
      "learning_rate": 0.00014945487681382598,
      "loss": 5.2578,
      "step": 4019
    },
    {
      "epoch": 0.402,
      "grad_norm": 3.352842092514038,
      "learning_rate": 0.00014942453474723935,
      "loss": 5.3151,
      "step": 4020
    },
    {
      "epoch": 0.4021,
      "grad_norm": 5.574296474456787,
      "learning_rate": 0.0001493941866584231,
      "loss": 4.8318,
      "step": 4021
    },
    {
      "epoch": 0.4022,
      "grad_norm": 3.753685474395752,
      "learning_rate": 0.00014936383255107505,
      "loss": 5.3092,
      "step": 4022
    },
    {
      "epoch": 0.4023,
      "grad_norm": 2.5550503730773926,
      "learning_rate": 0.0001493334724288937,
      "loss": 5.1419,
      "step": 4023
    },
    {
      "epoch": 0.4024,
      "grad_norm": 5.771059513092041,
      "learning_rate": 0.0001493031062955784,
      "loss": 5.3588,
      "step": 4024
    },
    {
      "epoch": 0.4025,
      "grad_norm": 3.4237449169158936,
      "learning_rate": 0.00014927273415482915,
      "loss": 5.3713,
      "step": 4025
    },
    {
      "epoch": 0.4026,
      "grad_norm": 2.6933209896087646,
      "learning_rate": 0.00014924235601034672,
      "loss": 4.9063,
      "step": 4026
    },
    {
      "epoch": 0.4027,
      "grad_norm": 3.3140668869018555,
      "learning_rate": 0.00014921197186583255,
      "loss": 4.6787,
      "step": 4027
    },
    {
      "epoch": 0.4028,
      "grad_norm": 3.9676754474639893,
      "learning_rate": 0.0001491815817249889,
      "loss": 5.022,
      "step": 4028
    },
    {
      "epoch": 0.4029,
      "grad_norm": 4.5106329917907715,
      "learning_rate": 0.0001491511855915187,
      "loss": 5.5425,
      "step": 4029
    },
    {
      "epoch": 0.403,
      "grad_norm": 2.189606189727783,
      "learning_rate": 0.00014912078346912563,
      "loss": 4.7621,
      "step": 4030
    },
    {
      "epoch": 0.4031,
      "grad_norm": 3.7519891262054443,
      "learning_rate": 0.00014909037536151409,
      "loss": 5.0174,
      "step": 4031
    },
    {
      "epoch": 0.4032,
      "grad_norm": 3.6548044681549072,
      "learning_rate": 0.0001490599612723892,
      "loss": 5.1493,
      "step": 4032
    },
    {
      "epoch": 0.4033,
      "grad_norm": 2.5294671058654785,
      "learning_rate": 0.00014902954120545687,
      "loss": 4.9024,
      "step": 4033
    },
    {
      "epoch": 0.4034,
      "grad_norm": 1.8745002746582031,
      "learning_rate": 0.00014899911516442365,
      "loss": 4.822,
      "step": 4034
    },
    {
      "epoch": 0.4035,
      "grad_norm": 2.7930006980895996,
      "learning_rate": 0.00014896868315299693,
      "loss": 5.2866,
      "step": 4035
    },
    {
      "epoch": 0.4036,
      "grad_norm": 7.225442409515381,
      "learning_rate": 0.00014893824517488464,
      "loss": 5.3044,
      "step": 4036
    },
    {
      "epoch": 0.4037,
      "grad_norm": 1.9372293949127197,
      "learning_rate": 0.00014890780123379564,
      "loss": 5.0082,
      "step": 4037
    },
    {
      "epoch": 0.4038,
      "grad_norm": 2.7710835933685303,
      "learning_rate": 0.0001488773513334394,
      "loss": 4.7923,
      "step": 4038
    },
    {
      "epoch": 0.4039,
      "grad_norm": 4.62398099899292,
      "learning_rate": 0.0001488468954775262,
      "loss": 5.0923,
      "step": 4039
    },
    {
      "epoch": 0.404,
      "grad_norm": 3.0589568614959717,
      "learning_rate": 0.00014881643366976692,
      "loss": 5.3323,
      "step": 4040
    },
    {
      "epoch": 0.4041,
      "grad_norm": 3.236536741256714,
      "learning_rate": 0.0001487859659138733,
      "loss": 4.913,
      "step": 4041
    },
    {
      "epoch": 0.4042,
      "grad_norm": 2.163938283920288,
      "learning_rate": 0.00014875549221355768,
      "loss": 4.8425,
      "step": 4042
    },
    {
      "epoch": 0.4043,
      "grad_norm": 2.9302873611450195,
      "learning_rate": 0.00014872501257253323,
      "loss": 4.999,
      "step": 4043
    },
    {
      "epoch": 0.4044,
      "grad_norm": 2.7009801864624023,
      "learning_rate": 0.00014869452699451383,
      "loss": 4.7401,
      "step": 4044
    },
    {
      "epoch": 0.4045,
      "grad_norm": 2.718425989151001,
      "learning_rate": 0.000148664035483214,
      "loss": 4.9194,
      "step": 4045
    },
    {
      "epoch": 0.4046,
      "grad_norm": 3.93976092338562,
      "learning_rate": 0.00014863353804234905,
      "loss": 5.0908,
      "step": 4046
    },
    {
      "epoch": 0.4047,
      "grad_norm": 2.925041913986206,
      "learning_rate": 0.00014860303467563503,
      "loss": 5.1764,
      "step": 4047
    },
    {
      "epoch": 0.4048,
      "grad_norm": 2.171557664871216,
      "learning_rate": 0.00014857252538678865,
      "loss": 4.9467,
      "step": 4048
    },
    {
      "epoch": 0.4049,
      "grad_norm": 4.900540351867676,
      "learning_rate": 0.0001485420101795274,
      "loss": 4.6963,
      "step": 4049
    },
    {
      "epoch": 0.405,
      "grad_norm": 7.081606388092041,
      "learning_rate": 0.00014851148905756947,
      "loss": 5.4451,
      "step": 4050
    },
    {
      "epoch": 0.4051,
      "grad_norm": 4.567622184753418,
      "learning_rate": 0.00014848096202463372,
      "loss": 4.6983,
      "step": 4051
    },
    {
      "epoch": 0.4052,
      "grad_norm": 4.744316577911377,
      "learning_rate": 0.0001484504290844398,
      "loss": 4.9326,
      "step": 4052
    },
    {
      "epoch": 0.4053,
      "grad_norm": 3.052912712097168,
      "learning_rate": 0.00014841989024070809,
      "loss": 5.1266,
      "step": 4053
    },
    {
      "epoch": 0.4054,
      "grad_norm": 4.309214115142822,
      "learning_rate": 0.00014838934549715963,
      "loss": 5.2648,
      "step": 4054
    },
    {
      "epoch": 0.4055,
      "grad_norm": 3.012064218521118,
      "learning_rate": 0.00014835879485751617,
      "loss": 4.793,
      "step": 4055
    },
    {
      "epoch": 0.4056,
      "grad_norm": 3.165146827697754,
      "learning_rate": 0.00014832823832550024,
      "loss": 5.0605,
      "step": 4056
    },
    {
      "epoch": 0.4057,
      "grad_norm": 8.709583282470703,
      "learning_rate": 0.00014829767590483506,
      "loss": 5.5924,
      "step": 4057
    },
    {
      "epoch": 0.4058,
      "grad_norm": 3.0664539337158203,
      "learning_rate": 0.0001482671075992446,
      "loss": 5.1719,
      "step": 4058
    },
    {
      "epoch": 0.4059,
      "grad_norm": 8.395539283752441,
      "learning_rate": 0.00014823653341245353,
      "loss": 4.8983,
      "step": 4059
    },
    {
      "epoch": 0.406,
      "grad_norm": 2.9812874794006348,
      "learning_rate": 0.00014820595334818712,
      "loss": 4.8026,
      "step": 4060
    },
    {
      "epoch": 0.4061,
      "grad_norm": 4.453608989715576,
      "learning_rate": 0.00014817536741017152,
      "loss": 5.0039,
      "step": 4061
    },
    {
      "epoch": 0.4062,
      "grad_norm": 6.846360683441162,
      "learning_rate": 0.00014814477560213358,
      "loss": 5.9471,
      "step": 4062
    },
    {
      "epoch": 0.4063,
      "grad_norm": 4.264512538909912,
      "learning_rate": 0.00014811417792780075,
      "loss": 4.924,
      "step": 4063
    },
    {
      "epoch": 0.4064,
      "grad_norm": 3.0738906860351562,
      "learning_rate": 0.00014808357439090127,
      "loss": 5.2661,
      "step": 4064
    },
    {
      "epoch": 0.4065,
      "grad_norm": 2.6499133110046387,
      "learning_rate": 0.00014805296499516407,
      "loss": 4.7414,
      "step": 4065
    },
    {
      "epoch": 0.4066,
      "grad_norm": 2.33486270904541,
      "learning_rate": 0.0001480223497443189,
      "loss": 4.8961,
      "step": 4066
    },
    {
      "epoch": 0.4067,
      "grad_norm": 2.4623613357543945,
      "learning_rate": 0.00014799172864209608,
      "loss": 5.255,
      "step": 4067
    },
    {
      "epoch": 0.4068,
      "grad_norm": 4.690597057342529,
      "learning_rate": 0.00014796110169222666,
      "loss": 4.8524,
      "step": 4068
    },
    {
      "epoch": 0.4069,
      "grad_norm": 2.9407544136047363,
      "learning_rate": 0.0001479304688984425,
      "loss": 5.0615,
      "step": 4069
    },
    {
      "epoch": 0.407,
      "grad_norm": 2.2412238121032715,
      "learning_rate": 0.00014789983026447612,
      "loss": 4.7768,
      "step": 4070
    },
    {
      "epoch": 0.4071,
      "grad_norm": 2.9954440593719482,
      "learning_rate": 0.0001478691857940607,
      "loss": 5.1715,
      "step": 4071
    },
    {
      "epoch": 0.4072,
      "grad_norm": 2.938633918762207,
      "learning_rate": 0.00014783853549093018,
      "loss": 4.9023,
      "step": 4072
    },
    {
      "epoch": 0.4073,
      "grad_norm": 3.3686575889587402,
      "learning_rate": 0.00014780787935881923,
      "loss": 6.5248,
      "step": 4073
    },
    {
      "epoch": 0.4074,
      "grad_norm": 1.914475917816162,
      "learning_rate": 0.0001477772174014632,
      "loss": 4.9537,
      "step": 4074
    },
    {
      "epoch": 0.4075,
      "grad_norm": 2.230748176574707,
      "learning_rate": 0.00014774654962259812,
      "loss": 5.1257,
      "step": 4075
    },
    {
      "epoch": 0.4076,
      "grad_norm": 2.6256511211395264,
      "learning_rate": 0.00014771587602596084,
      "loss": 5.5539,
      "step": 4076
    },
    {
      "epoch": 0.4077,
      "grad_norm": 12.477214813232422,
      "learning_rate": 0.0001476851966152888,
      "loss": 4.865,
      "step": 4077
    },
    {
      "epoch": 0.4078,
      "grad_norm": 3.513416290283203,
      "learning_rate": 0.0001476545113943202,
      "loss": 5.0162,
      "step": 4078
    },
    {
      "epoch": 0.4079,
      "grad_norm": 2.412984848022461,
      "learning_rate": 0.0001476238203667939,
      "loss": 5.2061,
      "step": 4079
    },
    {
      "epoch": 0.408,
      "grad_norm": 7.521999359130859,
      "learning_rate": 0.0001475931235364496,
      "loss": 4.5795,
      "step": 4080
    },
    {
      "epoch": 0.4081,
      "grad_norm": 3.228205680847168,
      "learning_rate": 0.00014756242090702756,
      "loss": 4.9291,
      "step": 4081
    },
    {
      "epoch": 0.4082,
      "grad_norm": 2.1659538745880127,
      "learning_rate": 0.00014753171248226875,
      "loss": 4.8618,
      "step": 4082
    },
    {
      "epoch": 0.4083,
      "grad_norm": 5.1164751052856445,
      "learning_rate": 0.00014750099826591498,
      "loss": 5.2513,
      "step": 4083
    },
    {
      "epoch": 0.4084,
      "grad_norm": 4.741237640380859,
      "learning_rate": 0.00014747027826170867,
      "loss": 4.4524,
      "step": 4084
    },
    {
      "epoch": 0.4085,
      "grad_norm": 2.5901577472686768,
      "learning_rate": 0.00014743955247339293,
      "loss": 5.5778,
      "step": 4085
    },
    {
      "epoch": 0.4086,
      "grad_norm": 2.441134452819824,
      "learning_rate": 0.0001474088209047116,
      "loss": 5.27,
      "step": 4086
    },
    {
      "epoch": 0.4087,
      "grad_norm": 2.450864315032959,
      "learning_rate": 0.00014737808355940932,
      "loss": 4.6699,
      "step": 4087
    },
    {
      "epoch": 0.4088,
      "grad_norm": 11.93677806854248,
      "learning_rate": 0.0001473473404412312,
      "loss": 5.2058,
      "step": 4088
    },
    {
      "epoch": 0.4089,
      "grad_norm": 2.5388023853302,
      "learning_rate": 0.00014731659155392332,
      "loss": 4.7763,
      "step": 4089
    },
    {
      "epoch": 0.409,
      "grad_norm": 3.8123881816864014,
      "learning_rate": 0.00014728583690123224,
      "loss": 5.7914,
      "step": 4090
    },
    {
      "epoch": 0.4091,
      "grad_norm": 2.814530611038208,
      "learning_rate": 0.00014725507648690543,
      "loss": 5.1636,
      "step": 4091
    },
    {
      "epoch": 0.4092,
      "grad_norm": 3.847078561782837,
      "learning_rate": 0.00014722431031469083,
      "loss": 5.089,
      "step": 4092
    },
    {
      "epoch": 0.4093,
      "grad_norm": 4.172993183135986,
      "learning_rate": 0.0001471935383883373,
      "loss": 5.2185,
      "step": 4093
    },
    {
      "epoch": 0.4094,
      "grad_norm": 2.9031856060028076,
      "learning_rate": 0.00014716276071159422,
      "loss": 4.874,
      "step": 4094
    },
    {
      "epoch": 0.4095,
      "grad_norm": 2.9634146690368652,
      "learning_rate": 0.00014713197728821183,
      "loss": 4.7625,
      "step": 4095
    },
    {
      "epoch": 0.4096,
      "grad_norm": 2.2291135787963867,
      "learning_rate": 0.000147101188121941,
      "loss": 5.3876,
      "step": 4096
    },
    {
      "epoch": 0.4097,
      "grad_norm": 3.682736396789551,
      "learning_rate": 0.0001470703932165333,
      "loss": 5.1383,
      "step": 4097
    },
    {
      "epoch": 0.4098,
      "grad_norm": 2.3433730602264404,
      "learning_rate": 0.0001470395925757409,
      "loss": 5.4018,
      "step": 4098
    },
    {
      "epoch": 0.4099,
      "grad_norm": 2.7669458389282227,
      "learning_rate": 0.00014700878620331684,
      "loss": 5.0183,
      "step": 4099
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.700169563293457,
      "learning_rate": 0.0001469779741030148,
      "loss": 4.8831,
      "step": 4100
    },
    {
      "epoch": 0.4101,
      "grad_norm": 5.043920993804932,
      "learning_rate": 0.00014694715627858908,
      "loss": 4.7203,
      "step": 4101
    },
    {
      "epoch": 0.4102,
      "grad_norm": 5.738803863525391,
      "learning_rate": 0.0001469163327337948,
      "loss": 5.5991,
      "step": 4102
    },
    {
      "epoch": 0.4103,
      "grad_norm": 2.937345266342163,
      "learning_rate": 0.0001468855034723877,
      "loss": 4.9128,
      "step": 4103
    },
    {
      "epoch": 0.4104,
      "grad_norm": 3.2656750679016113,
      "learning_rate": 0.00014685466849812418,
      "loss": 4.7321,
      "step": 4104
    },
    {
      "epoch": 0.4105,
      "grad_norm": 2.654353141784668,
      "learning_rate": 0.00014682382781476146,
      "loss": 5.0164,
      "step": 4105
    },
    {
      "epoch": 0.4106,
      "grad_norm": 2.6066949367523193,
      "learning_rate": 0.00014679298142605734,
      "loss": 4.9946,
      "step": 4106
    },
    {
      "epoch": 0.4107,
      "grad_norm": 6.463015079498291,
      "learning_rate": 0.0001467621293357704,
      "loss": 4.7857,
      "step": 4107
    },
    {
      "epoch": 0.4108,
      "grad_norm": 4.039083003997803,
      "learning_rate": 0.0001467312715476598,
      "loss": 5.417,
      "step": 4108
    },
    {
      "epoch": 0.4109,
      "grad_norm": 5.070590019226074,
      "learning_rate": 0.00014670040806548555,
      "loss": 5.7272,
      "step": 4109
    },
    {
      "epoch": 0.411,
      "grad_norm": 3.0557405948638916,
      "learning_rate": 0.0001466695388930082,
      "loss": 4.9559,
      "step": 4110
    },
    {
      "epoch": 0.4111,
      "grad_norm": 2.5316503047943115,
      "learning_rate": 0.00014663866403398913,
      "loss": 5.1496,
      "step": 4111
    },
    {
      "epoch": 0.4112,
      "grad_norm": 6.975402355194092,
      "learning_rate": 0.0001466077834921903,
      "loss": 4.6811,
      "step": 4112
    },
    {
      "epoch": 0.4113,
      "grad_norm": 4.7033772468566895,
      "learning_rate": 0.00014657689727137443,
      "loss": 5.457,
      "step": 4113
    },
    {
      "epoch": 0.4114,
      "grad_norm": 2.245906114578247,
      "learning_rate": 0.0001465460053753049,
      "loss": 4.6013,
      "step": 4114
    },
    {
      "epoch": 0.4115,
      "grad_norm": 9.264069557189941,
      "learning_rate": 0.00014651510780774583,
      "loss": 4.8599,
      "step": 4115
    },
    {
      "epoch": 0.4116,
      "grad_norm": 6.429627418518066,
      "learning_rate": 0.000146484204572462,
      "loss": 5.2208,
      "step": 4116
    },
    {
      "epoch": 0.4117,
      "grad_norm": 1.9331212043762207,
      "learning_rate": 0.0001464532956732188,
      "loss": 4.7583,
      "step": 4117
    },
    {
      "epoch": 0.4118,
      "grad_norm": 5.802704811096191,
      "learning_rate": 0.0001464223811137824,
      "loss": 4.7651,
      "step": 4118
    },
    {
      "epoch": 0.4119,
      "grad_norm": 2.4675140380859375,
      "learning_rate": 0.0001463914608979197,
      "loss": 5.6917,
      "step": 4119
    },
    {
      "epoch": 0.412,
      "grad_norm": 3.0371909141540527,
      "learning_rate": 0.00014636053502939823,
      "loss": 4.8984,
      "step": 4120
    },
    {
      "epoch": 0.4121,
      "grad_norm": 3.9780890941619873,
      "learning_rate": 0.00014632960351198618,
      "loss": 5.0293,
      "step": 4121
    },
    {
      "epoch": 0.4122,
      "grad_norm": 4.721348285675049,
      "learning_rate": 0.00014629866634945248,
      "loss": 5.6802,
      "step": 4122
    },
    {
      "epoch": 0.4123,
      "grad_norm": 3.4813029766082764,
      "learning_rate": 0.0001462677235455667,
      "loss": 5.0441,
      "step": 4123
    },
    {
      "epoch": 0.4124,
      "grad_norm": 2.976518154144287,
      "learning_rate": 0.00014623677510409918,
      "loss": 4.6201,
      "step": 4124
    },
    {
      "epoch": 0.4125,
      "grad_norm": 2.3840243816375732,
      "learning_rate": 0.00014620582102882089,
      "loss": 4.7022,
      "step": 4125
    },
    {
      "epoch": 0.4126,
      "grad_norm": 2.645148992538452,
      "learning_rate": 0.00014617486132350343,
      "loss": 5.1558,
      "step": 4126
    },
    {
      "epoch": 0.4127,
      "grad_norm": 3.6750411987304688,
      "learning_rate": 0.00014614389599191917,
      "loss": 4.9231,
      "step": 4127
    },
    {
      "epoch": 0.4128,
      "grad_norm": 3.376936435699463,
      "learning_rate": 0.00014611292503784117,
      "loss": 4.8655,
      "step": 4128
    },
    {
      "epoch": 0.4129,
      "grad_norm": 3.4537672996520996,
      "learning_rate": 0.0001460819484650431,
      "loss": 4.9964,
      "step": 4129
    },
    {
      "epoch": 0.413,
      "grad_norm": 5.113182544708252,
      "learning_rate": 0.0001460509662772994,
      "loss": 5.089,
      "step": 4130
    },
    {
      "epoch": 0.4131,
      "grad_norm": 2.223529100418091,
      "learning_rate": 0.00014601997847838518,
      "loss": 4.9053,
      "step": 4131
    },
    {
      "epoch": 0.4132,
      "grad_norm": 2.5935683250427246,
      "learning_rate": 0.00014598898507207615,
      "loss": 5.1055,
      "step": 4132
    },
    {
      "epoch": 0.4133,
      "grad_norm": 4.139305114746094,
      "learning_rate": 0.00014595798606214882,
      "loss": 5.9694,
      "step": 4133
    },
    {
      "epoch": 0.4134,
      "grad_norm": 1.9602690935134888,
      "learning_rate": 0.00014592698145238028,
      "loss": 5.2462,
      "step": 4134
    },
    {
      "epoch": 0.4135,
      "grad_norm": 3.054934024810791,
      "learning_rate": 0.00014589597124654833,
      "loss": 4.9007,
      "step": 4135
    },
    {
      "epoch": 0.4136,
      "grad_norm": 2.393120050430298,
      "learning_rate": 0.00014586495544843152,
      "loss": 4.8812,
      "step": 4136
    },
    {
      "epoch": 0.4137,
      "grad_norm": 5.621335029602051,
      "learning_rate": 0.000145833934061809,
      "loss": 5.8965,
      "step": 4137
    },
    {
      "epoch": 0.4138,
      "grad_norm": 3.2366087436676025,
      "learning_rate": 0.00014580290709046066,
      "loss": 5.1674,
      "step": 4138
    },
    {
      "epoch": 0.4139,
      "grad_norm": 3.3792924880981445,
      "learning_rate": 0.000145771874538167,
      "loss": 4.4539,
      "step": 4139
    },
    {
      "epoch": 0.414,
      "grad_norm": 3.0041022300720215,
      "learning_rate": 0.0001457408364087093,
      "loss": 5.6798,
      "step": 4140
    },
    {
      "epoch": 0.4141,
      "grad_norm": 2.1648027896881104,
      "learning_rate": 0.00014570979270586945,
      "loss": 4.8232,
      "step": 4141
    },
    {
      "epoch": 0.4142,
      "grad_norm": 5.445244312286377,
      "learning_rate": 0.00014567874343342997,
      "loss": 5.0286,
      "step": 4142
    },
    {
      "epoch": 0.4143,
      "grad_norm": 2.1667776107788086,
      "learning_rate": 0.00014564768859517418,
      "loss": 4.893,
      "step": 4143
    },
    {
      "epoch": 0.4144,
      "grad_norm": 2.57955002784729,
      "learning_rate": 0.00014561662819488597,
      "loss": 4.813,
      "step": 4144
    },
    {
      "epoch": 0.4145,
      "grad_norm": 2.978856325149536,
      "learning_rate": 0.00014558556223635003,
      "loss": 5.0868,
      "step": 4145
    },
    {
      "epoch": 0.4146,
      "grad_norm": 2.2246599197387695,
      "learning_rate": 0.00014555449072335157,
      "loss": 5.2371,
      "step": 4146
    },
    {
      "epoch": 0.4147,
      "grad_norm": 1.7733070850372314,
      "learning_rate": 0.00014552341365967658,
      "loss": 4.629,
      "step": 4147
    },
    {
      "epoch": 0.4148,
      "grad_norm": 3.0163633823394775,
      "learning_rate": 0.00014549233104911178,
      "loss": 4.5599,
      "step": 4148
    },
    {
      "epoch": 0.4149,
      "grad_norm": 2.0600383281707764,
      "learning_rate": 0.0001454612428954444,
      "loss": 4.8155,
      "step": 4149
    },
    {
      "epoch": 0.415,
      "grad_norm": 5.304862022399902,
      "learning_rate": 0.00014543014920246247,
      "loss": 5.147,
      "step": 4150
    },
    {
      "epoch": 0.4151,
      "grad_norm": 2.473590612411499,
      "learning_rate": 0.00014539904997395468,
      "loss": 5.0594,
      "step": 4151
    },
    {
      "epoch": 0.4152,
      "grad_norm": 2.4342246055603027,
      "learning_rate": 0.00014536794521371037,
      "loss": 4.8906,
      "step": 4152
    },
    {
      "epoch": 0.4153,
      "grad_norm": 2.3127167224884033,
      "learning_rate": 0.00014533683492551952,
      "loss": 5.3994,
      "step": 4153
    },
    {
      "epoch": 0.4154,
      "grad_norm": 2.6966347694396973,
      "learning_rate": 0.0001453057191131729,
      "loss": 5.1337,
      "step": 4154
    },
    {
      "epoch": 0.4155,
      "grad_norm": 2.3263769149780273,
      "learning_rate": 0.0001452745977804618,
      "loss": 5.2005,
      "step": 4155
    },
    {
      "epoch": 0.4156,
      "grad_norm": 2.9710288047790527,
      "learning_rate": 0.00014524347093117828,
      "loss": 4.6293,
      "step": 4156
    },
    {
      "epoch": 0.4157,
      "grad_norm": 2.423295021057129,
      "learning_rate": 0.00014521233856911508,
      "loss": 5.0535,
      "step": 4157
    },
    {
      "epoch": 0.4158,
      "grad_norm": 2.6867682933807373,
      "learning_rate": 0.00014518120069806557,
      "loss": 5.5303,
      "step": 4158
    },
    {
      "epoch": 0.4159,
      "grad_norm": 2.0550599098205566,
      "learning_rate": 0.00014515005732182383,
      "loss": 5.1211,
      "step": 4159
    },
    {
      "epoch": 0.416,
      "grad_norm": 5.190590858459473,
      "learning_rate": 0.00014511890844418453,
      "loss": 4.8865,
      "step": 4160
    },
    {
      "epoch": 0.4161,
      "grad_norm": 10.862092018127441,
      "learning_rate": 0.00014508775406894307,
      "loss": 5.6484,
      "step": 4161
    },
    {
      "epoch": 0.4162,
      "grad_norm": 6.157359600067139,
      "learning_rate": 0.0001450565941998956,
      "loss": 4.9519,
      "step": 4162
    },
    {
      "epoch": 0.4163,
      "grad_norm": 5.327624797821045,
      "learning_rate": 0.00014502542884083875,
      "loss": 5.6616,
      "step": 4163
    },
    {
      "epoch": 0.4164,
      "grad_norm": 7.581103324890137,
      "learning_rate": 0.00014499425799557,
      "loss": 5.2624,
      "step": 4164
    },
    {
      "epoch": 0.4165,
      "grad_norm": 3.03470778465271,
      "learning_rate": 0.0001449630816678874,
      "loss": 4.9687,
      "step": 4165
    },
    {
      "epoch": 0.4166,
      "grad_norm": 3.156670570373535,
      "learning_rate": 0.00014493189986158965,
      "loss": 5.8989,
      "step": 4166
    },
    {
      "epoch": 0.4167,
      "grad_norm": 2.475576400756836,
      "learning_rate": 0.00014490071258047623,
      "loss": 5.4577,
      "step": 4167
    },
    {
      "epoch": 0.4168,
      "grad_norm": 2.2756357192993164,
      "learning_rate": 0.0001448695198283472,
      "loss": 5.2097,
      "step": 4168
    },
    {
      "epoch": 0.4169,
      "grad_norm": 6.8662109375,
      "learning_rate": 0.00014483832160900326,
      "loss": 5.1938,
      "step": 4169
    },
    {
      "epoch": 0.417,
      "grad_norm": 2.598834276199341,
      "learning_rate": 0.0001448071179262458,
      "loss": 4.7651,
      "step": 4170
    },
    {
      "epoch": 0.4171,
      "grad_norm": 4.396021366119385,
      "learning_rate": 0.00014477590878387696,
      "loss": 5.309,
      "step": 4171
    },
    {
      "epoch": 0.4172,
      "grad_norm": 2.4157073497772217,
      "learning_rate": 0.0001447446941856995,
      "loss": 4.8121,
      "step": 4172
    },
    {
      "epoch": 0.4173,
      "grad_norm": 4.871314525604248,
      "learning_rate": 0.00014471347413551672,
      "loss": 4.4658,
      "step": 4173
    },
    {
      "epoch": 0.4174,
      "grad_norm": 4.735869884490967,
      "learning_rate": 0.00014468224863713278,
      "loss": 4.6879,
      "step": 4174
    },
    {
      "epoch": 0.4175,
      "grad_norm": 2.209685802459717,
      "learning_rate": 0.00014465101769435234,
      "loss": 5.0071,
      "step": 4175
    },
    {
      "epoch": 0.4176,
      "grad_norm": 3.1700308322906494,
      "learning_rate": 0.00014461978131098088,
      "loss": 5.2072,
      "step": 4176
    },
    {
      "epoch": 0.4177,
      "grad_norm": 1.8833427429199219,
      "learning_rate": 0.00014458853949082443,
      "loss": 4.7506,
      "step": 4177
    },
    {
      "epoch": 0.4178,
      "grad_norm": 3.8744068145751953,
      "learning_rate": 0.00014455729223768966,
      "loss": 5.402,
      "step": 4178
    },
    {
      "epoch": 0.4179,
      "grad_norm": 3.0460245609283447,
      "learning_rate": 0.00014452603955538397,
      "loss": 5.2696,
      "step": 4179
    },
    {
      "epoch": 0.418,
      "grad_norm": 1.9978291988372803,
      "learning_rate": 0.00014449478144771543,
      "loss": 4.6321,
      "step": 4180
    },
    {
      "epoch": 0.4181,
      "grad_norm": 1.9552379846572876,
      "learning_rate": 0.00014446351791849276,
      "loss": 5.2328,
      "step": 4181
    },
    {
      "epoch": 0.4182,
      "grad_norm": 5.9631781578063965,
      "learning_rate": 0.0001444322489715253,
      "loss": 5.1671,
      "step": 4182
    },
    {
      "epoch": 0.4183,
      "grad_norm": 3.3632991313934326,
      "learning_rate": 0.00014440097461062307,
      "loss": 5.1497,
      "step": 4183
    },
    {
      "epoch": 0.4184,
      "grad_norm": 3.715115785598755,
      "learning_rate": 0.00014436969483959676,
      "loss": 4.8275,
      "step": 4184
    },
    {
      "epoch": 0.4185,
      "grad_norm": 4.10849666595459,
      "learning_rate": 0.00014433840966225772,
      "loss": 5.2454,
      "step": 4185
    },
    {
      "epoch": 0.4186,
      "grad_norm": 9.031224250793457,
      "learning_rate": 0.00014430711908241798,
      "loss": 6.1855,
      "step": 4186
    },
    {
      "epoch": 0.4187,
      "grad_norm": 2.2069735527038574,
      "learning_rate": 0.0001442758231038902,
      "loss": 4.8756,
      "step": 4187
    },
    {
      "epoch": 0.4188,
      "grad_norm": 2.8955488204956055,
      "learning_rate": 0.0001442445217304876,
      "loss": 4.8325,
      "step": 4188
    },
    {
      "epoch": 0.4189,
      "grad_norm": 3.386857271194458,
      "learning_rate": 0.00014421321496602428,
      "loss": 5.2448,
      "step": 4189
    },
    {
      "epoch": 0.419,
      "grad_norm": 5.326070785522461,
      "learning_rate": 0.00014418190281431482,
      "loss": 4.389,
      "step": 4190
    },
    {
      "epoch": 0.4191,
      "grad_norm": 4.332716464996338,
      "learning_rate": 0.00014415058527917452,
      "loss": 5.7538,
      "step": 4191
    },
    {
      "epoch": 0.4192,
      "grad_norm": 2.866269826889038,
      "learning_rate": 0.00014411926236441934,
      "loss": 4.737,
      "step": 4192
    },
    {
      "epoch": 0.4193,
      "grad_norm": 3.2844512462615967,
      "learning_rate": 0.00014408793407386588,
      "loss": 5.0501,
      "step": 4193
    },
    {
      "epoch": 0.4194,
      "grad_norm": 4.323998928070068,
      "learning_rate": 0.00014405660041133132,
      "loss": 5.1716,
      "step": 4194
    },
    {
      "epoch": 0.4195,
      "grad_norm": 25.48111343383789,
      "learning_rate": 0.00014402526138063373,
      "loss": 5.6892,
      "step": 4195
    },
    {
      "epoch": 0.4196,
      "grad_norm": 17.35052490234375,
      "learning_rate": 0.00014399391698559152,
      "loss": 5.4979,
      "step": 4196
    },
    {
      "epoch": 0.4197,
      "grad_norm": 4.511854648590088,
      "learning_rate": 0.000143962567230024,
      "loss": 5.4525,
      "step": 4197
    },
    {
      "epoch": 0.4198,
      "grad_norm": 7.90883731842041,
      "learning_rate": 0.000143931212117751,
      "loss": 5.4397,
      "step": 4198
    },
    {
      "epoch": 0.4199,
      "grad_norm": 1.8736588954925537,
      "learning_rate": 0.00014389985165259308,
      "loss": 4.8693,
      "step": 4199
    },
    {
      "epoch": 0.42,
      "grad_norm": 4.364076137542725,
      "learning_rate": 0.0001438684858383714,
      "loss": 5.2494,
      "step": 4200
    },
    {
      "epoch": 0.4201,
      "grad_norm": 2.017749547958374,
      "learning_rate": 0.00014383711467890774,
      "loss": 4.6243,
      "step": 4201
    },
    {
      "epoch": 0.4202,
      "grad_norm": 5.705942153930664,
      "learning_rate": 0.00014380573817802467,
      "loss": 4.9202,
      "step": 4202
    },
    {
      "epoch": 0.4203,
      "grad_norm": 2.486443281173706,
      "learning_rate": 0.00014377435633954527,
      "loss": 4.946,
      "step": 4203
    },
    {
      "epoch": 0.4204,
      "grad_norm": 3.604522466659546,
      "learning_rate": 0.00014374296916729336,
      "loss": 6.2542,
      "step": 4204
    },
    {
      "epoch": 0.4205,
      "grad_norm": 2.547013759613037,
      "learning_rate": 0.0001437115766650933,
      "loss": 4.8211,
      "step": 4205
    },
    {
      "epoch": 0.4206,
      "grad_norm": 2.4177160263061523,
      "learning_rate": 0.00014368017883677024,
      "loss": 4.6795,
      "step": 4206
    },
    {
      "epoch": 0.4207,
      "grad_norm": 3.121065378189087,
      "learning_rate": 0.0001436487756861499,
      "loss": 5.0025,
      "step": 4207
    },
    {
      "epoch": 0.4208,
      "grad_norm": 2.07918381690979,
      "learning_rate": 0.0001436173672170586,
      "loss": 4.5855,
      "step": 4208
    },
    {
      "epoch": 0.4209,
      "grad_norm": 2.0390841960906982,
      "learning_rate": 0.00014358595343332342,
      "loss": 4.7408,
      "step": 4209
    },
    {
      "epoch": 0.421,
      "grad_norm": 3.4071013927459717,
      "learning_rate": 0.00014355453433877204,
      "loss": 5.6317,
      "step": 4210
    },
    {
      "epoch": 0.4211,
      "grad_norm": 2.493692636489868,
      "learning_rate": 0.00014352310993723277,
      "loss": 4.6641,
      "step": 4211
    },
    {
      "epoch": 0.4212,
      "grad_norm": 3.0606958866119385,
      "learning_rate": 0.00014349168023253456,
      "loss": 4.9002,
      "step": 4212
    },
    {
      "epoch": 0.4213,
      "grad_norm": 2.5728604793548584,
      "learning_rate": 0.00014346024522850703,
      "loss": 5.1609,
      "step": 4213
    },
    {
      "epoch": 0.4214,
      "grad_norm": 3.503826379776001,
      "learning_rate": 0.00014342880492898048,
      "loss": 5.6174,
      "step": 4214
    },
    {
      "epoch": 0.4215,
      "grad_norm": 3.8273327350616455,
      "learning_rate": 0.00014339735933778576,
      "loss": 5.6939,
      "step": 4215
    },
    {
      "epoch": 0.4216,
      "grad_norm": 6.188591003417969,
      "learning_rate": 0.00014336590845875446,
      "loss": 5.8361,
      "step": 4216
    },
    {
      "epoch": 0.4217,
      "grad_norm": 2.2473652362823486,
      "learning_rate": 0.00014333445229571873,
      "loss": 5.063,
      "step": 4217
    },
    {
      "epoch": 0.4218,
      "grad_norm": 4.85899019241333,
      "learning_rate": 0.00014330299085251144,
      "loss": 6.114,
      "step": 4218
    },
    {
      "epoch": 0.4219,
      "grad_norm": 2.304004430770874,
      "learning_rate": 0.00014327152413296608,
      "loss": 4.6266,
      "step": 4219
    },
    {
      "epoch": 0.422,
      "grad_norm": 4.884314060211182,
      "learning_rate": 0.00014324005214091676,
      "loss": 5.2977,
      "step": 4220
    },
    {
      "epoch": 0.4221,
      "grad_norm": 2.587482213973999,
      "learning_rate": 0.00014320857488019824,
      "loss": 4.7354,
      "step": 4221
    },
    {
      "epoch": 0.4222,
      "grad_norm": 3.470449447631836,
      "learning_rate": 0.00014317709235464593,
      "loss": 4.8,
      "step": 4222
    },
    {
      "epoch": 0.4223,
      "grad_norm": 3.361725091934204,
      "learning_rate": 0.0001431456045680959,
      "loss": 4.4733,
      "step": 4223
    },
    {
      "epoch": 0.4224,
      "grad_norm": 3.5648324489593506,
      "learning_rate": 0.00014311411152438482,
      "loss": 4.8965,
      "step": 4224
    },
    {
      "epoch": 0.4225,
      "grad_norm": 2.6036431789398193,
      "learning_rate": 0.00014308261322735005,
      "loss": 4.8334,
      "step": 4225
    },
    {
      "epoch": 0.4226,
      "grad_norm": 3.6999778747558594,
      "learning_rate": 0.00014305110968082952,
      "loss": 5.6307,
      "step": 4226
    },
    {
      "epoch": 0.4227,
      "grad_norm": 3.401167631149292,
      "learning_rate": 0.00014301960088866186,
      "loss": 5.2205,
      "step": 4227
    },
    {
      "epoch": 0.4228,
      "grad_norm": 2.0055460929870605,
      "learning_rate": 0.00014298808685468635,
      "loss": 5.1079,
      "step": 4228
    },
    {
      "epoch": 0.4229,
      "grad_norm": 4.944057464599609,
      "learning_rate": 0.00014295656758274284,
      "loss": 5.8042,
      "step": 4229
    },
    {
      "epoch": 0.423,
      "grad_norm": 3.0208210945129395,
      "learning_rate": 0.00014292504307667186,
      "loss": 4.5495,
      "step": 4230
    },
    {
      "epoch": 0.4231,
      "grad_norm": 2.2505030632019043,
      "learning_rate": 0.0001428935133403146,
      "loss": 5.0509,
      "step": 4231
    },
    {
      "epoch": 0.4232,
      "grad_norm": 8.172453880310059,
      "learning_rate": 0.00014286197837751286,
      "loss": 5.2934,
      "step": 4232
    },
    {
      "epoch": 0.4233,
      "grad_norm": 6.319840431213379,
      "learning_rate": 0.00014283043819210905,
      "loss": 5.6302,
      "step": 4233
    },
    {
      "epoch": 0.4234,
      "grad_norm": 3.102597713470459,
      "learning_rate": 0.00014279889278794627,
      "loss": 4.8255,
      "step": 4234
    },
    {
      "epoch": 0.4235,
      "grad_norm": 8.533182144165039,
      "learning_rate": 0.00014276734216886821,
      "loss": 6.6114,
      "step": 4235
    },
    {
      "epoch": 0.4236,
      "grad_norm": 3.3816308975219727,
      "learning_rate": 0.00014273578633871927,
      "loss": 5.2371,
      "step": 4236
    },
    {
      "epoch": 0.4237,
      "grad_norm": 2.971076726913452,
      "learning_rate": 0.00014270422530134432,
      "loss": 4.9231,
      "step": 4237
    },
    {
      "epoch": 0.4238,
      "grad_norm": 3.473162889480591,
      "learning_rate": 0.00014267265906058914,
      "loss": 5.1316,
      "step": 4238
    },
    {
      "epoch": 0.4239,
      "grad_norm": 5.557267665863037,
      "learning_rate": 0.0001426410876202999,
      "loss": 5.0546,
      "step": 4239
    },
    {
      "epoch": 0.424,
      "grad_norm": 4.972139358520508,
      "learning_rate": 0.00014260951098432343,
      "loss": 5.8384,
      "step": 4240
    },
    {
      "epoch": 0.4241,
      "grad_norm": 2.046356201171875,
      "learning_rate": 0.00014257792915650728,
      "loss": 5.0239,
      "step": 4241
    },
    {
      "epoch": 0.4242,
      "grad_norm": 2.653825283050537,
      "learning_rate": 0.00014254634214069963,
      "loss": 4.7236,
      "step": 4242
    },
    {
      "epoch": 0.4243,
      "grad_norm": 2.5631282329559326,
      "learning_rate": 0.00014251474994074928,
      "loss": 4.7409,
      "step": 4243
    },
    {
      "epoch": 0.4244,
      "grad_norm": 3.911489963531494,
      "learning_rate": 0.00014248315256050557,
      "loss": 5.0813,
      "step": 4244
    },
    {
      "epoch": 0.4245,
      "grad_norm": 3.698472023010254,
      "learning_rate": 0.0001424515500038186,
      "loss": 5.0495,
      "step": 4245
    },
    {
      "epoch": 0.4246,
      "grad_norm": 1.796613097190857,
      "learning_rate": 0.00014241994227453901,
      "loss": 5.0445,
      "step": 4246
    },
    {
      "epoch": 0.4247,
      "grad_norm": 3.3749449253082275,
      "learning_rate": 0.00014238832937651816,
      "loss": 4.8998,
      "step": 4247
    },
    {
      "epoch": 0.4248,
      "grad_norm": 4.540114402770996,
      "learning_rate": 0.00014235671131360798,
      "loss": 5.0929,
      "step": 4248
    },
    {
      "epoch": 0.4249,
      "grad_norm": 2.6910064220428467,
      "learning_rate": 0.00014232508808966098,
      "loss": 5.0963,
      "step": 4249
    },
    {
      "epoch": 0.425,
      "grad_norm": 2.9471757411956787,
      "learning_rate": 0.00014229345970853032,
      "loss": 4.6466,
      "step": 4250
    },
    {
      "epoch": 0.4251,
      "grad_norm": 2.8972439765930176,
      "learning_rate": 0.00014226182617406996,
      "loss": 5.188,
      "step": 4251
    },
    {
      "epoch": 0.4252,
      "grad_norm": 5.198869705200195,
      "learning_rate": 0.00014223018749013423,
      "loss": 5.3915,
      "step": 4252
    },
    {
      "epoch": 0.4253,
      "grad_norm": 5.034494876861572,
      "learning_rate": 0.0001421985436605783,
      "loss": 5.3579,
      "step": 4253
    },
    {
      "epoch": 0.4254,
      "grad_norm": 3.3608686923980713,
      "learning_rate": 0.0001421668946892578,
      "loss": 5.1536,
      "step": 4254
    },
    {
      "epoch": 0.4255,
      "grad_norm": 3.0403285026550293,
      "learning_rate": 0.0001421352405800291,
      "loss": 4.876,
      "step": 4255
    },
    {
      "epoch": 0.4256,
      "grad_norm": 2.3115618228912354,
      "learning_rate": 0.00014210358133674912,
      "loss": 4.9997,
      "step": 4256
    },
    {
      "epoch": 0.4257,
      "grad_norm": 2.971242666244507,
      "learning_rate": 0.00014207191696327548,
      "loss": 4.9789,
      "step": 4257
    },
    {
      "epoch": 0.4258,
      "grad_norm": 6.545262813568115,
      "learning_rate": 0.00014204024746346637,
      "loss": 4.5238,
      "step": 4258
    },
    {
      "epoch": 0.4259,
      "grad_norm": 2.087585210800171,
      "learning_rate": 0.00014200857284118066,
      "loss": 5.2582,
      "step": 4259
    },
    {
      "epoch": 0.426,
      "grad_norm": 2.9625823497772217,
      "learning_rate": 0.00014197689310027772,
      "loss": 5.5173,
      "step": 4260
    },
    {
      "epoch": 0.4261,
      "grad_norm": 3.411710023880005,
      "learning_rate": 0.00014194520824461771,
      "loss": 5.1994,
      "step": 4261
    },
    {
      "epoch": 0.4262,
      "grad_norm": 2.3999557495117188,
      "learning_rate": 0.00014191351827806133,
      "loss": 4.9448,
      "step": 4262
    },
    {
      "epoch": 0.4263,
      "grad_norm": 2.6709063053131104,
      "learning_rate": 0.00014188182320446985,
      "loss": 5.2738,
      "step": 4263
    },
    {
      "epoch": 0.4264,
      "grad_norm": 3.824464797973633,
      "learning_rate": 0.00014185012302770527,
      "loss": 6.034,
      "step": 4264
    },
    {
      "epoch": 0.4265,
      "grad_norm": 2.635446071624756,
      "learning_rate": 0.00014181841775163013,
      "loss": 5.0189,
      "step": 4265
    },
    {
      "epoch": 0.4266,
      "grad_norm": 3.6146750450134277,
      "learning_rate": 0.0001417867073801077,
      "loss": 5.2065,
      "step": 4266
    },
    {
      "epoch": 0.4267,
      "grad_norm": 3.024960994720459,
      "learning_rate": 0.00014175499191700167,
      "loss": 5.6824,
      "step": 4267
    },
    {
      "epoch": 0.4268,
      "grad_norm": 4.852941036224365,
      "learning_rate": 0.00014172327136617656,
      "loss": 5.723,
      "step": 4268
    },
    {
      "epoch": 0.4269,
      "grad_norm": 4.369772911071777,
      "learning_rate": 0.00014169154573149737,
      "loss": 5.2212,
      "step": 4269
    },
    {
      "epoch": 0.427,
      "grad_norm": 2.700202465057373,
      "learning_rate": 0.0001416598150168298,
      "loss": 4.8375,
      "step": 4270
    },
    {
      "epoch": 0.4271,
      "grad_norm": 2.9254045486450195,
      "learning_rate": 0.00014162807922604012,
      "loss": 5.568,
      "step": 4271
    },
    {
      "epoch": 0.4272,
      "grad_norm": 3.1988437175750732,
      "learning_rate": 0.00014159633836299527,
      "loss": 4.7438,
      "step": 4272
    },
    {
      "epoch": 0.4273,
      "grad_norm": 2.378045082092285,
      "learning_rate": 0.0001415645924315628,
      "loss": 4.783,
      "step": 4273
    },
    {
      "epoch": 0.4274,
      "grad_norm": 3.5385255813598633,
      "learning_rate": 0.0001415328414356108,
      "loss": 5.6747,
      "step": 4274
    },
    {
      "epoch": 0.4275,
      "grad_norm": 4.935507774353027,
      "learning_rate": 0.00014150108537900805,
      "loss": 5.1808,
      "step": 4275
    },
    {
      "epoch": 0.4276,
      "grad_norm": 3.4887516498565674,
      "learning_rate": 0.00014146932426562392,
      "loss": 5.7163,
      "step": 4276
    },
    {
      "epoch": 0.4277,
      "grad_norm": 5.720993995666504,
      "learning_rate": 0.00014143755809932845,
      "loss": 5.2773,
      "step": 4277
    },
    {
      "epoch": 0.4278,
      "grad_norm": 2.5807182788848877,
      "learning_rate": 0.00014140578688399218,
      "loss": 4.8477,
      "step": 4278
    },
    {
      "epoch": 0.4279,
      "grad_norm": 3.185429096221924,
      "learning_rate": 0.00014137401062348638,
      "loss": 4.7729,
      "step": 4279
    },
    {
      "epoch": 0.428,
      "grad_norm": 2.927767038345337,
      "learning_rate": 0.0001413422293216829,
      "loss": 4.9136,
      "step": 4280
    },
    {
      "epoch": 0.4281,
      "grad_norm": 4.444439888000488,
      "learning_rate": 0.0001413104429824542,
      "loss": 5.2555,
      "step": 4281
    },
    {
      "epoch": 0.4282,
      "grad_norm": 2.778923988342285,
      "learning_rate": 0.0001412786516096733,
      "loss": 5.0951,
      "step": 4282
    },
    {
      "epoch": 0.4283,
      "grad_norm": 2.6715681552886963,
      "learning_rate": 0.00014124685520721392,
      "loss": 5.3247,
      "step": 4283
    },
    {
      "epoch": 0.4284,
      "grad_norm": 4.6849470138549805,
      "learning_rate": 0.00014121505377895038,
      "loss": 5.6127,
      "step": 4284
    },
    {
      "epoch": 0.4285,
      "grad_norm": 2.939182996749878,
      "learning_rate": 0.00014118324732875748,
      "loss": 4.6431,
      "step": 4285
    },
    {
      "epoch": 0.4286,
      "grad_norm": 3.2252519130706787,
      "learning_rate": 0.00014115143586051088,
      "loss": 4.8798,
      "step": 4286
    },
    {
      "epoch": 0.4287,
      "grad_norm": 2.3576693534851074,
      "learning_rate": 0.00014111961937808665,
      "loss": 4.9756,
      "step": 4287
    },
    {
      "epoch": 0.4288,
      "grad_norm": 2.0043563842773438,
      "learning_rate": 0.0001410877978853615,
      "loss": 4.9042,
      "step": 4288
    },
    {
      "epoch": 0.4289,
      "grad_norm": 2.564906120300293,
      "learning_rate": 0.0001410559713862128,
      "loss": 5.3158,
      "step": 4289
    },
    {
      "epoch": 0.429,
      "grad_norm": 8.103261947631836,
      "learning_rate": 0.00014102413988451856,
      "loss": 4.9532,
      "step": 4290
    },
    {
      "epoch": 0.4291,
      "grad_norm": 2.2829105854034424,
      "learning_rate": 0.00014099230338415728,
      "loss": 5.1542,
      "step": 4291
    },
    {
      "epoch": 0.4292,
      "grad_norm": 4.063913822174072,
      "learning_rate": 0.00014096046188900822,
      "loss": 4.6422,
      "step": 4292
    },
    {
      "epoch": 0.4293,
      "grad_norm": 2.718841314315796,
      "learning_rate": 0.00014092861540295108,
      "loss": 4.6528,
      "step": 4293
    },
    {
      "epoch": 0.4294,
      "grad_norm": 3.00895094871521,
      "learning_rate": 0.0001408967639298663,
      "loss": 4.4253,
      "step": 4294
    },
    {
      "epoch": 0.4295,
      "grad_norm": 4.405314922332764,
      "learning_rate": 0.00014086490747363493,
      "loss": 5.8466,
      "step": 4295
    },
    {
      "epoch": 0.4296,
      "grad_norm": 2.566592216491699,
      "learning_rate": 0.00014083304603813848,
      "loss": 4.8411,
      "step": 4296
    },
    {
      "epoch": 0.4297,
      "grad_norm": 3.2199065685272217,
      "learning_rate": 0.0001408011796272593,
      "loss": 5.203,
      "step": 4297
    },
    {
      "epoch": 0.4298,
      "grad_norm": 9.752431869506836,
      "learning_rate": 0.0001407693082448801,
      "loss": 5.2772,
      "step": 4298
    },
    {
      "epoch": 0.4299,
      "grad_norm": 3.512995481491089,
      "learning_rate": 0.00014073743189488435,
      "loss": 5.2814,
      "step": 4299
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.4843592643737793,
      "learning_rate": 0.00014070555058115614,
      "loss": 4.4638,
      "step": 4300
    },
    {
      "epoch": 0.4301,
      "grad_norm": 7.728984355926514,
      "learning_rate": 0.00014067366430758004,
      "loss": 5.6284,
      "step": 4301
    },
    {
      "epoch": 0.4302,
      "grad_norm": 4.077146053314209,
      "learning_rate": 0.00014064177307804133,
      "loss": 4.8211,
      "step": 4302
    },
    {
      "epoch": 0.4303,
      "grad_norm": 5.723459243774414,
      "learning_rate": 0.00014060987689642581,
      "loss": 4.894,
      "step": 4303
    },
    {
      "epoch": 0.4304,
      "grad_norm": 2.5143914222717285,
      "learning_rate": 0.00014057797576662,
      "loss": 4.8694,
      "step": 4304
    },
    {
      "epoch": 0.4305,
      "grad_norm": 3.4436938762664795,
      "learning_rate": 0.00014054606969251095,
      "loss": 5.7511,
      "step": 4305
    },
    {
      "epoch": 0.4306,
      "grad_norm": 7.453980445861816,
      "learning_rate": 0.00014051415867798628,
      "loss": 5.2459,
      "step": 4306
    },
    {
      "epoch": 0.4307,
      "grad_norm": 5.3914995193481445,
      "learning_rate": 0.00014048224272693424,
      "loss": 5.1545,
      "step": 4307
    },
    {
      "epoch": 0.4308,
      "grad_norm": 5.3301897048950195,
      "learning_rate": 0.00014045032184324373,
      "loss": 5.7805,
      "step": 4308
    },
    {
      "epoch": 0.4309,
      "grad_norm": 3.006605386734009,
      "learning_rate": 0.00014041839603080422,
      "loss": 4.8828,
      "step": 4309
    },
    {
      "epoch": 0.431,
      "grad_norm": 3.317019462585449,
      "learning_rate": 0.00014038646529350579,
      "loss": 4.9148,
      "step": 4310
    },
    {
      "epoch": 0.4311,
      "grad_norm": 2.703815460205078,
      "learning_rate": 0.00014035452963523902,
      "loss": 4.8058,
      "step": 4311
    },
    {
      "epoch": 0.4312,
      "grad_norm": 3.700493097305298,
      "learning_rate": 0.00014032258905989522,
      "loss": 5.7927,
      "step": 4312
    },
    {
      "epoch": 0.4313,
      "grad_norm": 4.466899871826172,
      "learning_rate": 0.00014029064357136628,
      "loss": 5.846,
      "step": 4313
    },
    {
      "epoch": 0.4314,
      "grad_norm": 2.353220224380493,
      "learning_rate": 0.00014025869317354462,
      "loss": 5.1369,
      "step": 4314
    },
    {
      "epoch": 0.4315,
      "grad_norm": 2.475616693496704,
      "learning_rate": 0.00014022673787032332,
      "loss": 5.1662,
      "step": 4315
    },
    {
      "epoch": 0.4316,
      "grad_norm": 3.363023281097412,
      "learning_rate": 0.00014019477766559604,
      "loss": 4.7797,
      "step": 4316
    },
    {
      "epoch": 0.4317,
      "grad_norm": 6.976577281951904,
      "learning_rate": 0.000140162812563257,
      "loss": 4.9607,
      "step": 4317
    },
    {
      "epoch": 0.4318,
      "grad_norm": 3.1696321964263916,
      "learning_rate": 0.00014013084256720107,
      "loss": 5.1192,
      "step": 4318
    },
    {
      "epoch": 0.4319,
      "grad_norm": 7.132504940032959,
      "learning_rate": 0.00014009886768132375,
      "loss": 5.229,
      "step": 4319
    },
    {
      "epoch": 0.432,
      "grad_norm": 4.1928582191467285,
      "learning_rate": 0.000140066887909521,
      "loss": 4.8774,
      "step": 4320
    },
    {
      "epoch": 0.4321,
      "grad_norm": 2.5700037479400635,
      "learning_rate": 0.00014003490325568954,
      "loss": 4.8867,
      "step": 4321
    },
    {
      "epoch": 0.4322,
      "grad_norm": 2.524662733078003,
      "learning_rate": 0.00014000291372372647,
      "loss": 4.9714,
      "step": 4322
    },
    {
      "epoch": 0.4323,
      "grad_norm": 12.67235279083252,
      "learning_rate": 0.00013997091931752977,
      "loss": 5.3769,
      "step": 4323
    },
    {
      "epoch": 0.4324,
      "grad_norm": 3.3916993141174316,
      "learning_rate": 0.00013993892004099777,
      "loss": 4.7674,
      "step": 4324
    },
    {
      "epoch": 0.4325,
      "grad_norm": 3.1083240509033203,
      "learning_rate": 0.00013990691589802954,
      "loss": 5.2118,
      "step": 4325
    },
    {
      "epoch": 0.4326,
      "grad_norm": 3.270618438720703,
      "learning_rate": 0.00013987490689252463,
      "loss": 5.3751,
      "step": 4326
    },
    {
      "epoch": 0.4327,
      "grad_norm": 2.1952147483825684,
      "learning_rate": 0.00013984289302838328,
      "loss": 4.3544,
      "step": 4327
    },
    {
      "epoch": 0.4328,
      "grad_norm": 3.9438469409942627,
      "learning_rate": 0.00013981087430950628,
      "loss": 6.3532,
      "step": 4328
    },
    {
      "epoch": 0.4329,
      "grad_norm": 2.5046167373657227,
      "learning_rate": 0.000139778850739795,
      "loss": 4.5774,
      "step": 4329
    },
    {
      "epoch": 0.433,
      "grad_norm": 4.254671096801758,
      "learning_rate": 0.0001397468223231514,
      "loss": 5.6035,
      "step": 4330
    },
    {
      "epoch": 0.4331,
      "grad_norm": 2.6140644550323486,
      "learning_rate": 0.00013971478906347806,
      "loss": 4.9508,
      "step": 4331
    },
    {
      "epoch": 0.4332,
      "grad_norm": 3.365114450454712,
      "learning_rate": 0.0001396827509646782,
      "loss": 5.1658,
      "step": 4332
    },
    {
      "epoch": 0.4333,
      "grad_norm": 4.110029220581055,
      "learning_rate": 0.00013965070803065543,
      "loss": 5.177,
      "step": 4333
    },
    {
      "epoch": 0.4334,
      "grad_norm": 3.2771835327148438,
      "learning_rate": 0.00013961866026531417,
      "loss": 5.1924,
      "step": 4334
    },
    {
      "epoch": 0.4335,
      "grad_norm": 4.525790691375732,
      "learning_rate": 0.00013958660767255938,
      "loss": 4.875,
      "step": 4335
    },
    {
      "epoch": 0.4336,
      "grad_norm": 2.8782236576080322,
      "learning_rate": 0.00013955455025629651,
      "loss": 5.0595,
      "step": 4336
    },
    {
      "epoch": 0.4337,
      "grad_norm": 4.313908100128174,
      "learning_rate": 0.00013952248802043165,
      "loss": 5.5826,
      "step": 4337
    },
    {
      "epoch": 0.4338,
      "grad_norm": 2.0734975337982178,
      "learning_rate": 0.00013949042096887153,
      "loss": 4.8664,
      "step": 4338
    },
    {
      "epoch": 0.4339,
      "grad_norm": 2.2555148601531982,
      "learning_rate": 0.0001394583491055234,
      "loss": 5.482,
      "step": 4339
    },
    {
      "epoch": 0.434,
      "grad_norm": 2.8123576641082764,
      "learning_rate": 0.00013942627243429512,
      "loss": 5.4159,
      "step": 4340
    },
    {
      "epoch": 0.4341,
      "grad_norm": 5.734133243560791,
      "learning_rate": 0.00013939419095909512,
      "loss": 5.1453,
      "step": 4341
    },
    {
      "epoch": 0.4342,
      "grad_norm": 4.583925247192383,
      "learning_rate": 0.00013936210468383246,
      "loss": 5.6415,
      "step": 4342
    },
    {
      "epoch": 0.4343,
      "grad_norm": 3.131018877029419,
      "learning_rate": 0.00013933001361241673,
      "loss": 4.565,
      "step": 4343
    },
    {
      "epoch": 0.4344,
      "grad_norm": 3.097849130630493,
      "learning_rate": 0.00013929791774875815,
      "loss": 5.2202,
      "step": 4344
    },
    {
      "epoch": 0.4345,
      "grad_norm": 2.0084002017974854,
      "learning_rate": 0.00013926581709676751,
      "loss": 4.7001,
      "step": 4345
    },
    {
      "epoch": 0.4346,
      "grad_norm": 3.7408249378204346,
      "learning_rate": 0.00013923371166035616,
      "loss": 4.607,
      "step": 4346
    },
    {
      "epoch": 0.4347,
      "grad_norm": 3.0506436824798584,
      "learning_rate": 0.00013920160144343603,
      "loss": 5.3782,
      "step": 4347
    },
    {
      "epoch": 0.4348,
      "grad_norm": 2.828325033187866,
      "learning_rate": 0.0001391694864499197,
      "loss": 5.3868,
      "step": 4348
    },
    {
      "epoch": 0.4349,
      "grad_norm": 2.229884386062622,
      "learning_rate": 0.00013913736668372026,
      "loss": 5.0123,
      "step": 4349
    },
    {
      "epoch": 0.435,
      "grad_norm": 12.29251480102539,
      "learning_rate": 0.00013910524214875137,
      "loss": 6.4821,
      "step": 4350
    },
    {
      "epoch": 0.4351,
      "grad_norm": 2.689028263092041,
      "learning_rate": 0.00013907311284892736,
      "loss": 5.4106,
      "step": 4351
    },
    {
      "epoch": 0.4352,
      "grad_norm": 4.0768208503723145,
      "learning_rate": 0.00013904097878816312,
      "loss": 5.1112,
      "step": 4352
    },
    {
      "epoch": 0.4353,
      "grad_norm": 2.6715610027313232,
      "learning_rate": 0.00013900883997037397,
      "loss": 5.4625,
      "step": 4353
    },
    {
      "epoch": 0.4354,
      "grad_norm": 3.5138533115386963,
      "learning_rate": 0.00013897669639947606,
      "loss": 5.0134,
      "step": 4354
    },
    {
      "epoch": 0.4355,
      "grad_norm": 2.5471134185791016,
      "learning_rate": 0.00013894454807938586,
      "loss": 5.163,
      "step": 4355
    },
    {
      "epoch": 0.4356,
      "grad_norm": 4.254251956939697,
      "learning_rate": 0.0001389123950140206,
      "loss": 5.0234,
      "step": 4356
    },
    {
      "epoch": 0.4357,
      "grad_norm": 10.22170352935791,
      "learning_rate": 0.0001388802372072981,
      "loss": 5.1365,
      "step": 4357
    },
    {
      "epoch": 0.4358,
      "grad_norm": 5.689600467681885,
      "learning_rate": 0.00013884807466313663,
      "loss": 4.9306,
      "step": 4358
    },
    {
      "epoch": 0.4359,
      "grad_norm": 3.828493118286133,
      "learning_rate": 0.00013881590738545508,
      "loss": 4.7606,
      "step": 4359
    },
    {
      "epoch": 0.436,
      "grad_norm": 4.191102504730225,
      "learning_rate": 0.00013878373537817292,
      "loss": 5.0827,
      "step": 4360
    },
    {
      "epoch": 0.4361,
      "grad_norm": 3.143944025039673,
      "learning_rate": 0.0001387515586452103,
      "loss": 4.7489,
      "step": 4361
    },
    {
      "epoch": 0.4362,
      "grad_norm": 5.6285576820373535,
      "learning_rate": 0.00013871937719048779,
      "loss": 5.2544,
      "step": 4362
    },
    {
      "epoch": 0.4363,
      "grad_norm": 3.3018267154693604,
      "learning_rate": 0.00013868719101792665,
      "loss": 5.69,
      "step": 4363
    },
    {
      "epoch": 0.4364,
      "grad_norm": 3.035914659500122,
      "learning_rate": 0.00013865500013144857,
      "loss": 5.0202,
      "step": 4364
    },
    {
      "epoch": 0.4365,
      "grad_norm": 4.252862453460693,
      "learning_rate": 0.000138622804534976,
      "loss": 4.8385,
      "step": 4365
    },
    {
      "epoch": 0.4366,
      "grad_norm": 2.931455612182617,
      "learning_rate": 0.00013859060423243187,
      "loss": 4.6991,
      "step": 4366
    },
    {
      "epoch": 0.4367,
      "grad_norm": 5.308396339416504,
      "learning_rate": 0.00013855839922773968,
      "loss": 5.176,
      "step": 4367
    },
    {
      "epoch": 0.4368,
      "grad_norm": 4.624040603637695,
      "learning_rate": 0.00013852618952482347,
      "loss": 5.4429,
      "step": 4368
    },
    {
      "epoch": 0.4369,
      "grad_norm": 3.0166072845458984,
      "learning_rate": 0.00013849397512760795,
      "loss": 4.8667,
      "step": 4369
    },
    {
      "epoch": 0.437,
      "grad_norm": 3.0207529067993164,
      "learning_rate": 0.0001384617560400183,
      "loss": 5.8251,
      "step": 4370
    },
    {
      "epoch": 0.4371,
      "grad_norm": 3.5607504844665527,
      "learning_rate": 0.00013842953226598037,
      "loss": 5.2123,
      "step": 4371
    },
    {
      "epoch": 0.4372,
      "grad_norm": 5.736277103424072,
      "learning_rate": 0.00013839730380942053,
      "loss": 5.2967,
      "step": 4372
    },
    {
      "epoch": 0.4373,
      "grad_norm": 2.705404281616211,
      "learning_rate": 0.00013836507067426564,
      "loss": 5.2927,
      "step": 4373
    },
    {
      "epoch": 0.4374,
      "grad_norm": 3.482083797454834,
      "learning_rate": 0.00013833283286444328,
      "loss": 4.582,
      "step": 4374
    },
    {
      "epoch": 0.4375,
      "grad_norm": 2.811985969543457,
      "learning_rate": 0.0001383005903838815,
      "loss": 4.9915,
      "step": 4375
    },
    {
      "epoch": 0.4376,
      "grad_norm": 2.6694228649139404,
      "learning_rate": 0.000138268343236509,
      "loss": 5.285,
      "step": 4376
    },
    {
      "epoch": 0.4377,
      "grad_norm": 5.787441253662109,
      "learning_rate": 0.0001382360914262549,
      "loss": 5.1908,
      "step": 4377
    },
    {
      "epoch": 0.4378,
      "grad_norm": 1.929758906364441,
      "learning_rate": 0.0001382038349570491,
      "loss": 4.8511,
      "step": 4378
    },
    {
      "epoch": 0.4379,
      "grad_norm": 6.089313983917236,
      "learning_rate": 0.00013817157383282184,
      "loss": 4.6925,
      "step": 4379
    },
    {
      "epoch": 0.438,
      "grad_norm": 9.377686500549316,
      "learning_rate": 0.00013813930805750413,
      "loss": 4.7891,
      "step": 4380
    },
    {
      "epoch": 0.4381,
      "grad_norm": 4.364232540130615,
      "learning_rate": 0.00013810703763502744,
      "loss": 5.1315,
      "step": 4381
    },
    {
      "epoch": 0.4382,
      "grad_norm": 3.3886218070983887,
      "learning_rate": 0.00013807476256932376,
      "loss": 4.9517,
      "step": 4382
    },
    {
      "epoch": 0.4383,
      "grad_norm": 4.3907084465026855,
      "learning_rate": 0.00013804248286432578,
      "loss": 5.047,
      "step": 4383
    },
    {
      "epoch": 0.4384,
      "grad_norm": 6.572762966156006,
      "learning_rate": 0.00013801019852396665,
      "loss": 5.3586,
      "step": 4384
    },
    {
      "epoch": 0.4385,
      "grad_norm": 4.704990386962891,
      "learning_rate": 0.00013797790955218014,
      "loss": 5.0261,
      "step": 4385
    },
    {
      "epoch": 0.4386,
      "grad_norm": 3.843437910079956,
      "learning_rate": 0.00013794561595290052,
      "loss": 5.6865,
      "step": 4386
    },
    {
      "epoch": 0.4387,
      "grad_norm": 2.8469247817993164,
      "learning_rate": 0.0001379133177300627,
      "loss": 4.5894,
      "step": 4387
    },
    {
      "epoch": 0.4388,
      "grad_norm": 2.648911952972412,
      "learning_rate": 0.00013788101488760215,
      "loss": 4.6103,
      "step": 4388
    },
    {
      "epoch": 0.4389,
      "grad_norm": 3.693490743637085,
      "learning_rate": 0.00013784870742945482,
      "loss": 5.4865,
      "step": 4389
    },
    {
      "epoch": 0.439,
      "grad_norm": 5.26721715927124,
      "learning_rate": 0.00013781639535955732,
      "loss": 5.1956,
      "step": 4390
    },
    {
      "epoch": 0.4391,
      "grad_norm": 3.260960817337036,
      "learning_rate": 0.00013778407868184672,
      "loss": 4.8581,
      "step": 4391
    },
    {
      "epoch": 0.4392,
      "grad_norm": 2.194485902786255,
      "learning_rate": 0.00013775175740026078,
      "loss": 4.7901,
      "step": 4392
    },
    {
      "epoch": 0.4393,
      "grad_norm": 2.8884730339050293,
      "learning_rate": 0.00013771943151873767,
      "loss": 4.7748,
      "step": 4393
    },
    {
      "epoch": 0.4394,
      "grad_norm": 3.134455680847168,
      "learning_rate": 0.00013768710104121627,
      "loss": 4.8511,
      "step": 4394
    },
    {
      "epoch": 0.4395,
      "grad_norm": 2.673513889312744,
      "learning_rate": 0.00013765476597163594,
      "loss": 4.5637,
      "step": 4395
    },
    {
      "epoch": 0.4396,
      "grad_norm": 1.9774264097213745,
      "learning_rate": 0.00013762242631393655,
      "loss": 4.9701,
      "step": 4396
    },
    {
      "epoch": 0.4397,
      "grad_norm": 3.479719638824463,
      "learning_rate": 0.0001375900820720587,
      "loss": 4.5487,
      "step": 4397
    },
    {
      "epoch": 0.4398,
      "grad_norm": 10.150960922241211,
      "learning_rate": 0.0001375577332499433,
      "loss": 5.8444,
      "step": 4398
    },
    {
      "epoch": 0.4399,
      "grad_norm": 6.618307590484619,
      "learning_rate": 0.0001375253798515321,
      "loss": 5.7149,
      "step": 4399
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.6495611667633057,
      "learning_rate": 0.00013749302188076717,
      "loss": 4.9601,
      "step": 4400
    },
    {
      "epoch": 0.4401,
      "grad_norm": 2.445164203643799,
      "learning_rate": 0.00013746065934159123,
      "loss": 5.1248,
      "step": 4401
    },
    {
      "epoch": 0.4402,
      "grad_norm": 3.6941757202148438,
      "learning_rate": 0.00013742829223794759,
      "loss": 5.4715,
      "step": 4402
    },
    {
      "epoch": 0.4403,
      "grad_norm": 2.582294225692749,
      "learning_rate": 0.00013739592057378003,
      "loss": 4.832,
      "step": 4403
    },
    {
      "epoch": 0.4404,
      "grad_norm": 4.69214391708374,
      "learning_rate": 0.00013736354435303305,
      "loss": 5.7138,
      "step": 4404
    },
    {
      "epoch": 0.4405,
      "grad_norm": 2.3501460552215576,
      "learning_rate": 0.0001373311635796515,
      "loss": 4.8994,
      "step": 4405
    },
    {
      "epoch": 0.4406,
      "grad_norm": 2.7886040210723877,
      "learning_rate": 0.0001372987782575809,
      "loss": 4.9437,
      "step": 4406
    },
    {
      "epoch": 0.4407,
      "grad_norm": 2.6696739196777344,
      "learning_rate": 0.0001372663883907673,
      "loss": 5.8026,
      "step": 4407
    },
    {
      "epoch": 0.4408,
      "grad_norm": 9.716922760009766,
      "learning_rate": 0.00013723399398315734,
      "loss": 6.2957,
      "step": 4408
    },
    {
      "epoch": 0.4409,
      "grad_norm": 1.7785007953643799,
      "learning_rate": 0.00013720159503869815,
      "loss": 4.4836,
      "step": 4409
    },
    {
      "epoch": 0.441,
      "grad_norm": 3.119108200073242,
      "learning_rate": 0.00013716919156133746,
      "loss": 5.5177,
      "step": 4410
    },
    {
      "epoch": 0.4411,
      "grad_norm": 3.781236410140991,
      "learning_rate": 0.00013713678355502351,
      "loss": 4.6851,
      "step": 4411
    },
    {
      "epoch": 0.4412,
      "grad_norm": 2.593754768371582,
      "learning_rate": 0.0001371043710237051,
      "loss": 4.9863,
      "step": 4412
    },
    {
      "epoch": 0.4413,
      "grad_norm": 2.4118812084198,
      "learning_rate": 0.00013707195397133165,
      "loss": 4.4354,
      "step": 4413
    },
    {
      "epoch": 0.4414,
      "grad_norm": 2.678340435028076,
      "learning_rate": 0.0001370395324018531,
      "loss": 5.0774,
      "step": 4414
    },
    {
      "epoch": 0.4415,
      "grad_norm": 2.2538905143737793,
      "learning_rate": 0.00013700710631921984,
      "loss": 4.765,
      "step": 4415
    },
    {
      "epoch": 0.4416,
      "grad_norm": 3.149444818496704,
      "learning_rate": 0.00013697467572738295,
      "loss": 5.2277,
      "step": 4416
    },
    {
      "epoch": 0.4417,
      "grad_norm": 3.7121834754943848,
      "learning_rate": 0.00013694224063029396,
      "loss": 5.4252,
      "step": 4417
    },
    {
      "epoch": 0.4418,
      "grad_norm": 3.1267378330230713,
      "learning_rate": 0.00013690980103190503,
      "loss": 5.6647,
      "step": 4418
    },
    {
      "epoch": 0.4419,
      "grad_norm": 3.3361213207244873,
      "learning_rate": 0.00013687735693616876,
      "loss": 5.4759,
      "step": 4419
    },
    {
      "epoch": 0.442,
      "grad_norm": 3.4468166828155518,
      "learning_rate": 0.00013684490834703843,
      "loss": 4.9048,
      "step": 4420
    },
    {
      "epoch": 0.4421,
      "grad_norm": 2.547630786895752,
      "learning_rate": 0.00013681245526846783,
      "loss": 4.9952,
      "step": 4421
    },
    {
      "epoch": 0.4422,
      "grad_norm": 3.679051160812378,
      "learning_rate": 0.00013677999770441115,
      "loss": 5.3252,
      "step": 4422
    },
    {
      "epoch": 0.4423,
      "grad_norm": 3.491093635559082,
      "learning_rate": 0.00013674753565882334,
      "loss": 5.5578,
      "step": 4423
    },
    {
      "epoch": 0.4424,
      "grad_norm": 4.1005144119262695,
      "learning_rate": 0.0001367150691356598,
      "loss": 5.1721,
      "step": 4424
    },
    {
      "epoch": 0.4425,
      "grad_norm": 16.33206558227539,
      "learning_rate": 0.00013668259813887643,
      "loss": 5.4644,
      "step": 4425
    },
    {
      "epoch": 0.4426,
      "grad_norm": 2.3436102867126465,
      "learning_rate": 0.00013665012267242974,
      "loss": 4.8266,
      "step": 4426
    },
    {
      "epoch": 0.4427,
      "grad_norm": 2.1355109214782715,
      "learning_rate": 0.00013661764274027678,
      "loss": 4.8364,
      "step": 4427
    },
    {
      "epoch": 0.4428,
      "grad_norm": 2.3901703357696533,
      "learning_rate": 0.00013658515834637512,
      "loss": 5.0535,
      "step": 4428
    },
    {
      "epoch": 0.4429,
      "grad_norm": 5.075170516967773,
      "learning_rate": 0.0001365526694946829,
      "loss": 5.4758,
      "step": 4429
    },
    {
      "epoch": 0.443,
      "grad_norm": 2.0689947605133057,
      "learning_rate": 0.0001365201761891588,
      "loss": 5.0972,
      "step": 4430
    },
    {
      "epoch": 0.4431,
      "grad_norm": 2.3696227073669434,
      "learning_rate": 0.00013648767843376196,
      "loss": 4.744,
      "step": 4431
    },
    {
      "epoch": 0.4432,
      "grad_norm": 2.3493492603302,
      "learning_rate": 0.0001364551762324522,
      "loss": 4.6941,
      "step": 4432
    },
    {
      "epoch": 0.4433,
      "grad_norm": 4.5541300773620605,
      "learning_rate": 0.00013642266958918984,
      "loss": 4.9356,
      "step": 4433
    },
    {
      "epoch": 0.4434,
      "grad_norm": 1.8745697736740112,
      "learning_rate": 0.00013639015850793563,
      "loss": 4.7111,
      "step": 4434
    },
    {
      "epoch": 0.4435,
      "grad_norm": 3.719759464263916,
      "learning_rate": 0.000136357642992651,
      "loss": 4.9224,
      "step": 4435
    },
    {
      "epoch": 0.4436,
      "grad_norm": 2.3263700008392334,
      "learning_rate": 0.00013632512304729785,
      "loss": 4.7516,
      "step": 4436
    },
    {
      "epoch": 0.4437,
      "grad_norm": 2.900296688079834,
      "learning_rate": 0.00013629259867583863,
      "loss": 5.2879,
      "step": 4437
    },
    {
      "epoch": 0.4438,
      "grad_norm": 2.57181978225708,
      "learning_rate": 0.00013626006988223636,
      "loss": 4.5172,
      "step": 4438
    },
    {
      "epoch": 0.4439,
      "grad_norm": 4.969818592071533,
      "learning_rate": 0.00013622753667045457,
      "loss": 4.7633,
      "step": 4439
    },
    {
      "epoch": 0.444,
      "grad_norm": 3.00455641746521,
      "learning_rate": 0.00013619499904445734,
      "loss": 4.5917,
      "step": 4440
    },
    {
      "epoch": 0.4441,
      "grad_norm": 2.5814285278320312,
      "learning_rate": 0.00013616245700820922,
      "loss": 4.8746,
      "step": 4441
    },
    {
      "epoch": 0.4442,
      "grad_norm": 8.264567375183105,
      "learning_rate": 0.00013612991056567543,
      "loss": 5.2742,
      "step": 4442
    },
    {
      "epoch": 0.4443,
      "grad_norm": 3.171640157699585,
      "learning_rate": 0.00013609735972082166,
      "loss": 5.1077,
      "step": 4443
    },
    {
      "epoch": 0.4444,
      "grad_norm": 4.071385860443115,
      "learning_rate": 0.00013606480447761409,
      "loss": 5.1106,
      "step": 4444
    },
    {
      "epoch": 0.4445,
      "grad_norm": 4.30679988861084,
      "learning_rate": 0.00013603224484001948,
      "loss": 4.7812,
      "step": 4445
    },
    {
      "epoch": 0.4446,
      "grad_norm": 3.4273793697357178,
      "learning_rate": 0.00013599968081200514,
      "loss": 5.2883,
      "step": 4446
    },
    {
      "epoch": 0.4447,
      "grad_norm": 2.7225379943847656,
      "learning_rate": 0.0001359671123975389,
      "loss": 5.4068,
      "step": 4447
    },
    {
      "epoch": 0.4448,
      "grad_norm": 3.5412187576293945,
      "learning_rate": 0.00013593453960058908,
      "loss": 5.0233,
      "step": 4448
    },
    {
      "epoch": 0.4449,
      "grad_norm": 3.439924955368042,
      "learning_rate": 0.00013590196242512463,
      "loss": 4.6701,
      "step": 4449
    },
    {
      "epoch": 0.445,
      "grad_norm": 4.241576671600342,
      "learning_rate": 0.00013586938087511494,
      "loss": 4.961,
      "step": 4450
    },
    {
      "epoch": 0.4451,
      "grad_norm": 3.348606824874878,
      "learning_rate": 0.00013583679495453,
      "loss": 4.7783,
      "step": 4451
    },
    {
      "epoch": 0.4452,
      "grad_norm": 4.814090251922607,
      "learning_rate": 0.00013580420466734037,
      "loss": 5.7149,
      "step": 4452
    },
    {
      "epoch": 0.4453,
      "grad_norm": 2.064643621444702,
      "learning_rate": 0.00013577161001751694,
      "loss": 4.7717,
      "step": 4453
    },
    {
      "epoch": 0.4454,
      "grad_norm": 3.3059186935424805,
      "learning_rate": 0.00013573901100903134,
      "loss": 4.9257,
      "step": 4454
    },
    {
      "epoch": 0.4455,
      "grad_norm": 2.8091654777526855,
      "learning_rate": 0.00013570640764585566,
      "loss": 4.5622,
      "step": 4455
    },
    {
      "epoch": 0.4456,
      "grad_norm": 2.439967632293701,
      "learning_rate": 0.00013567379993196252,
      "loss": 5.8653,
      "step": 4456
    },
    {
      "epoch": 0.4457,
      "grad_norm": 2.1443593502044678,
      "learning_rate": 0.00013564118787132506,
      "loss": 5.6995,
      "step": 4457
    },
    {
      "epoch": 0.4458,
      "grad_norm": 1.9473141431808472,
      "learning_rate": 0.000135608571467917,
      "loss": 5.042,
      "step": 4458
    },
    {
      "epoch": 0.4459,
      "grad_norm": 6.500906467437744,
      "learning_rate": 0.0001355759507257125,
      "loss": 4.6682,
      "step": 4459
    },
    {
      "epoch": 0.446,
      "grad_norm": 7.02877140045166,
      "learning_rate": 0.0001355433256486863,
      "loss": 5.8089,
      "step": 4460
    },
    {
      "epoch": 0.4461,
      "grad_norm": 9.407878875732422,
      "learning_rate": 0.0001355106962408137,
      "loss": 4.969,
      "step": 4461
    },
    {
      "epoch": 0.4462,
      "grad_norm": 3.2439353466033936,
      "learning_rate": 0.0001354780625060705,
      "loss": 4.8681,
      "step": 4462
    },
    {
      "epoch": 0.4463,
      "grad_norm": 2.4278180599212646,
      "learning_rate": 0.00013544542444843299,
      "loss": 5.1437,
      "step": 4463
    },
    {
      "epoch": 0.4464,
      "grad_norm": 3.6893389225006104,
      "learning_rate": 0.000135412782071878,
      "loss": 4.5328,
      "step": 4464
    },
    {
      "epoch": 0.4465,
      "grad_norm": 2.7085583209991455,
      "learning_rate": 0.00013538013538038295,
      "loss": 4.7771,
      "step": 4465
    },
    {
      "epoch": 0.4466,
      "grad_norm": 2.4535248279571533,
      "learning_rate": 0.00013534748437792573,
      "loss": 5.1648,
      "step": 4466
    },
    {
      "epoch": 0.4467,
      "grad_norm": 3.2181286811828613,
      "learning_rate": 0.00013531482906848475,
      "loss": 4.8712,
      "step": 4467
    },
    {
      "epoch": 0.4468,
      "grad_norm": 3.2877144813537598,
      "learning_rate": 0.000135282169456039,
      "loss": 6.1317,
      "step": 4468
    },
    {
      "epoch": 0.4469,
      "grad_norm": 2.7903966903686523,
      "learning_rate": 0.00013524950554456784,
      "loss": 5.4707,
      "step": 4469
    },
    {
      "epoch": 0.447,
      "grad_norm": 2.170713186264038,
      "learning_rate": 0.00013521683733805145,
      "loss": 5.055,
      "step": 4470
    },
    {
      "epoch": 0.4471,
      "grad_norm": 3.4468464851379395,
      "learning_rate": 0.00013518416484047018,
      "loss": 5.0334,
      "step": 4471
    },
    {
      "epoch": 0.4472,
      "grad_norm": 2.6723990440368652,
      "learning_rate": 0.0001351514880558052,
      "loss": 4.7622,
      "step": 4472
    },
    {
      "epoch": 0.4473,
      "grad_norm": 6.4998555183410645,
      "learning_rate": 0.00013511880698803801,
      "loss": 5.2411,
      "step": 4473
    },
    {
      "epoch": 0.4474,
      "grad_norm": 2.3764967918395996,
      "learning_rate": 0.00013508612164115068,
      "loss": 4.9826,
      "step": 4474
    },
    {
      "epoch": 0.4475,
      "grad_norm": 2.390089750289917,
      "learning_rate": 0.0001350534320191259,
      "loss": 5.4658,
      "step": 4475
    },
    {
      "epoch": 0.4476,
      "grad_norm": 3.6314492225646973,
      "learning_rate": 0.00013502073812594675,
      "loss": 5.228,
      "step": 4476
    },
    {
      "epoch": 0.4477,
      "grad_norm": 2.331752300262451,
      "learning_rate": 0.00013498803996559692,
      "loss": 5.0363,
      "step": 4477
    },
    {
      "epoch": 0.4478,
      "grad_norm": 2.197211503982544,
      "learning_rate": 0.0001349553375420605,
      "loss": 4.6888,
      "step": 4478
    },
    {
      "epoch": 0.4479,
      "grad_norm": 2.6859822273254395,
      "learning_rate": 0.00013492263085932224,
      "loss": 5.0141,
      "step": 4479
    },
    {
      "epoch": 0.448,
      "grad_norm": 3.739854335784912,
      "learning_rate": 0.00013488991992136734,
      "loss": 4.8461,
      "step": 4480
    },
    {
      "epoch": 0.4481,
      "grad_norm": 8.959406852722168,
      "learning_rate": 0.00013485720473218154,
      "loss": 4.8689,
      "step": 4481
    },
    {
      "epoch": 0.4482,
      "grad_norm": 3.1924734115600586,
      "learning_rate": 0.00013482448529575106,
      "loss": 5.0646,
      "step": 4482
    },
    {
      "epoch": 0.4483,
      "grad_norm": 6.617849349975586,
      "learning_rate": 0.0001347917616160627,
      "loss": 5.2711,
      "step": 4483
    },
    {
      "epoch": 0.4484,
      "grad_norm": 4.49888801574707,
      "learning_rate": 0.0001347590336971037,
      "loss": 5.2549,
      "step": 4484
    },
    {
      "epoch": 0.4485,
      "grad_norm": 4.155158042907715,
      "learning_rate": 0.0001347263015428619,
      "loss": 4.7328,
      "step": 4485
    },
    {
      "epoch": 0.4486,
      "grad_norm": 2.106078624725342,
      "learning_rate": 0.00013469356515732558,
      "loss": 4.7862,
      "step": 4486
    },
    {
      "epoch": 0.4487,
      "grad_norm": 2.2562012672424316,
      "learning_rate": 0.00013466082454448362,
      "loss": 5.1765,
      "step": 4487
    },
    {
      "epoch": 0.4488,
      "grad_norm": 4.66193962097168,
      "learning_rate": 0.0001346280797083253,
      "loss": 4.948,
      "step": 4488
    },
    {
      "epoch": 0.4489,
      "grad_norm": 2.085111379623413,
      "learning_rate": 0.0001345953306528405,
      "loss": 5.0454,
      "step": 4489
    },
    {
      "epoch": 0.449,
      "grad_norm": 2.8357481956481934,
      "learning_rate": 0.00013456257738201957,
      "loss": 4.7062,
      "step": 4490
    },
    {
      "epoch": 0.4491,
      "grad_norm": 4.18306303024292,
      "learning_rate": 0.00013452981989985348,
      "loss": 4.4423,
      "step": 4491
    },
    {
      "epoch": 0.4492,
      "grad_norm": 2.3040058612823486,
      "learning_rate": 0.00013449705821033355,
      "loss": 4.6639,
      "step": 4492
    },
    {
      "epoch": 0.4493,
      "grad_norm": 2.812925100326538,
      "learning_rate": 0.0001344642923174517,
      "loss": 4.8339,
      "step": 4493
    },
    {
      "epoch": 0.4494,
      "grad_norm": 3.4842119216918945,
      "learning_rate": 0.00013443152222520038,
      "loss": 4.7039,
      "step": 4494
    },
    {
      "epoch": 0.4495,
      "grad_norm": 2.2863945960998535,
      "learning_rate": 0.00013439874793757254,
      "loss": 4.7969,
      "step": 4495
    },
    {
      "epoch": 0.4496,
      "grad_norm": 2.695615530014038,
      "learning_rate": 0.00013436596945856164,
      "loss": 5.4206,
      "step": 4496
    },
    {
      "epoch": 0.4497,
      "grad_norm": 2.7848591804504395,
      "learning_rate": 0.00013433318679216153,
      "loss": 4.8928,
      "step": 4497
    },
    {
      "epoch": 0.4498,
      "grad_norm": 3.7426652908325195,
      "learning_rate": 0.00013430039994236678,
      "loss": 4.5767,
      "step": 4498
    },
    {
      "epoch": 0.4499,
      "grad_norm": 2.2581968307495117,
      "learning_rate": 0.00013426760891317236,
      "loss": 4.8389,
      "step": 4499
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.3084065914154053,
      "learning_rate": 0.00013423481370857375,
      "loss": 4.7982,
      "step": 4500
    },
    {
      "epoch": 0.4501,
      "grad_norm": 4.484655380249023,
      "learning_rate": 0.00013420201433256689,
      "loss": 6.005,
      "step": 4501
    },
    {
      "epoch": 0.4502,
      "grad_norm": 5.474215984344482,
      "learning_rate": 0.00013416921078914835,
      "loss": 4.7838,
      "step": 4502
    },
    {
      "epoch": 0.4503,
      "grad_norm": 7.26628303527832,
      "learning_rate": 0.0001341364030823151,
      "loss": 4.6508,
      "step": 4503
    },
    {
      "epoch": 0.4504,
      "grad_norm": 3.4742772579193115,
      "learning_rate": 0.0001341035912160647,
      "loss": 4.6415,
      "step": 4504
    },
    {
      "epoch": 0.4505,
      "grad_norm": 2.7096099853515625,
      "learning_rate": 0.0001340707751943952,
      "loss": 4.9439,
      "step": 4505
    },
    {
      "epoch": 0.4506,
      "grad_norm": 4.070446491241455,
      "learning_rate": 0.00013403795502130503,
      "loss": 4.4533,
      "step": 4506
    },
    {
      "epoch": 0.4507,
      "grad_norm": 1.9775433540344238,
      "learning_rate": 0.0001340051307007933,
      "loss": 5.2018,
      "step": 4507
    },
    {
      "epoch": 0.4508,
      "grad_norm": 3.7639050483703613,
      "learning_rate": 0.00013397230223685956,
      "loss": 4.6921,
      "step": 4508
    },
    {
      "epoch": 0.4509,
      "grad_norm": 2.578042507171631,
      "learning_rate": 0.00013393946963350382,
      "loss": 4.7808,
      "step": 4509
    },
    {
      "epoch": 0.451,
      "grad_norm": 3.914198637008667,
      "learning_rate": 0.00013390663289472666,
      "loss": 5.0708,
      "step": 4510
    },
    {
      "epoch": 0.4511,
      "grad_norm": 3.0787954330444336,
      "learning_rate": 0.00013387379202452917,
      "loss": 5.0974,
      "step": 4511
    },
    {
      "epoch": 0.4512,
      "grad_norm": 2.287778377532959,
      "learning_rate": 0.00013384094702691281,
      "loss": 4.9753,
      "step": 4512
    },
    {
      "epoch": 0.4513,
      "grad_norm": 3.4871296882629395,
      "learning_rate": 0.00013380809790587974,
      "loss": 4.8566,
      "step": 4513
    },
    {
      "epoch": 0.4514,
      "grad_norm": 3.5687522888183594,
      "learning_rate": 0.00013377524466543248,
      "loss": 4.6443,
      "step": 4514
    },
    {
      "epoch": 0.4515,
      "grad_norm": 3.6031365394592285,
      "learning_rate": 0.00013374238730957412,
      "loss": 5.259,
      "step": 4515
    },
    {
      "epoch": 0.4516,
      "grad_norm": 3.930617094039917,
      "learning_rate": 0.00013370952584230823,
      "loss": 5.615,
      "step": 4516
    },
    {
      "epoch": 0.4517,
      "grad_norm": 5.735434532165527,
      "learning_rate": 0.00013367666026763882,
      "loss": 5.9394,
      "step": 4517
    },
    {
      "epoch": 0.4518,
      "grad_norm": 5.4739813804626465,
      "learning_rate": 0.00013364379058957056,
      "loss": 5.5387,
      "step": 4518
    },
    {
      "epoch": 0.4519,
      "grad_norm": 3.578657627105713,
      "learning_rate": 0.00013361091681210845,
      "loss": 5.5364,
      "step": 4519
    },
    {
      "epoch": 0.452,
      "grad_norm": 3.787731647491455,
      "learning_rate": 0.00013357803893925807,
      "loss": 5.0711,
      "step": 4520
    },
    {
      "epoch": 0.4521,
      "grad_norm": 8.31989574432373,
      "learning_rate": 0.00013354515697502553,
      "loss": 4.8183,
      "step": 4521
    },
    {
      "epoch": 0.4522,
      "grad_norm": 4.1772918701171875,
      "learning_rate": 0.00013351227092341732,
      "loss": 5.9171,
      "step": 4522
    },
    {
      "epoch": 0.4523,
      "grad_norm": 13.342952728271484,
      "learning_rate": 0.0001334793807884406,
      "loss": 4.9541,
      "step": 4523
    },
    {
      "epoch": 0.4524,
      "grad_norm": 5.651301860809326,
      "learning_rate": 0.00013344648657410282,
      "loss": 5.4538,
      "step": 4524
    },
    {
      "epoch": 0.4525,
      "grad_norm": 3.9770402908325195,
      "learning_rate": 0.00013341358828441218,
      "loss": 5.2551,
      "step": 4525
    },
    {
      "epoch": 0.4526,
      "grad_norm": 12.739694595336914,
      "learning_rate": 0.0001333806859233771,
      "loss": 5.5451,
      "step": 4526
    },
    {
      "epoch": 0.4527,
      "grad_norm": 2.1419808864593506,
      "learning_rate": 0.00013334777949500673,
      "loss": 4.9427,
      "step": 4527
    },
    {
      "epoch": 0.4528,
      "grad_norm": 6.716566562652588,
      "learning_rate": 0.00013331486900331057,
      "loss": 4.9202,
      "step": 4528
    },
    {
      "epoch": 0.4529,
      "grad_norm": 2.675017833709717,
      "learning_rate": 0.00013328195445229868,
      "loss": 5.0987,
      "step": 4529
    },
    {
      "epoch": 0.453,
      "grad_norm": 3.913407564163208,
      "learning_rate": 0.0001332490358459816,
      "loss": 5.5155,
      "step": 4530
    },
    {
      "epoch": 0.4531,
      "grad_norm": 2.206069231033325,
      "learning_rate": 0.00013321611318837032,
      "loss": 4.8832,
      "step": 4531
    },
    {
      "epoch": 0.4532,
      "grad_norm": 3.4990780353546143,
      "learning_rate": 0.00013318318648347646,
      "loss": 4.9573,
      "step": 4532
    },
    {
      "epoch": 0.4533,
      "grad_norm": 3.495326280593872,
      "learning_rate": 0.00013315025573531198,
      "loss": 4.992,
      "step": 4533
    },
    {
      "epoch": 0.4534,
      "grad_norm": 3.2406346797943115,
      "learning_rate": 0.00013311732094788937,
      "loss": 5.2341,
      "step": 4534
    },
    {
      "epoch": 0.4535,
      "grad_norm": 2.3283393383026123,
      "learning_rate": 0.00013308438212522163,
      "loss": 4.8143,
      "step": 4535
    },
    {
      "epoch": 0.4536,
      "grad_norm": 3.4315104484558105,
      "learning_rate": 0.0001330514392713223,
      "loss": 4.9522,
      "step": 4536
    },
    {
      "epoch": 0.4537,
      "grad_norm": 3.055168390274048,
      "learning_rate": 0.00013301849239020537,
      "loss": 4.783,
      "step": 4537
    },
    {
      "epoch": 0.4538,
      "grad_norm": 2.879873752593994,
      "learning_rate": 0.00013298554148588528,
      "loss": 5.0248,
      "step": 4538
    },
    {
      "epoch": 0.4539,
      "grad_norm": 3.780299186706543,
      "learning_rate": 0.00013295258656237702,
      "loss": 4.6447,
      "step": 4539
    },
    {
      "epoch": 0.454,
      "grad_norm": 2.1001758575439453,
      "learning_rate": 0.00013291962762369608,
      "loss": 4.7804,
      "step": 4540
    },
    {
      "epoch": 0.4541,
      "grad_norm": 3.5401251316070557,
      "learning_rate": 0.00013288666467385833,
      "loss": 5.0109,
      "step": 4541
    },
    {
      "epoch": 0.4542,
      "grad_norm": 3.046821355819702,
      "learning_rate": 0.00013285369771688025,
      "loss": 5.0928,
      "step": 4542
    },
    {
      "epoch": 0.4543,
      "grad_norm": 3.497878313064575,
      "learning_rate": 0.0001328207267567788,
      "loss": 4.7534,
      "step": 4543
    },
    {
      "epoch": 0.4544,
      "grad_norm": 3.1388843059539795,
      "learning_rate": 0.00013278775179757131,
      "loss": 4.8313,
      "step": 4544
    },
    {
      "epoch": 0.4545,
      "grad_norm": 2.95028018951416,
      "learning_rate": 0.0001327547728432757,
      "loss": 5.2732,
      "step": 4545
    },
    {
      "epoch": 0.4546,
      "grad_norm": 2.424560785293579,
      "learning_rate": 0.0001327217898979104,
      "loss": 5.4186,
      "step": 4546
    },
    {
      "epoch": 0.4547,
      "grad_norm": 2.8907859325408936,
      "learning_rate": 0.00013268880296549425,
      "loss": 4.9922,
      "step": 4547
    },
    {
      "epoch": 0.4548,
      "grad_norm": 2.3381290435791016,
      "learning_rate": 0.00013265581205004661,
      "loss": 5.0811,
      "step": 4548
    },
    {
      "epoch": 0.4549,
      "grad_norm": 5.904923915863037,
      "learning_rate": 0.00013262281715558736,
      "loss": 5.6034,
      "step": 4549
    },
    {
      "epoch": 0.455,
      "grad_norm": 2.097555637359619,
      "learning_rate": 0.00013258981828613678,
      "loss": 4.8908,
      "step": 4550
    },
    {
      "epoch": 0.4551,
      "grad_norm": 2.427426338195801,
      "learning_rate": 0.00013255681544571568,
      "loss": 5.1703,
      "step": 4551
    },
    {
      "epoch": 0.4552,
      "grad_norm": 6.727451324462891,
      "learning_rate": 0.0001325238086383454,
      "loss": 4.6592,
      "step": 4552
    },
    {
      "epoch": 0.4553,
      "grad_norm": 4.7562713623046875,
      "learning_rate": 0.00013249079786804765,
      "loss": 4.8157,
      "step": 4553
    },
    {
      "epoch": 0.4554,
      "grad_norm": 2.1229946613311768,
      "learning_rate": 0.00013245778313884478,
      "loss": 5.1329,
      "step": 4554
    },
    {
      "epoch": 0.4555,
      "grad_norm": 2.195054292678833,
      "learning_rate": 0.00013242476445475944,
      "loss": 5.1622,
      "step": 4555
    },
    {
      "epoch": 0.4556,
      "grad_norm": 2.910449743270874,
      "learning_rate": 0.00013239174181981495,
      "loss": 4.5939,
      "step": 4556
    },
    {
      "epoch": 0.4557,
      "grad_norm": 6.986539840698242,
      "learning_rate": 0.00013235871523803496,
      "loss": 4.5619,
      "step": 4557
    },
    {
      "epoch": 0.4558,
      "grad_norm": 3.1768431663513184,
      "learning_rate": 0.0001323256847134437,
      "loss": 4.6967,
      "step": 4558
    },
    {
      "epoch": 0.4559,
      "grad_norm": 1.941521406173706,
      "learning_rate": 0.00013229265025006576,
      "loss": 4.482,
      "step": 4559
    },
    {
      "epoch": 0.456,
      "grad_norm": 4.052547454833984,
      "learning_rate": 0.00013225961185192638,
      "loss": 5.4819,
      "step": 4560
    },
    {
      "epoch": 0.4561,
      "grad_norm": 2.4633100032806396,
      "learning_rate": 0.00013222656952305113,
      "loss": 5.0584,
      "step": 4561
    },
    {
      "epoch": 0.4562,
      "grad_norm": 3.3567259311676025,
      "learning_rate": 0.00013219352326746613,
      "loss": 4.7419,
      "step": 4562
    },
    {
      "epoch": 0.4563,
      "grad_norm": 1.901278018951416,
      "learning_rate": 0.00013216047308919798,
      "loss": 4.8265,
      "step": 4563
    },
    {
      "epoch": 0.4564,
      "grad_norm": 2.6541073322296143,
      "learning_rate": 0.00013212741899227374,
      "loss": 4.8962,
      "step": 4564
    },
    {
      "epoch": 0.4565,
      "grad_norm": 3.0921897888183594,
      "learning_rate": 0.00013209436098072095,
      "loss": 5.1733,
      "step": 4565
    },
    {
      "epoch": 0.4566,
      "grad_norm": 4.2030792236328125,
      "learning_rate": 0.00013206129905856765,
      "loss": 4.7177,
      "step": 4566
    },
    {
      "epoch": 0.4567,
      "grad_norm": 2.844991445541382,
      "learning_rate": 0.00013202823322984228,
      "loss": 5.3855,
      "step": 4567
    },
    {
      "epoch": 0.4568,
      "grad_norm": 2.8866078853607178,
      "learning_rate": 0.00013199516349857385,
      "loss": 5.2057,
      "step": 4568
    },
    {
      "epoch": 0.4569,
      "grad_norm": 3.17895245552063,
      "learning_rate": 0.0001319620898687918,
      "loss": 4.8265,
      "step": 4569
    },
    {
      "epoch": 0.457,
      "grad_norm": 11.000997543334961,
      "learning_rate": 0.00013192901234452607,
      "loss": 5.1122,
      "step": 4570
    },
    {
      "epoch": 0.4571,
      "grad_norm": 5.354388236999512,
      "learning_rate": 0.00013189593092980702,
      "loss": 5.5103,
      "step": 4571
    },
    {
      "epoch": 0.4572,
      "grad_norm": 3.4925601482391357,
      "learning_rate": 0.00013186284562866554,
      "loss": 6.1579,
      "step": 4572
    },
    {
      "epoch": 0.4573,
      "grad_norm": 6.952093124389648,
      "learning_rate": 0.00013182975644513296,
      "loss": 4.7503,
      "step": 4573
    },
    {
      "epoch": 0.4574,
      "grad_norm": 6.031211853027344,
      "learning_rate": 0.00013179666338324108,
      "loss": 4.5621,
      "step": 4574
    },
    {
      "epoch": 0.4575,
      "grad_norm": 4.729159832000732,
      "learning_rate": 0.00013176356644702225,
      "loss": 4.7912,
      "step": 4575
    },
    {
      "epoch": 0.4576,
      "grad_norm": 6.744852066040039,
      "learning_rate": 0.00013173046564050924,
      "loss": 4.4212,
      "step": 4576
    },
    {
      "epoch": 0.4577,
      "grad_norm": 4.572323799133301,
      "learning_rate": 0.0001316973609677352,
      "loss": 4.9375,
      "step": 4577
    },
    {
      "epoch": 0.4578,
      "grad_norm": 2.9619598388671875,
      "learning_rate": 0.00013166425243273385,
      "loss": 5.1874,
      "step": 4578
    },
    {
      "epoch": 0.4579,
      "grad_norm": 3.684473752975464,
      "learning_rate": 0.0001316311400395394,
      "loss": 5.0515,
      "step": 4579
    },
    {
      "epoch": 0.458,
      "grad_norm": 2.7845072746276855,
      "learning_rate": 0.0001315980237921865,
      "loss": 5.3088,
      "step": 4580
    },
    {
      "epoch": 0.4581,
      "grad_norm": 4.2275519371032715,
      "learning_rate": 0.00013156490369471027,
      "loss": 4.7971,
      "step": 4581
    },
    {
      "epoch": 0.4582,
      "grad_norm": 5.492747783660889,
      "learning_rate": 0.00013153177975114625,
      "loss": 4.8834,
      "step": 4582
    },
    {
      "epoch": 0.4583,
      "grad_norm": 2.0675506591796875,
      "learning_rate": 0.0001314986519655305,
      "loss": 4.7233,
      "step": 4583
    },
    {
      "epoch": 0.4584,
      "grad_norm": 2.4807686805725098,
      "learning_rate": 0.00013146552034189955,
      "loss": 4.9979,
      "step": 4584
    },
    {
      "epoch": 0.4585,
      "grad_norm": 7.431777000427246,
      "learning_rate": 0.0001314323848842904,
      "loss": 5.1978,
      "step": 4585
    },
    {
      "epoch": 0.4586,
      "grad_norm": 2.2975213527679443,
      "learning_rate": 0.00013139924559674052,
      "loss": 5.0545,
      "step": 4586
    },
    {
      "epoch": 0.4587,
      "grad_norm": 3.656280040740967,
      "learning_rate": 0.00013136610248328778,
      "loss": 5.2021,
      "step": 4587
    },
    {
      "epoch": 0.4588,
      "grad_norm": 3.894408941268921,
      "learning_rate": 0.0001313329555479706,
      "loss": 5.3065,
      "step": 4588
    },
    {
      "epoch": 0.4589,
      "grad_norm": 2.5660736560821533,
      "learning_rate": 0.00013129980479482782,
      "loss": 4.8819,
      "step": 4589
    },
    {
      "epoch": 0.459,
      "grad_norm": 2.5684597492218018,
      "learning_rate": 0.00013126665022789879,
      "loss": 4.9379,
      "step": 4590
    },
    {
      "epoch": 0.4591,
      "grad_norm": 5.280359268188477,
      "learning_rate": 0.00013123349185122327,
      "loss": 4.8267,
      "step": 4591
    },
    {
      "epoch": 0.4592,
      "grad_norm": 3.514431953430176,
      "learning_rate": 0.0001312003296688415,
      "loss": 5.2426,
      "step": 4592
    },
    {
      "epoch": 0.4593,
      "grad_norm": 3.655137538909912,
      "learning_rate": 0.0001311671636847942,
      "loss": 5.1546,
      "step": 4593
    },
    {
      "epoch": 0.4594,
      "grad_norm": 3.5458128452301025,
      "learning_rate": 0.00013113399390312256,
      "loss": 5.3635,
      "step": 4594
    },
    {
      "epoch": 0.4595,
      "grad_norm": 2.999589681625366,
      "learning_rate": 0.0001311008203278682,
      "loss": 5.5841,
      "step": 4595
    },
    {
      "epoch": 0.4596,
      "grad_norm": 2.555783271789551,
      "learning_rate": 0.0001310676429630732,
      "loss": 5.1191,
      "step": 4596
    },
    {
      "epoch": 0.4597,
      "grad_norm": 3.9886362552642822,
      "learning_rate": 0.00013103446181278013,
      "loss": 4.8868,
      "step": 4597
    },
    {
      "epoch": 0.4598,
      "grad_norm": 4.138184070587158,
      "learning_rate": 0.00013100127688103205,
      "loss": 5.4806,
      "step": 4598
    },
    {
      "epoch": 0.4599,
      "grad_norm": 3.933178424835205,
      "learning_rate": 0.00013096808817187242,
      "loss": 4.8402,
      "step": 4599
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.071746349334717,
      "learning_rate": 0.0001309348956893452,
      "loss": 4.7205,
      "step": 4600
    },
    {
      "epoch": 0.4601,
      "grad_norm": 16.345203399658203,
      "learning_rate": 0.00013090169943749476,
      "loss": 6.3974,
      "step": 4601
    },
    {
      "epoch": 0.4602,
      "grad_norm": 3.823220729827881,
      "learning_rate": 0.00013086849942036595,
      "loss": 5.1634,
      "step": 4602
    },
    {
      "epoch": 0.4603,
      "grad_norm": 6.965376853942871,
      "learning_rate": 0.00013083529564200417,
      "loss": 6.3581,
      "step": 4603
    },
    {
      "epoch": 0.4604,
      "grad_norm": 3.00107741355896,
      "learning_rate": 0.00013080208810645514,
      "loss": 4.9231,
      "step": 4604
    },
    {
      "epoch": 0.4605,
      "grad_norm": 5.583538055419922,
      "learning_rate": 0.0001307688768177651,
      "loss": 5.4983,
      "step": 4605
    },
    {
      "epoch": 0.4606,
      "grad_norm": 4.620184898376465,
      "learning_rate": 0.00013073566177998074,
      "loss": 4.9445,
      "step": 4606
    },
    {
      "epoch": 0.4607,
      "grad_norm": 2.8599114418029785,
      "learning_rate": 0.0001307024429971492,
      "loss": 5.1068,
      "step": 4607
    },
    {
      "epoch": 0.4608,
      "grad_norm": 3.6943726539611816,
      "learning_rate": 0.00013066922047331813,
      "loss": 4.7823,
      "step": 4608
    },
    {
      "epoch": 0.4609,
      "grad_norm": 2.806788682937622,
      "learning_rate": 0.00013063599421253558,
      "loss": 4.587,
      "step": 4609
    },
    {
      "epoch": 0.461,
      "grad_norm": 2.008420467376709,
      "learning_rate": 0.0001306027642188501,
      "loss": 4.2172,
      "step": 4610
    },
    {
      "epoch": 0.4611,
      "grad_norm": 3.222940444946289,
      "learning_rate": 0.00013056953049631057,
      "loss": 4.9249,
      "step": 4611
    },
    {
      "epoch": 0.4612,
      "grad_norm": 2.795884370803833,
      "learning_rate": 0.0001305362930489665,
      "loss": 5.2934,
      "step": 4612
    },
    {
      "epoch": 0.4613,
      "grad_norm": 4.210395812988281,
      "learning_rate": 0.0001305030518808678,
      "loss": 5.2048,
      "step": 4613
    },
    {
      "epoch": 0.4614,
      "grad_norm": 4.129330635070801,
      "learning_rate": 0.00013046980699606467,
      "loss": 5.7933,
      "step": 4614
    },
    {
      "epoch": 0.4615,
      "grad_norm": 2.0133297443389893,
      "learning_rate": 0.00013043655839860803,
      "loss": 5.1006,
      "step": 4615
    },
    {
      "epoch": 0.4616,
      "grad_norm": 2.4375321865081787,
      "learning_rate": 0.00013040330609254903,
      "loss": 5.3542,
      "step": 4616
    },
    {
      "epoch": 0.4617,
      "grad_norm": 6.068768501281738,
      "learning_rate": 0.00013037005008193942,
      "loss": 5.1006,
      "step": 4617
    },
    {
      "epoch": 0.4618,
      "grad_norm": 2.297502040863037,
      "learning_rate": 0.00013033679037083138,
      "loss": 5.2811,
      "step": 4618
    },
    {
      "epoch": 0.4619,
      "grad_norm": 4.366490840911865,
      "learning_rate": 0.00013030352696327742,
      "loss": 5.2479,
      "step": 4619
    },
    {
      "epoch": 0.462,
      "grad_norm": 3.7294270992279053,
      "learning_rate": 0.0001302702598633306,
      "loss": 5.0971,
      "step": 4620
    },
    {
      "epoch": 0.4621,
      "grad_norm": 7.243568420410156,
      "learning_rate": 0.00013023698907504446,
      "loss": 4.5496,
      "step": 4621
    },
    {
      "epoch": 0.4622,
      "grad_norm": 2.7457382678985596,
      "learning_rate": 0.0001302037146024729,
      "loss": 4.8113,
      "step": 4622
    },
    {
      "epoch": 0.4623,
      "grad_norm": 9.516593933105469,
      "learning_rate": 0.00013017043644967036,
      "loss": 4.9868,
      "step": 4623
    },
    {
      "epoch": 0.4624,
      "grad_norm": 2.445350408554077,
      "learning_rate": 0.00013013715462069166,
      "loss": 5.2424,
      "step": 4624
    },
    {
      "epoch": 0.4625,
      "grad_norm": 2.5148138999938965,
      "learning_rate": 0.00013010386911959206,
      "loss": 5.2881,
      "step": 4625
    },
    {
      "epoch": 0.4626,
      "grad_norm": 5.5328593254089355,
      "learning_rate": 0.00013007057995042732,
      "loss": 5.0254,
      "step": 4626
    },
    {
      "epoch": 0.4627,
      "grad_norm": 2.573657274246216,
      "learning_rate": 0.0001300372871172536,
      "loss": 4.9162,
      "step": 4627
    },
    {
      "epoch": 0.4628,
      "grad_norm": 3.462028741836548,
      "learning_rate": 0.00013000399062412763,
      "loss": 5.2441,
      "step": 4628
    },
    {
      "epoch": 0.4629,
      "grad_norm": 5.39613151550293,
      "learning_rate": 0.0001299706904751064,
      "loss": 5.137,
      "step": 4629
    },
    {
      "epoch": 0.463,
      "grad_norm": 2.31024169921875,
      "learning_rate": 0.0001299373866742474,
      "loss": 4.7737,
      "step": 4630
    },
    {
      "epoch": 0.4631,
      "grad_norm": 3.164882183074951,
      "learning_rate": 0.00012990407922560868,
      "loss": 5.0295,
      "step": 4631
    },
    {
      "epoch": 0.4632,
      "grad_norm": 7.598518371582031,
      "learning_rate": 0.00012987076813324858,
      "loss": 5.5544,
      "step": 4632
    },
    {
      "epoch": 0.4633,
      "grad_norm": 3.4492580890655518,
      "learning_rate": 0.00012983745340122604,
      "loss": 4.634,
      "step": 4633
    },
    {
      "epoch": 0.4634,
      "grad_norm": 4.900599002838135,
      "learning_rate": 0.0001298041350336003,
      "loss": 5.0635,
      "step": 4634
    },
    {
      "epoch": 0.4635,
      "grad_norm": 2.684541702270508,
      "learning_rate": 0.00012977081303443107,
      "loss": 4.8936,
      "step": 4635
    },
    {
      "epoch": 0.4636,
      "grad_norm": 2.488908529281616,
      "learning_rate": 0.0001297374874077786,
      "loss": 4.9596,
      "step": 4636
    },
    {
      "epoch": 0.4637,
      "grad_norm": 2.6916873455047607,
      "learning_rate": 0.0001297041581577035,
      "loss": 4.7704,
      "step": 4637
    },
    {
      "epoch": 0.4638,
      "grad_norm": 3.5672736167907715,
      "learning_rate": 0.00012967082528826684,
      "loss": 4.7903,
      "step": 4638
    },
    {
      "epoch": 0.4639,
      "grad_norm": 5.2342095375061035,
      "learning_rate": 0.00012963748880353011,
      "loss": 4.9397,
      "step": 4639
    },
    {
      "epoch": 0.464,
      "grad_norm": 4.723827362060547,
      "learning_rate": 0.00012960414870755524,
      "loss": 4.9551,
      "step": 4640
    },
    {
      "epoch": 0.4641,
      "grad_norm": 2.39870023727417,
      "learning_rate": 0.00012957080500440468,
      "loss": 4.8946,
      "step": 4641
    },
    {
      "epoch": 0.4642,
      "grad_norm": 4.415536880493164,
      "learning_rate": 0.00012953745769814123,
      "loss": 5.3641,
      "step": 4642
    },
    {
      "epoch": 0.4643,
      "grad_norm": 7.006067276000977,
      "learning_rate": 0.00012950410679282815,
      "loss": 5.4158,
      "step": 4643
    },
    {
      "epoch": 0.4644,
      "grad_norm": 2.5810112953186035,
      "learning_rate": 0.00012947075229252915,
      "loss": 5.0875,
      "step": 4644
    },
    {
      "epoch": 0.4645,
      "grad_norm": 3.3689801692962646,
      "learning_rate": 0.00012943739420130836,
      "loss": 5.1203,
      "step": 4645
    },
    {
      "epoch": 0.4646,
      "grad_norm": 2.0109214782714844,
      "learning_rate": 0.0001294040325232304,
      "loss": 4.9705,
      "step": 4646
    },
    {
      "epoch": 0.4647,
      "grad_norm": 2.318631410598755,
      "learning_rate": 0.0001293706672623603,
      "loss": 4.5031,
      "step": 4647
    },
    {
      "epoch": 0.4648,
      "grad_norm": 4.062061309814453,
      "learning_rate": 0.00012933729842276343,
      "loss": 4.9582,
      "step": 4648
    },
    {
      "epoch": 0.4649,
      "grad_norm": 4.148094654083252,
      "learning_rate": 0.00012930392600850573,
      "loss": 5.0578,
      "step": 4649
    },
    {
      "epoch": 0.465,
      "grad_norm": 2.1024067401885986,
      "learning_rate": 0.0001292705500236536,
      "loss": 4.8318,
      "step": 4650
    },
    {
      "epoch": 0.4651,
      "grad_norm": 2.359062433242798,
      "learning_rate": 0.00012923717047227368,
      "loss": 5.3051,
      "step": 4651
    },
    {
      "epoch": 0.4652,
      "grad_norm": 2.9861814975738525,
      "learning_rate": 0.00012920378735843327,
      "loss": 5.0444,
      "step": 4652
    },
    {
      "epoch": 0.4653,
      "grad_norm": 2.7693769931793213,
      "learning_rate": 0.0001291704006861999,
      "loss": 5.0809,
      "step": 4653
    },
    {
      "epoch": 0.4654,
      "grad_norm": 3.3338422775268555,
      "learning_rate": 0.00012913701045964173,
      "loss": 5.4615,
      "step": 4654
    },
    {
      "epoch": 0.4655,
      "grad_norm": 2.932871103286743,
      "learning_rate": 0.00012910361668282719,
      "loss": 5.674,
      "step": 4655
    },
    {
      "epoch": 0.4656,
      "grad_norm": 2.944760799407959,
      "learning_rate": 0.00012907021935982526,
      "loss": 4.8384,
      "step": 4656
    },
    {
      "epoch": 0.4657,
      "grad_norm": 2.1871109008789062,
      "learning_rate": 0.00012903681849470527,
      "loss": 5.0856,
      "step": 4657
    },
    {
      "epoch": 0.4658,
      "grad_norm": 13.385909080505371,
      "learning_rate": 0.00012900341409153703,
      "loss": 7.0015,
      "step": 4658
    },
    {
      "epoch": 0.4659,
      "grad_norm": 8.287419319152832,
      "learning_rate": 0.00012897000615439075,
      "loss": 5.6013,
      "step": 4659
    },
    {
      "epoch": 0.466,
      "grad_norm": 2.6945888996124268,
      "learning_rate": 0.0001289365946873371,
      "loss": 4.8009,
      "step": 4660
    },
    {
      "epoch": 0.4661,
      "grad_norm": 12.501625061035156,
      "learning_rate": 0.00012890317969444716,
      "loss": 6.593,
      "step": 4661
    },
    {
      "epoch": 0.4662,
      "grad_norm": 3.8193225860595703,
      "learning_rate": 0.00012886976117979247,
      "loss": 4.8939,
      "step": 4662
    },
    {
      "epoch": 0.4663,
      "grad_norm": 4.124384880065918,
      "learning_rate": 0.00012883633914744492,
      "loss": 5.3767,
      "step": 4663
    },
    {
      "epoch": 0.4664,
      "grad_norm": 5.508857727050781,
      "learning_rate": 0.00012880291360147693,
      "loss": 4.898,
      "step": 4664
    },
    {
      "epoch": 0.4665,
      "grad_norm": 2.8480942249298096,
      "learning_rate": 0.00012876948454596128,
      "loss": 4.6585,
      "step": 4665
    },
    {
      "epoch": 0.4666,
      "grad_norm": 3.6257426738739014,
      "learning_rate": 0.00012873605198497124,
      "loss": 5.7565,
      "step": 4666
    },
    {
      "epoch": 0.4667,
      "grad_norm": 4.568176746368408,
      "learning_rate": 0.00012870261592258037,
      "loss": 5.9819,
      "step": 4667
    },
    {
      "epoch": 0.4668,
      "grad_norm": 2.9081928730010986,
      "learning_rate": 0.00012866917636286286,
      "loss": 5.082,
      "step": 4668
    },
    {
      "epoch": 0.4669,
      "grad_norm": 5.553041458129883,
      "learning_rate": 0.00012863573330989313,
      "loss": 5.7119,
      "step": 4669
    },
    {
      "epoch": 0.467,
      "grad_norm": 5.082150936126709,
      "learning_rate": 0.0001286022867677462,
      "loss": 5.9497,
      "step": 4670
    },
    {
      "epoch": 0.4671,
      "grad_norm": 2.164533853530884,
      "learning_rate": 0.00012856883674049736,
      "loss": 4.822,
      "step": 4671
    },
    {
      "epoch": 0.4672,
      "grad_norm": 5.26478910446167,
      "learning_rate": 0.00012853538323222242,
      "loss": 4.7727,
      "step": 4672
    },
    {
      "epoch": 0.4673,
      "grad_norm": 3.100634813308716,
      "learning_rate": 0.0001285019262469976,
      "loss": 5.3877,
      "step": 4673
    },
    {
      "epoch": 0.4674,
      "grad_norm": 3.603137969970703,
      "learning_rate": 0.00012846846578889955,
      "loss": 5.1937,
      "step": 4674
    },
    {
      "epoch": 0.4675,
      "grad_norm": 4.071640491485596,
      "learning_rate": 0.00012843500186200527,
      "loss": 5.3596,
      "step": 4675
    },
    {
      "epoch": 0.4676,
      "grad_norm": 4.0607781410217285,
      "learning_rate": 0.00012840153447039228,
      "loss": 5.8621,
      "step": 4676
    },
    {
      "epoch": 0.4677,
      "grad_norm": 9.819579124450684,
      "learning_rate": 0.00012836806361813844,
      "loss": 5.2105,
      "step": 4677
    },
    {
      "epoch": 0.4678,
      "grad_norm": 2.3684825897216797,
      "learning_rate": 0.00012833458930932212,
      "loss": 5.1576,
      "step": 4678
    },
    {
      "epoch": 0.4679,
      "grad_norm": 9.362683296203613,
      "learning_rate": 0.00012830111154802203,
      "loss": 4.9683,
      "step": 4679
    },
    {
      "epoch": 0.468,
      "grad_norm": 7.103567123413086,
      "learning_rate": 0.00012826763033831735,
      "loss": 4.8583,
      "step": 4680
    },
    {
      "epoch": 0.4681,
      "grad_norm": 2.8162803649902344,
      "learning_rate": 0.00012823414568428768,
      "loss": 4.6093,
      "step": 4681
    },
    {
      "epoch": 0.4682,
      "grad_norm": 3.9651405811309814,
      "learning_rate": 0.00012820065759001293,
      "loss": 4.882,
      "step": 4682
    },
    {
      "epoch": 0.4683,
      "grad_norm": 4.796658515930176,
      "learning_rate": 0.00012816716605957367,
      "loss": 5.1426,
      "step": 4683
    },
    {
      "epoch": 0.4684,
      "grad_norm": 7.672381401062012,
      "learning_rate": 0.00012813367109705063,
      "loss": 5.1531,
      "step": 4684
    },
    {
      "epoch": 0.4685,
      "grad_norm": 3.671056032180786,
      "learning_rate": 0.00012810017270652513,
      "loss": 5.1705,
      "step": 4685
    },
    {
      "epoch": 0.4686,
      "grad_norm": 2.689619302749634,
      "learning_rate": 0.0001280666708920788,
      "loss": 4.9925,
      "step": 4686
    },
    {
      "epoch": 0.4687,
      "grad_norm": 2.4301249980926514,
      "learning_rate": 0.00012803316565779377,
      "loss": 4.6453,
      "step": 4687
    },
    {
      "epoch": 0.4688,
      "grad_norm": 3.127243995666504,
      "learning_rate": 0.0001279996570077525,
      "loss": 4.7321,
      "step": 4688
    },
    {
      "epoch": 0.4689,
      "grad_norm": 3.698385715484619,
      "learning_rate": 0.000127966144946038,
      "loss": 4.8917,
      "step": 4689
    },
    {
      "epoch": 0.469,
      "grad_norm": 3.909945249557495,
      "learning_rate": 0.00012793262947673355,
      "loss": 5.2826,
      "step": 4690
    },
    {
      "epoch": 0.4691,
      "grad_norm": 3.7508411407470703,
      "learning_rate": 0.00012789911060392294,
      "loss": 5.1287,
      "step": 4691
    },
    {
      "epoch": 0.4692,
      "grad_norm": 4.310873508453369,
      "learning_rate": 0.00012786558833169031,
      "loss": 4.8022,
      "step": 4692
    },
    {
      "epoch": 0.4693,
      "grad_norm": 4.97828483581543,
      "learning_rate": 0.0001278320626641203,
      "loss": 5.8635,
      "step": 4693
    },
    {
      "epoch": 0.4694,
      "grad_norm": 2.6634445190429688,
      "learning_rate": 0.00012779853360529785,
      "loss": 4.9257,
      "step": 4694
    },
    {
      "epoch": 0.4695,
      "grad_norm": 4.504347801208496,
      "learning_rate": 0.00012776500115930842,
      "loss": 6.0869,
      "step": 4695
    },
    {
      "epoch": 0.4696,
      "grad_norm": 7.720456123352051,
      "learning_rate": 0.00012773146533023782,
      "loss": 5.7282,
      "step": 4696
    },
    {
      "epoch": 0.4697,
      "grad_norm": 2.1541075706481934,
      "learning_rate": 0.00012769792612217224,
      "loss": 4.6363,
      "step": 4697
    },
    {
      "epoch": 0.4698,
      "grad_norm": 2.2334043979644775,
      "learning_rate": 0.0001276643835391984,
      "loss": 4.7902,
      "step": 4698
    },
    {
      "epoch": 0.4699,
      "grad_norm": 2.554163932800293,
      "learning_rate": 0.00012763083758540337,
      "loss": 5.1938,
      "step": 4699
    },
    {
      "epoch": 0.47,
      "grad_norm": 2.19870662689209,
      "learning_rate": 0.0001275972882648746,
      "loss": 5.1911,
      "step": 4700
    },
    {
      "epoch": 0.4701,
      "grad_norm": 7.6685075759887695,
      "learning_rate": 0.0001275637355816999,
      "loss": 5.5855,
      "step": 4701
    },
    {
      "epoch": 0.4702,
      "grad_norm": 3.015491008758545,
      "learning_rate": 0.0001275301795399677,
      "loss": 5.4289,
      "step": 4702
    },
    {
      "epoch": 0.4703,
      "grad_norm": 3.445246458053589,
      "learning_rate": 0.00012749662014376663,
      "loss": 5.5405,
      "step": 4703
    },
    {
      "epoch": 0.4704,
      "grad_norm": 1.9877740144729614,
      "learning_rate": 0.00012746305739718577,
      "loss": 5.1655,
      "step": 4704
    },
    {
      "epoch": 0.4705,
      "grad_norm": 11.165088653564453,
      "learning_rate": 0.00012742949130431467,
      "loss": 4.6577,
      "step": 4705
    },
    {
      "epoch": 0.4706,
      "grad_norm": 7.8819580078125,
      "learning_rate": 0.00012739592186924328,
      "loss": 5.3741,
      "step": 4706
    },
    {
      "epoch": 0.4707,
      "grad_norm": 10.647436141967773,
      "learning_rate": 0.00012736234909606185,
      "loss": 5.919,
      "step": 4707
    },
    {
      "epoch": 0.4708,
      "grad_norm": 6.314175128936768,
      "learning_rate": 0.00012732877298886124,
      "loss": 4.7874,
      "step": 4708
    },
    {
      "epoch": 0.4709,
      "grad_norm": 7.5525102615356445,
      "learning_rate": 0.00012729519355173254,
      "loss": 4.7848,
      "step": 4709
    },
    {
      "epoch": 0.471,
      "grad_norm": 4.509529113769531,
      "learning_rate": 0.0001272616107887673,
      "loss": 5.0841,
      "step": 4710
    },
    {
      "epoch": 0.4711,
      "grad_norm": 2.8838956356048584,
      "learning_rate": 0.00012722802470405744,
      "loss": 5.231,
      "step": 4711
    },
    {
      "epoch": 0.4712,
      "grad_norm": 7.29749870300293,
      "learning_rate": 0.0001271944353016954,
      "loss": 5.4658,
      "step": 4712
    },
    {
      "epoch": 0.4713,
      "grad_norm": 4.791306018829346,
      "learning_rate": 0.00012716084258577388,
      "loss": 4.6544,
      "step": 4713
    },
    {
      "epoch": 0.4714,
      "grad_norm": 2.437469005584717,
      "learning_rate": 0.0001271272465603861,
      "loss": 5.4741,
      "step": 4714
    },
    {
      "epoch": 0.4715,
      "grad_norm": 2.6286163330078125,
      "learning_rate": 0.0001270936472296256,
      "loss": 4.9589,
      "step": 4715
    },
    {
      "epoch": 0.4716,
      "grad_norm": 4.114535808563232,
      "learning_rate": 0.00012706004459758636,
      "loss": 5.702,
      "step": 4716
    },
    {
      "epoch": 0.4717,
      "grad_norm": 7.2534685134887695,
      "learning_rate": 0.00012702643866836278,
      "loss": 4.6311,
      "step": 4717
    },
    {
      "epoch": 0.4718,
      "grad_norm": 1.9636229276657104,
      "learning_rate": 0.00012699282944604967,
      "loss": 4.6688,
      "step": 4718
    },
    {
      "epoch": 0.4719,
      "grad_norm": 3.0871617794036865,
      "learning_rate": 0.00012695921693474212,
      "loss": 4.8165,
      "step": 4719
    },
    {
      "epoch": 0.472,
      "grad_norm": 3.8091378211975098,
      "learning_rate": 0.00012692560113853574,
      "loss": 4.9048,
      "step": 4720
    },
    {
      "epoch": 0.4721,
      "grad_norm": 4.2893805503845215,
      "learning_rate": 0.00012689198206152657,
      "loss": 5.0149,
      "step": 4721
    },
    {
      "epoch": 0.4722,
      "grad_norm": 3.773608446121216,
      "learning_rate": 0.00012685835970781097,
      "loss": 5.0383,
      "step": 4722
    },
    {
      "epoch": 0.4723,
      "grad_norm": 2.6504530906677246,
      "learning_rate": 0.0001268247340814857,
      "loss": 4.9353,
      "step": 4723
    },
    {
      "epoch": 0.4724,
      "grad_norm": 6.077789306640625,
      "learning_rate": 0.00012679110518664795,
      "loss": 5.1581,
      "step": 4724
    },
    {
      "epoch": 0.4725,
      "grad_norm": 3.301248788833618,
      "learning_rate": 0.00012675747302739527,
      "loss": 5.2755,
      "step": 4725
    },
    {
      "epoch": 0.4726,
      "grad_norm": 3.4665842056274414,
      "learning_rate": 0.00012672383760782568,
      "loss": 4.7995,
      "step": 4726
    },
    {
      "epoch": 0.4727,
      "grad_norm": 2.6451659202575684,
      "learning_rate": 0.00012669019893203759,
      "loss": 4.8072,
      "step": 4727
    },
    {
      "epoch": 0.4728,
      "grad_norm": 2.308215856552124,
      "learning_rate": 0.00012665655700412967,
      "loss": 4.7808,
      "step": 4728
    },
    {
      "epoch": 0.4729,
      "grad_norm": 2.0535480976104736,
      "learning_rate": 0.00012662291182820114,
      "loss": 4.8423,
      "step": 4729
    },
    {
      "epoch": 0.473,
      "grad_norm": 2.299574136734009,
      "learning_rate": 0.00012658926340835156,
      "loss": 5.1926,
      "step": 4730
    },
    {
      "epoch": 0.4731,
      "grad_norm": 1.9899940490722656,
      "learning_rate": 0.00012655561174868088,
      "loss": 4.9261,
      "step": 4731
    },
    {
      "epoch": 0.4732,
      "grad_norm": 2.3321728706359863,
      "learning_rate": 0.00012652195685328946,
      "loss": 5.1823,
      "step": 4732
    },
    {
      "epoch": 0.4733,
      "grad_norm": 6.093174457550049,
      "learning_rate": 0.00012648829872627807,
      "loss": 5.0109,
      "step": 4733
    },
    {
      "epoch": 0.4734,
      "grad_norm": 8.367283821105957,
      "learning_rate": 0.00012645463737174782,
      "loss": 6.439,
      "step": 4734
    },
    {
      "epoch": 0.4735,
      "grad_norm": 7.974514961242676,
      "learning_rate": 0.00012642097279380027,
      "loss": 4.8377,
      "step": 4735
    },
    {
      "epoch": 0.4736,
      "grad_norm": 3.818570613861084,
      "learning_rate": 0.0001263873049965373,
      "loss": 5.0046,
      "step": 4736
    },
    {
      "epoch": 0.4737,
      "grad_norm": 5.0252604484558105,
      "learning_rate": 0.00012635363398406127,
      "loss": 5.7193,
      "step": 4737
    },
    {
      "epoch": 0.4738,
      "grad_norm": 5.194836616516113,
      "learning_rate": 0.00012631995976047488,
      "loss": 4.9649,
      "step": 4738
    },
    {
      "epoch": 0.4739,
      "grad_norm": 6.576800346374512,
      "learning_rate": 0.0001262862823298812,
      "loss": 5.873,
      "step": 4739
    },
    {
      "epoch": 0.474,
      "grad_norm": 6.03289794921875,
      "learning_rate": 0.00012625260169638378,
      "loss": 5.1836,
      "step": 4740
    },
    {
      "epoch": 0.4741,
      "grad_norm": 2.4072673320770264,
      "learning_rate": 0.00012621891786408648,
      "loss": 5.2224,
      "step": 4741
    },
    {
      "epoch": 0.4742,
      "grad_norm": 2.583235740661621,
      "learning_rate": 0.00012618523083709357,
      "loss": 5.4653,
      "step": 4742
    },
    {
      "epoch": 0.4743,
      "grad_norm": 3.2359237670898438,
      "learning_rate": 0.00012615154061950968,
      "loss": 4.6748,
      "step": 4743
    },
    {
      "epoch": 0.4744,
      "grad_norm": 4.161806106567383,
      "learning_rate": 0.00012611784721543995,
      "loss": 4.958,
      "step": 4744
    },
    {
      "epoch": 0.4745,
      "grad_norm": 3.276212692260742,
      "learning_rate": 0.00012608415062898972,
      "loss": 5.5845,
      "step": 4745
    },
    {
      "epoch": 0.4746,
      "grad_norm": 2.2946364879608154,
      "learning_rate": 0.00012605045086426487,
      "loss": 5.1709,
      "step": 4746
    },
    {
      "epoch": 0.4747,
      "grad_norm": 3.711672306060791,
      "learning_rate": 0.00012601674792537157,
      "loss": 5.2985,
      "step": 4747
    },
    {
      "epoch": 0.4748,
      "grad_norm": 3.1831235885620117,
      "learning_rate": 0.00012598304181641646,
      "loss": 4.6215,
      "step": 4748
    },
    {
      "epoch": 0.4749,
      "grad_norm": 5.441218376159668,
      "learning_rate": 0.00012594933254150655,
      "loss": 5.6404,
      "step": 4749
    },
    {
      "epoch": 0.475,
      "grad_norm": 3.936042547225952,
      "learning_rate": 0.00012591562010474914,
      "loss": 4.7964,
      "step": 4750
    },
    {
      "epoch": 0.4751,
      "grad_norm": 4.337837219238281,
      "learning_rate": 0.00012588190451025207,
      "loss": 4.5195,
      "step": 4751
    },
    {
      "epoch": 0.4752,
      "grad_norm": 4.3235697746276855,
      "learning_rate": 0.00012584818576212345,
      "loss": 5.0616,
      "step": 4752
    },
    {
      "epoch": 0.4753,
      "grad_norm": 2.1085214614868164,
      "learning_rate": 0.0001258144638644718,
      "loss": 4.7598,
      "step": 4753
    },
    {
      "epoch": 0.4754,
      "grad_norm": 2.51017689704895,
      "learning_rate": 0.000125780738821406,
      "loss": 4.3657,
      "step": 4754
    },
    {
      "epoch": 0.4755,
      "grad_norm": 4.492839813232422,
      "learning_rate": 0.0001257470106370354,
      "loss": 4.5387,
      "step": 4755
    },
    {
      "epoch": 0.4756,
      "grad_norm": 3.8681256771087646,
      "learning_rate": 0.00012571327931546963,
      "loss": 5.6806,
      "step": 4756
    },
    {
      "epoch": 0.4757,
      "grad_norm": 5.916121482849121,
      "learning_rate": 0.00012567954486081878,
      "loss": 4.9409,
      "step": 4757
    },
    {
      "epoch": 0.4758,
      "grad_norm": 2.0847582817077637,
      "learning_rate": 0.0001256458072771933,
      "loss": 4.8257,
      "step": 4758
    },
    {
      "epoch": 0.4759,
      "grad_norm": 3.7158405780792236,
      "learning_rate": 0.00012561206656870396,
      "loss": 5.891,
      "step": 4759
    },
    {
      "epoch": 0.476,
      "grad_norm": 2.796931505203247,
      "learning_rate": 0.00012557832273946205,
      "loss": 5.294,
      "step": 4760
    },
    {
      "epoch": 0.4761,
      "grad_norm": 2.282628297805786,
      "learning_rate": 0.00012554457579357905,
      "loss": 4.8943,
      "step": 4761
    },
    {
      "epoch": 0.4762,
      "grad_norm": 2.6288487911224365,
      "learning_rate": 0.00012551082573516705,
      "loss": 5.2542,
      "step": 4762
    },
    {
      "epoch": 0.4763,
      "grad_norm": 3.6782705783843994,
      "learning_rate": 0.00012547707256833823,
      "loss": 5.1136,
      "step": 4763
    },
    {
      "epoch": 0.4764,
      "grad_norm": 2.4793782234191895,
      "learning_rate": 0.00012544331629720543,
      "loss": 4.5269,
      "step": 4764
    },
    {
      "epoch": 0.4765,
      "grad_norm": 2.032851219177246,
      "learning_rate": 0.00012540955692588173,
      "loss": 5.0929,
      "step": 4765
    },
    {
      "epoch": 0.4766,
      "grad_norm": 3.9417061805725098,
      "learning_rate": 0.00012537579445848058,
      "loss": 5.6039,
      "step": 4766
    },
    {
      "epoch": 0.4767,
      "grad_norm": 2.331897020339966,
      "learning_rate": 0.00012534202889911584,
      "loss": 4.9513,
      "step": 4767
    },
    {
      "epoch": 0.4768,
      "grad_norm": 3.3849196434020996,
      "learning_rate": 0.00012530826025190174,
      "loss": 5.5662,
      "step": 4768
    },
    {
      "epoch": 0.4769,
      "grad_norm": 1.9855934381484985,
      "learning_rate": 0.00012527448852095295,
      "loss": 4.2911,
      "step": 4769
    },
    {
      "epoch": 0.477,
      "grad_norm": 4.066205024719238,
      "learning_rate": 0.00012524071371038434,
      "loss": 5.0182,
      "step": 4770
    },
    {
      "epoch": 0.4771,
      "grad_norm": 2.5596423149108887,
      "learning_rate": 0.0001252069358243114,
      "loss": 4.9422,
      "step": 4771
    },
    {
      "epoch": 0.4772,
      "grad_norm": 2.2367467880249023,
      "learning_rate": 0.00012517315486684972,
      "loss": 4.7065,
      "step": 4772
    },
    {
      "epoch": 0.4773,
      "grad_norm": 2.2307612895965576,
      "learning_rate": 0.0001251393708421155,
      "loss": 5.2554,
      "step": 4773
    },
    {
      "epoch": 0.4774,
      "grad_norm": 2.586764097213745,
      "learning_rate": 0.00012510558375422522,
      "loss": 5.0416,
      "step": 4774
    },
    {
      "epoch": 0.4775,
      "grad_norm": 2.680516242980957,
      "learning_rate": 0.0001250717936072957,
      "loss": 4.9306,
      "step": 4775
    },
    {
      "epoch": 0.4776,
      "grad_norm": 2.814789295196533,
      "learning_rate": 0.00012503800040544416,
      "loss": 5.8631,
      "step": 4776
    },
    {
      "epoch": 0.4777,
      "grad_norm": 3.5075111389160156,
      "learning_rate": 0.00012500420415278822,
      "loss": 4.8366,
      "step": 4777
    },
    {
      "epoch": 0.4778,
      "grad_norm": 2.585418939590454,
      "learning_rate": 0.00012497040485344584,
      "loss": 5.1955,
      "step": 4778
    },
    {
      "epoch": 0.4779,
      "grad_norm": 3.2078211307525635,
      "learning_rate": 0.0001249366025115354,
      "loss": 5.8421,
      "step": 4779
    },
    {
      "epoch": 0.478,
      "grad_norm": 2.7910213470458984,
      "learning_rate": 0.0001249027971311756,
      "loss": 5.0926,
      "step": 4780
    },
    {
      "epoch": 0.4781,
      "grad_norm": 3.7762699127197266,
      "learning_rate": 0.0001248689887164855,
      "loss": 5.6055,
      "step": 4781
    },
    {
      "epoch": 0.4782,
      "grad_norm": 2.1931543350219727,
      "learning_rate": 0.00012483517727158454,
      "loss": 5.3214,
      "step": 4782
    },
    {
      "epoch": 0.4783,
      "grad_norm": 4.013397216796875,
      "learning_rate": 0.00012480136280059256,
      "loss": 5.6217,
      "step": 4783
    },
    {
      "epoch": 0.4784,
      "grad_norm": 2.5854787826538086,
      "learning_rate": 0.00012476754530762977,
      "loss": 4.9281,
      "step": 4784
    },
    {
      "epoch": 0.4785,
      "grad_norm": 2.516181468963623,
      "learning_rate": 0.00012473372479681672,
      "loss": 5.2728,
      "step": 4785
    },
    {
      "epoch": 0.4786,
      "grad_norm": 3.64013671875,
      "learning_rate": 0.00012469990127227432,
      "loss": 4.7491,
      "step": 4786
    },
    {
      "epoch": 0.4787,
      "grad_norm": 1.9202715158462524,
      "learning_rate": 0.00012466607473812387,
      "loss": 5.2833,
      "step": 4787
    },
    {
      "epoch": 0.4788,
      "grad_norm": 4.027298927307129,
      "learning_rate": 0.000124632245198487,
      "loss": 4.6963,
      "step": 4788
    },
    {
      "epoch": 0.4789,
      "grad_norm": 3.1005194187164307,
      "learning_rate": 0.0001245984126574858,
      "loss": 5.0426,
      "step": 4789
    },
    {
      "epoch": 0.479,
      "grad_norm": 2.9278481006622314,
      "learning_rate": 0.00012456457711924266,
      "loss": 4.951,
      "step": 4790
    },
    {
      "epoch": 0.4791,
      "grad_norm": 3.0348124504089355,
      "learning_rate": 0.00012453073858788026,
      "loss": 4.8436,
      "step": 4791
    },
    {
      "epoch": 0.4792,
      "grad_norm": 2.4012582302093506,
      "learning_rate": 0.00012449689706752178,
      "loss": 5.0919,
      "step": 4792
    },
    {
      "epoch": 0.4793,
      "grad_norm": 2.4893274307250977,
      "learning_rate": 0.00012446305256229073,
      "loss": 4.78,
      "step": 4793
    },
    {
      "epoch": 0.4794,
      "grad_norm": 6.5770463943481445,
      "learning_rate": 0.00012442920507631092,
      "loss": 5.2875,
      "step": 4794
    },
    {
      "epoch": 0.4795,
      "grad_norm": 8.00265884399414,
      "learning_rate": 0.00012439535461370658,
      "loss": 5.8239,
      "step": 4795
    },
    {
      "epoch": 0.4796,
      "grad_norm": 2.7428390979766846,
      "learning_rate": 0.00012436150117860225,
      "loss": 4.8786,
      "step": 4796
    },
    {
      "epoch": 0.4797,
      "grad_norm": 2.437455654144287,
      "learning_rate": 0.00012432764477512292,
      "loss": 4.8899,
      "step": 4797
    },
    {
      "epoch": 0.4798,
      "grad_norm": 3.8792660236358643,
      "learning_rate": 0.00012429378540739386,
      "loss": 5.0908,
      "step": 4798
    },
    {
      "epoch": 0.4799,
      "grad_norm": 2.9274051189422607,
      "learning_rate": 0.00012425992307954075,
      "loss": 4.3465,
      "step": 4799
    },
    {
      "epoch": 0.48,
      "grad_norm": 2.110602617263794,
      "learning_rate": 0.0001242260577956896,
      "loss": 4.76,
      "step": 4800
    },
    {
      "epoch": 0.4801,
      "grad_norm": 3.0873563289642334,
      "learning_rate": 0.00012419218955996676,
      "loss": 4.8917,
      "step": 4801
    },
    {
      "epoch": 0.4802,
      "grad_norm": 3.2305052280426025,
      "learning_rate": 0.00012415831837649905,
      "loss": 4.5518,
      "step": 4802
    },
    {
      "epoch": 0.4803,
      "grad_norm": 2.426723003387451,
      "learning_rate": 0.0001241244442494135,
      "loss": 4.7148,
      "step": 4803
    },
    {
      "epoch": 0.4804,
      "grad_norm": 2.484395980834961,
      "learning_rate": 0.0001240905671828376,
      "loss": 4.9559,
      "step": 4804
    },
    {
      "epoch": 0.4805,
      "grad_norm": 2.6021668910980225,
      "learning_rate": 0.00012405668718089917,
      "loss": 4.9616,
      "step": 4805
    },
    {
      "epoch": 0.4806,
      "grad_norm": 3.4943172931671143,
      "learning_rate": 0.00012402280424772639,
      "loss": 5.4627,
      "step": 4806
    },
    {
      "epoch": 0.4807,
      "grad_norm": 2.6045889854431152,
      "learning_rate": 0.00012398891838744778,
      "loss": 4.7944,
      "step": 4807
    },
    {
      "epoch": 0.4808,
      "grad_norm": 3.772183895111084,
      "learning_rate": 0.0001239550296041922,
      "loss": 5.6275,
      "step": 4808
    },
    {
      "epoch": 0.4809,
      "grad_norm": 1.9896140098571777,
      "learning_rate": 0.00012392113790208895,
      "loss": 4.6092,
      "step": 4809
    },
    {
      "epoch": 0.481,
      "grad_norm": 2.896233558654785,
      "learning_rate": 0.0001238872432852676,
      "loss": 5.2607,
      "step": 4810
    },
    {
      "epoch": 0.4811,
      "grad_norm": 3.450072765350342,
      "learning_rate": 0.0001238533457578581,
      "loss": 5.2184,
      "step": 4811
    },
    {
      "epoch": 0.4812,
      "grad_norm": 14.179388999938965,
      "learning_rate": 0.0001238194453239908,
      "loss": 5.4071,
      "step": 4812
    },
    {
      "epoch": 0.4813,
      "grad_norm": 3.129210948944092,
      "learning_rate": 0.0001237855419877963,
      "loss": 4.9881,
      "step": 4813
    },
    {
      "epoch": 0.4814,
      "grad_norm": 5.527695655822754,
      "learning_rate": 0.0001237516357534057,
      "loss": 4.7068,
      "step": 4814
    },
    {
      "epoch": 0.4815,
      "grad_norm": 8.013246536254883,
      "learning_rate": 0.0001237177266249503,
      "loss": 4.4829,
      "step": 4815
    },
    {
      "epoch": 0.4816,
      "grad_norm": 4.016153335571289,
      "learning_rate": 0.0001236838146065619,
      "loss": 4.8288,
      "step": 4816
    },
    {
      "epoch": 0.4817,
      "grad_norm": 2.8315212726593018,
      "learning_rate": 0.00012364989970237248,
      "loss": 4.4453,
      "step": 4817
    },
    {
      "epoch": 0.4818,
      "grad_norm": 2.748802900314331,
      "learning_rate": 0.00012361598191651453,
      "loss": 5.7524,
      "step": 4818
    },
    {
      "epoch": 0.4819,
      "grad_norm": 4.618260860443115,
      "learning_rate": 0.00012358206125312083,
      "loss": 6.3267,
      "step": 4819
    },
    {
      "epoch": 0.482,
      "grad_norm": 3.0869901180267334,
      "learning_rate": 0.00012354813771632447,
      "loss": 5.1337,
      "step": 4820
    },
    {
      "epoch": 0.4821,
      "grad_norm": 4.691171169281006,
      "learning_rate": 0.000123514211310259,
      "loss": 5.4598,
      "step": 4821
    },
    {
      "epoch": 0.4822,
      "grad_norm": 6.028543472290039,
      "learning_rate": 0.00012348028203905818,
      "loss": 5.4033,
      "step": 4822
    },
    {
      "epoch": 0.4823,
      "grad_norm": 2.942174196243286,
      "learning_rate": 0.00012344634990685624,
      "loss": 5.3128,
      "step": 4823
    },
    {
      "epoch": 0.4824,
      "grad_norm": 4.653453350067139,
      "learning_rate": 0.0001234124149177877,
      "loss": 5.2041,
      "step": 4824
    },
    {
      "epoch": 0.4825,
      "grad_norm": 2.1572859287261963,
      "learning_rate": 0.00012337847707598738,
      "loss": 4.9493,
      "step": 4825
    },
    {
      "epoch": 0.4826,
      "grad_norm": 2.9391188621520996,
      "learning_rate": 0.00012334453638559057,
      "loss": 4.8009,
      "step": 4826
    },
    {
      "epoch": 0.4827,
      "grad_norm": 3.2572438716888428,
      "learning_rate": 0.0001233105928507328,
      "loss": 4.8488,
      "step": 4827
    },
    {
      "epoch": 0.4828,
      "grad_norm": 1.8434728384017944,
      "learning_rate": 0.00012327664647554998,
      "loss": 4.7048,
      "step": 4828
    },
    {
      "epoch": 0.4829,
      "grad_norm": 2.1146228313446045,
      "learning_rate": 0.0001232426972641784,
      "loss": 5.01,
      "step": 4829
    },
    {
      "epoch": 0.483,
      "grad_norm": 2.9208927154541016,
      "learning_rate": 0.00012320874522075468,
      "loss": 4.5106,
      "step": 4830
    },
    {
      "epoch": 0.4831,
      "grad_norm": 2.6312384605407715,
      "learning_rate": 0.00012317479034941573,
      "loss": 4.6589,
      "step": 4831
    },
    {
      "epoch": 0.4832,
      "grad_norm": 3.0062544345855713,
      "learning_rate": 0.0001231408326542989,
      "loss": 4.981,
      "step": 4832
    },
    {
      "epoch": 0.4833,
      "grad_norm": 2.0318076610565186,
      "learning_rate": 0.00012310687213954182,
      "loss": 4.7788,
      "step": 4833
    },
    {
      "epoch": 0.4834,
      "grad_norm": 1.9198836088180542,
      "learning_rate": 0.0001230729088092824,
      "loss": 4.544,
      "step": 4834
    },
    {
      "epoch": 0.4835,
      "grad_norm": 3.2295708656311035,
      "learning_rate": 0.00012303894266765908,
      "loss": 4.6338,
      "step": 4835
    },
    {
      "epoch": 0.4836,
      "grad_norm": 2.056459665298462,
      "learning_rate": 0.00012300497371881046,
      "loss": 4.6171,
      "step": 4836
    },
    {
      "epoch": 0.4837,
      "grad_norm": 2.718320846557617,
      "learning_rate": 0.00012297100196687557,
      "loss": 4.7432,
      "step": 4837
    },
    {
      "epoch": 0.4838,
      "grad_norm": 2.3535115718841553,
      "learning_rate": 0.00012293702741599378,
      "loss": 5.0179,
      "step": 4838
    },
    {
      "epoch": 0.4839,
      "grad_norm": 2.444675922393799,
      "learning_rate": 0.00012290305007030478,
      "loss": 4.6869,
      "step": 4839
    },
    {
      "epoch": 0.484,
      "grad_norm": 3.244028091430664,
      "learning_rate": 0.00012286906993394856,
      "loss": 4.2573,
      "step": 4840
    },
    {
      "epoch": 0.4841,
      "grad_norm": 2.7715399265289307,
      "learning_rate": 0.00012283508701106557,
      "loss": 4.7954,
      "step": 4841
    },
    {
      "epoch": 0.4842,
      "grad_norm": 2.3858344554901123,
      "learning_rate": 0.0001228011013057965,
      "loss": 4.8839,
      "step": 4842
    },
    {
      "epoch": 0.4843,
      "grad_norm": 2.0472829341888428,
      "learning_rate": 0.0001227671128222824,
      "loss": 4.5326,
      "step": 4843
    },
    {
      "epoch": 0.4844,
      "grad_norm": 3.8324074745178223,
      "learning_rate": 0.00012273312156466464,
      "loss": 5.1071,
      "step": 4844
    },
    {
      "epoch": 0.4845,
      "grad_norm": 2.7959628105163574,
      "learning_rate": 0.00012269912753708502,
      "loss": 5.0424,
      "step": 4845
    },
    {
      "epoch": 0.4846,
      "grad_norm": 9.004178047180176,
      "learning_rate": 0.0001226651307436855,
      "loss": 5.0314,
      "step": 4846
    },
    {
      "epoch": 0.4847,
      "grad_norm": 11.682735443115234,
      "learning_rate": 0.0001226311311886086,
      "loss": 4.8875,
      "step": 4847
    },
    {
      "epoch": 0.4848,
      "grad_norm": 3.678985834121704,
      "learning_rate": 0.00012259712887599698,
      "loss": 4.8073,
      "step": 4848
    },
    {
      "epoch": 0.4849,
      "grad_norm": 5.857907295227051,
      "learning_rate": 0.00012256312380999376,
      "loss": 4.6136,
      "step": 4849
    },
    {
      "epoch": 0.485,
      "grad_norm": 2.109318971633911,
      "learning_rate": 0.00012252911599474237,
      "loss": 5.2826,
      "step": 4850
    },
    {
      "epoch": 0.4851,
      "grad_norm": 2.169940233230591,
      "learning_rate": 0.0001224951054343865,
      "loss": 4.9475,
      "step": 4851
    },
    {
      "epoch": 0.4852,
      "grad_norm": 3.2414610385894775,
      "learning_rate": 0.0001224610921330703,
      "loss": 5.271,
      "step": 4852
    },
    {
      "epoch": 0.4853,
      "grad_norm": 1.9551329612731934,
      "learning_rate": 0.00012242707609493814,
      "loss": 4.5524,
      "step": 4853
    },
    {
      "epoch": 0.4854,
      "grad_norm": 2.1778080463409424,
      "learning_rate": 0.00012239305732413477,
      "loss": 5.5129,
      "step": 4854
    },
    {
      "epoch": 0.4855,
      "grad_norm": 3.659738540649414,
      "learning_rate": 0.0001223590358248053,
      "loss": 5.2609,
      "step": 4855
    },
    {
      "epoch": 0.4856,
      "grad_norm": 5.926359176635742,
      "learning_rate": 0.00012232501160109514,
      "loss": 5.814,
      "step": 4856
    },
    {
      "epoch": 0.4857,
      "grad_norm": 3.0799055099487305,
      "learning_rate": 0.00012229098465715006,
      "loss": 5.1902,
      "step": 4857
    },
    {
      "epoch": 0.4858,
      "grad_norm": 5.1428542137146,
      "learning_rate": 0.00012225695499711606,
      "loss": 5.3653,
      "step": 4858
    },
    {
      "epoch": 0.4859,
      "grad_norm": 1.890869379043579,
      "learning_rate": 0.00012222292262513965,
      "loss": 5.1247,
      "step": 4859
    },
    {
      "epoch": 0.486,
      "grad_norm": 2.448237895965576,
      "learning_rate": 0.0001221888875453675,
      "loss": 4.9479,
      "step": 4860
    },
    {
      "epoch": 0.4861,
      "grad_norm": 1.6368783712387085,
      "learning_rate": 0.00012215484976194676,
      "loss": 4.6071,
      "step": 4861
    },
    {
      "epoch": 0.4862,
      "grad_norm": 2.4537720680236816,
      "learning_rate": 0.00012212080927902474,
      "loss": 4.6557,
      "step": 4862
    },
    {
      "epoch": 0.4863,
      "grad_norm": 3.158629894256592,
      "learning_rate": 0.00012208676610074917,
      "loss": 4.8364,
      "step": 4863
    },
    {
      "epoch": 0.4864,
      "grad_norm": 2.2754218578338623,
      "learning_rate": 0.00012205272023126821,
      "loss": 5.0438,
      "step": 4864
    },
    {
      "epoch": 0.4865,
      "grad_norm": 3.0274178981781006,
      "learning_rate": 0.00012201867167473015,
      "loss": 5.364,
      "step": 4865
    },
    {
      "epoch": 0.4866,
      "grad_norm": 2.135620355606079,
      "learning_rate": 0.00012198462043528376,
      "loss": 4.8108,
      "step": 4866
    },
    {
      "epoch": 0.4867,
      "grad_norm": 2.3864424228668213,
      "learning_rate": 0.00012195056651707806,
      "loss": 4.4766,
      "step": 4867
    },
    {
      "epoch": 0.4868,
      "grad_norm": 2.246500015258789,
      "learning_rate": 0.00012191650992426238,
      "loss": 5.0911,
      "step": 4868
    },
    {
      "epoch": 0.4869,
      "grad_norm": 3.9290666580200195,
      "learning_rate": 0.00012188245066098647,
      "loss": 4.7382,
      "step": 4869
    },
    {
      "epoch": 0.487,
      "grad_norm": 3.0472609996795654,
      "learning_rate": 0.00012184838873140032,
      "loss": 4.8719,
      "step": 4870
    },
    {
      "epoch": 0.4871,
      "grad_norm": 2.667097806930542,
      "learning_rate": 0.00012181432413965428,
      "loss": 4.9532,
      "step": 4871
    },
    {
      "epoch": 0.4872,
      "grad_norm": 2.6634678840637207,
      "learning_rate": 0.000121780256889899,
      "loss": 4.865,
      "step": 4872
    },
    {
      "epoch": 0.4873,
      "grad_norm": 2.27132511138916,
      "learning_rate": 0.00012174618698628549,
      "loss": 4.7455,
      "step": 4873
    },
    {
      "epoch": 0.4874,
      "grad_norm": 8.379716873168945,
      "learning_rate": 0.00012171211443296505,
      "loss": 6.018,
      "step": 4874
    },
    {
      "epoch": 0.4875,
      "grad_norm": 5.919673919677734,
      "learning_rate": 0.00012167803923408934,
      "loss": 5.3593,
      "step": 4875
    },
    {
      "epoch": 0.4876,
      "grad_norm": 2.371420383453369,
      "learning_rate": 0.00012164396139381029,
      "loss": 4.834,
      "step": 4876
    },
    {
      "epoch": 0.4877,
      "grad_norm": 3.0776593685150146,
      "learning_rate": 0.00012160988091628022,
      "loss": 5.015,
      "step": 4877
    },
    {
      "epoch": 0.4878,
      "grad_norm": 2.5432448387145996,
      "learning_rate": 0.00012157579780565169,
      "loss": 5.4241,
      "step": 4878
    },
    {
      "epoch": 0.4879,
      "grad_norm": 2.74983286857605,
      "learning_rate": 0.00012154171206607764,
      "loss": 4.941,
      "step": 4879
    },
    {
      "epoch": 0.488,
      "grad_norm": 3.3531837463378906,
      "learning_rate": 0.00012150762370171136,
      "loss": 5.2325,
      "step": 4880
    },
    {
      "epoch": 0.4881,
      "grad_norm": 5.012828350067139,
      "learning_rate": 0.00012147353271670634,
      "loss": 5.5605,
      "step": 4881
    },
    {
      "epoch": 0.4882,
      "grad_norm": 2.622945785522461,
      "learning_rate": 0.00012143943911521647,
      "loss": 4.5208,
      "step": 4882
    },
    {
      "epoch": 0.4883,
      "grad_norm": 2.0334019660949707,
      "learning_rate": 0.000121405342901396,
      "loss": 5.3141,
      "step": 4883
    },
    {
      "epoch": 0.4884,
      "grad_norm": 2.9935319423675537,
      "learning_rate": 0.00012137124407939943,
      "loss": 5.3797,
      "step": 4884
    },
    {
      "epoch": 0.4885,
      "grad_norm": 3.5483617782592773,
      "learning_rate": 0.00012133714265338161,
      "loss": 4.6227,
      "step": 4885
    },
    {
      "epoch": 0.4886,
      "grad_norm": 3.018944025039673,
      "learning_rate": 0.00012130303862749767,
      "loss": 5.2752,
      "step": 4886
    },
    {
      "epoch": 0.4887,
      "grad_norm": 3.6635243892669678,
      "learning_rate": 0.00012126893200590308,
      "loss": 5.4485,
      "step": 4887
    },
    {
      "epoch": 0.4888,
      "grad_norm": 6.192577838897705,
      "learning_rate": 0.00012123482279275365,
      "loss": 5.4288,
      "step": 4888
    },
    {
      "epoch": 0.4889,
      "grad_norm": 2.3717844486236572,
      "learning_rate": 0.00012120071099220549,
      "loss": 4.7142,
      "step": 4889
    },
    {
      "epoch": 0.489,
      "grad_norm": 3.733391046524048,
      "learning_rate": 0.00012116659660841499,
      "loss": 5.1946,
      "step": 4890
    },
    {
      "epoch": 0.4891,
      "grad_norm": 2.5217678546905518,
      "learning_rate": 0.00012113247964553888,
      "loss": 4.7486,
      "step": 4891
    },
    {
      "epoch": 0.4892,
      "grad_norm": 9.217755317687988,
      "learning_rate": 0.00012109836010773424,
      "loss": 4.7303,
      "step": 4892
    },
    {
      "epoch": 0.4893,
      "grad_norm": 2.6160099506378174,
      "learning_rate": 0.0001210642379991584,
      "loss": 5.9137,
      "step": 4893
    },
    {
      "epoch": 0.4894,
      "grad_norm": 3.3584916591644287,
      "learning_rate": 0.00012103011332396908,
      "loss": 5.0296,
      "step": 4894
    },
    {
      "epoch": 0.4895,
      "grad_norm": 3.911177158355713,
      "learning_rate": 0.00012099598608632428,
      "loss": 6.0023,
      "step": 4895
    },
    {
      "epoch": 0.4896,
      "grad_norm": 2.108734607696533,
      "learning_rate": 0.0001209618562903822,
      "loss": 4.8222,
      "step": 4896
    },
    {
      "epoch": 0.4897,
      "grad_norm": 2.1341590881347656,
      "learning_rate": 0.00012092772394030152,
      "loss": 5.0495,
      "step": 4897
    },
    {
      "epoch": 0.4898,
      "grad_norm": 3.249053478240967,
      "learning_rate": 0.00012089358904024117,
      "loss": 4.9966,
      "step": 4898
    },
    {
      "epoch": 0.4899,
      "grad_norm": 4.907408237457275,
      "learning_rate": 0.00012085945159436038,
      "loss": 4.1406,
      "step": 4899
    },
    {
      "epoch": 0.49,
      "grad_norm": 4.1233439445495605,
      "learning_rate": 0.00012082531160681869,
      "loss": 4.7257,
      "step": 4900
    },
    {
      "epoch": 0.4901,
      "grad_norm": 2.6316258907318115,
      "learning_rate": 0.00012079116908177593,
      "loss": 5.1549,
      "step": 4901
    },
    {
      "epoch": 0.4902,
      "grad_norm": 2.418120861053467,
      "learning_rate": 0.00012075702402339231,
      "loss": 4.9499,
      "step": 4902
    },
    {
      "epoch": 0.4903,
      "grad_norm": 2.9266650676727295,
      "learning_rate": 0.00012072287643582825,
      "loss": 4.7591,
      "step": 4903
    },
    {
      "epoch": 0.4904,
      "grad_norm": 3.3835222721099854,
      "learning_rate": 0.0001206887263232446,
      "loss": 4.752,
      "step": 4904
    },
    {
      "epoch": 0.4905,
      "grad_norm": 3.0978472232818604,
      "learning_rate": 0.00012065457368980236,
      "loss": 5.3939,
      "step": 4905
    },
    {
      "epoch": 0.4906,
      "grad_norm": 5.697460651397705,
      "learning_rate": 0.00012062041853966298,
      "loss": 4.8107,
      "step": 4906
    },
    {
      "epoch": 0.4907,
      "grad_norm": 3.268723249435425,
      "learning_rate": 0.00012058626087698814,
      "loss": 4.4453,
      "step": 4907
    },
    {
      "epoch": 0.4908,
      "grad_norm": 3.1713366508483887,
      "learning_rate": 0.00012055210070593988,
      "loss": 5.4707,
      "step": 4908
    },
    {
      "epoch": 0.4909,
      "grad_norm": 2.6965456008911133,
      "learning_rate": 0.00012051793803068046,
      "loss": 4.4714,
      "step": 4909
    },
    {
      "epoch": 0.491,
      "grad_norm": 3.5864765644073486,
      "learning_rate": 0.00012048377285537256,
      "loss": 5.0494,
      "step": 4910
    },
    {
      "epoch": 0.4911,
      "grad_norm": 2.2198383808135986,
      "learning_rate": 0.00012044960518417903,
      "loss": 4.7352,
      "step": 4911
    },
    {
      "epoch": 0.4912,
      "grad_norm": 2.982361078262329,
      "learning_rate": 0.00012041543502126318,
      "loss": 4.8476,
      "step": 4912
    },
    {
      "epoch": 0.4913,
      "grad_norm": 2.9540412425994873,
      "learning_rate": 0.0001203812623707885,
      "loss": 4.6411,
      "step": 4913
    },
    {
      "epoch": 0.4914,
      "grad_norm": 4.005640506744385,
      "learning_rate": 0.00012034708723691881,
      "loss": 4.6144,
      "step": 4914
    },
    {
      "epoch": 0.4915,
      "grad_norm": 3.819263219833374,
      "learning_rate": 0.00012031290962381823,
      "loss": 5.832,
      "step": 4915
    },
    {
      "epoch": 0.4916,
      "grad_norm": 4.36256742477417,
      "learning_rate": 0.00012027872953565125,
      "loss": 5.7277,
      "step": 4916
    },
    {
      "epoch": 0.4917,
      "grad_norm": 2.987086057662964,
      "learning_rate": 0.00012024454697658261,
      "loss": 4.5695,
      "step": 4917
    },
    {
      "epoch": 0.4918,
      "grad_norm": 2.2818424701690674,
      "learning_rate": 0.00012021036195077731,
      "loss": 4.9024,
      "step": 4918
    },
    {
      "epoch": 0.4919,
      "grad_norm": 2.981339931488037,
      "learning_rate": 0.0001201761744624007,
      "loss": 4.8822,
      "step": 4919
    },
    {
      "epoch": 0.492,
      "grad_norm": 3.433734178543091,
      "learning_rate": 0.00012014198451561842,
      "loss": 4.6834,
      "step": 4920
    },
    {
      "epoch": 0.4921,
      "grad_norm": 5.027676582336426,
      "learning_rate": 0.00012010779211459648,
      "loss": 4.2194,
      "step": 4921
    },
    {
      "epoch": 0.4922,
      "grad_norm": 3.107717275619507,
      "learning_rate": 0.00012007359726350105,
      "loss": 6.0258,
      "step": 4922
    },
    {
      "epoch": 0.4923,
      "grad_norm": 2.843733072280884,
      "learning_rate": 0.00012003939996649865,
      "loss": 5.6467,
      "step": 4923
    },
    {
      "epoch": 0.4924,
      "grad_norm": 2.9290971755981445,
      "learning_rate": 0.00012000520022775617,
      "loss": 5.0139,
      "step": 4924
    },
    {
      "epoch": 0.4925,
      "grad_norm": 3.2461395263671875,
      "learning_rate": 0.00011997099805144069,
      "loss": 5.0892,
      "step": 4925
    },
    {
      "epoch": 0.4926,
      "grad_norm": 2.8921313285827637,
      "learning_rate": 0.00011993679344171973,
      "loss": 5.2313,
      "step": 4926
    },
    {
      "epoch": 0.4927,
      "grad_norm": 3.3276569843292236,
      "learning_rate": 0.00011990258640276094,
      "loss": 4.8312,
      "step": 4927
    },
    {
      "epoch": 0.4928,
      "grad_norm": 4.44606876373291,
      "learning_rate": 0.00011986837693873237,
      "loss": 5.5901,
      "step": 4928
    },
    {
      "epoch": 0.4929,
      "grad_norm": 3.9995920658111572,
      "learning_rate": 0.00011983416505380234,
      "loss": 5.369,
      "step": 4929
    },
    {
      "epoch": 0.493,
      "grad_norm": 2.481658697128296,
      "learning_rate": 0.00011979995075213946,
      "loss": 4.8307,
      "step": 4930
    },
    {
      "epoch": 0.4931,
      "grad_norm": 3.3066508769989014,
      "learning_rate": 0.00011976573403791262,
      "loss": 4.738,
      "step": 4931
    },
    {
      "epoch": 0.4932,
      "grad_norm": 2.3699305057525635,
      "learning_rate": 0.00011973151491529107,
      "loss": 4.9266,
      "step": 4932
    },
    {
      "epoch": 0.4933,
      "grad_norm": 2.316655397415161,
      "learning_rate": 0.00011969729338844429,
      "loss": 4.5342,
      "step": 4933
    },
    {
      "epoch": 0.4934,
      "grad_norm": 1.6117398738861084,
      "learning_rate": 0.000119663069461542,
      "loss": 4.7484,
      "step": 4934
    },
    {
      "epoch": 0.4935,
      "grad_norm": 3.4555912017822266,
      "learning_rate": 0.0001196288431387544,
      "loss": 6.0919,
      "step": 4935
    },
    {
      "epoch": 0.4936,
      "grad_norm": 2.422483205795288,
      "learning_rate": 0.00011959461442425177,
      "loss": 4.5343,
      "step": 4936
    },
    {
      "epoch": 0.4937,
      "grad_norm": 2.2364673614501953,
      "learning_rate": 0.00011956038332220483,
      "loss": 4.7541,
      "step": 4937
    },
    {
      "epoch": 0.4938,
      "grad_norm": 2.6670517921447754,
      "learning_rate": 0.00011952614983678452,
      "loss": 5.0481,
      "step": 4938
    },
    {
      "epoch": 0.4939,
      "grad_norm": 2.071582078933716,
      "learning_rate": 0.00011949191397216206,
      "loss": 4.8796,
      "step": 4939
    },
    {
      "epoch": 0.494,
      "grad_norm": 2.7773704528808594,
      "learning_rate": 0.00011945767573250903,
      "loss": 5.1985,
      "step": 4940
    },
    {
      "epoch": 0.4941,
      "grad_norm": 11.301085472106934,
      "learning_rate": 0.0001194234351219972,
      "loss": 5.2509,
      "step": 4941
    },
    {
      "epoch": 0.4942,
      "grad_norm": 2.823653221130371,
      "learning_rate": 0.00011938919214479876,
      "loss": 4.9079,
      "step": 4942
    },
    {
      "epoch": 0.4943,
      "grad_norm": 4.341390609741211,
      "learning_rate": 0.00011935494680508606,
      "loss": 5.1798,
      "step": 4943
    },
    {
      "epoch": 0.4944,
      "grad_norm": 2.48140811920166,
      "learning_rate": 0.00011932069910703176,
      "loss": 5.4186,
      "step": 4944
    },
    {
      "epoch": 0.4945,
      "grad_norm": 5.887575149536133,
      "learning_rate": 0.0001192864490548089,
      "loss": 5.4662,
      "step": 4945
    },
    {
      "epoch": 0.4946,
      "grad_norm": 2.4830727577209473,
      "learning_rate": 0.00011925219665259075,
      "loss": 4.3281,
      "step": 4946
    },
    {
      "epoch": 0.4947,
      "grad_norm": 3.20198917388916,
      "learning_rate": 0.00011921794190455082,
      "loss": 5.3228,
      "step": 4947
    },
    {
      "epoch": 0.4948,
      "grad_norm": 4.6519246101379395,
      "learning_rate": 0.00011918368481486297,
      "loss": 5.1634,
      "step": 4948
    },
    {
      "epoch": 0.4949,
      "grad_norm": 4.230257034301758,
      "learning_rate": 0.00011914942538770131,
      "loss": 4.9572,
      "step": 4949
    },
    {
      "epoch": 0.495,
      "grad_norm": 2.1922645568847656,
      "learning_rate": 0.00011911516362724024,
      "loss": 5.221,
      "step": 4950
    },
    {
      "epoch": 0.4951,
      "grad_norm": 2.1267495155334473,
      "learning_rate": 0.00011908089953765449,
      "loss": 4.7267,
      "step": 4951
    },
    {
      "epoch": 0.4952,
      "grad_norm": 2.4641733169555664,
      "learning_rate": 0.00011904663312311901,
      "loss": 5.0673,
      "step": 4952
    },
    {
      "epoch": 0.4953,
      "grad_norm": 2.283445119857788,
      "learning_rate": 0.00011901236438780902,
      "loss": 5.0093,
      "step": 4953
    },
    {
      "epoch": 0.4954,
      "grad_norm": 3.4752743244171143,
      "learning_rate": 0.00011897809333590014,
      "loss": 4.9175,
      "step": 4954
    },
    {
      "epoch": 0.4955,
      "grad_norm": 3.387819528579712,
      "learning_rate": 0.00011894381997156813,
      "loss": 4.9986,
      "step": 4955
    },
    {
      "epoch": 0.4956,
      "grad_norm": 2.4482810497283936,
      "learning_rate": 0.00011890954429898912,
      "loss": 4.957,
      "step": 4956
    },
    {
      "epoch": 0.4957,
      "grad_norm": 2.4773921966552734,
      "learning_rate": 0.00011887526632233954,
      "loss": 4.6247,
      "step": 4957
    },
    {
      "epoch": 0.4958,
      "grad_norm": 2.6392762660980225,
      "learning_rate": 0.00011884098604579597,
      "loss": 4.559,
      "step": 4958
    },
    {
      "epoch": 0.4959,
      "grad_norm": 2.023937702178955,
      "learning_rate": 0.00011880670347353539,
      "loss": 4.6492,
      "step": 4959
    },
    {
      "epoch": 0.496,
      "grad_norm": 2.164858341217041,
      "learning_rate": 0.00011877241860973507,
      "loss": 5.3458,
      "step": 4960
    },
    {
      "epoch": 0.4961,
      "grad_norm": 5.6744842529296875,
      "learning_rate": 0.00011873813145857249,
      "loss": 4.7125,
      "step": 4961
    },
    {
      "epoch": 0.4962,
      "grad_norm": 2.919757604598999,
      "learning_rate": 0.0001187038420242254,
      "loss": 5.0287,
      "step": 4962
    },
    {
      "epoch": 0.4963,
      "grad_norm": 5.008179664611816,
      "learning_rate": 0.0001186695503108719,
      "loss": 4.681,
      "step": 4963
    },
    {
      "epoch": 0.4964,
      "grad_norm": 4.873087406158447,
      "learning_rate": 0.00011863525632269032,
      "loss": 5.2699,
      "step": 4964
    },
    {
      "epoch": 0.4965,
      "grad_norm": 4.111302375793457,
      "learning_rate": 0.0001186009600638593,
      "loss": 5.3459,
      "step": 4965
    },
    {
      "epoch": 0.4966,
      "grad_norm": 3.8457441329956055,
      "learning_rate": 0.00011856666153855776,
      "loss": 4.4284,
      "step": 4966
    },
    {
      "epoch": 0.4967,
      "grad_norm": 4.197775363922119,
      "learning_rate": 0.00011853236075096474,
      "loss": 4.6924,
      "step": 4967
    },
    {
      "epoch": 0.4968,
      "grad_norm": 2.9448797702789307,
      "learning_rate": 0.00011849805770525983,
      "loss": 4.936,
      "step": 4968
    },
    {
      "epoch": 0.4969,
      "grad_norm": 3.7674131393432617,
      "learning_rate": 0.0001184637524056227,
      "loss": 4.8192,
      "step": 4969
    },
    {
      "epoch": 0.497,
      "grad_norm": 2.0034499168395996,
      "learning_rate": 0.00011842944485623335,
      "loss": 4.5121,
      "step": 4970
    },
    {
      "epoch": 0.4971,
      "grad_norm": 2.4077091217041016,
      "learning_rate": 0.00011839513506127203,
      "loss": 5.1868,
      "step": 4971
    },
    {
      "epoch": 0.4972,
      "grad_norm": 2.0205204486846924,
      "learning_rate": 0.0001183608230249193,
      "loss": 4.4917,
      "step": 4972
    },
    {
      "epoch": 0.4973,
      "grad_norm": 1.8754981756210327,
      "learning_rate": 0.00011832650875135598,
      "loss": 4.5543,
      "step": 4973
    },
    {
      "epoch": 0.4974,
      "grad_norm": 4.492429256439209,
      "learning_rate": 0.00011829219224476318,
      "loss": 4.5765,
      "step": 4974
    },
    {
      "epoch": 0.4975,
      "grad_norm": 2.2653753757476807,
      "learning_rate": 0.00011825787350932222,
      "loss": 4.4569,
      "step": 4975
    },
    {
      "epoch": 0.4976,
      "grad_norm": 2.5028133392333984,
      "learning_rate": 0.00011822355254921478,
      "loss": 5.7617,
      "step": 4976
    },
    {
      "epoch": 0.4977,
      "grad_norm": 3.859727144241333,
      "learning_rate": 0.00011818922936862269,
      "loss": 5.0855,
      "step": 4977
    },
    {
      "epoch": 0.4978,
      "grad_norm": 2.990004062652588,
      "learning_rate": 0.00011815490397172821,
      "loss": 5.1119,
      "step": 4978
    },
    {
      "epoch": 0.4979,
      "grad_norm": 2.5834126472473145,
      "learning_rate": 0.00011812057636271374,
      "loss": 4.7495,
      "step": 4979
    },
    {
      "epoch": 0.498,
      "grad_norm": 2.39286208152771,
      "learning_rate": 0.00011808624654576202,
      "loss": 4.8496,
      "step": 4980
    },
    {
      "epoch": 0.4981,
      "grad_norm": 2.2606890201568604,
      "learning_rate": 0.00011805191452505602,
      "loss": 4.6404,
      "step": 4981
    },
    {
      "epoch": 0.4982,
      "grad_norm": 2.910456657409668,
      "learning_rate": 0.00011801758030477897,
      "loss": 5.0601,
      "step": 4982
    },
    {
      "epoch": 0.4983,
      "grad_norm": 2.684546947479248,
      "learning_rate": 0.00011798324388911444,
      "loss": 5.5687,
      "step": 4983
    },
    {
      "epoch": 0.4984,
      "grad_norm": 2.0667717456817627,
      "learning_rate": 0.00011794890528224618,
      "loss": 5.0922,
      "step": 4984
    },
    {
      "epoch": 0.4985,
      "grad_norm": 2.9432523250579834,
      "learning_rate": 0.00011791456448835825,
      "loss": 4.5208,
      "step": 4985
    },
    {
      "epoch": 0.4986,
      "grad_norm": 2.7403383255004883,
      "learning_rate": 0.00011788022151163495,
      "loss": 5.0856,
      "step": 4986
    },
    {
      "epoch": 0.4987,
      "grad_norm": 2.445810079574585,
      "learning_rate": 0.00011784587635626094,
      "loss": 4.7452,
      "step": 4987
    },
    {
      "epoch": 0.4988,
      "grad_norm": 2.7646374702453613,
      "learning_rate": 0.000117811529026421,
      "loss": 4.7274,
      "step": 4988
    },
    {
      "epoch": 0.4989,
      "grad_norm": 2.97550368309021,
      "learning_rate": 0.00011777717952630031,
      "loss": 4.9399,
      "step": 4989
    },
    {
      "epoch": 0.499,
      "grad_norm": 4.3508381843566895,
      "learning_rate": 0.00011774282786008422,
      "loss": 5.0856,
      "step": 4990
    },
    {
      "epoch": 0.4991,
      "grad_norm": 2.334636926651001,
      "learning_rate": 0.00011770847403195834,
      "loss": 5.2918,
      "step": 4991
    },
    {
      "epoch": 0.4992,
      "grad_norm": 2.279707670211792,
      "learning_rate": 0.00011767411804610864,
      "loss": 4.7018,
      "step": 4992
    },
    {
      "epoch": 0.4993,
      "grad_norm": 2.7558107376098633,
      "learning_rate": 0.00011763975990672125,
      "loss": 4.5936,
      "step": 4993
    },
    {
      "epoch": 0.4994,
      "grad_norm": 2.7103822231292725,
      "learning_rate": 0.00011760539961798262,
      "loss": 4.8384,
      "step": 4994
    },
    {
      "epoch": 0.4995,
      "grad_norm": 2.484658718109131,
      "learning_rate": 0.00011757103718407947,
      "loss": 5.1806,
      "step": 4995
    },
    {
      "epoch": 0.4996,
      "grad_norm": 2.7403926849365234,
      "learning_rate": 0.00011753667260919872,
      "loss": 5.0326,
      "step": 4996
    },
    {
      "epoch": 0.4997,
      "grad_norm": 4.221589088439941,
      "learning_rate": 0.00011750230589752762,
      "loss": 4.772,
      "step": 4997
    },
    {
      "epoch": 0.4998,
      "grad_norm": 1.9203424453735352,
      "learning_rate": 0.00011746793705325363,
      "loss": 4.7126,
      "step": 4998
    },
    {
      "epoch": 0.4999,
      "grad_norm": 2.4649789333343506,
      "learning_rate": 0.00011743356608056449,
      "loss": 4.7522,
      "step": 4999
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.14591121673584,
      "learning_rate": 0.0001173991929836482,
      "loss": 5.3436,
      "step": 5000
    },
    {
      "epoch": 0.5001,
      "grad_norm": 2.8451600074768066,
      "learning_rate": 0.00011736481776669306,
      "loss": 5.2284,
      "step": 5001
    },
    {
      "epoch": 0.5002,
      "grad_norm": 3.3560030460357666,
      "learning_rate": 0.00011733044043388752,
      "loss": 4.6536,
      "step": 5002
    },
    {
      "epoch": 0.5003,
      "grad_norm": 7.708975791931152,
      "learning_rate": 0.00011729606098942039,
      "loss": 5.0056,
      "step": 5003
    },
    {
      "epoch": 0.5004,
      "grad_norm": 2.9192917346954346,
      "learning_rate": 0.00011726167943748067,
      "loss": 5.0889,
      "step": 5004
    },
    {
      "epoch": 0.5005,
      "grad_norm": 4.045736312866211,
      "learning_rate": 0.00011722729578225769,
      "loss": 5.115,
      "step": 5005
    },
    {
      "epoch": 0.5006,
      "grad_norm": 5.1102213859558105,
      "learning_rate": 0.00011719291002794096,
      "loss": 4.8729,
      "step": 5006
    },
    {
      "epoch": 0.5007,
      "grad_norm": 2.505596160888672,
      "learning_rate": 0.0001171585221787203,
      "loss": 4.7551,
      "step": 5007
    },
    {
      "epoch": 0.5008,
      "grad_norm": 2.3833024501800537,
      "learning_rate": 0.00011712413223878578,
      "loss": 4.8379,
      "step": 5008
    },
    {
      "epoch": 0.5009,
      "grad_norm": 3.372441053390503,
      "learning_rate": 0.00011708974021232769,
      "loss": 5.3406,
      "step": 5009
    },
    {
      "epoch": 0.501,
      "grad_norm": 3.5426454544067383,
      "learning_rate": 0.00011705534610353657,
      "loss": 4.9527,
      "step": 5010
    },
    {
      "epoch": 0.5011,
      "grad_norm": 2.74198842048645,
      "learning_rate": 0.00011702094991660326,
      "loss": 4.7336,
      "step": 5011
    },
    {
      "epoch": 0.5012,
      "grad_norm": 2.254774808883667,
      "learning_rate": 0.00011698655165571886,
      "loss": 5.0044,
      "step": 5012
    },
    {
      "epoch": 0.5013,
      "grad_norm": 2.925022602081299,
      "learning_rate": 0.00011695215132507464,
      "loss": 5.2192,
      "step": 5013
    },
    {
      "epoch": 0.5014,
      "grad_norm": 3.2504115104675293,
      "learning_rate": 0.00011691774892886222,
      "loss": 4.6316,
      "step": 5014
    },
    {
      "epoch": 0.5015,
      "grad_norm": 4.667217254638672,
      "learning_rate": 0.00011688334447127338,
      "loss": 5.7521,
      "step": 5015
    },
    {
      "epoch": 0.5016,
      "grad_norm": 2.811882734298706,
      "learning_rate": 0.00011684893795650027,
      "loss": 4.4376,
      "step": 5016
    },
    {
      "epoch": 0.5017,
      "grad_norm": 3.495797872543335,
      "learning_rate": 0.00011681452938873516,
      "loss": 4.8015,
      "step": 5017
    },
    {
      "epoch": 0.5018,
      "grad_norm": 4.252373695373535,
      "learning_rate": 0.00011678011877217065,
      "loss": 4.8528,
      "step": 5018
    },
    {
      "epoch": 0.5019,
      "grad_norm": 3.331357479095459,
      "learning_rate": 0.00011674570611099955,
      "loss": 4.6531,
      "step": 5019
    },
    {
      "epoch": 0.502,
      "grad_norm": 3.4586868286132812,
      "learning_rate": 0.00011671129140941499,
      "loss": 4.7263,
      "step": 5020
    },
    {
      "epoch": 0.5021,
      "grad_norm": 2.5030298233032227,
      "learning_rate": 0.00011667687467161024,
      "loss": 4.8988,
      "step": 5021
    },
    {
      "epoch": 0.5022,
      "grad_norm": 3.3324036598205566,
      "learning_rate": 0.00011664245590177892,
      "loss": 4.5944,
      "step": 5022
    },
    {
      "epoch": 0.5023,
      "grad_norm": 2.7645883560180664,
      "learning_rate": 0.0001166080351041148,
      "loss": 5.0468,
      "step": 5023
    },
    {
      "epoch": 0.5024,
      "grad_norm": 2.9634878635406494,
      "learning_rate": 0.00011657361228281199,
      "loss": 5.1913,
      "step": 5024
    },
    {
      "epoch": 0.5025,
      "grad_norm": 3.0980849266052246,
      "learning_rate": 0.00011653918744206478,
      "loss": 5.703,
      "step": 5025
    },
    {
      "epoch": 0.5026,
      "grad_norm": 3.987279176712036,
      "learning_rate": 0.00011650476058606777,
      "loss": 5.5342,
      "step": 5026
    },
    {
      "epoch": 0.5027,
      "grad_norm": 2.197049140930176,
      "learning_rate": 0.00011647033171901573,
      "loss": 4.4372,
      "step": 5027
    },
    {
      "epoch": 0.5028,
      "grad_norm": 1.8245807886123657,
      "learning_rate": 0.00011643590084510379,
      "loss": 4.6808,
      "step": 5028
    },
    {
      "epoch": 0.5029,
      "grad_norm": 2.2848262786865234,
      "learning_rate": 0.00011640146796852711,
      "loss": 4.5173,
      "step": 5029
    },
    {
      "epoch": 0.503,
      "grad_norm": 2.455059766769409,
      "learning_rate": 0.00011636703309348133,
      "loss": 5.0271,
      "step": 5030
    },
    {
      "epoch": 0.5031,
      "grad_norm": 3.7837586402893066,
      "learning_rate": 0.00011633259622416224,
      "loss": 5.3993,
      "step": 5031
    },
    {
      "epoch": 0.5032,
      "grad_norm": 2.1076014041900635,
      "learning_rate": 0.00011629815736476581,
      "loss": 4.7669,
      "step": 5032
    },
    {
      "epoch": 0.5033,
      "grad_norm": 2.507875919342041,
      "learning_rate": 0.00011626371651948838,
      "loss": 4.4948,
      "step": 5033
    },
    {
      "epoch": 0.5034,
      "grad_norm": 2.5080716609954834,
      "learning_rate": 0.00011622927369252638,
      "loss": 4.4482,
      "step": 5034
    },
    {
      "epoch": 0.5035,
      "grad_norm": 4.49135160446167,
      "learning_rate": 0.00011619482888807662,
      "loss": 5.4216,
      "step": 5035
    },
    {
      "epoch": 0.5036,
      "grad_norm": 5.0831732749938965,
      "learning_rate": 0.00011616038211033613,
      "loss": 4.5766,
      "step": 5036
    },
    {
      "epoch": 0.5037,
      "grad_norm": 2.7157046794891357,
      "learning_rate": 0.00011612593336350208,
      "loss": 4.862,
      "step": 5037
    },
    {
      "epoch": 0.5038,
      "grad_norm": 2.0702767372131348,
      "learning_rate": 0.00011609148265177193,
      "loss": 4.942,
      "step": 5038
    },
    {
      "epoch": 0.5039,
      "grad_norm": 4.730004787445068,
      "learning_rate": 0.00011605702997934345,
      "loss": 4.7977,
      "step": 5039
    },
    {
      "epoch": 0.504,
      "grad_norm": 3.691871166229248,
      "learning_rate": 0.00011602257535041459,
      "loss": 4.5446,
      "step": 5040
    },
    {
      "epoch": 0.5041,
      "grad_norm": 2.9457318782806396,
      "learning_rate": 0.0001159881187691835,
      "loss": 5.0347,
      "step": 5041
    },
    {
      "epoch": 0.5042,
      "grad_norm": 3.7831826210021973,
      "learning_rate": 0.00011595366023984864,
      "loss": 5.0627,
      "step": 5042
    },
    {
      "epoch": 0.5043,
      "grad_norm": 2.3857853412628174,
      "learning_rate": 0.00011591919976660868,
      "loss": 5.085,
      "step": 5043
    },
    {
      "epoch": 0.5044,
      "grad_norm": 2.155697822570801,
      "learning_rate": 0.00011588473735366249,
      "loss": 4.5819,
      "step": 5044
    },
    {
      "epoch": 0.5045,
      "grad_norm": 2.4268903732299805,
      "learning_rate": 0.0001158502730052093,
      "loss": 4.7569,
      "step": 5045
    },
    {
      "epoch": 0.5046,
      "grad_norm": 5.963418960571289,
      "learning_rate": 0.00011581580672544838,
      "loss": 5.02,
      "step": 5046
    },
    {
      "epoch": 0.5047,
      "grad_norm": 2.442901849746704,
      "learning_rate": 0.0001157813385185794,
      "loss": 4.316,
      "step": 5047
    },
    {
      "epoch": 0.5048,
      "grad_norm": 3.6661458015441895,
      "learning_rate": 0.00011574686838880215,
      "loss": 4.6278,
      "step": 5048
    },
    {
      "epoch": 0.5049,
      "grad_norm": 5.693005084991455,
      "learning_rate": 0.00011571239634031679,
      "loss": 6.065,
      "step": 5049
    },
    {
      "epoch": 0.505,
      "grad_norm": 2.9857017993927,
      "learning_rate": 0.00011567792237732358,
      "loss": 5.2417,
      "step": 5050
    },
    {
      "epoch": 0.5051,
      "grad_norm": 2.729140043258667,
      "learning_rate": 0.0001156434465040231,
      "loss": 5.7351,
      "step": 5051
    },
    {
      "epoch": 0.5052,
      "grad_norm": 3.6620161533355713,
      "learning_rate": 0.0001156089687246161,
      "loss": 5.3935,
      "step": 5052
    },
    {
      "epoch": 0.5053,
      "grad_norm": 2.436286449432373,
      "learning_rate": 0.00011557448904330362,
      "loss": 5.3555,
      "step": 5053
    },
    {
      "epoch": 0.5054,
      "grad_norm": 5.728464126586914,
      "learning_rate": 0.0001155400074642869,
      "loss": 5.1032,
      "step": 5054
    },
    {
      "epoch": 0.5055,
      "grad_norm": 5.030971527099609,
      "learning_rate": 0.00011550552399176739,
      "loss": 5.0303,
      "step": 5055
    },
    {
      "epoch": 0.5056,
      "grad_norm": 3.0096194744110107,
      "learning_rate": 0.00011547103862994684,
      "loss": 4.6314,
      "step": 5056
    },
    {
      "epoch": 0.5057,
      "grad_norm": 2.5301170349121094,
      "learning_rate": 0.00011543655138302714,
      "loss": 4.2339,
      "step": 5057
    },
    {
      "epoch": 0.5058,
      "grad_norm": 17.05284309387207,
      "learning_rate": 0.00011540206225521046,
      "loss": 5.4242,
      "step": 5058
    },
    {
      "epoch": 0.5059,
      "grad_norm": 3.372722864151001,
      "learning_rate": 0.00011536757125069923,
      "loss": 5.4426,
      "step": 5059
    },
    {
      "epoch": 0.506,
      "grad_norm": 2.4192373752593994,
      "learning_rate": 0.00011533307837369607,
      "loss": 5.5033,
      "step": 5060
    },
    {
      "epoch": 0.5061,
      "grad_norm": 3.3812389373779297,
      "learning_rate": 0.00011529858362840382,
      "loss": 4.7023,
      "step": 5061
    },
    {
      "epoch": 0.5062,
      "grad_norm": 2.94621205329895,
      "learning_rate": 0.00011526408701902556,
      "loss": 4.7676,
      "step": 5062
    },
    {
      "epoch": 0.5063,
      "grad_norm": 4.642330169677734,
      "learning_rate": 0.00011522958854976458,
      "loss": 5.7677,
      "step": 5063
    },
    {
      "epoch": 0.5064,
      "grad_norm": 2.5625953674316406,
      "learning_rate": 0.00011519508822482446,
      "loss": 4.2805,
      "step": 5064
    },
    {
      "epoch": 0.5065,
      "grad_norm": 5.776980400085449,
      "learning_rate": 0.00011516058604840891,
      "loss": 5.763,
      "step": 5065
    },
    {
      "epoch": 0.5066,
      "grad_norm": 7.6262664794921875,
      "learning_rate": 0.00011512608202472194,
      "loss": 6.5375,
      "step": 5066
    },
    {
      "epoch": 0.5067,
      "grad_norm": 4.7243242263793945,
      "learning_rate": 0.00011509157615796776,
      "loss": 4.674,
      "step": 5067
    },
    {
      "epoch": 0.5068,
      "grad_norm": 2.3028485774993896,
      "learning_rate": 0.00011505706845235078,
      "loss": 4.4855,
      "step": 5068
    },
    {
      "epoch": 0.5069,
      "grad_norm": 2.2853102684020996,
      "learning_rate": 0.00011502255891207572,
      "loss": 4.7581,
      "step": 5069
    },
    {
      "epoch": 0.507,
      "grad_norm": 3.6515040397644043,
      "learning_rate": 0.0001149880475413474,
      "loss": 4.9924,
      "step": 5070
    },
    {
      "epoch": 0.5071,
      "grad_norm": 2.5292279720306396,
      "learning_rate": 0.00011495353434437098,
      "loss": 5.2638,
      "step": 5071
    },
    {
      "epoch": 0.5072,
      "grad_norm": 8.069612503051758,
      "learning_rate": 0.00011491901932535172,
      "loss": 5.5975,
      "step": 5072
    },
    {
      "epoch": 0.5073,
      "grad_norm": 4.674898624420166,
      "learning_rate": 0.00011488450248849522,
      "loss": 5.4025,
      "step": 5073
    },
    {
      "epoch": 0.5074,
      "grad_norm": 4.054290294647217,
      "learning_rate": 0.00011484998383800726,
      "loss": 4.6782,
      "step": 5074
    },
    {
      "epoch": 0.5075,
      "grad_norm": 5.124908924102783,
      "learning_rate": 0.00011481546337809381,
      "loss": 4.6242,
      "step": 5075
    },
    {
      "epoch": 0.5076,
      "grad_norm": 2.5155999660491943,
      "learning_rate": 0.00011478094111296109,
      "loss": 4.3974,
      "step": 5076
    },
    {
      "epoch": 0.5077,
      "grad_norm": 3.5744380950927734,
      "learning_rate": 0.0001147464170468155,
      "loss": 4.6753,
      "step": 5077
    },
    {
      "epoch": 0.5078,
      "grad_norm": 2.0337471961975098,
      "learning_rate": 0.00011471189118386375,
      "loss": 4.7506,
      "step": 5078
    },
    {
      "epoch": 0.5079,
      "grad_norm": 2.720656394958496,
      "learning_rate": 0.00011467736352831266,
      "loss": 4.5058,
      "step": 5079
    },
    {
      "epoch": 0.508,
      "grad_norm": 3.0553536415100098,
      "learning_rate": 0.00011464283408436938,
      "loss": 4.8935,
      "step": 5080
    },
    {
      "epoch": 0.5081,
      "grad_norm": 2.7661356925964355,
      "learning_rate": 0.00011460830285624118,
      "loss": 4.9841,
      "step": 5081
    },
    {
      "epoch": 0.5082,
      "grad_norm": 3.14349102973938,
      "learning_rate": 0.00011457376984813557,
      "loss": 4.9125,
      "step": 5082
    },
    {
      "epoch": 0.5083,
      "grad_norm": 6.624685764312744,
      "learning_rate": 0.00011453923506426032,
      "loss": 5.4318,
      "step": 5083
    },
    {
      "epoch": 0.5084,
      "grad_norm": 2.3148417472839355,
      "learning_rate": 0.00011450469850882337,
      "loss": 5.3543,
      "step": 5084
    },
    {
      "epoch": 0.5085,
      "grad_norm": 3.438167095184326,
      "learning_rate": 0.00011447016018603292,
      "loss": 5.8574,
      "step": 5085
    },
    {
      "epoch": 0.5086,
      "grad_norm": 3.7067909240722656,
      "learning_rate": 0.00011443562010009731,
      "loss": 5.0886,
      "step": 5086
    },
    {
      "epoch": 0.5087,
      "grad_norm": 2.303342580795288,
      "learning_rate": 0.00011440107825522521,
      "loss": 5.1018,
      "step": 5087
    },
    {
      "epoch": 0.5088,
      "grad_norm": 2.7623560428619385,
      "learning_rate": 0.00011436653465562542,
      "loss": 4.4556,
      "step": 5088
    },
    {
      "epoch": 0.5089,
      "grad_norm": 5.558529376983643,
      "learning_rate": 0.00011433198930550695,
      "loss": 5.3755,
      "step": 5089
    },
    {
      "epoch": 0.509,
      "grad_norm": 2.8604164123535156,
      "learning_rate": 0.00011429744220907903,
      "loss": 4.5823,
      "step": 5090
    },
    {
      "epoch": 0.5091,
      "grad_norm": 2.7095723152160645,
      "learning_rate": 0.00011426289337055119,
      "loss": 4.8684,
      "step": 5091
    },
    {
      "epoch": 0.5092,
      "grad_norm": 3.9518239498138428,
      "learning_rate": 0.00011422834279413301,
      "loss": 5.5566,
      "step": 5092
    },
    {
      "epoch": 0.5093,
      "grad_norm": 3.374332904815674,
      "learning_rate": 0.00011419379048403444,
      "loss": 5.5479,
      "step": 5093
    },
    {
      "epoch": 0.5094,
      "grad_norm": 2.710293769836426,
      "learning_rate": 0.00011415923644446557,
      "loss": 5.5712,
      "step": 5094
    },
    {
      "epoch": 0.5095,
      "grad_norm": 3.0935232639312744,
      "learning_rate": 0.00011412468067963669,
      "loss": 4.7488,
      "step": 5095
    },
    {
      "epoch": 0.5096,
      "grad_norm": 2.8939826488494873,
      "learning_rate": 0.00011409012319375827,
      "loss": 5.7943,
      "step": 5096
    },
    {
      "epoch": 0.5097,
      "grad_norm": 2.1383652687072754,
      "learning_rate": 0.00011405556399104109,
      "loss": 5.0203,
      "step": 5097
    },
    {
      "epoch": 0.5098,
      "grad_norm": 5.429567813873291,
      "learning_rate": 0.00011402100307569612,
      "loss": 5.5345,
      "step": 5098
    },
    {
      "epoch": 0.5099,
      "grad_norm": 4.2227067947387695,
      "learning_rate": 0.00011398644045193444,
      "loss": 5.1133,
      "step": 5099
    },
    {
      "epoch": 0.51,
      "grad_norm": 6.965172290802002,
      "learning_rate": 0.00011395187612396738,
      "loss": 4.7527,
      "step": 5100
    },
    {
      "epoch": 0.5101,
      "grad_norm": 9.469866752624512,
      "learning_rate": 0.00011391731009600654,
      "loss": 5.0547,
      "step": 5101
    },
    {
      "epoch": 0.5102,
      "grad_norm": 2.3819377422332764,
      "learning_rate": 0.00011388274237226371,
      "loss": 4.8447,
      "step": 5102
    },
    {
      "epoch": 0.5103,
      "grad_norm": 4.953355312347412,
      "learning_rate": 0.00011384817295695083,
      "loss": 5.2031,
      "step": 5103
    },
    {
      "epoch": 0.5104,
      "grad_norm": 2.7308642864227295,
      "learning_rate": 0.00011381360185428007,
      "loss": 4.8326,
      "step": 5104
    },
    {
      "epoch": 0.5105,
      "grad_norm": 3.1627557277679443,
      "learning_rate": 0.0001137790290684638,
      "loss": 4.4114,
      "step": 5105
    },
    {
      "epoch": 0.5106,
      "grad_norm": 3.506584882736206,
      "learning_rate": 0.00011374445460371466,
      "loss": 4.9995,
      "step": 5106
    },
    {
      "epoch": 0.5107,
      "grad_norm": 2.5314619541168213,
      "learning_rate": 0.00011370987846424546,
      "loss": 4.5643,
      "step": 5107
    },
    {
      "epoch": 0.5108,
      "grad_norm": 4.769519805908203,
      "learning_rate": 0.0001136753006542691,
      "loss": 5.3526,
      "step": 5108
    },
    {
      "epoch": 0.5109,
      "grad_norm": 4.850581169128418,
      "learning_rate": 0.00011364072117799885,
      "loss": 5.3574,
      "step": 5109
    },
    {
      "epoch": 0.511,
      "grad_norm": 3.5454142093658447,
      "learning_rate": 0.00011360614003964809,
      "loss": 4.8762,
      "step": 5110
    },
    {
      "epoch": 0.5111,
      "grad_norm": 2.331784725189209,
      "learning_rate": 0.00011357155724343045,
      "loss": 4.6004,
      "step": 5111
    },
    {
      "epoch": 0.5112,
      "grad_norm": 8.631500244140625,
      "learning_rate": 0.00011353697279355973,
      "loss": 5.9196,
      "step": 5112
    },
    {
      "epoch": 0.5113,
      "grad_norm": 3.061845302581787,
      "learning_rate": 0.00011350238669424993,
      "loss": 5.0399,
      "step": 5113
    },
    {
      "epoch": 0.5114,
      "grad_norm": 2.7822988033294678,
      "learning_rate": 0.00011346779894971527,
      "loss": 4.8099,
      "step": 5114
    },
    {
      "epoch": 0.5115,
      "grad_norm": 2.4719231128692627,
      "learning_rate": 0.00011343320956417014,
      "loss": 4.9269,
      "step": 5115
    },
    {
      "epoch": 0.5116,
      "grad_norm": 2.3403687477111816,
      "learning_rate": 0.00011339861854182922,
      "loss": 4.472,
      "step": 5116
    },
    {
      "epoch": 0.5117,
      "grad_norm": 3.9665801525115967,
      "learning_rate": 0.00011336402588690726,
      "loss": 4.9086,
      "step": 5117
    },
    {
      "epoch": 0.5118,
      "grad_norm": 2.955031394958496,
      "learning_rate": 0.00011332943160361926,
      "loss": 4.8094,
      "step": 5118
    },
    {
      "epoch": 0.5119,
      "grad_norm": 2.4145820140838623,
      "learning_rate": 0.00011329483569618045,
      "loss": 5.0626,
      "step": 5119
    },
    {
      "epoch": 0.512,
      "grad_norm": 2.525526762008667,
      "learning_rate": 0.00011326023816880625,
      "loss": 4.8237,
      "step": 5120
    },
    {
      "epoch": 0.5121,
      "grad_norm": 5.732985973358154,
      "learning_rate": 0.00011322563902571226,
      "loss": 4.932,
      "step": 5121
    },
    {
      "epoch": 0.5122,
      "grad_norm": 1.955767273902893,
      "learning_rate": 0.00011319103827111426,
      "loss": 5.1159,
      "step": 5122
    },
    {
      "epoch": 0.5123,
      "grad_norm": 2.2978763580322266,
      "learning_rate": 0.00011315643590922827,
      "loss": 5.3634,
      "step": 5123
    },
    {
      "epoch": 0.5124,
      "grad_norm": 2.104496717453003,
      "learning_rate": 0.00011312183194427046,
      "loss": 4.7807,
      "step": 5124
    },
    {
      "epoch": 0.5125,
      "grad_norm": 2.6799731254577637,
      "learning_rate": 0.00011308722638045724,
      "loss": 5.084,
      "step": 5125
    },
    {
      "epoch": 0.5126,
      "grad_norm": 2.190539598464966,
      "learning_rate": 0.00011305261922200519,
      "loss": 4.2919,
      "step": 5126
    },
    {
      "epoch": 0.5127,
      "grad_norm": 2.4367268085479736,
      "learning_rate": 0.00011301801047313105,
      "loss": 5.4218,
      "step": 5127
    },
    {
      "epoch": 0.5128,
      "grad_norm": 2.610581636428833,
      "learning_rate": 0.00011298340013805184,
      "loss": 5.3071,
      "step": 5128
    },
    {
      "epoch": 0.5129,
      "grad_norm": 2.33589506149292,
      "learning_rate": 0.00011294878822098469,
      "loss": 4.8649,
      "step": 5129
    },
    {
      "epoch": 0.513,
      "grad_norm": 2.166128158569336,
      "learning_rate": 0.000112914174726147,
      "loss": 4.4697,
      "step": 5130
    },
    {
      "epoch": 0.5131,
      "grad_norm": 3.230576276779175,
      "learning_rate": 0.0001128795596577563,
      "loss": 4.65,
      "step": 5131
    },
    {
      "epoch": 0.5132,
      "grad_norm": 3.923985481262207,
      "learning_rate": 0.0001128449430200303,
      "loss": 5.1177,
      "step": 5132
    },
    {
      "epoch": 0.5133,
      "grad_norm": 2.2185051441192627,
      "learning_rate": 0.00011281032481718697,
      "loss": 4.2841,
      "step": 5133
    },
    {
      "epoch": 0.5134,
      "grad_norm": 4.8393073081970215,
      "learning_rate": 0.0001127757050534444,
      "loss": 4.6713,
      "step": 5134
    },
    {
      "epoch": 0.5135,
      "grad_norm": 3.59246826171875,
      "learning_rate": 0.00011274108373302095,
      "loss": 4.7466,
      "step": 5135
    },
    {
      "epoch": 0.5136,
      "grad_norm": 2.418388605117798,
      "learning_rate": 0.00011270646086013505,
      "loss": 4.8543,
      "step": 5136
    },
    {
      "epoch": 0.5137,
      "grad_norm": 3.0980305671691895,
      "learning_rate": 0.00011267183643900548,
      "loss": 4.7553,
      "step": 5137
    },
    {
      "epoch": 0.5138,
      "grad_norm": 2.1017017364501953,
      "learning_rate": 0.00011263721047385105,
      "loss": 4.9682,
      "step": 5138
    },
    {
      "epoch": 0.5139,
      "grad_norm": 2.4388303756713867,
      "learning_rate": 0.00011260258296889086,
      "loss": 5.293,
      "step": 5139
    },
    {
      "epoch": 0.514,
      "grad_norm": 2.353564739227295,
      "learning_rate": 0.00011256795392834419,
      "loss": 4.8763,
      "step": 5140
    },
    {
      "epoch": 0.5141,
      "grad_norm": 6.001526355743408,
      "learning_rate": 0.00011253332335643043,
      "loss": 5.7536,
      "step": 5141
    },
    {
      "epoch": 0.5142,
      "grad_norm": 13.632649421691895,
      "learning_rate": 0.00011249869125736925,
      "loss": 5.4306,
      "step": 5142
    },
    {
      "epoch": 0.5143,
      "grad_norm": 2.635683298110962,
      "learning_rate": 0.00011246405763538046,
      "loss": 4.3209,
      "step": 5143
    },
    {
      "epoch": 0.5144,
      "grad_norm": 2.9843504428863525,
      "learning_rate": 0.00011242942249468402,
      "loss": 5.0912,
      "step": 5144
    },
    {
      "epoch": 0.5145,
      "grad_norm": 2.0680184364318848,
      "learning_rate": 0.00011239478583950018,
      "loss": 5.2907,
      "step": 5145
    },
    {
      "epoch": 0.5146,
      "grad_norm": 17.353801727294922,
      "learning_rate": 0.00011236014767404927,
      "loss": 5.8839,
      "step": 5146
    },
    {
      "epoch": 0.5147,
      "grad_norm": 2.844663143157959,
      "learning_rate": 0.00011232550800255188,
      "loss": 4.3735,
      "step": 5147
    },
    {
      "epoch": 0.5148,
      "grad_norm": 3.786489486694336,
      "learning_rate": 0.00011229086682922869,
      "loss": 4.8814,
      "step": 5148
    },
    {
      "epoch": 0.5149,
      "grad_norm": 6.3488945960998535,
      "learning_rate": 0.00011225622415830068,
      "loss": 5.9874,
      "step": 5149
    },
    {
      "epoch": 0.515,
      "grad_norm": 7.318120956420898,
      "learning_rate": 0.00011222157999398895,
      "loss": 5.1104,
      "step": 5150
    },
    {
      "epoch": 0.5151,
      "grad_norm": 3.6326911449432373,
      "learning_rate": 0.00011218693434051475,
      "loss": 5.0461,
      "step": 5151
    },
    {
      "epoch": 0.5152,
      "grad_norm": 4.679540157318115,
      "learning_rate": 0.00011215228720209958,
      "loss": 5.3151,
      "step": 5152
    },
    {
      "epoch": 0.5153,
      "grad_norm": 5.120943069458008,
      "learning_rate": 0.00011211763858296507,
      "loss": 5.036,
      "step": 5153
    },
    {
      "epoch": 0.5154,
      "grad_norm": 2.9242546558380127,
      "learning_rate": 0.00011208298848733305,
      "loss": 4.8008,
      "step": 5154
    },
    {
      "epoch": 0.5155,
      "grad_norm": 4.279149532318115,
      "learning_rate": 0.00011204833691942553,
      "loss": 5.3089,
      "step": 5155
    },
    {
      "epoch": 0.5156,
      "grad_norm": 2.929394245147705,
      "learning_rate": 0.00011201368388346471,
      "loss": 4.9864,
      "step": 5156
    },
    {
      "epoch": 0.5157,
      "grad_norm": 3.0709099769592285,
      "learning_rate": 0.00011197902938367298,
      "loss": 5.5794,
      "step": 5157
    },
    {
      "epoch": 0.5158,
      "grad_norm": 2.060927629470825,
      "learning_rate": 0.0001119443734242728,
      "loss": 4.9593,
      "step": 5158
    },
    {
      "epoch": 0.5159,
      "grad_norm": 5.14162015914917,
      "learning_rate": 0.00011190971600948699,
      "loss": 6.2288,
      "step": 5159
    },
    {
      "epoch": 0.516,
      "grad_norm": 2.766749143600464,
      "learning_rate": 0.0001118750571435384,
      "loss": 4.7315,
      "step": 5160
    },
    {
      "epoch": 0.5161,
      "grad_norm": 7.801217079162598,
      "learning_rate": 0.00011184039683065013,
      "loss": 4.908,
      "step": 5161
    },
    {
      "epoch": 0.5162,
      "grad_norm": 9.24281120300293,
      "learning_rate": 0.00011180573507504537,
      "loss": 5.4281,
      "step": 5162
    },
    {
      "epoch": 0.5163,
      "grad_norm": 4.800714492797852,
      "learning_rate": 0.00011177107188094764,
      "loss": 4.8329,
      "step": 5163
    },
    {
      "epoch": 0.5164,
      "grad_norm": 5.13341760635376,
      "learning_rate": 0.00011173640725258052,
      "loss": 4.4661,
      "step": 5164
    },
    {
      "epoch": 0.5165,
      "grad_norm": 2.548807382583618,
      "learning_rate": 0.00011170174119416776,
      "loss": 5.0583,
      "step": 5165
    },
    {
      "epoch": 0.5166,
      "grad_norm": 8.550806045532227,
      "learning_rate": 0.00011166707370993333,
      "loss": 4.8557,
      "step": 5166
    },
    {
      "epoch": 0.5167,
      "grad_norm": 2.3448147773742676,
      "learning_rate": 0.00011163240480410135,
      "loss": 5.0377,
      "step": 5167
    },
    {
      "epoch": 0.5168,
      "grad_norm": 4.405214786529541,
      "learning_rate": 0.00011159773448089614,
      "loss": 4.622,
      "step": 5168
    },
    {
      "epoch": 0.5169,
      "grad_norm": 3.5411484241485596,
      "learning_rate": 0.00011156306274454218,
      "loss": 5.3303,
      "step": 5169
    },
    {
      "epoch": 0.517,
      "grad_norm": 13.245482444763184,
      "learning_rate": 0.00011152838959926408,
      "loss": 5.4786,
      "step": 5170
    },
    {
      "epoch": 0.5171,
      "grad_norm": 3.105025053024292,
      "learning_rate": 0.00011149371504928668,
      "loss": 4.7528,
      "step": 5171
    },
    {
      "epoch": 0.5172,
      "grad_norm": 2.910844087600708,
      "learning_rate": 0.00011145903909883495,
      "loss": 5.1451,
      "step": 5172
    },
    {
      "epoch": 0.5173,
      "grad_norm": 3.8131513595581055,
      "learning_rate": 0.00011142436175213409,
      "loss": 5.5096,
      "step": 5173
    },
    {
      "epoch": 0.5174,
      "grad_norm": 4.742544174194336,
      "learning_rate": 0.0001113896830134094,
      "loss": 5.1273,
      "step": 5174
    },
    {
      "epoch": 0.5175,
      "grad_norm": 3.7574384212493896,
      "learning_rate": 0.00011135500288688636,
      "loss": 4.862,
      "step": 5175
    },
    {
      "epoch": 0.5176,
      "grad_norm": 2.866393566131592,
      "learning_rate": 0.0001113203213767907,
      "loss": 5.4729,
      "step": 5176
    },
    {
      "epoch": 0.5177,
      "grad_norm": 2.8581764698028564,
      "learning_rate": 0.00011128563848734816,
      "loss": 4.9712,
      "step": 5177
    },
    {
      "epoch": 0.5178,
      "grad_norm": 4.252046585083008,
      "learning_rate": 0.00011125095422278486,
      "loss": 5.1161,
      "step": 5178
    },
    {
      "epoch": 0.5179,
      "grad_norm": 4.335312843322754,
      "learning_rate": 0.0001112162685873269,
      "loss": 5.1421,
      "step": 5179
    },
    {
      "epoch": 0.518,
      "grad_norm": 3.961906909942627,
      "learning_rate": 0.00011118158158520064,
      "loss": 4.9415,
      "step": 5180
    },
    {
      "epoch": 0.5181,
      "grad_norm": 2.686519145965576,
      "learning_rate": 0.00011114689322063255,
      "loss": 4.3925,
      "step": 5181
    },
    {
      "epoch": 0.5182,
      "grad_norm": 2.5249457359313965,
      "learning_rate": 0.00011111220349784937,
      "loss": 4.6997,
      "step": 5182
    },
    {
      "epoch": 0.5183,
      "grad_norm": 2.6636061668395996,
      "learning_rate": 0.00011107751242107787,
      "loss": 4.8635,
      "step": 5183
    },
    {
      "epoch": 0.5184,
      "grad_norm": 2.5308427810668945,
      "learning_rate": 0.00011104281999454511,
      "loss": 5.1123,
      "step": 5184
    },
    {
      "epoch": 0.5185,
      "grad_norm": 2.4471559524536133,
      "learning_rate": 0.00011100812622247822,
      "loss": 5.7052,
      "step": 5185
    },
    {
      "epoch": 0.5186,
      "grad_norm": 4.256524085998535,
      "learning_rate": 0.00011097343110910452,
      "loss": 4.9875,
      "step": 5186
    },
    {
      "epoch": 0.5187,
      "grad_norm": 3.612743616104126,
      "learning_rate": 0.00011093873465865157,
      "loss": 4.4879,
      "step": 5187
    },
    {
      "epoch": 0.5188,
      "grad_norm": 2.2443206310272217,
      "learning_rate": 0.00011090403687534697,
      "loss": 4.7463,
      "step": 5188
    },
    {
      "epoch": 0.5189,
      "grad_norm": 3.660733699798584,
      "learning_rate": 0.00011086933776341852,
      "loss": 4.3671,
      "step": 5189
    },
    {
      "epoch": 0.519,
      "grad_norm": 3.882718324661255,
      "learning_rate": 0.00011083463732709425,
      "loss": 5.458,
      "step": 5190
    },
    {
      "epoch": 0.5191,
      "grad_norm": 3.01058030128479,
      "learning_rate": 0.0001107999355706023,
      "loss": 4.7954,
      "step": 5191
    },
    {
      "epoch": 0.5192,
      "grad_norm": 9.388501167297363,
      "learning_rate": 0.00011076523249817094,
      "loss": 6.7379,
      "step": 5192
    },
    {
      "epoch": 0.5193,
      "grad_norm": 5.326094150543213,
      "learning_rate": 0.00011073052811402867,
      "loss": 6.3145,
      "step": 5193
    },
    {
      "epoch": 0.5194,
      "grad_norm": 5.771568775177002,
      "learning_rate": 0.0001106958224224041,
      "loss": 4.7376,
      "step": 5194
    },
    {
      "epoch": 0.5195,
      "grad_norm": 2.3775172233581543,
      "learning_rate": 0.000110661115427526,
      "loss": 4.4904,
      "step": 5195
    },
    {
      "epoch": 0.5196,
      "grad_norm": 9.373706817626953,
      "learning_rate": 0.00011062640713362333,
      "loss": 5.0486,
      "step": 5196
    },
    {
      "epoch": 0.5197,
      "grad_norm": 2.3305540084838867,
      "learning_rate": 0.0001105916975449252,
      "loss": 5.0308,
      "step": 5197
    },
    {
      "epoch": 0.5198,
      "grad_norm": 4.556834697723389,
      "learning_rate": 0.00011055698666566084,
      "loss": 5.3313,
      "step": 5198
    },
    {
      "epoch": 0.5199,
      "grad_norm": 3.6782267093658447,
      "learning_rate": 0.00011052227450005967,
      "loss": 5.1161,
      "step": 5199
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.36562180519104,
      "learning_rate": 0.00011048756105235125,
      "loss": 4.7798,
      "step": 5200
    },
    {
      "epoch": 0.5201,
      "grad_norm": 3.205207347869873,
      "learning_rate": 0.00011045284632676536,
      "loss": 4.9447,
      "step": 5201
    },
    {
      "epoch": 0.5202,
      "grad_norm": 5.352639675140381,
      "learning_rate": 0.00011041813032753183,
      "loss": 5.4349,
      "step": 5202
    },
    {
      "epoch": 0.5203,
      "grad_norm": 3.960451126098633,
      "learning_rate": 0.00011038341305888074,
      "loss": 4.8174,
      "step": 5203
    },
    {
      "epoch": 0.5204,
      "grad_norm": 1.9321945905685425,
      "learning_rate": 0.00011034869452504226,
      "loss": 4.6604,
      "step": 5204
    },
    {
      "epoch": 0.5205,
      "grad_norm": 5.208124160766602,
      "learning_rate": 0.00011031397473024674,
      "loss": 5.1862,
      "step": 5205
    },
    {
      "epoch": 0.5206,
      "grad_norm": 2.7555127143859863,
      "learning_rate": 0.00011027925367872469,
      "loss": 5.4593,
      "step": 5206
    },
    {
      "epoch": 0.5207,
      "grad_norm": 2.9215240478515625,
      "learning_rate": 0.00011024453137470677,
      "loss": 4.4833,
      "step": 5207
    },
    {
      "epoch": 0.5208,
      "grad_norm": 2.992326498031616,
      "learning_rate": 0.00011020980782242376,
      "loss": 5.0906,
      "step": 5208
    },
    {
      "epoch": 0.5209,
      "grad_norm": 2.683954954147339,
      "learning_rate": 0.00011017508302610664,
      "loss": 5.0799,
      "step": 5209
    },
    {
      "epoch": 0.521,
      "grad_norm": 16.802440643310547,
      "learning_rate": 0.00011014035698998651,
      "loss": 5.3552,
      "step": 5210
    },
    {
      "epoch": 0.5211,
      "grad_norm": 3.5417420864105225,
      "learning_rate": 0.00011010562971829463,
      "loss": 4.7894,
      "step": 5211
    },
    {
      "epoch": 0.5212,
      "grad_norm": 4.126340389251709,
      "learning_rate": 0.00011007090121526245,
      "loss": 5.3308,
      "step": 5212
    },
    {
      "epoch": 0.5213,
      "grad_norm": 7.821913719177246,
      "learning_rate": 0.00011003617148512149,
      "loss": 5.4283,
      "step": 5213
    },
    {
      "epoch": 0.5214,
      "grad_norm": 4.4477033615112305,
      "learning_rate": 0.00011000144053210348,
      "loss": 4.7425,
      "step": 5214
    },
    {
      "epoch": 0.5215,
      "grad_norm": 2.365215301513672,
      "learning_rate": 0.0001099667083604403,
      "loss": 5.0364,
      "step": 5215
    },
    {
      "epoch": 0.5216,
      "grad_norm": 3.882765054702759,
      "learning_rate": 0.00010993197497436391,
      "loss": 5.1196,
      "step": 5216
    },
    {
      "epoch": 0.5217,
      "grad_norm": 3.2980494499206543,
      "learning_rate": 0.00010989724037810652,
      "loss": 5.2754,
      "step": 5217
    },
    {
      "epoch": 0.5218,
      "grad_norm": 2.315361499786377,
      "learning_rate": 0.00010986250457590039,
      "loss": 4.8782,
      "step": 5218
    },
    {
      "epoch": 0.5219,
      "grad_norm": 3.4980499744415283,
      "learning_rate": 0.00010982776757197799,
      "loss": 5.1034,
      "step": 5219
    },
    {
      "epoch": 0.522,
      "grad_norm": 4.10078763961792,
      "learning_rate": 0.00010979302937057192,
      "loss": 4.9711,
      "step": 5220
    },
    {
      "epoch": 0.5221,
      "grad_norm": 4.518370628356934,
      "learning_rate": 0.00010975828997591495,
      "loss": 5.0493,
      "step": 5221
    },
    {
      "epoch": 0.5222,
      "grad_norm": 3.3651623725891113,
      "learning_rate": 0.00010972354939223996,
      "loss": 5.1067,
      "step": 5222
    },
    {
      "epoch": 0.5223,
      "grad_norm": 2.0387847423553467,
      "learning_rate": 0.00010968880762377993,
      "loss": 4.4981,
      "step": 5223
    },
    {
      "epoch": 0.5224,
      "grad_norm": 2.2132973670959473,
      "learning_rate": 0.00010965406467476808,
      "loss": 5.4583,
      "step": 5224
    },
    {
      "epoch": 0.5225,
      "grad_norm": 11.397340774536133,
      "learning_rate": 0.00010961932054943778,
      "loss": 4.9116,
      "step": 5225
    },
    {
      "epoch": 0.5226,
      "grad_norm": 4.2591986656188965,
      "learning_rate": 0.00010958457525202241,
      "loss": 4.9875,
      "step": 5226
    },
    {
      "epoch": 0.5227,
      "grad_norm": 5.338444232940674,
      "learning_rate": 0.00010954982878675563,
      "loss": 5.4443,
      "step": 5227
    },
    {
      "epoch": 0.5228,
      "grad_norm": 3.479802131652832,
      "learning_rate": 0.00010951508115787119,
      "loss": 4.5677,
      "step": 5228
    },
    {
      "epoch": 0.5229,
      "grad_norm": 2.2432374954223633,
      "learning_rate": 0.00010948033236960294,
      "loss": 4.5473,
      "step": 5229
    },
    {
      "epoch": 0.523,
      "grad_norm": 2.5992190837860107,
      "learning_rate": 0.00010944558242618496,
      "loss": 5.0629,
      "step": 5230
    },
    {
      "epoch": 0.5231,
      "grad_norm": 2.7180707454681396,
      "learning_rate": 0.00010941083133185146,
      "loss": 5.2715,
      "step": 5231
    },
    {
      "epoch": 0.5232,
      "grad_norm": 2.3750813007354736,
      "learning_rate": 0.00010937607909083667,
      "loss": 4.9841,
      "step": 5232
    },
    {
      "epoch": 0.5233,
      "grad_norm": 2.0752251148223877,
      "learning_rate": 0.00010934132570737507,
      "loss": 5.3232,
      "step": 5233
    },
    {
      "epoch": 0.5234,
      "grad_norm": 3.7973008155822754,
      "learning_rate": 0.00010930657118570126,
      "loss": 5.3004,
      "step": 5234
    },
    {
      "epoch": 0.5235,
      "grad_norm": 2.496412754058838,
      "learning_rate": 0.00010927181553005002,
      "loss": 4.9344,
      "step": 5235
    },
    {
      "epoch": 0.5236,
      "grad_norm": 4.662930965423584,
      "learning_rate": 0.00010923705874465618,
      "loss": 5.2871,
      "step": 5236
    },
    {
      "epoch": 0.5237,
      "grad_norm": 2.089168071746826,
      "learning_rate": 0.00010920230083375473,
      "loss": 4.8861,
      "step": 5237
    },
    {
      "epoch": 0.5238,
      "grad_norm": 3.0770039558410645,
      "learning_rate": 0.00010916754180158082,
      "loss": 5.8151,
      "step": 5238
    },
    {
      "epoch": 0.5239,
      "grad_norm": 9.532428741455078,
      "learning_rate": 0.00010913278165236978,
      "loss": 7.2424,
      "step": 5239
    },
    {
      "epoch": 0.524,
      "grad_norm": 2.325252056121826,
      "learning_rate": 0.00010909802039035701,
      "loss": 5.4207,
      "step": 5240
    },
    {
      "epoch": 0.5241,
      "grad_norm": 2.961077928543091,
      "learning_rate": 0.00010906325801977804,
      "loss": 5.1619,
      "step": 5241
    },
    {
      "epoch": 0.5242,
      "grad_norm": 3.665371894836426,
      "learning_rate": 0.00010902849454486856,
      "loss": 4.824,
      "step": 5242
    },
    {
      "epoch": 0.5243,
      "grad_norm": 3.425252676010132,
      "learning_rate": 0.00010899372996986439,
      "loss": 5.2703,
      "step": 5243
    },
    {
      "epoch": 0.5244,
      "grad_norm": 3.093675374984741,
      "learning_rate": 0.00010895896429900154,
      "loss": 4.9961,
      "step": 5244
    },
    {
      "epoch": 0.5245,
      "grad_norm": 6.924033164978027,
      "learning_rate": 0.00010892419753651606,
      "loss": 4.7782,
      "step": 5245
    },
    {
      "epoch": 0.5246,
      "grad_norm": 2.7353272438049316,
      "learning_rate": 0.00010888942968664417,
      "loss": 4.8727,
      "step": 5246
    },
    {
      "epoch": 0.5247,
      "grad_norm": 4.507065296173096,
      "learning_rate": 0.00010885466075362223,
      "loss": 4.978,
      "step": 5247
    },
    {
      "epoch": 0.5248,
      "grad_norm": 5.201504707336426,
      "learning_rate": 0.00010881989074168673,
      "loss": 4.8287,
      "step": 5248
    },
    {
      "epoch": 0.5249,
      "grad_norm": 2.1956634521484375,
      "learning_rate": 0.00010878511965507434,
      "loss": 5.356,
      "step": 5249
    },
    {
      "epoch": 0.525,
      "grad_norm": 2.8380465507507324,
      "learning_rate": 0.00010875034749802173,
      "loss": 4.5341,
      "step": 5250
    },
    {
      "epoch": 0.5251,
      "grad_norm": 3.208770751953125,
      "learning_rate": 0.00010871557427476583,
      "loss": 4.7789,
      "step": 5251
    },
    {
      "epoch": 0.5252,
      "grad_norm": 3.5865390300750732,
      "learning_rate": 0.00010868079998954364,
      "loss": 5.2376,
      "step": 5252
    },
    {
      "epoch": 0.5253,
      "grad_norm": 2.737034797668457,
      "learning_rate": 0.0001086460246465923,
      "loss": 5.5534,
      "step": 5253
    },
    {
      "epoch": 0.5254,
      "grad_norm": 3.521530866622925,
      "learning_rate": 0.00010861124825014908,
      "loss": 4.6061,
      "step": 5254
    },
    {
      "epoch": 0.5255,
      "grad_norm": 2.342454195022583,
      "learning_rate": 0.00010857647080445139,
      "loss": 4.6091,
      "step": 5255
    },
    {
      "epoch": 0.5256,
      "grad_norm": 3.327662944793701,
      "learning_rate": 0.00010854169231373676,
      "loss": 4.3818,
      "step": 5256
    },
    {
      "epoch": 0.5257,
      "grad_norm": 4.06915807723999,
      "learning_rate": 0.00010850691278224281,
      "loss": 5.0765,
      "step": 5257
    },
    {
      "epoch": 0.5258,
      "grad_norm": 1.9118151664733887,
      "learning_rate": 0.00010847213221420736,
      "loss": 4.6702,
      "step": 5258
    },
    {
      "epoch": 0.5259,
      "grad_norm": 2.0120692253112793,
      "learning_rate": 0.00010843735061386828,
      "loss": 4.4354,
      "step": 5259
    },
    {
      "epoch": 0.526,
      "grad_norm": 2.598719358444214,
      "learning_rate": 0.00010840256798546364,
      "loss": 4.8463,
      "step": 5260
    },
    {
      "epoch": 0.5261,
      "grad_norm": 2.235013484954834,
      "learning_rate": 0.00010836778433323158,
      "loss": 4.7094,
      "step": 5261
    },
    {
      "epoch": 0.5262,
      "grad_norm": 2.29376482963562,
      "learning_rate": 0.00010833299966141035,
      "loss": 5.1115,
      "step": 5262
    },
    {
      "epoch": 0.5263,
      "grad_norm": 2.389396905899048,
      "learning_rate": 0.0001082982139742384,
      "loss": 4.8539,
      "step": 5263
    },
    {
      "epoch": 0.5264,
      "grad_norm": 2.033965826034546,
      "learning_rate": 0.00010826342727595426,
      "loss": 5.1432,
      "step": 5264
    },
    {
      "epoch": 0.5265,
      "grad_norm": 2.612853765487671,
      "learning_rate": 0.00010822863957079656,
      "loss": 4.9443,
      "step": 5265
    },
    {
      "epoch": 0.5266,
      "grad_norm": 2.897850751876831,
      "learning_rate": 0.0001081938508630041,
      "loss": 5.4167,
      "step": 5266
    },
    {
      "epoch": 0.5267,
      "grad_norm": 2.851834774017334,
      "learning_rate": 0.00010815906115681578,
      "loss": 4.9398,
      "step": 5267
    },
    {
      "epoch": 0.5268,
      "grad_norm": 2.399059295654297,
      "learning_rate": 0.00010812427045647058,
      "loss": 4.7491,
      "step": 5268
    },
    {
      "epoch": 0.5269,
      "grad_norm": 4.4248366355896,
      "learning_rate": 0.00010808947876620767,
      "loss": 5.1171,
      "step": 5269
    },
    {
      "epoch": 0.527,
      "grad_norm": 2.608915328979492,
      "learning_rate": 0.00010805468609026632,
      "loss": 4.2554,
      "step": 5270
    },
    {
      "epoch": 0.5271,
      "grad_norm": 4.992733001708984,
      "learning_rate": 0.00010801989243288589,
      "loss": 4.9422,
      "step": 5271
    },
    {
      "epoch": 0.5272,
      "grad_norm": 2.6402902603149414,
      "learning_rate": 0.0001079850977983059,
      "loss": 4.6034,
      "step": 5272
    },
    {
      "epoch": 0.5273,
      "grad_norm": 5.626682758331299,
      "learning_rate": 0.00010795030219076599,
      "loss": 5.77,
      "step": 5273
    },
    {
      "epoch": 0.5274,
      "grad_norm": 2.434175729751587,
      "learning_rate": 0.00010791550561450586,
      "loss": 4.8421,
      "step": 5274
    },
    {
      "epoch": 0.5275,
      "grad_norm": 5.854641437530518,
      "learning_rate": 0.00010788070807376536,
      "loss": 4.7841,
      "step": 5275
    },
    {
      "epoch": 0.5276,
      "grad_norm": 2.720086097717285,
      "learning_rate": 0.0001078459095727845,
      "loss": 4.9462,
      "step": 5276
    },
    {
      "epoch": 0.5277,
      "grad_norm": 5.835578918457031,
      "learning_rate": 0.00010781111011580336,
      "loss": 4.2333,
      "step": 5277
    },
    {
      "epoch": 0.5278,
      "grad_norm": 13.477553367614746,
      "learning_rate": 0.00010777630970706217,
      "loss": 4.9594,
      "step": 5278
    },
    {
      "epoch": 0.5279,
      "grad_norm": 4.650624752044678,
      "learning_rate": 0.00010774150835080119,
      "loss": 4.7016,
      "step": 5279
    },
    {
      "epoch": 0.528,
      "grad_norm": 2.8810157775878906,
      "learning_rate": 0.00010770670605126092,
      "loss": 4.7481,
      "step": 5280
    },
    {
      "epoch": 0.5281,
      "grad_norm": 1.8442530632019043,
      "learning_rate": 0.00010767190281268187,
      "loss": 4.7113,
      "step": 5281
    },
    {
      "epoch": 0.5282,
      "grad_norm": 3.224052906036377,
      "learning_rate": 0.00010763709863930476,
      "loss": 4.8739,
      "step": 5282
    },
    {
      "epoch": 0.5283,
      "grad_norm": 3.030360221862793,
      "learning_rate": 0.00010760229353537033,
      "loss": 4.6108,
      "step": 5283
    },
    {
      "epoch": 0.5284,
      "grad_norm": 2.4443018436431885,
      "learning_rate": 0.00010756748750511953,
      "loss": 4.7173,
      "step": 5284
    },
    {
      "epoch": 0.5285,
      "grad_norm": 3.8157622814178467,
      "learning_rate": 0.00010753268055279329,
      "loss": 5.4791,
      "step": 5285
    },
    {
      "epoch": 0.5286,
      "grad_norm": 4.073340892791748,
      "learning_rate": 0.00010749787268263279,
      "loss": 5.2069,
      "step": 5286
    },
    {
      "epoch": 0.5287,
      "grad_norm": 5.230839729309082,
      "learning_rate": 0.00010746306389887924,
      "loss": 5.0986,
      "step": 5287
    },
    {
      "epoch": 0.5288,
      "grad_norm": 3.3232710361480713,
      "learning_rate": 0.000107428254205774,
      "loss": 4.7829,
      "step": 5288
    },
    {
      "epoch": 0.5289,
      "grad_norm": 3.8769242763519287,
      "learning_rate": 0.00010739344360755852,
      "loss": 5.4127,
      "step": 5289
    },
    {
      "epoch": 0.529,
      "grad_norm": 1.9333198070526123,
      "learning_rate": 0.00010735863210847433,
      "loss": 4.9048,
      "step": 5290
    },
    {
      "epoch": 0.5291,
      "grad_norm": 2.913815498352051,
      "learning_rate": 0.00010732381971276318,
      "loss": 4.8862,
      "step": 5291
    },
    {
      "epoch": 0.5292,
      "grad_norm": 6.353015422821045,
      "learning_rate": 0.0001072890064246668,
      "loss": 5.583,
      "step": 5292
    },
    {
      "epoch": 0.5293,
      "grad_norm": 2.8387093544006348,
      "learning_rate": 0.0001072541922484271,
      "loss": 4.8919,
      "step": 5293
    },
    {
      "epoch": 0.5294,
      "grad_norm": 3.8525240421295166,
      "learning_rate": 0.0001072193771882861,
      "loss": 5.4929,
      "step": 5294
    },
    {
      "epoch": 0.5295,
      "grad_norm": 2.8215060234069824,
      "learning_rate": 0.00010718456124848583,
      "loss": 5.0597,
      "step": 5295
    },
    {
      "epoch": 0.5296,
      "grad_norm": 2.444279432296753,
      "learning_rate": 0.0001071497444332686,
      "loss": 4.4679,
      "step": 5296
    },
    {
      "epoch": 0.5297,
      "grad_norm": 3.204582452774048,
      "learning_rate": 0.00010711492674687671,
      "loss": 4.8579,
      "step": 5297
    },
    {
      "epoch": 0.5298,
      "grad_norm": 2.2024831771850586,
      "learning_rate": 0.00010708010819355256,
      "loss": 4.7449,
      "step": 5298
    },
    {
      "epoch": 0.5299,
      "grad_norm": 10.240447998046875,
      "learning_rate": 0.0001070452887775387,
      "loss": 6.915,
      "step": 5299
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.0075409412384033,
      "learning_rate": 0.00010701046850307777,
      "loss": 4.8887,
      "step": 5300
    },
    {
      "epoch": 0.5301,
      "grad_norm": 2.940943956375122,
      "learning_rate": 0.00010697564737441252,
      "loss": 4.8239,
      "step": 5301
    },
    {
      "epoch": 0.5302,
      "grad_norm": 7.015295505523682,
      "learning_rate": 0.00010694082539578585,
      "loss": 4.754,
      "step": 5302
    },
    {
      "epoch": 0.5303,
      "grad_norm": 2.5157277584075928,
      "learning_rate": 0.00010690600257144061,
      "loss": 4.5702,
      "step": 5303
    },
    {
      "epoch": 0.5304,
      "grad_norm": 6.074969291687012,
      "learning_rate": 0.00010687117890561988,
      "loss": 5.3686,
      "step": 5304
    },
    {
      "epoch": 0.5305,
      "grad_norm": 2.7708399295806885,
      "learning_rate": 0.00010683635440256687,
      "loss": 5.1589,
      "step": 5305
    },
    {
      "epoch": 0.5306,
      "grad_norm": 4.6926116943359375,
      "learning_rate": 0.00010680152906652483,
      "loss": 4.8803,
      "step": 5306
    },
    {
      "epoch": 0.5307,
      "grad_norm": 2.399260997772217,
      "learning_rate": 0.00010676670290173709,
      "loss": 5.0068,
      "step": 5307
    },
    {
      "epoch": 0.5308,
      "grad_norm": 5.023684501647949,
      "learning_rate": 0.00010673187591244714,
      "loss": 5.1162,
      "step": 5308
    },
    {
      "epoch": 0.5309,
      "grad_norm": 7.70731258392334,
      "learning_rate": 0.00010669704810289851,
      "loss": 5.0553,
      "step": 5309
    },
    {
      "epoch": 0.531,
      "grad_norm": 5.805966377258301,
      "learning_rate": 0.00010666221947733486,
      "loss": 5.0691,
      "step": 5310
    },
    {
      "epoch": 0.5311,
      "grad_norm": 6.960228443145752,
      "learning_rate": 0.00010662739004000005,
      "loss": 5.1387,
      "step": 5311
    },
    {
      "epoch": 0.5312,
      "grad_norm": 2.6895415782928467,
      "learning_rate": 0.0001065925597951378,
      "loss": 4.7855,
      "step": 5312
    },
    {
      "epoch": 0.5313,
      "grad_norm": 5.557896614074707,
      "learning_rate": 0.00010655772874699217,
      "loss": 5.3666,
      "step": 5313
    },
    {
      "epoch": 0.5314,
      "grad_norm": 4.848421096801758,
      "learning_rate": 0.00010652289689980714,
      "loss": 5.0824,
      "step": 5314
    },
    {
      "epoch": 0.5315,
      "grad_norm": 9.797986030578613,
      "learning_rate": 0.00010648806425782695,
      "loss": 5.5657,
      "step": 5315
    },
    {
      "epoch": 0.5316,
      "grad_norm": 3.763331651687622,
      "learning_rate": 0.00010645323082529581,
      "loss": 5.3141,
      "step": 5316
    },
    {
      "epoch": 0.5317,
      "grad_norm": 2.2141101360321045,
      "learning_rate": 0.00010641839660645805,
      "loss": 5.0285,
      "step": 5317
    },
    {
      "epoch": 0.5318,
      "grad_norm": 2.9871461391448975,
      "learning_rate": 0.00010638356160555816,
      "loss": 4.455,
      "step": 5318
    },
    {
      "epoch": 0.5319,
      "grad_norm": 4.904662132263184,
      "learning_rate": 0.00010634872582684061,
      "loss": 5.0945,
      "step": 5319
    },
    {
      "epoch": 0.532,
      "grad_norm": 3.8617589473724365,
      "learning_rate": 0.00010631388927455013,
      "loss": 5.4873,
      "step": 5320
    },
    {
      "epoch": 0.5321,
      "grad_norm": 3.2280662059783936,
      "learning_rate": 0.00010627905195293135,
      "loss": 5.4887,
      "step": 5321
    },
    {
      "epoch": 0.5322,
      "grad_norm": 3.3078908920288086,
      "learning_rate": 0.00010624421386622916,
      "loss": 4.6311,
      "step": 5322
    },
    {
      "epoch": 0.5323,
      "grad_norm": 3.364835739135742,
      "learning_rate": 0.00010620937501868841,
      "loss": 5.56,
      "step": 5323
    },
    {
      "epoch": 0.5324,
      "grad_norm": 6.377293586730957,
      "learning_rate": 0.0001061745354145542,
      "loss": 4.9897,
      "step": 5324
    },
    {
      "epoch": 0.5325,
      "grad_norm": 2.487792730331421,
      "learning_rate": 0.00010613969505807156,
      "loss": 4.5717,
      "step": 5325
    },
    {
      "epoch": 0.5326,
      "grad_norm": 2.0318827629089355,
      "learning_rate": 0.00010610485395348571,
      "loss": 4.5748,
      "step": 5326
    },
    {
      "epoch": 0.5327,
      "grad_norm": 2.266846179962158,
      "learning_rate": 0.00010607001210504191,
      "loss": 4.5874,
      "step": 5327
    },
    {
      "epoch": 0.5328,
      "grad_norm": 3.207592487335205,
      "learning_rate": 0.00010603516951698556,
      "loss": 5.2913,
      "step": 5328
    },
    {
      "epoch": 0.5329,
      "grad_norm": 4.453176498413086,
      "learning_rate": 0.00010600032619356209,
      "loss": 5.2777,
      "step": 5329
    },
    {
      "epoch": 0.533,
      "grad_norm": 4.013986110687256,
      "learning_rate": 0.00010596548213901708,
      "loss": 5.6042,
      "step": 5330
    },
    {
      "epoch": 0.5331,
      "grad_norm": 7.149019718170166,
      "learning_rate": 0.00010593063735759618,
      "loss": 5.1964,
      "step": 5331
    },
    {
      "epoch": 0.5332,
      "grad_norm": 2.816434621810913,
      "learning_rate": 0.0001058957918535451,
      "loss": 5.0504,
      "step": 5332
    },
    {
      "epoch": 0.5333,
      "grad_norm": 4.311955451965332,
      "learning_rate": 0.00010586094563110964,
      "loss": 5.1636,
      "step": 5333
    },
    {
      "epoch": 0.5334,
      "grad_norm": 2.5574333667755127,
      "learning_rate": 0.00010582609869453577,
      "loss": 5.3943,
      "step": 5334
    },
    {
      "epoch": 0.5335,
      "grad_norm": 2.3613429069519043,
      "learning_rate": 0.00010579125104806944,
      "loss": 4.6226,
      "step": 5335
    },
    {
      "epoch": 0.5336,
      "grad_norm": 4.957475662231445,
      "learning_rate": 0.00010575640269595675,
      "loss": 4.5357,
      "step": 5336
    },
    {
      "epoch": 0.5337,
      "grad_norm": 3.952322483062744,
      "learning_rate": 0.00010572155364244382,
      "loss": 4.8879,
      "step": 5337
    },
    {
      "epoch": 0.5338,
      "grad_norm": 2.7104058265686035,
      "learning_rate": 0.00010568670389177696,
      "loss": 5.2095,
      "step": 5338
    },
    {
      "epoch": 0.5339,
      "grad_norm": 4.68414831161499,
      "learning_rate": 0.00010565185344820247,
      "loss": 4.7942,
      "step": 5339
    },
    {
      "epoch": 0.534,
      "grad_norm": 7.130825996398926,
      "learning_rate": 0.00010561700231596678,
      "loss": 5.4238,
      "step": 5340
    },
    {
      "epoch": 0.5341,
      "grad_norm": 4.578100204467773,
      "learning_rate": 0.00010558215049931638,
      "loss": 5.2043,
      "step": 5341
    },
    {
      "epoch": 0.5342,
      "grad_norm": 3.7825305461883545,
      "learning_rate": 0.00010554729800249792,
      "loss": 4.945,
      "step": 5342
    },
    {
      "epoch": 0.5343,
      "grad_norm": 5.477085113525391,
      "learning_rate": 0.00010551244482975798,
      "loss": 5.1729,
      "step": 5343
    },
    {
      "epoch": 0.5344,
      "grad_norm": 3.010256767272949,
      "learning_rate": 0.00010547759098534335,
      "loss": 4.6934,
      "step": 5344
    },
    {
      "epoch": 0.5345,
      "grad_norm": 4.0623860359191895,
      "learning_rate": 0.00010544273647350092,
      "loss": 5.1096,
      "step": 5345
    },
    {
      "epoch": 0.5346,
      "grad_norm": 2.06535005569458,
      "learning_rate": 0.00010540788129847756,
      "loss": 4.5993,
      "step": 5346
    },
    {
      "epoch": 0.5347,
      "grad_norm": 2.6360714435577393,
      "learning_rate": 0.00010537302546452022,
      "loss": 5.2173,
      "step": 5347
    },
    {
      "epoch": 0.5348,
      "grad_norm": 3.7533345222473145,
      "learning_rate": 0.00010533816897587606,
      "loss": 5.0702,
      "step": 5348
    },
    {
      "epoch": 0.5349,
      "grad_norm": 4.485620498657227,
      "learning_rate": 0.00010530331183679218,
      "loss": 4.9692,
      "step": 5349
    },
    {
      "epoch": 0.535,
      "grad_norm": 3.438960313796997,
      "learning_rate": 0.00010526845405151586,
      "loss": 5.1874,
      "step": 5350
    },
    {
      "epoch": 0.5351,
      "grad_norm": 2.1453323364257812,
      "learning_rate": 0.0001052335956242944,
      "loss": 5.0817,
      "step": 5351
    },
    {
      "epoch": 0.5352,
      "grad_norm": 3.880995750427246,
      "learning_rate": 0.00010519873655937516,
      "loss": 4.818,
      "step": 5352
    },
    {
      "epoch": 0.5353,
      "grad_norm": 3.106341600418091,
      "learning_rate": 0.00010516387686100566,
      "loss": 4.8444,
      "step": 5353
    },
    {
      "epoch": 0.5354,
      "grad_norm": 3.3331570625305176,
      "learning_rate": 0.00010512901653343344,
      "loss": 5.3154,
      "step": 5354
    },
    {
      "epoch": 0.5355,
      "grad_norm": 2.195869207382202,
      "learning_rate": 0.00010509415558090609,
      "loss": 4.6596,
      "step": 5355
    },
    {
      "epoch": 0.5356,
      "grad_norm": 2.361971378326416,
      "learning_rate": 0.00010505929400767134,
      "loss": 5.1756,
      "step": 5356
    },
    {
      "epoch": 0.5357,
      "grad_norm": 9.926926612854004,
      "learning_rate": 0.00010502443181797697,
      "loss": 5.8473,
      "step": 5357
    },
    {
      "epoch": 0.5358,
      "grad_norm": 3.9198083877563477,
      "learning_rate": 0.00010498956901607083,
      "loss": 4.8432,
      "step": 5358
    },
    {
      "epoch": 0.5359,
      "grad_norm": 5.041543483734131,
      "learning_rate": 0.00010495470560620083,
      "loss": 4.5741,
      "step": 5359
    },
    {
      "epoch": 0.536,
      "grad_norm": 5.302394390106201,
      "learning_rate": 0.00010491984159261496,
      "loss": 4.7426,
      "step": 5360
    },
    {
      "epoch": 0.5361,
      "grad_norm": 2.2588648796081543,
      "learning_rate": 0.00010488497697956135,
      "loss": 4.6209,
      "step": 5361
    },
    {
      "epoch": 0.5362,
      "grad_norm": 1.9454554319381714,
      "learning_rate": 0.00010485011177128807,
      "loss": 4.7983,
      "step": 5362
    },
    {
      "epoch": 0.5363,
      "grad_norm": 3.7604289054870605,
      "learning_rate": 0.00010481524597204342,
      "loss": 4.8638,
      "step": 5363
    },
    {
      "epoch": 0.5364,
      "grad_norm": 2.1428732872009277,
      "learning_rate": 0.00010478037958607568,
      "loss": 4.6438,
      "step": 5364
    },
    {
      "epoch": 0.5365,
      "grad_norm": 1.9544117450714111,
      "learning_rate": 0.00010474551261763314,
      "loss": 4.7677,
      "step": 5365
    },
    {
      "epoch": 0.5366,
      "grad_norm": 2.927248001098633,
      "learning_rate": 0.00010471064507096426,
      "loss": 5.2396,
      "step": 5366
    },
    {
      "epoch": 0.5367,
      "grad_norm": 2.7970314025878906,
      "learning_rate": 0.00010467577695031762,
      "loss": 4.6367,
      "step": 5367
    },
    {
      "epoch": 0.5368,
      "grad_norm": 4.60487699508667,
      "learning_rate": 0.00010464090825994173,
      "loss": 4.7081,
      "step": 5368
    },
    {
      "epoch": 0.5369,
      "grad_norm": 2.901648759841919,
      "learning_rate": 0.00010460603900408523,
      "loss": 5.1729,
      "step": 5369
    },
    {
      "epoch": 0.537,
      "grad_norm": 2.4109737873077393,
      "learning_rate": 0.00010457116918699688,
      "loss": 3.9021,
      "step": 5370
    },
    {
      "epoch": 0.5371,
      "grad_norm": 11.750842094421387,
      "learning_rate": 0.00010453629881292538,
      "loss": 4.9769,
      "step": 5371
    },
    {
      "epoch": 0.5372,
      "grad_norm": 3.1280436515808105,
      "learning_rate": 0.00010450142788611965,
      "loss": 5.0335,
      "step": 5372
    },
    {
      "epoch": 0.5373,
      "grad_norm": 2.5917510986328125,
      "learning_rate": 0.00010446655641082862,
      "loss": 5.2681,
      "step": 5373
    },
    {
      "epoch": 0.5374,
      "grad_norm": 3.1811251640319824,
      "learning_rate": 0.00010443168439130122,
      "loss": 4.7453,
      "step": 5374
    },
    {
      "epoch": 0.5375,
      "grad_norm": 2.713888168334961,
      "learning_rate": 0.0001043968118317865,
      "loss": 5.4084,
      "step": 5375
    },
    {
      "epoch": 0.5376,
      "grad_norm": 2.4831137657165527,
      "learning_rate": 0.00010436193873653361,
      "loss": 4.9614,
      "step": 5376
    },
    {
      "epoch": 0.5377,
      "grad_norm": 2.907500982284546,
      "learning_rate": 0.00010432706510979171,
      "loss": 5.1259,
      "step": 5377
    },
    {
      "epoch": 0.5378,
      "grad_norm": 3.8611819744110107,
      "learning_rate": 0.00010429219095581007,
      "loss": 4.9729,
      "step": 5378
    },
    {
      "epoch": 0.5379,
      "grad_norm": 2.1207258701324463,
      "learning_rate": 0.00010425731627883797,
      "loss": 4.5178,
      "step": 5379
    },
    {
      "epoch": 0.538,
      "grad_norm": 4.775862693786621,
      "learning_rate": 0.0001042224410831248,
      "loss": 4.9776,
      "step": 5380
    },
    {
      "epoch": 0.5381,
      "grad_norm": 2.526689052581787,
      "learning_rate": 0.00010418756537291996,
      "loss": 5.1121,
      "step": 5381
    },
    {
      "epoch": 0.5382,
      "grad_norm": 1.9997766017913818,
      "learning_rate": 0.00010415268915247303,
      "loss": 4.5585,
      "step": 5382
    },
    {
      "epoch": 0.5383,
      "grad_norm": 2.0393881797790527,
      "learning_rate": 0.00010411781242603352,
      "loss": 4.8031,
      "step": 5383
    },
    {
      "epoch": 0.5384,
      "grad_norm": 3.5653650760650635,
      "learning_rate": 0.00010408293519785101,
      "loss": 5.1594,
      "step": 5384
    },
    {
      "epoch": 0.5385,
      "grad_norm": 5.660174369812012,
      "learning_rate": 0.00010404805747217526,
      "loss": 5.4031,
      "step": 5385
    },
    {
      "epoch": 0.5386,
      "grad_norm": 2.397306203842163,
      "learning_rate": 0.00010401317925325598,
      "loss": 4.9282,
      "step": 5386
    },
    {
      "epoch": 0.5387,
      "grad_norm": 2.778876304626465,
      "learning_rate": 0.000103978300545343,
      "loss": 4.6851,
      "step": 5387
    },
    {
      "epoch": 0.5388,
      "grad_norm": 5.910252094268799,
      "learning_rate": 0.00010394342135268613,
      "loss": 5.2573,
      "step": 5388
    },
    {
      "epoch": 0.5389,
      "grad_norm": 4.080316066741943,
      "learning_rate": 0.00010390854167953537,
      "loss": 4.8872,
      "step": 5389
    },
    {
      "epoch": 0.539,
      "grad_norm": 2.3069581985473633,
      "learning_rate": 0.00010387366153014062,
      "loss": 4.5098,
      "step": 5390
    },
    {
      "epoch": 0.5391,
      "grad_norm": 4.288173198699951,
      "learning_rate": 0.00010383878090875201,
      "loss": 5.379,
      "step": 5391
    },
    {
      "epoch": 0.5392,
      "grad_norm": 4.860734939575195,
      "learning_rate": 0.00010380389981961958,
      "loss": 6.0078,
      "step": 5392
    },
    {
      "epoch": 0.5393,
      "grad_norm": 2.2490439414978027,
      "learning_rate": 0.00010376901826699348,
      "loss": 4.5648,
      "step": 5393
    },
    {
      "epoch": 0.5394,
      "grad_norm": 2.7251343727111816,
      "learning_rate": 0.00010373413625512394,
      "loss": 4.2628,
      "step": 5394
    },
    {
      "epoch": 0.5395,
      "grad_norm": 3.8785035610198975,
      "learning_rate": 0.0001036992537882612,
      "loss": 4.4524,
      "step": 5395
    },
    {
      "epoch": 0.5396,
      "grad_norm": 1.7137807607650757,
      "learning_rate": 0.00010366437087065564,
      "loss": 4.7427,
      "step": 5396
    },
    {
      "epoch": 0.5397,
      "grad_norm": 2.8969204425811768,
      "learning_rate": 0.00010362948750655759,
      "loss": 4.9971,
      "step": 5397
    },
    {
      "epoch": 0.5398,
      "grad_norm": 2.8494153022766113,
      "learning_rate": 0.0001035946037002175,
      "loss": 5.7734,
      "step": 5398
    },
    {
      "epoch": 0.5399,
      "grad_norm": 2.932854175567627,
      "learning_rate": 0.00010355971945588585,
      "loss": 5.0655,
      "step": 5399
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.4352924823760986,
      "learning_rate": 0.0001035248347778132,
      "loss": 4.9547,
      "step": 5400
    },
    {
      "epoch": 0.5401,
      "grad_norm": 1.8641595840454102,
      "learning_rate": 0.00010348994967025012,
      "loss": 5.1608,
      "step": 5401
    },
    {
      "epoch": 0.5402,
      "grad_norm": 4.816633701324463,
      "learning_rate": 0.00010345506413744726,
      "loss": 4.8487,
      "step": 5402
    },
    {
      "epoch": 0.5403,
      "grad_norm": 2.113842010498047,
      "learning_rate": 0.0001034201781836553,
      "loss": 4.834,
      "step": 5403
    },
    {
      "epoch": 0.5404,
      "grad_norm": 2.217216730117798,
      "learning_rate": 0.00010338529181312497,
      "loss": 5.0235,
      "step": 5404
    },
    {
      "epoch": 0.5405,
      "grad_norm": 2.90484881401062,
      "learning_rate": 0.00010335040503010716,
      "loss": 4.582,
      "step": 5405
    },
    {
      "epoch": 0.5406,
      "grad_norm": 3.251386880874634,
      "learning_rate": 0.00010331551783885263,
      "loss": 4.9771,
      "step": 5406
    },
    {
      "epoch": 0.5407,
      "grad_norm": 3.1037144660949707,
      "learning_rate": 0.00010328063024361232,
      "loss": 5.4424,
      "step": 5407
    },
    {
      "epoch": 0.5408,
      "grad_norm": 2.1232728958129883,
      "learning_rate": 0.00010324574224863717,
      "loss": 4.874,
      "step": 5408
    },
    {
      "epoch": 0.5409,
      "grad_norm": 3.2304816246032715,
      "learning_rate": 0.00010321085385817817,
      "loss": 4.9596,
      "step": 5409
    },
    {
      "epoch": 0.541,
      "grad_norm": 3.9440231323242188,
      "learning_rate": 0.00010317596507648636,
      "loss": 5.4005,
      "step": 5410
    },
    {
      "epoch": 0.5411,
      "grad_norm": 2.481675624847412,
      "learning_rate": 0.00010314107590781284,
      "loss": 4.9502,
      "step": 5411
    },
    {
      "epoch": 0.5412,
      "grad_norm": 2.0023462772369385,
      "learning_rate": 0.00010310618635640876,
      "loss": 4.4126,
      "step": 5412
    },
    {
      "epoch": 0.5413,
      "grad_norm": 2.372957944869995,
      "learning_rate": 0.00010307129642652528,
      "loss": 4.9213,
      "step": 5413
    },
    {
      "epoch": 0.5414,
      "grad_norm": 2.0349316596984863,
      "learning_rate": 0.00010303640612241363,
      "loss": 4.7514,
      "step": 5414
    },
    {
      "epoch": 0.5415,
      "grad_norm": 5.17565393447876,
      "learning_rate": 0.00010300151544832512,
      "loss": 5.1234,
      "step": 5415
    },
    {
      "epoch": 0.5416,
      "grad_norm": 3.5146210193634033,
      "learning_rate": 0.00010296662440851108,
      "loss": 5.0507,
      "step": 5416
    },
    {
      "epoch": 0.5417,
      "grad_norm": 2.461639642715454,
      "learning_rate": 0.00010293173300722285,
      "loss": 4.5974,
      "step": 5417
    },
    {
      "epoch": 0.5418,
      "grad_norm": 2.10895037651062,
      "learning_rate": 0.00010289684124871182,
      "loss": 5.1251,
      "step": 5418
    },
    {
      "epoch": 0.5419,
      "grad_norm": 2.1038269996643066,
      "learning_rate": 0.00010286194913722948,
      "loss": 4.7383,
      "step": 5419
    },
    {
      "epoch": 0.542,
      "grad_norm": 2.5333874225616455,
      "learning_rate": 0.00010282705667702734,
      "loss": 4.8412,
      "step": 5420
    },
    {
      "epoch": 0.5421,
      "grad_norm": 2.4532675743103027,
      "learning_rate": 0.0001027921638723569,
      "loss": 5.6665,
      "step": 5421
    },
    {
      "epoch": 0.5422,
      "grad_norm": 2.7544991970062256,
      "learning_rate": 0.00010275727072746977,
      "loss": 5.7579,
      "step": 5422
    },
    {
      "epoch": 0.5423,
      "grad_norm": 2.3407392501831055,
      "learning_rate": 0.00010272237724661753,
      "loss": 4.6677,
      "step": 5423
    },
    {
      "epoch": 0.5424,
      "grad_norm": 6.748663425445557,
      "learning_rate": 0.00010268748343405192,
      "loss": 5.8323,
      "step": 5424
    },
    {
      "epoch": 0.5425,
      "grad_norm": 2.2446610927581787,
      "learning_rate": 0.0001026525892940246,
      "loss": 4.4429,
      "step": 5425
    },
    {
      "epoch": 0.5426,
      "grad_norm": 2.2463929653167725,
      "learning_rate": 0.00010261769483078733,
      "loss": 5.1713,
      "step": 5426
    },
    {
      "epoch": 0.5427,
      "grad_norm": 1.8202298879623413,
      "learning_rate": 0.00010258280004859188,
      "loss": 4.4829,
      "step": 5427
    },
    {
      "epoch": 0.5428,
      "grad_norm": 3.3366217613220215,
      "learning_rate": 0.00010254790495169006,
      "loss": 4.4889,
      "step": 5428
    },
    {
      "epoch": 0.5429,
      "grad_norm": 4.647078037261963,
      "learning_rate": 0.00010251300954433376,
      "loss": 4.7126,
      "step": 5429
    },
    {
      "epoch": 0.543,
      "grad_norm": 7.147397041320801,
      "learning_rate": 0.00010247811383077487,
      "loss": 4.7505,
      "step": 5430
    },
    {
      "epoch": 0.5431,
      "grad_norm": 2.478344440460205,
      "learning_rate": 0.00010244321781526533,
      "loss": 4.4229,
      "step": 5431
    },
    {
      "epoch": 0.5432,
      "grad_norm": 2.3989787101745605,
      "learning_rate": 0.0001024083215020571,
      "loss": 4.6689,
      "step": 5432
    },
    {
      "epoch": 0.5433,
      "grad_norm": 5.054302215576172,
      "learning_rate": 0.0001023734248954022,
      "loss": 4.9783,
      "step": 5433
    },
    {
      "epoch": 0.5434,
      "grad_norm": 2.6915669441223145,
      "learning_rate": 0.00010233852799955268,
      "loss": 4.7656,
      "step": 5434
    },
    {
      "epoch": 0.5435,
      "grad_norm": 3.057892322540283,
      "learning_rate": 0.00010230363081876064,
      "loss": 4.8554,
      "step": 5435
    },
    {
      "epoch": 0.5436,
      "grad_norm": 4.034045696258545,
      "learning_rate": 0.00010226873335727816,
      "loss": 4.8515,
      "step": 5436
    },
    {
      "epoch": 0.5437,
      "grad_norm": 6.79312801361084,
      "learning_rate": 0.00010223383561935738,
      "loss": 5.9131,
      "step": 5437
    },
    {
      "epoch": 0.5438,
      "grad_norm": 3.3730053901672363,
      "learning_rate": 0.00010219893760925052,
      "loss": 5.2988,
      "step": 5438
    },
    {
      "epoch": 0.5439,
      "grad_norm": 2.21000599861145,
      "learning_rate": 0.00010216403933120979,
      "loss": 4.7429,
      "step": 5439
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.8770246505737305,
      "learning_rate": 0.0001021291407894874,
      "loss": 4.6617,
      "step": 5440
    },
    {
      "epoch": 0.5441,
      "grad_norm": 3.1746840476989746,
      "learning_rate": 0.0001020942419883357,
      "loss": 5.729,
      "step": 5441
    },
    {
      "epoch": 0.5442,
      "grad_norm": 2.072951316833496,
      "learning_rate": 0.00010205934293200696,
      "loss": 4.8806,
      "step": 5442
    },
    {
      "epoch": 0.5443,
      "grad_norm": 3.5451176166534424,
      "learning_rate": 0.00010202444362475352,
      "loss": 5.0636,
      "step": 5443
    },
    {
      "epoch": 0.5444,
      "grad_norm": 2.5082621574401855,
      "learning_rate": 0.0001019895440708278,
      "loss": 4.8026,
      "step": 5444
    },
    {
      "epoch": 0.5445,
      "grad_norm": 2.2520806789398193,
      "learning_rate": 0.00010195464427448213,
      "loss": 4.8646,
      "step": 5445
    },
    {
      "epoch": 0.5446,
      "grad_norm": 5.635747909545898,
      "learning_rate": 0.000101919744239969,
      "loss": 5.2113,
      "step": 5446
    },
    {
      "epoch": 0.5447,
      "grad_norm": 3.1332550048828125,
      "learning_rate": 0.00010188484397154084,
      "loss": 4.5684,
      "step": 5447
    },
    {
      "epoch": 0.5448,
      "grad_norm": 3.570986032485962,
      "learning_rate": 0.00010184994347345016,
      "loss": 4.3715,
      "step": 5448
    },
    {
      "epoch": 0.5449,
      "grad_norm": 1.9895484447479248,
      "learning_rate": 0.00010181504274994948,
      "loss": 4.9142,
      "step": 5449
    },
    {
      "epoch": 0.545,
      "grad_norm": 2.5852954387664795,
      "learning_rate": 0.00010178014180529136,
      "loss": 4.9499,
      "step": 5450
    },
    {
      "epoch": 0.5451,
      "grad_norm": 2.17303204536438,
      "learning_rate": 0.00010174524064372837,
      "loss": 4.8774,
      "step": 5451
    },
    {
      "epoch": 0.5452,
      "grad_norm": 3.7809360027313232,
      "learning_rate": 0.00010171033926951305,
      "loss": 5.2838,
      "step": 5452
    },
    {
      "epoch": 0.5453,
      "grad_norm": 2.571255683898926,
      "learning_rate": 0.00010167543768689815,
      "loss": 4.879,
      "step": 5453
    },
    {
      "epoch": 0.5454,
      "grad_norm": 3.9816768169403076,
      "learning_rate": 0.00010164053590013622,
      "loss": 4.9912,
      "step": 5454
    },
    {
      "epoch": 0.5455,
      "grad_norm": 3.358883857727051,
      "learning_rate": 0.00010160563391347996,
      "loss": 5.2664,
      "step": 5455
    },
    {
      "epoch": 0.5456,
      "grad_norm": 2.3042232990264893,
      "learning_rate": 0.00010157073173118208,
      "loss": 4.9558,
      "step": 5456
    },
    {
      "epoch": 0.5457,
      "grad_norm": 3.2075490951538086,
      "learning_rate": 0.0001015358293574953,
      "loss": 4.7403,
      "step": 5457
    },
    {
      "epoch": 0.5458,
      "grad_norm": 3.9475176334381104,
      "learning_rate": 0.00010150092679667238,
      "loss": 4.6195,
      "step": 5458
    },
    {
      "epoch": 0.5459,
      "grad_norm": 8.127921104431152,
      "learning_rate": 0.00010146602405296607,
      "loss": 4.9404,
      "step": 5459
    },
    {
      "epoch": 0.546,
      "grad_norm": 5.45184326171875,
      "learning_rate": 0.00010143112113062918,
      "loss": 5.0294,
      "step": 5460
    },
    {
      "epoch": 0.5461,
      "grad_norm": 4.205347537994385,
      "learning_rate": 0.00010139621803391455,
      "loss": 4.8152,
      "step": 5461
    },
    {
      "epoch": 0.5462,
      "grad_norm": 2.877962589263916,
      "learning_rate": 0.00010136131476707496,
      "loss": 5.0794,
      "step": 5462
    },
    {
      "epoch": 0.5463,
      "grad_norm": 2.8160202503204346,
      "learning_rate": 0.0001013264113343633,
      "loss": 4.6917,
      "step": 5463
    },
    {
      "epoch": 0.5464,
      "grad_norm": 8.540634155273438,
      "learning_rate": 0.00010129150774003245,
      "loss": 4.7849,
      "step": 5464
    },
    {
      "epoch": 0.5465,
      "grad_norm": 2.7958195209503174,
      "learning_rate": 0.00010125660398833528,
      "loss": 4.3703,
      "step": 5465
    },
    {
      "epoch": 0.5466,
      "grad_norm": 3.7226243019104004,
      "learning_rate": 0.00010122170008352472,
      "loss": 5.3141,
      "step": 5466
    },
    {
      "epoch": 0.5467,
      "grad_norm": 3.967045783996582,
      "learning_rate": 0.00010118679602985373,
      "loss": 4.7689,
      "step": 5467
    },
    {
      "epoch": 0.5468,
      "grad_norm": 4.374542713165283,
      "learning_rate": 0.00010115189183157523,
      "loss": 5.4186,
      "step": 5468
    },
    {
      "epoch": 0.5469,
      "grad_norm": 2.6537137031555176,
      "learning_rate": 0.00010111698749294223,
      "loss": 5.0631,
      "step": 5469
    },
    {
      "epoch": 0.547,
      "grad_norm": 3.8440916538238525,
      "learning_rate": 0.00010108208301820767,
      "loss": 5.0268,
      "step": 5470
    },
    {
      "epoch": 0.5471,
      "grad_norm": 2.8020198345184326,
      "learning_rate": 0.00010104717841162458,
      "loss": 4.466,
      "step": 5471
    },
    {
      "epoch": 0.5472,
      "grad_norm": 3.4328839778900146,
      "learning_rate": 0.000101012273677446,
      "loss": 4.9923,
      "step": 5472
    },
    {
      "epoch": 0.5473,
      "grad_norm": 2.50728440284729,
      "learning_rate": 0.00010097736881992492,
      "loss": 5.3602,
      "step": 5473
    },
    {
      "epoch": 0.5474,
      "grad_norm": 2.285681962966919,
      "learning_rate": 0.00010094246384331442,
      "loss": 4.7607,
      "step": 5474
    },
    {
      "epoch": 0.5475,
      "grad_norm": 2.736659526824951,
      "learning_rate": 0.00010090755875186753,
      "loss": 5.0055,
      "step": 5475
    },
    {
      "epoch": 0.5476,
      "grad_norm": 2.9079861640930176,
      "learning_rate": 0.0001008726535498374,
      "loss": 4.8123,
      "step": 5476
    },
    {
      "epoch": 0.5477,
      "grad_norm": 2.288050413131714,
      "learning_rate": 0.00010083774824147708,
      "loss": 4.5998,
      "step": 5477
    },
    {
      "epoch": 0.5478,
      "grad_norm": 4.716646194458008,
      "learning_rate": 0.00010080284283103965,
      "loss": 4.9601,
      "step": 5478
    },
    {
      "epoch": 0.5479,
      "grad_norm": 3.3166017532348633,
      "learning_rate": 0.00010076793732277829,
      "loss": 5.0239,
      "step": 5479
    },
    {
      "epoch": 0.548,
      "grad_norm": 2.7117583751678467,
      "learning_rate": 0.00010073303172094606,
      "loss": 4.4816,
      "step": 5480
    },
    {
      "epoch": 0.5481,
      "grad_norm": 7.101806163787842,
      "learning_rate": 0.00010069812602979615,
      "loss": 4.1952,
      "step": 5481
    },
    {
      "epoch": 0.5482,
      "grad_norm": 2.4026577472686768,
      "learning_rate": 0.00010066322025358173,
      "loss": 4.9619,
      "step": 5482
    },
    {
      "epoch": 0.5483,
      "grad_norm": 4.094573020935059,
      "learning_rate": 0.00010062831439655591,
      "loss": 4.5931,
      "step": 5483
    },
    {
      "epoch": 0.5484,
      "grad_norm": 2.0137746334075928,
      "learning_rate": 0.00010059340846297189,
      "loss": 4.8004,
      "step": 5484
    },
    {
      "epoch": 0.5485,
      "grad_norm": 5.389651298522949,
      "learning_rate": 0.00010055850245708283,
      "loss": 5.1472,
      "step": 5485
    },
    {
      "epoch": 0.5486,
      "grad_norm": 3.739625930786133,
      "learning_rate": 0.00010052359638314195,
      "loss": 4.5212,
      "step": 5486
    },
    {
      "epoch": 0.5487,
      "grad_norm": 2.484733819961548,
      "learning_rate": 0.00010048869024540247,
      "loss": 4.6448,
      "step": 5487
    },
    {
      "epoch": 0.5488,
      "grad_norm": 2.273024797439575,
      "learning_rate": 0.00010045378404811756,
      "loss": 4.5523,
      "step": 5488
    },
    {
      "epoch": 0.5489,
      "grad_norm": 3.4581215381622314,
      "learning_rate": 0.0001004188777955404,
      "loss": 4.8009,
      "step": 5489
    },
    {
      "epoch": 0.549,
      "grad_norm": 2.540560483932495,
      "learning_rate": 0.00010038397149192426,
      "loss": 4.8824,
      "step": 5490
    },
    {
      "epoch": 0.5491,
      "grad_norm": 2.783998727798462,
      "learning_rate": 0.00010034906514152238,
      "loss": 5.0902,
      "step": 5491
    },
    {
      "epoch": 0.5492,
      "grad_norm": 2.721076488494873,
      "learning_rate": 0.00010031415874858797,
      "loss": 4.6011,
      "step": 5492
    },
    {
      "epoch": 0.5493,
      "grad_norm": 2.6707189083099365,
      "learning_rate": 0.00010027925231737428,
      "loss": 4.7144,
      "step": 5493
    },
    {
      "epoch": 0.5494,
      "grad_norm": 3.345188617706299,
      "learning_rate": 0.00010024434585213451,
      "loss": 5.6022,
      "step": 5494
    },
    {
      "epoch": 0.5495,
      "grad_norm": 3.073892593383789,
      "learning_rate": 0.00010020943935712192,
      "loss": 4.7421,
      "step": 5495
    },
    {
      "epoch": 0.5496,
      "grad_norm": 2.888950824737549,
      "learning_rate": 0.00010017453283658984,
      "loss": 4.7851,
      "step": 5496
    },
    {
      "epoch": 0.5497,
      "grad_norm": 5.000915050506592,
      "learning_rate": 0.00010013962629479146,
      "loss": 4.6231,
      "step": 5497
    },
    {
      "epoch": 0.5498,
      "grad_norm": 2.5680391788482666,
      "learning_rate": 0.00010010471973598002,
      "loss": 4.6807,
      "step": 5498
    },
    {
      "epoch": 0.5499,
      "grad_norm": 2.148725748062134,
      "learning_rate": 0.00010006981316440875,
      "loss": 4.8392,
      "step": 5499
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.072357416152954,
      "learning_rate": 0.00010003490658433101,
      "loss": 4.9985,
      "step": 5500
    },
    {
      "epoch": 0.5501,
      "grad_norm": 9.570509910583496,
      "learning_rate": 0.0001,
      "loss": 7.5144,
      "step": 5501
    },
    {
      "epoch": 0.5502,
      "grad_norm": 2.5566024780273438,
      "learning_rate": 9.996509341566904e-05,
      "loss": 4.9515,
      "step": 5502
    },
    {
      "epoch": 0.5503,
      "grad_norm": 2.7192368507385254,
      "learning_rate": 9.993018683559126e-05,
      "loss": 4.8933,
      "step": 5503
    },
    {
      "epoch": 0.5504,
      "grad_norm": 2.910677671432495,
      "learning_rate": 9.989528026402003e-05,
      "loss": 5.5641,
      "step": 5504
    },
    {
      "epoch": 0.5505,
      "grad_norm": 4.380270004272461,
      "learning_rate": 9.986037370520857e-05,
      "loss": 5.0816,
      "step": 5505
    },
    {
      "epoch": 0.5506,
      "grad_norm": 2.6305439472198486,
      "learning_rate": 9.98254671634102e-05,
      "loss": 5.2449,
      "step": 5506
    },
    {
      "epoch": 0.5507,
      "grad_norm": 4.751548767089844,
      "learning_rate": 9.979056064287806e-05,
      "loss": 5.2408,
      "step": 5507
    },
    {
      "epoch": 0.5508,
      "grad_norm": 2.5671803951263428,
      "learning_rate": 9.975565414786551e-05,
      "loss": 4.2961,
      "step": 5508
    },
    {
      "epoch": 0.5509,
      "grad_norm": 2.3327105045318604,
      "learning_rate": 9.972074768262576e-05,
      "loss": 5.0216,
      "step": 5509
    },
    {
      "epoch": 0.551,
      "grad_norm": 2.4502203464508057,
      "learning_rate": 9.968584125141204e-05,
      "loss": 5.118,
      "step": 5510
    },
    {
      "epoch": 0.5511,
      "grad_norm": 2.308328866958618,
      "learning_rate": 9.965093485847767e-05,
      "loss": 4.9239,
      "step": 5511
    },
    {
      "epoch": 0.5512,
      "grad_norm": 3.3715786933898926,
      "learning_rate": 9.961602850807576e-05,
      "loss": 4.6335,
      "step": 5512
    },
    {
      "epoch": 0.5513,
      "grad_norm": 2.768501043319702,
      "learning_rate": 9.958112220445963e-05,
      "loss": 4.8012,
      "step": 5513
    },
    {
      "epoch": 0.5514,
      "grad_norm": 2.6973087787628174,
      "learning_rate": 9.954621595188247e-05,
      "loss": 4.8184,
      "step": 5514
    },
    {
      "epoch": 0.5515,
      "grad_norm": 2.6463544368743896,
      "learning_rate": 9.951130975459757e-05,
      "loss": 4.852,
      "step": 5515
    },
    {
      "epoch": 0.5516,
      "grad_norm": 3.523364782333374,
      "learning_rate": 9.947640361685804e-05,
      "loss": 4.7645,
      "step": 5516
    },
    {
      "epoch": 0.5517,
      "grad_norm": 2.2609589099884033,
      "learning_rate": 9.94414975429172e-05,
      "loss": 5.0953,
      "step": 5517
    },
    {
      "epoch": 0.5518,
      "grad_norm": 2.9991774559020996,
      "learning_rate": 9.940659153702813e-05,
      "loss": 4.7286,
      "step": 5518
    },
    {
      "epoch": 0.5519,
      "grad_norm": 3.408844470977783,
      "learning_rate": 9.937168560344412e-05,
      "loss": 5.0414,
      "step": 5519
    },
    {
      "epoch": 0.552,
      "grad_norm": 2.6903512477874756,
      "learning_rate": 9.933677974641831e-05,
      "loss": 5.381,
      "step": 5520
    },
    {
      "epoch": 0.5521,
      "grad_norm": 1.9170001745224,
      "learning_rate": 9.930187397020386e-05,
      "loss": 4.5091,
      "step": 5521
    },
    {
      "epoch": 0.5522,
      "grad_norm": 3.9716107845306396,
      "learning_rate": 9.926696827905395e-05,
      "loss": 4.5726,
      "step": 5522
    },
    {
      "epoch": 0.5523,
      "grad_norm": 2.7756850719451904,
      "learning_rate": 9.923206267722173e-05,
      "loss": 5.0199,
      "step": 5523
    },
    {
      "epoch": 0.5524,
      "grad_norm": 3.45843243598938,
      "learning_rate": 9.919715716896036e-05,
      "loss": 4.9622,
      "step": 5524
    },
    {
      "epoch": 0.5525,
      "grad_norm": 3.1712679862976074,
      "learning_rate": 9.916225175852293e-05,
      "loss": 5.0652,
      "step": 5525
    },
    {
      "epoch": 0.5526,
      "grad_norm": 3.457044839859009,
      "learning_rate": 9.912734645016263e-05,
      "loss": 5.2878,
      "step": 5526
    },
    {
      "epoch": 0.5527,
      "grad_norm": 4.247889041900635,
      "learning_rate": 9.909244124813247e-05,
      "loss": 5.3884,
      "step": 5527
    },
    {
      "epoch": 0.5528,
      "grad_norm": 2.6642394065856934,
      "learning_rate": 9.90575361566856e-05,
      "loss": 4.9064,
      "step": 5528
    },
    {
      "epoch": 0.5529,
      "grad_norm": 5.28337287902832,
      "learning_rate": 9.902263118007513e-05,
      "loss": 5.0295,
      "step": 5529
    },
    {
      "epoch": 0.553,
      "grad_norm": 3.328256845474243,
      "learning_rate": 9.898772632255403e-05,
      "loss": 4.8347,
      "step": 5530
    },
    {
      "epoch": 0.5531,
      "grad_norm": 2.6703734397888184,
      "learning_rate": 9.895282158837545e-05,
      "loss": 4.8761,
      "step": 5531
    },
    {
      "epoch": 0.5532,
      "grad_norm": 2.2018253803253174,
      "learning_rate": 9.891791698179235e-05,
      "loss": 4.8134,
      "step": 5532
    },
    {
      "epoch": 0.5533,
      "grad_norm": 3.9586400985717773,
      "learning_rate": 9.888301250705779e-05,
      "loss": 5.5821,
      "step": 5533
    },
    {
      "epoch": 0.5534,
      "grad_norm": 2.2015411853790283,
      "learning_rate": 9.884810816842475e-05,
      "loss": 5.0802,
      "step": 5534
    },
    {
      "epoch": 0.5535,
      "grad_norm": 2.5014121532440186,
      "learning_rate": 9.881320397014629e-05,
      "loss": 5.1103,
      "step": 5535
    },
    {
      "epoch": 0.5536,
      "grad_norm": 2.579294443130493,
      "learning_rate": 9.877829991647528e-05,
      "loss": 5.034,
      "step": 5536
    },
    {
      "epoch": 0.5537,
      "grad_norm": 4.45060920715332,
      "learning_rate": 9.874339601166473e-05,
      "loss": 5.1524,
      "step": 5537
    },
    {
      "epoch": 0.5538,
      "grad_norm": 4.501793384552002,
      "learning_rate": 9.87084922599676e-05,
      "loss": 4.9823,
      "step": 5538
    },
    {
      "epoch": 0.5539,
      "grad_norm": 4.09404182434082,
      "learning_rate": 9.867358866563673e-05,
      "loss": 5.7911,
      "step": 5539
    },
    {
      "epoch": 0.554,
      "grad_norm": 2.640573263168335,
      "learning_rate": 9.863868523292509e-05,
      "loss": 4.6573,
      "step": 5540
    },
    {
      "epoch": 0.5541,
      "grad_norm": 2.054227590560913,
      "learning_rate": 9.860378196608549e-05,
      "loss": 4.7104,
      "step": 5541
    },
    {
      "epoch": 0.5542,
      "grad_norm": 2.6972739696502686,
      "learning_rate": 9.856887886937083e-05,
      "loss": 5.109,
      "step": 5542
    },
    {
      "epoch": 0.5543,
      "grad_norm": 3.987403631210327,
      "learning_rate": 9.853397594703394e-05,
      "loss": 5.7229,
      "step": 5543
    },
    {
      "epoch": 0.5544,
      "grad_norm": 2.6666409969329834,
      "learning_rate": 9.849907320332766e-05,
      "loss": 4.8227,
      "step": 5544
    },
    {
      "epoch": 0.5545,
      "grad_norm": 2.828993082046509,
      "learning_rate": 9.846417064250471e-05,
      "loss": 4.8617,
      "step": 5545
    },
    {
      "epoch": 0.5546,
      "grad_norm": 2.8326570987701416,
      "learning_rate": 9.842926826881796e-05,
      "loss": 5.3067,
      "step": 5546
    },
    {
      "epoch": 0.5547,
      "grad_norm": 3.0119056701660156,
      "learning_rate": 9.839436608652007e-05,
      "loss": 5.1885,
      "step": 5547
    },
    {
      "epoch": 0.5548,
      "grad_norm": 4.476268768310547,
      "learning_rate": 9.83594640998638e-05,
      "loss": 5.4156,
      "step": 5548
    },
    {
      "epoch": 0.5549,
      "grad_norm": 3.883227825164795,
      "learning_rate": 9.832456231310189e-05,
      "loss": 4.8426,
      "step": 5549
    },
    {
      "epoch": 0.555,
      "grad_norm": 4.467809200286865,
      "learning_rate": 9.828966073048693e-05,
      "loss": 5.2983,
      "step": 5550
    },
    {
      "epoch": 0.5551,
      "grad_norm": 6.28291654586792,
      "learning_rate": 9.825475935627165e-05,
      "loss": 4.8627,
      "step": 5551
    },
    {
      "epoch": 0.5552,
      "grad_norm": 5.0624847412109375,
      "learning_rate": 9.821985819470863e-05,
      "loss": 4.5269,
      "step": 5552
    },
    {
      "epoch": 0.5553,
      "grad_norm": 2.6101622581481934,
      "learning_rate": 9.818495725005054e-05,
      "loss": 5.1471,
      "step": 5553
    },
    {
      "epoch": 0.5554,
      "grad_norm": 2.28783917427063,
      "learning_rate": 9.815005652654985e-05,
      "loss": 4.7161,
      "step": 5554
    },
    {
      "epoch": 0.5555,
      "grad_norm": 4.800546646118164,
      "learning_rate": 9.81151560284592e-05,
      "loss": 5.6544,
      "step": 5555
    },
    {
      "epoch": 0.5556,
      "grad_norm": 6.042082786560059,
      "learning_rate": 9.808025576003104e-05,
      "loss": 5.0103,
      "step": 5556
    },
    {
      "epoch": 0.5557,
      "grad_norm": 3.6525604724884033,
      "learning_rate": 9.804535572551789e-05,
      "loss": 4.6385,
      "step": 5557
    },
    {
      "epoch": 0.5558,
      "grad_norm": 2.613635301589966,
      "learning_rate": 9.801045592917226e-05,
      "loss": 5.0291,
      "step": 5558
    },
    {
      "epoch": 0.5559,
      "grad_norm": 2.215414524078369,
      "learning_rate": 9.797555637524649e-05,
      "loss": 4.9845,
      "step": 5559
    },
    {
      "epoch": 0.556,
      "grad_norm": 2.8859102725982666,
      "learning_rate": 9.794065706799306e-05,
      "loss": 4.7937,
      "step": 5560
    },
    {
      "epoch": 0.5561,
      "grad_norm": 2.5716028213500977,
      "learning_rate": 9.790575801166432e-05,
      "loss": 5.0102,
      "step": 5561
    },
    {
      "epoch": 0.5562,
      "grad_norm": 2.613760471343994,
      "learning_rate": 9.78708592105126e-05,
      "loss": 4.8437,
      "step": 5562
    },
    {
      "epoch": 0.5563,
      "grad_norm": 2.218503475189209,
      "learning_rate": 9.783596066879022e-05,
      "loss": 4.0696,
      "step": 5563
    },
    {
      "epoch": 0.5564,
      "grad_norm": 2.227290630340576,
      "learning_rate": 9.78010623907495e-05,
      "loss": 4.7281,
      "step": 5564
    },
    {
      "epoch": 0.5565,
      "grad_norm": 2.082629442214966,
      "learning_rate": 9.776616438064264e-05,
      "loss": 5.1614,
      "step": 5565
    },
    {
      "epoch": 0.5566,
      "grad_norm": 2.2992022037506104,
      "learning_rate": 9.773126664272186e-05,
      "loss": 4.5309,
      "step": 5566
    },
    {
      "epoch": 0.5567,
      "grad_norm": 2.7072947025299072,
      "learning_rate": 9.76963691812394e-05,
      "loss": 5.4292,
      "step": 5567
    },
    {
      "epoch": 0.5568,
      "grad_norm": 6.985541343688965,
      "learning_rate": 9.766147200044732e-05,
      "loss": 5.512,
      "step": 5568
    },
    {
      "epoch": 0.5569,
      "grad_norm": 1.9755759239196777,
      "learning_rate": 9.762657510459783e-05,
      "loss": 4.7422,
      "step": 5569
    },
    {
      "epoch": 0.557,
      "grad_norm": 3.9953012466430664,
      "learning_rate": 9.759167849794292e-05,
      "loss": 4.4465,
      "step": 5570
    },
    {
      "epoch": 0.5571,
      "grad_norm": 3.36859130859375,
      "learning_rate": 9.755678218473469e-05,
      "loss": 5.0388,
      "step": 5571
    },
    {
      "epoch": 0.5572,
      "grad_norm": 2.775139093399048,
      "learning_rate": 9.752188616922518e-05,
      "loss": 4.4624,
      "step": 5572
    },
    {
      "epoch": 0.5573,
      "grad_norm": 2.948092460632324,
      "learning_rate": 9.748699045566626e-05,
      "loss": 4.7053,
      "step": 5573
    },
    {
      "epoch": 0.5574,
      "grad_norm": 3.047785758972168,
      "learning_rate": 9.745209504830996e-05,
      "loss": 4.6865,
      "step": 5574
    },
    {
      "epoch": 0.5575,
      "grad_norm": 3.667297124862671,
      "learning_rate": 9.741719995140814e-05,
      "loss": 5.1102,
      "step": 5575
    },
    {
      "epoch": 0.5576,
      "grad_norm": 10.263474464416504,
      "learning_rate": 9.73823051692127e-05,
      "loss": 7.8553,
      "step": 5576
    },
    {
      "epoch": 0.5577,
      "grad_norm": 2.475595712661743,
      "learning_rate": 9.734741070597539e-05,
      "loss": 4.9798,
      "step": 5577
    },
    {
      "epoch": 0.5578,
      "grad_norm": 3.2376809120178223,
      "learning_rate": 9.73125165659481e-05,
      "loss": 4.9301,
      "step": 5578
    },
    {
      "epoch": 0.5579,
      "grad_norm": 4.418052673339844,
      "learning_rate": 9.727762275338246e-05,
      "loss": 4.7404,
      "step": 5579
    },
    {
      "epoch": 0.558,
      "grad_norm": 5.379616737365723,
      "learning_rate": 9.724272927253025e-05,
      "loss": 5.0068,
      "step": 5580
    },
    {
      "epoch": 0.5581,
      "grad_norm": 2.178351640701294,
      "learning_rate": 9.720783612764314e-05,
      "loss": 4.7024,
      "step": 5581
    },
    {
      "epoch": 0.5582,
      "grad_norm": 4.073245048522949,
      "learning_rate": 9.717294332297268e-05,
      "loss": 4.9006,
      "step": 5582
    },
    {
      "epoch": 0.5583,
      "grad_norm": 3.1810836791992188,
      "learning_rate": 9.713805086277054e-05,
      "loss": 4.4886,
      "step": 5583
    },
    {
      "epoch": 0.5584,
      "grad_norm": 2.1909148693084717,
      "learning_rate": 9.710315875128819e-05,
      "loss": 5.0913,
      "step": 5584
    },
    {
      "epoch": 0.5585,
      "grad_norm": 3.000108003616333,
      "learning_rate": 9.706826699277718e-05,
      "loss": 4.376,
      "step": 5585
    },
    {
      "epoch": 0.5586,
      "grad_norm": 2.9048538208007812,
      "learning_rate": 9.703337559148892e-05,
      "loss": 4.7748,
      "step": 5586
    },
    {
      "epoch": 0.5587,
      "grad_norm": 2.99444842338562,
      "learning_rate": 9.699848455167489e-05,
      "loss": 4.9996,
      "step": 5587
    },
    {
      "epoch": 0.5588,
      "grad_norm": 3.7299411296844482,
      "learning_rate": 9.696359387758636e-05,
      "loss": 5.6566,
      "step": 5588
    },
    {
      "epoch": 0.5589,
      "grad_norm": 4.527307510375977,
      "learning_rate": 9.692870357347473e-05,
      "loss": 4.5367,
      "step": 5589
    },
    {
      "epoch": 0.559,
      "grad_norm": 2.628405809402466,
      "learning_rate": 9.689381364359129e-05,
      "loss": 4.764,
      "step": 5590
    },
    {
      "epoch": 0.5591,
      "grad_norm": 2.8575034141540527,
      "learning_rate": 9.685892409218717e-05,
      "loss": 4.6367,
      "step": 5591
    },
    {
      "epoch": 0.5592,
      "grad_norm": 5.3343610763549805,
      "learning_rate": 9.682403492351369e-05,
      "loss": 5.1276,
      "step": 5592
    },
    {
      "epoch": 0.5593,
      "grad_norm": 5.128260612487793,
      "learning_rate": 9.678914614182185e-05,
      "loss": 4.8304,
      "step": 5593
    },
    {
      "epoch": 0.5594,
      "grad_norm": 3.905325412750244,
      "learning_rate": 9.675425775136286e-05,
      "loss": 5.3439,
      "step": 5594
    },
    {
      "epoch": 0.5595,
      "grad_norm": 3.1525397300720215,
      "learning_rate": 9.671936975638768e-05,
      "loss": 4.5072,
      "step": 5595
    },
    {
      "epoch": 0.5596,
      "grad_norm": 2.437095880508423,
      "learning_rate": 9.668448216114739e-05,
      "loss": 4.5515,
      "step": 5596
    },
    {
      "epoch": 0.5597,
      "grad_norm": 2.8166236877441406,
      "learning_rate": 9.664959496989285e-05,
      "loss": 4.7755,
      "step": 5597
    },
    {
      "epoch": 0.5598,
      "grad_norm": 2.1157102584838867,
      "learning_rate": 9.661470818687503e-05,
      "loss": 4.7428,
      "step": 5598
    },
    {
      "epoch": 0.5599,
      "grad_norm": 2.4750490188598633,
      "learning_rate": 9.657982181634475e-05,
      "loss": 4.8734,
      "step": 5599
    },
    {
      "epoch": 0.56,
      "grad_norm": 6.362810134887695,
      "learning_rate": 9.654493586255278e-05,
      "loss": 5.5587,
      "step": 5600
    },
    {
      "epoch": 0.5601,
      "grad_norm": 2.8367035388946533,
      "learning_rate": 9.651005032974994e-05,
      "loss": 5.1412,
      "step": 5601
    },
    {
      "epoch": 0.5602,
      "grad_norm": 3.3103437423706055,
      "learning_rate": 9.647516522218683e-05,
      "loss": 4.609,
      "step": 5602
    },
    {
      "epoch": 0.5603,
      "grad_norm": 8.979642868041992,
      "learning_rate": 9.644028054411416e-05,
      "loss": 5.1513,
      "step": 5603
    },
    {
      "epoch": 0.5604,
      "grad_norm": 2.3761205673217773,
      "learning_rate": 9.64053962997825e-05,
      "loss": 4.5415,
      "step": 5604
    },
    {
      "epoch": 0.5605,
      "grad_norm": 3.6219184398651123,
      "learning_rate": 9.637051249344243e-05,
      "loss": 4.977,
      "step": 5605
    },
    {
      "epoch": 0.5606,
      "grad_norm": 3.0494961738586426,
      "learning_rate": 9.633562912934436e-05,
      "loss": 5.0288,
      "step": 5606
    },
    {
      "epoch": 0.5607,
      "grad_norm": 6.788987636566162,
      "learning_rate": 9.630074621173883e-05,
      "loss": 6.3926,
      "step": 5607
    },
    {
      "epoch": 0.5608,
      "grad_norm": 2.6008169651031494,
      "learning_rate": 9.62658637448761e-05,
      "loss": 4.751,
      "step": 5608
    },
    {
      "epoch": 0.5609,
      "grad_norm": 2.0926172733306885,
      "learning_rate": 9.623098173300654e-05,
      "loss": 4.9171,
      "step": 5609
    },
    {
      "epoch": 0.561,
      "grad_norm": 1.9987387657165527,
      "learning_rate": 9.619610018038048e-05,
      "loss": 4.9593,
      "step": 5610
    },
    {
      "epoch": 0.5611,
      "grad_norm": 4.311956882476807,
      "learning_rate": 9.616121909124801e-05,
      "loss": 5.4455,
      "step": 5611
    },
    {
      "epoch": 0.5612,
      "grad_norm": 3.252964496612549,
      "learning_rate": 9.612633846985941e-05,
      "loss": 4.8768,
      "step": 5612
    },
    {
      "epoch": 0.5613,
      "grad_norm": 2.205819845199585,
      "learning_rate": 9.609145832046465e-05,
      "loss": 4.5106,
      "step": 5613
    },
    {
      "epoch": 0.5614,
      "grad_norm": 5.105076313018799,
      "learning_rate": 9.605657864731388e-05,
      "loss": 4.938,
      "step": 5614
    },
    {
      "epoch": 0.5615,
      "grad_norm": 2.120903968811035,
      "learning_rate": 9.602169945465702e-05,
      "loss": 4.6277,
      "step": 5615
    },
    {
      "epoch": 0.5616,
      "grad_norm": 3.1241936683654785,
      "learning_rate": 9.598682074674405e-05,
      "loss": 5.0902,
      "step": 5616
    },
    {
      "epoch": 0.5617,
      "grad_norm": 6.743006229400635,
      "learning_rate": 9.595194252782477e-05,
      "loss": 5.3185,
      "step": 5617
    },
    {
      "epoch": 0.5618,
      "grad_norm": 2.699793815612793,
      "learning_rate": 9.591706480214901e-05,
      "loss": 4.5857,
      "step": 5618
    },
    {
      "epoch": 0.5619,
      "grad_norm": 2.3653974533081055,
      "learning_rate": 9.588218757396655e-05,
      "loss": 4.6986,
      "step": 5619
    },
    {
      "epoch": 0.562,
      "grad_norm": 5.223864555358887,
      "learning_rate": 9.584731084752699e-05,
      "loss": 4.6182,
      "step": 5620
    },
    {
      "epoch": 0.5621,
      "grad_norm": 2.4828813076019287,
      "learning_rate": 9.581243462708006e-05,
      "loss": 5.0166,
      "step": 5621
    },
    {
      "epoch": 0.5622,
      "grad_norm": 2.353539228439331,
      "learning_rate": 9.577755891687523e-05,
      "loss": 4.3329,
      "step": 5622
    },
    {
      "epoch": 0.5623,
      "grad_norm": 1.8154183626174927,
      "learning_rate": 9.574268372116205e-05,
      "loss": 4.7356,
      "step": 5623
    },
    {
      "epoch": 0.5624,
      "grad_norm": 2.635411262512207,
      "learning_rate": 9.570780904418993e-05,
      "loss": 4.5851,
      "step": 5624
    },
    {
      "epoch": 0.5625,
      "grad_norm": 7.395214557647705,
      "learning_rate": 9.567293489020831e-05,
      "loss": 5.1768,
      "step": 5625
    },
    {
      "epoch": 0.5626,
      "grad_norm": 2.556908369064331,
      "learning_rate": 9.563806126346642e-05,
      "loss": 4.859,
      "step": 5626
    },
    {
      "epoch": 0.5627,
      "grad_norm": 2.557927370071411,
      "learning_rate": 9.560318816821353e-05,
      "loss": 4.4376,
      "step": 5627
    },
    {
      "epoch": 0.5628,
      "grad_norm": 2.8435287475585938,
      "learning_rate": 9.556831560869882e-05,
      "loss": 4.6217,
      "step": 5628
    },
    {
      "epoch": 0.5629,
      "grad_norm": 4.444569110870361,
      "learning_rate": 9.55334435891714e-05,
      "loss": 4.907,
      "step": 5629
    },
    {
      "epoch": 0.563,
      "grad_norm": 2.380498170852661,
      "learning_rate": 9.549857211388037e-05,
      "loss": 4.3308,
      "step": 5630
    },
    {
      "epoch": 0.5631,
      "grad_norm": 2.1652660369873047,
      "learning_rate": 9.546370118707463e-05,
      "loss": 4.391,
      "step": 5631
    },
    {
      "epoch": 0.5632,
      "grad_norm": 3.9296414852142334,
      "learning_rate": 9.542883081300316e-05,
      "loss": 5.1628,
      "step": 5632
    },
    {
      "epoch": 0.5633,
      "grad_norm": 11.568896293640137,
      "learning_rate": 9.539396099591476e-05,
      "loss": 7.5437,
      "step": 5633
    },
    {
      "epoch": 0.5634,
      "grad_norm": 3.750544548034668,
      "learning_rate": 9.53590917400583e-05,
      "loss": 5.8262,
      "step": 5634
    },
    {
      "epoch": 0.5635,
      "grad_norm": 4.129705429077148,
      "learning_rate": 9.532422304968243e-05,
      "loss": 4.8999,
      "step": 5635
    },
    {
      "epoch": 0.5636,
      "grad_norm": 2.735687017440796,
      "learning_rate": 9.528935492903575e-05,
      "loss": 4.8688,
      "step": 5636
    },
    {
      "epoch": 0.5637,
      "grad_norm": 3.2715775966644287,
      "learning_rate": 9.525448738236691e-05,
      "loss": 4.9495,
      "step": 5637
    },
    {
      "epoch": 0.5638,
      "grad_norm": 2.6421515941619873,
      "learning_rate": 9.521962041392436e-05,
      "loss": 4.9671,
      "step": 5638
    },
    {
      "epoch": 0.5639,
      "grad_norm": 3.1443099975585938,
      "learning_rate": 9.518475402795661e-05,
      "loss": 5.3694,
      "step": 5639
    },
    {
      "epoch": 0.564,
      "grad_norm": 3.132920026779175,
      "learning_rate": 9.514988822871193e-05,
      "loss": 5.1241,
      "step": 5640
    },
    {
      "epoch": 0.5641,
      "grad_norm": 3.8182461261749268,
      "learning_rate": 9.511502302043868e-05,
      "loss": 4.5022,
      "step": 5641
    },
    {
      "epoch": 0.5642,
      "grad_norm": 2.5584805011749268,
      "learning_rate": 9.508015840738503e-05,
      "loss": 4.7598,
      "step": 5642
    },
    {
      "epoch": 0.5643,
      "grad_norm": 6.916779041290283,
      "learning_rate": 9.504529439379921e-05,
      "loss": 4.8551,
      "step": 5643
    },
    {
      "epoch": 0.5644,
      "grad_norm": 3.72162127494812,
      "learning_rate": 9.501043098392924e-05,
      "loss": 4.6447,
      "step": 5644
    },
    {
      "epoch": 0.5645,
      "grad_norm": 2.7273385524749756,
      "learning_rate": 9.497556818202306e-05,
      "loss": 4.6765,
      "step": 5645
    },
    {
      "epoch": 0.5646,
      "grad_norm": 3.8584089279174805,
      "learning_rate": 9.494070599232868e-05,
      "loss": 5.0829,
      "step": 5646
    },
    {
      "epoch": 0.5647,
      "grad_norm": 1.9082965850830078,
      "learning_rate": 9.490584441909392e-05,
      "loss": 4.4439,
      "step": 5647
    },
    {
      "epoch": 0.5648,
      "grad_norm": 2.9442992210388184,
      "learning_rate": 9.48709834665666e-05,
      "loss": 4.5653,
      "step": 5648
    },
    {
      "epoch": 0.5649,
      "grad_norm": 2.7254562377929688,
      "learning_rate": 9.483612313899435e-05,
      "loss": 5.5395,
      "step": 5649
    },
    {
      "epoch": 0.565,
      "grad_norm": 2.2038912773132324,
      "learning_rate": 9.480126344062487e-05,
      "loss": 4.6232,
      "step": 5650
    },
    {
      "epoch": 0.5651,
      "grad_norm": 3.30456805229187,
      "learning_rate": 9.476640437570562e-05,
      "loss": 4.8683,
      "step": 5651
    },
    {
      "epoch": 0.5652,
      "grad_norm": 1.9127570390701294,
      "learning_rate": 9.473154594848415e-05,
      "loss": 4.5689,
      "step": 5652
    },
    {
      "epoch": 0.5653,
      "grad_norm": 2.571654796600342,
      "learning_rate": 9.469668816320784e-05,
      "loss": 4.9324,
      "step": 5653
    },
    {
      "epoch": 0.5654,
      "grad_norm": 2.38254976272583,
      "learning_rate": 9.466183102412395e-05,
      "loss": 4.8246,
      "step": 5654
    },
    {
      "epoch": 0.5655,
      "grad_norm": 2.675060749053955,
      "learning_rate": 9.462697453547979e-05,
      "loss": 5.0536,
      "step": 5655
    },
    {
      "epoch": 0.5656,
      "grad_norm": 4.387674331665039,
      "learning_rate": 9.459211870152245e-05,
      "loss": 5.9103,
      "step": 5656
    },
    {
      "epoch": 0.5657,
      "grad_norm": 2.6054906845092773,
      "learning_rate": 9.455726352649911e-05,
      "loss": 4.7912,
      "step": 5657
    },
    {
      "epoch": 0.5658,
      "grad_norm": 3.6427478790283203,
      "learning_rate": 9.452240901465663e-05,
      "loss": 4.7966,
      "step": 5658
    },
    {
      "epoch": 0.5659,
      "grad_norm": 2.8119654655456543,
      "learning_rate": 9.448755517024206e-05,
      "loss": 5.3833,
      "step": 5659
    },
    {
      "epoch": 0.566,
      "grad_norm": 3.904181957244873,
      "learning_rate": 9.445270199750212e-05,
      "loss": 4.5162,
      "step": 5660
    },
    {
      "epoch": 0.5661,
      "grad_norm": 3.4606354236602783,
      "learning_rate": 9.441784950068362e-05,
      "loss": 5.8194,
      "step": 5661
    },
    {
      "epoch": 0.5662,
      "grad_norm": 8.633391380310059,
      "learning_rate": 9.438299768403327e-05,
      "loss": 5.3782,
      "step": 5662
    },
    {
      "epoch": 0.5663,
      "grad_norm": 3.5687923431396484,
      "learning_rate": 9.434814655179755e-05,
      "loss": 4.7115,
      "step": 5663
    },
    {
      "epoch": 0.5664,
      "grad_norm": 3.321997880935669,
      "learning_rate": 9.43132961082231e-05,
      "loss": 4.5516,
      "step": 5664
    },
    {
      "epoch": 0.5665,
      "grad_norm": 2.58662486076355,
      "learning_rate": 9.427844635755619e-05,
      "loss": 4.4754,
      "step": 5665
    },
    {
      "epoch": 0.5666,
      "grad_norm": 3.8349759578704834,
      "learning_rate": 9.424359730404329e-05,
      "loss": 4.9144,
      "step": 5666
    },
    {
      "epoch": 0.5667,
      "grad_norm": 3.6315598487854004,
      "learning_rate": 9.420874895193056e-05,
      "loss": 5.0122,
      "step": 5667
    },
    {
      "epoch": 0.5668,
      "grad_norm": 5.022012710571289,
      "learning_rate": 9.417390130546426e-05,
      "loss": 5.2402,
      "step": 5668
    },
    {
      "epoch": 0.5669,
      "grad_norm": 4.009657859802246,
      "learning_rate": 9.413905436889035e-05,
      "loss": 5.7075,
      "step": 5669
    },
    {
      "epoch": 0.567,
      "grad_norm": 4.065012454986572,
      "learning_rate": 9.410420814645493e-05,
      "loss": 4.9864,
      "step": 5670
    },
    {
      "epoch": 0.5671,
      "grad_norm": 8.151076316833496,
      "learning_rate": 9.406936264240386e-05,
      "loss": 5.9914,
      "step": 5671
    },
    {
      "epoch": 0.5672,
      "grad_norm": 2.026172161102295,
      "learning_rate": 9.403451786098294e-05,
      "loss": 4.6344,
      "step": 5672
    },
    {
      "epoch": 0.5673,
      "grad_norm": 2.920797109603882,
      "learning_rate": 9.399967380643796e-05,
      "loss": 4.3212,
      "step": 5673
    },
    {
      "epoch": 0.5674,
      "grad_norm": 2.2597532272338867,
      "learning_rate": 9.396483048301448e-05,
      "loss": 4.5859,
      "step": 5674
    },
    {
      "epoch": 0.5675,
      "grad_norm": 3.8646204471588135,
      "learning_rate": 9.392998789495811e-05,
      "loss": 5.1296,
      "step": 5675
    },
    {
      "epoch": 0.5676,
      "grad_norm": 2.619368076324463,
      "learning_rate": 9.38951460465143e-05,
      "loss": 5.0798,
      "step": 5676
    },
    {
      "epoch": 0.5677,
      "grad_norm": 4.650434970855713,
      "learning_rate": 9.386030494192846e-05,
      "loss": 5.0525,
      "step": 5677
    },
    {
      "epoch": 0.5678,
      "grad_norm": 2.1046643257141113,
      "learning_rate": 9.382546458544582e-05,
      "loss": 4.3089,
      "step": 5678
    },
    {
      "epoch": 0.5679,
      "grad_norm": 2.7886972427368164,
      "learning_rate": 9.37906249813116e-05,
      "loss": 5.0376,
      "step": 5679
    },
    {
      "epoch": 0.568,
      "grad_norm": 2.392195701599121,
      "learning_rate": 9.375578613377089e-05,
      "loss": 4.5302,
      "step": 5680
    },
    {
      "epoch": 0.5681,
      "grad_norm": 3.928952693939209,
      "learning_rate": 9.372094804706867e-05,
      "loss": 4.1734,
      "step": 5681
    },
    {
      "epoch": 0.5682,
      "grad_norm": 7.594268798828125,
      "learning_rate": 9.368611072544992e-05,
      "loss": 5.7292,
      "step": 5682
    },
    {
      "epoch": 0.5683,
      "grad_norm": 3.587920904159546,
      "learning_rate": 9.36512741731594e-05,
      "loss": 4.5714,
      "step": 5683
    },
    {
      "epoch": 0.5684,
      "grad_norm": 2.7494494915008545,
      "learning_rate": 9.361643839444188e-05,
      "loss": 5.1303,
      "step": 5684
    },
    {
      "epoch": 0.5685,
      "grad_norm": 4.247781753540039,
      "learning_rate": 9.358160339354194e-05,
      "loss": 4.4628,
      "step": 5685
    },
    {
      "epoch": 0.5686,
      "grad_norm": 5.530324459075928,
      "learning_rate": 9.354676917470422e-05,
      "loss": 5.3033,
      "step": 5686
    },
    {
      "epoch": 0.5687,
      "grad_norm": 2.8032352924346924,
      "learning_rate": 9.351193574217306e-05,
      "loss": 4.4851,
      "step": 5687
    },
    {
      "epoch": 0.5688,
      "grad_norm": 4.302609920501709,
      "learning_rate": 9.347710310019288e-05,
      "loss": 5.11,
      "step": 5688
    },
    {
      "epoch": 0.5689,
      "grad_norm": 4.035916328430176,
      "learning_rate": 9.344227125300788e-05,
      "loss": 4.4639,
      "step": 5689
    },
    {
      "epoch": 0.569,
      "grad_norm": 10.619770050048828,
      "learning_rate": 9.340744020486222e-05,
      "loss": 5.8932,
      "step": 5690
    },
    {
      "epoch": 0.5691,
      "grad_norm": 2.81027889251709,
      "learning_rate": 9.337260996000002e-05,
      "loss": 4.9516,
      "step": 5691
    },
    {
      "epoch": 0.5692,
      "grad_norm": 4.275862693786621,
      "learning_rate": 9.333778052266513e-05,
      "loss": 5.0219,
      "step": 5692
    },
    {
      "epoch": 0.5693,
      "grad_norm": 3.915952682495117,
      "learning_rate": 9.330295189710152e-05,
      "loss": 4.7871,
      "step": 5693
    },
    {
      "epoch": 0.5694,
      "grad_norm": 2.160630941390991,
      "learning_rate": 9.32681240875529e-05,
      "loss": 4.7252,
      "step": 5694
    },
    {
      "epoch": 0.5695,
      "grad_norm": 2.9896392822265625,
      "learning_rate": 9.323329709826294e-05,
      "loss": 5.0401,
      "step": 5695
    },
    {
      "epoch": 0.5696,
      "grad_norm": 2.89589786529541,
      "learning_rate": 9.319847093347522e-05,
      "loss": 4.6558,
      "step": 5696
    },
    {
      "epoch": 0.5697,
      "grad_norm": 2.335172176361084,
      "learning_rate": 9.316364559743314e-05,
      "loss": 4.4044,
      "step": 5697
    },
    {
      "epoch": 0.5698,
      "grad_norm": 2.728975534439087,
      "learning_rate": 9.312882109438013e-05,
      "loss": 4.5471,
      "step": 5698
    },
    {
      "epoch": 0.5699,
      "grad_norm": 4.748671531677246,
      "learning_rate": 9.309399742855942e-05,
      "loss": 5.1328,
      "step": 5699
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.9837777614593506,
      "learning_rate": 9.30591746042142e-05,
      "loss": 4.7994,
      "step": 5700
    },
    {
      "epoch": 0.5701,
      "grad_norm": 3.2760419845581055,
      "learning_rate": 9.302435262558747e-05,
      "loss": 5.2215,
      "step": 5701
    },
    {
      "epoch": 0.5702,
      "grad_norm": 3.4630000591278076,
      "learning_rate": 9.298953149692225e-05,
      "loss": 4.9998,
      "step": 5702
    },
    {
      "epoch": 0.5703,
      "grad_norm": 4.430224418640137,
      "learning_rate": 9.295471122246131e-05,
      "loss": 5.3131,
      "step": 5703
    },
    {
      "epoch": 0.5704,
      "grad_norm": 2.8670644760131836,
      "learning_rate": 9.291989180644747e-05,
      "loss": 5.0179,
      "step": 5704
    },
    {
      "epoch": 0.5705,
      "grad_norm": 2.248974561691284,
      "learning_rate": 9.288507325312335e-05,
      "loss": 4.8104,
      "step": 5705
    },
    {
      "epoch": 0.5706,
      "grad_norm": 3.785498857498169,
      "learning_rate": 9.285025556673141e-05,
      "loss": 4.8008,
      "step": 5706
    },
    {
      "epoch": 0.5707,
      "grad_norm": 2.7336883544921875,
      "learning_rate": 9.281543875151419e-05,
      "loss": 4.7122,
      "step": 5707
    },
    {
      "epoch": 0.5708,
      "grad_norm": 2.2793161869049072,
      "learning_rate": 9.278062281171393e-05,
      "loss": 4.5992,
      "step": 5708
    },
    {
      "epoch": 0.5709,
      "grad_norm": 3.911947727203369,
      "learning_rate": 9.274580775157294e-05,
      "loss": 4.4989,
      "step": 5709
    },
    {
      "epoch": 0.571,
      "grad_norm": 2.5245442390441895,
      "learning_rate": 9.271099357533321e-05,
      "loss": 4.3,
      "step": 5710
    },
    {
      "epoch": 0.5711,
      "grad_norm": 2.1482126712799072,
      "learning_rate": 9.267618028723686e-05,
      "loss": 4.7976,
      "step": 5711
    },
    {
      "epoch": 0.5712,
      "grad_norm": 2.02308988571167,
      "learning_rate": 9.264136789152567e-05,
      "loss": 4.6019,
      "step": 5712
    },
    {
      "epoch": 0.5713,
      "grad_norm": 6.513894081115723,
      "learning_rate": 9.26065563924415e-05,
      "loss": 5.1851,
      "step": 5713
    },
    {
      "epoch": 0.5714,
      "grad_norm": 7.705225944519043,
      "learning_rate": 9.257174579422605e-05,
      "loss": 4.4102,
      "step": 5714
    },
    {
      "epoch": 0.5715,
      "grad_norm": 4.16671895980835,
      "learning_rate": 9.253693610112078e-05,
      "loss": 4.7118,
      "step": 5715
    },
    {
      "epoch": 0.5716,
      "grad_norm": 3.300677537918091,
      "learning_rate": 9.250212731736726e-05,
      "loss": 4.726,
      "step": 5716
    },
    {
      "epoch": 0.5717,
      "grad_norm": 2.1330504417419434,
      "learning_rate": 9.246731944720675e-05,
      "loss": 4.8802,
      "step": 5717
    },
    {
      "epoch": 0.5718,
      "grad_norm": 2.3464431762695312,
      "learning_rate": 9.243251249488052e-05,
      "loss": 5.0474,
      "step": 5718
    },
    {
      "epoch": 0.5719,
      "grad_norm": 2.260089874267578,
      "learning_rate": 9.239770646462968e-05,
      "loss": 5.0189,
      "step": 5719
    },
    {
      "epoch": 0.572,
      "grad_norm": 4.9637956619262695,
      "learning_rate": 9.236290136069528e-05,
      "loss": 5.3906,
      "step": 5720
    },
    {
      "epoch": 0.5721,
      "grad_norm": 4.201067924499512,
      "learning_rate": 9.232809718731814e-05,
      "loss": 4.769,
      "step": 5721
    },
    {
      "epoch": 0.5722,
      "grad_norm": 2.700578212738037,
      "learning_rate": 9.229329394873911e-05,
      "loss": 4.2542,
      "step": 5722
    },
    {
      "epoch": 0.5723,
      "grad_norm": 5.2327375411987305,
      "learning_rate": 9.225849164919885e-05,
      "loss": 4.5023,
      "step": 5723
    },
    {
      "epoch": 0.5724,
      "grad_norm": 4.826581001281738,
      "learning_rate": 9.222369029293787e-05,
      "loss": 5.4896,
      "step": 5724
    },
    {
      "epoch": 0.5725,
      "grad_norm": 2.5440917015075684,
      "learning_rate": 9.218888988419668e-05,
      "loss": 4.9816,
      "step": 5725
    },
    {
      "epoch": 0.5726,
      "grad_norm": 2.861053705215454,
      "learning_rate": 9.215409042721552e-05,
      "loss": 4.7534,
      "step": 5726
    },
    {
      "epoch": 0.5727,
      "grad_norm": 3.3211565017700195,
      "learning_rate": 9.211929192623467e-05,
      "loss": 4.5782,
      "step": 5727
    },
    {
      "epoch": 0.5728,
      "grad_norm": 2.455784320831299,
      "learning_rate": 9.208449438549415e-05,
      "loss": 5.0187,
      "step": 5728
    },
    {
      "epoch": 0.5729,
      "grad_norm": 3.844161033630371,
      "learning_rate": 9.204969780923403e-05,
      "loss": 5.0995,
      "step": 5729
    },
    {
      "epoch": 0.573,
      "grad_norm": 2.2664272785186768,
      "learning_rate": 9.201490220169408e-05,
      "loss": 4.8212,
      "step": 5730
    },
    {
      "epoch": 0.5731,
      "grad_norm": 2.133730173110962,
      "learning_rate": 9.198010756711412e-05,
      "loss": 4.9378,
      "step": 5731
    },
    {
      "epoch": 0.5732,
      "grad_norm": 1.778021216392517,
      "learning_rate": 9.194531390973371e-05,
      "loss": 4.7836,
      "step": 5732
    },
    {
      "epoch": 0.5733,
      "grad_norm": 9.418067932128906,
      "learning_rate": 9.191052123379234e-05,
      "loss": 4.8187,
      "step": 5733
    },
    {
      "epoch": 0.5734,
      "grad_norm": 4.537530422210693,
      "learning_rate": 9.187572954352947e-05,
      "loss": 6.3388,
      "step": 5734
    },
    {
      "epoch": 0.5735,
      "grad_norm": 3.2016444206237793,
      "learning_rate": 9.184093884318425e-05,
      "loss": 4.9048,
      "step": 5735
    },
    {
      "epoch": 0.5736,
      "grad_norm": 3.3084490299224854,
      "learning_rate": 9.180614913699592e-05,
      "loss": 5.0346,
      "step": 5736
    },
    {
      "epoch": 0.5737,
      "grad_norm": 4.69877815246582,
      "learning_rate": 9.177136042920344e-05,
      "loss": 4.8835,
      "step": 5737
    },
    {
      "epoch": 0.5738,
      "grad_norm": 3.6143646240234375,
      "learning_rate": 9.173657272404576e-05,
      "loss": 5.6103,
      "step": 5738
    },
    {
      "epoch": 0.5739,
      "grad_norm": 2.2903926372528076,
      "learning_rate": 9.17017860257616e-05,
      "loss": 4.622,
      "step": 5739
    },
    {
      "epoch": 0.574,
      "grad_norm": 3.0217838287353516,
      "learning_rate": 9.166700033858969e-05,
      "loss": 5.4284,
      "step": 5740
    },
    {
      "epoch": 0.5741,
      "grad_norm": 3.0407519340515137,
      "learning_rate": 9.163221566676847e-05,
      "loss": 5.3802,
      "step": 5741
    },
    {
      "epoch": 0.5742,
      "grad_norm": 4.632083415985107,
      "learning_rate": 9.159743201453638e-05,
      "loss": 4.7591,
      "step": 5742
    },
    {
      "epoch": 0.5743,
      "grad_norm": 3.445807695388794,
      "learning_rate": 9.156264938613174e-05,
      "loss": 4.5984,
      "step": 5743
    },
    {
      "epoch": 0.5744,
      "grad_norm": 2.682847738265991,
      "learning_rate": 9.152786778579267e-05,
      "loss": 4.4275,
      "step": 5744
    },
    {
      "epoch": 0.5745,
      "grad_norm": 5.584566116333008,
      "learning_rate": 9.14930872177572e-05,
      "loss": 4.613,
      "step": 5745
    },
    {
      "epoch": 0.5746,
      "grad_norm": 5.209883213043213,
      "learning_rate": 9.145830768626327e-05,
      "loss": 5.2392,
      "step": 5746
    },
    {
      "epoch": 0.5747,
      "grad_norm": 2.4683005809783936,
      "learning_rate": 9.142352919554862e-05,
      "loss": 4.8332,
      "step": 5747
    },
    {
      "epoch": 0.5748,
      "grad_norm": 1.9796162843704224,
      "learning_rate": 9.138875174985091e-05,
      "loss": 4.7099,
      "step": 5748
    },
    {
      "epoch": 0.5749,
      "grad_norm": 2.7786521911621094,
      "learning_rate": 9.135397535340773e-05,
      "loss": 5.3012,
      "step": 5749
    },
    {
      "epoch": 0.575,
      "grad_norm": 5.810111045837402,
      "learning_rate": 9.131920001045638e-05,
      "loss": 4.8171,
      "step": 5750
    },
    {
      "epoch": 0.5751,
      "grad_norm": 3.4948008060455322,
      "learning_rate": 9.128442572523417e-05,
      "loss": 4.3414,
      "step": 5751
    },
    {
      "epoch": 0.5752,
      "grad_norm": 3.0550591945648193,
      "learning_rate": 9.12496525019783e-05,
      "loss": 5.0068,
      "step": 5752
    },
    {
      "epoch": 0.5753,
      "grad_norm": 3.227062225341797,
      "learning_rate": 9.121488034492569e-05,
      "loss": 4.8815,
      "step": 5753
    },
    {
      "epoch": 0.5754,
      "grad_norm": 2.7805519104003906,
      "learning_rate": 9.11801092583133e-05,
      "loss": 4.1848,
      "step": 5754
    },
    {
      "epoch": 0.5755,
      "grad_norm": 3.8567309379577637,
      "learning_rate": 9.114533924637778e-05,
      "loss": 4.8448,
      "step": 5755
    },
    {
      "epoch": 0.5756,
      "grad_norm": 2.78935170173645,
      "learning_rate": 9.111057031335585e-05,
      "loss": 4.8279,
      "step": 5756
    },
    {
      "epoch": 0.5757,
      "grad_norm": 2.3062546253204346,
      "learning_rate": 9.107580246348395e-05,
      "loss": 4.7555,
      "step": 5757
    },
    {
      "epoch": 0.5758,
      "grad_norm": 3.2284393310546875,
      "learning_rate": 9.104103570099848e-05,
      "loss": 5.4264,
      "step": 5758
    },
    {
      "epoch": 0.5759,
      "grad_norm": 4.672474384307861,
      "learning_rate": 9.100627003013562e-05,
      "loss": 4.0802,
      "step": 5759
    },
    {
      "epoch": 0.576,
      "grad_norm": 4.467238903045654,
      "learning_rate": 9.097150545513145e-05,
      "loss": 4.9929,
      "step": 5760
    },
    {
      "epoch": 0.5761,
      "grad_norm": 3.4955992698669434,
      "learning_rate": 9.093674198022201e-05,
      "loss": 4.972,
      "step": 5761
    },
    {
      "epoch": 0.5762,
      "grad_norm": 2.3081514835357666,
      "learning_rate": 9.090197960964301e-05,
      "loss": 4.8156,
      "step": 5762
    },
    {
      "epoch": 0.5763,
      "grad_norm": 3.4781880378723145,
      "learning_rate": 9.086721834763024e-05,
      "loss": 4.5636,
      "step": 5763
    },
    {
      "epoch": 0.5764,
      "grad_norm": 3.77581787109375,
      "learning_rate": 9.083245819841918e-05,
      "loss": 5.2049,
      "step": 5764
    },
    {
      "epoch": 0.5765,
      "grad_norm": 2.1789255142211914,
      "learning_rate": 9.07976991662453e-05,
      "loss": 4.3106,
      "step": 5765
    },
    {
      "epoch": 0.5766,
      "grad_norm": 3.50543212890625,
      "learning_rate": 9.076294125534383e-05,
      "loss": 4.9792,
      "step": 5766
    },
    {
      "epoch": 0.5767,
      "grad_norm": 2.423548698425293,
      "learning_rate": 9.072818446994999e-05,
      "loss": 5.1476,
      "step": 5767
    },
    {
      "epoch": 0.5768,
      "grad_norm": 3.017194986343384,
      "learning_rate": 9.069342881429876e-05,
      "loss": 4.9879,
      "step": 5768
    },
    {
      "epoch": 0.5769,
      "grad_norm": 2.4076104164123535,
      "learning_rate": 9.065867429262496e-05,
      "loss": 5.0509,
      "step": 5769
    },
    {
      "epoch": 0.577,
      "grad_norm": 2.432236671447754,
      "learning_rate": 9.062392090916337e-05,
      "loss": 4.6923,
      "step": 5770
    },
    {
      "epoch": 0.5771,
      "grad_norm": 4.679635524749756,
      "learning_rate": 9.058916866814858e-05,
      "loss": 4.9822,
      "step": 5771
    },
    {
      "epoch": 0.5772,
      "grad_norm": 2.5118472576141357,
      "learning_rate": 9.055441757381506e-05,
      "loss": 4.6464,
      "step": 5772
    },
    {
      "epoch": 0.5773,
      "grad_norm": 4.099633693695068,
      "learning_rate": 9.051966763039707e-05,
      "loss": 4.603,
      "step": 5773
    },
    {
      "epoch": 0.5774,
      "grad_norm": 2.0654103755950928,
      "learning_rate": 9.048491884212884e-05,
      "loss": 5.0328,
      "step": 5774
    },
    {
      "epoch": 0.5775,
      "grad_norm": 2.5790724754333496,
      "learning_rate": 9.045017121324438e-05,
      "loss": 4.4814,
      "step": 5775
    },
    {
      "epoch": 0.5776,
      "grad_norm": 2.706047296524048,
      "learning_rate": 9.04154247479776e-05,
      "loss": 5.116,
      "step": 5776
    },
    {
      "epoch": 0.5777,
      "grad_norm": 2.220155715942383,
      "learning_rate": 9.038067945056227e-05,
      "loss": 4.6247,
      "step": 5777
    },
    {
      "epoch": 0.5778,
      "grad_norm": 1.9902760982513428,
      "learning_rate": 9.034593532523193e-05,
      "loss": 4.6762,
      "step": 5778
    },
    {
      "epoch": 0.5779,
      "grad_norm": 2.661778450012207,
      "learning_rate": 9.03111923762201e-05,
      "loss": 5.0751,
      "step": 5779
    },
    {
      "epoch": 0.578,
      "grad_norm": 1.7678375244140625,
      "learning_rate": 9.027645060776006e-05,
      "loss": 4.6206,
      "step": 5780
    },
    {
      "epoch": 0.5781,
      "grad_norm": 1.8947266340255737,
      "learning_rate": 9.024171002408506e-05,
      "loss": 5.094,
      "step": 5781
    },
    {
      "epoch": 0.5782,
      "grad_norm": 2.0835154056549072,
      "learning_rate": 9.020697062942807e-05,
      "loss": 4.4832,
      "step": 5782
    },
    {
      "epoch": 0.5783,
      "grad_norm": 2.5419044494628906,
      "learning_rate": 9.017223242802204e-05,
      "loss": 5.1045,
      "step": 5783
    },
    {
      "epoch": 0.5784,
      "grad_norm": 2.186368227005005,
      "learning_rate": 9.013749542409963e-05,
      "loss": 4.9282,
      "step": 5784
    },
    {
      "epoch": 0.5785,
      "grad_norm": 2.752305269241333,
      "learning_rate": 9.01027596218935e-05,
      "loss": 4.792,
      "step": 5785
    },
    {
      "epoch": 0.5786,
      "grad_norm": 2.811737298965454,
      "learning_rate": 9.006802502563612e-05,
      "loss": 4.5489,
      "step": 5786
    },
    {
      "epoch": 0.5787,
      "grad_norm": 3.9564168453216553,
      "learning_rate": 9.003329163955972e-05,
      "loss": 4.4651,
      "step": 5787
    },
    {
      "epoch": 0.5788,
      "grad_norm": 6.752963542938232,
      "learning_rate": 8.999855946789653e-05,
      "loss": 4.4145,
      "step": 5788
    },
    {
      "epoch": 0.5789,
      "grad_norm": 3.9631330966949463,
      "learning_rate": 8.99638285148785e-05,
      "loss": 4.4016,
      "step": 5789
    },
    {
      "epoch": 0.579,
      "grad_norm": 3.0115835666656494,
      "learning_rate": 8.992909878473758e-05,
      "loss": 5.816,
      "step": 5790
    },
    {
      "epoch": 0.5791,
      "grad_norm": 3.744901418685913,
      "learning_rate": 8.989437028170537e-05,
      "loss": 5.2389,
      "step": 5791
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.8727577924728394,
      "learning_rate": 8.985964301001353e-05,
      "loss": 4.3976,
      "step": 5792
    },
    {
      "epoch": 0.5793,
      "grad_norm": 3.1671087741851807,
      "learning_rate": 8.982491697389338e-05,
      "loss": 4.7437,
      "step": 5793
    },
    {
      "epoch": 0.5794,
      "grad_norm": 2.3158414363861084,
      "learning_rate": 8.979019217757625e-05,
      "loss": 4.6987,
      "step": 5794
    },
    {
      "epoch": 0.5795,
      "grad_norm": 3.0083534717559814,
      "learning_rate": 8.975546862529328e-05,
      "loss": 4.4301,
      "step": 5795
    },
    {
      "epoch": 0.5796,
      "grad_norm": 2.853339195251465,
      "learning_rate": 8.972074632127533e-05,
      "loss": 5.1708,
      "step": 5796
    },
    {
      "epoch": 0.5797,
      "grad_norm": 3.0535051822662354,
      "learning_rate": 8.96860252697533e-05,
      "loss": 4.2851,
      "step": 5797
    },
    {
      "epoch": 0.5798,
      "grad_norm": 2.4209468364715576,
      "learning_rate": 8.965130547495776e-05,
      "loss": 4.3514,
      "step": 5798
    },
    {
      "epoch": 0.5799,
      "grad_norm": 2.060234308242798,
      "learning_rate": 8.961658694111929e-05,
      "loss": 4.1556,
      "step": 5799
    },
    {
      "epoch": 0.58,
      "grad_norm": 2.463620662689209,
      "learning_rate": 8.958186967246816e-05,
      "loss": 4.434,
      "step": 5800
    },
    {
      "epoch": 0.5801,
      "grad_norm": 4.292128086090088,
      "learning_rate": 8.954715367323468e-05,
      "loss": 4.9122,
      "step": 5801
    },
    {
      "epoch": 0.5802,
      "grad_norm": 2.1726646423339844,
      "learning_rate": 8.951243894764876e-05,
      "loss": 4.5575,
      "step": 5802
    },
    {
      "epoch": 0.5803,
      "grad_norm": 2.2619471549987793,
      "learning_rate": 8.947772549994035e-05,
      "loss": 4.8239,
      "step": 5803
    },
    {
      "epoch": 0.5804,
      "grad_norm": 2.5940842628479004,
      "learning_rate": 8.944301333433922e-05,
      "loss": 4.6091,
      "step": 5804
    },
    {
      "epoch": 0.5805,
      "grad_norm": 2.259291172027588,
      "learning_rate": 8.940830245507483e-05,
      "loss": 4.7541,
      "step": 5805
    },
    {
      "epoch": 0.5806,
      "grad_norm": 2.8091423511505127,
      "learning_rate": 8.93735928663767e-05,
      "loss": 4.781,
      "step": 5806
    },
    {
      "epoch": 0.5807,
      "grad_norm": 5.5960774421691895,
      "learning_rate": 8.933888457247402e-05,
      "loss": 5.3168,
      "step": 5807
    },
    {
      "epoch": 0.5808,
      "grad_norm": 2.292833089828491,
      "learning_rate": 8.930417757759592e-05,
      "loss": 4.3472,
      "step": 5808
    },
    {
      "epoch": 0.5809,
      "grad_norm": 2.0597546100616455,
      "learning_rate": 8.926947188597134e-05,
      "loss": 4.5326,
      "step": 5809
    },
    {
      "epoch": 0.581,
      "grad_norm": 2.000739574432373,
      "learning_rate": 8.923476750182908e-05,
      "loss": 4.5159,
      "step": 5810
    },
    {
      "epoch": 0.5811,
      "grad_norm": 2.893308401107788,
      "learning_rate": 8.920006442939772e-05,
      "loss": 4.7607,
      "step": 5811
    },
    {
      "epoch": 0.5812,
      "grad_norm": 2.3997349739074707,
      "learning_rate": 8.916536267290578e-05,
      "loss": 4.4695,
      "step": 5812
    },
    {
      "epoch": 0.5813,
      "grad_norm": 2.064173936843872,
      "learning_rate": 8.913066223658151e-05,
      "loss": 4.3183,
      "step": 5813
    },
    {
      "epoch": 0.5814,
      "grad_norm": 2.7728893756866455,
      "learning_rate": 8.909596312465306e-05,
      "loss": 5.3245,
      "step": 5814
    },
    {
      "epoch": 0.5815,
      "grad_norm": 2.246063709259033,
      "learning_rate": 8.906126534134848e-05,
      "loss": 4.8477,
      "step": 5815
    },
    {
      "epoch": 0.5816,
      "grad_norm": 2.5084431171417236,
      "learning_rate": 8.902656889089548e-05,
      "loss": 5.3561,
      "step": 5816
    },
    {
      "epoch": 0.5817,
      "grad_norm": 3.5547614097595215,
      "learning_rate": 8.89918737775218e-05,
      "loss": 4.8052,
      "step": 5817
    },
    {
      "epoch": 0.5818,
      "grad_norm": 2.992640495300293,
      "learning_rate": 8.895718000545489e-05,
      "loss": 4.6099,
      "step": 5818
    },
    {
      "epoch": 0.5819,
      "grad_norm": 6.807314872741699,
      "learning_rate": 8.892248757892214e-05,
      "loss": 4.621,
      "step": 5819
    },
    {
      "epoch": 0.582,
      "grad_norm": 2.6076581478118896,
      "learning_rate": 8.888779650215068e-05,
      "loss": 4.7106,
      "step": 5820
    },
    {
      "epoch": 0.5821,
      "grad_norm": 2.0228142738342285,
      "learning_rate": 8.885310677936746e-05,
      "loss": 4.9188,
      "step": 5821
    },
    {
      "epoch": 0.5822,
      "grad_norm": 2.5634679794311523,
      "learning_rate": 8.88184184147994e-05,
      "loss": 5.028,
      "step": 5822
    },
    {
      "epoch": 0.5823,
      "grad_norm": 2.6093196868896484,
      "learning_rate": 8.878373141267311e-05,
      "loss": 4.8949,
      "step": 5823
    },
    {
      "epoch": 0.5824,
      "grad_norm": 2.4860992431640625,
      "learning_rate": 8.874904577721518e-05,
      "loss": 4.7129,
      "step": 5824
    },
    {
      "epoch": 0.5825,
      "grad_norm": 3.0705349445343018,
      "learning_rate": 8.871436151265184e-05,
      "loss": 5.2129,
      "step": 5825
    },
    {
      "epoch": 0.5826,
      "grad_norm": 4.530531406402588,
      "learning_rate": 8.867967862320934e-05,
      "loss": 5.0014,
      "step": 5826
    },
    {
      "epoch": 0.5827,
      "grad_norm": 3.5282325744628906,
      "learning_rate": 8.864499711311362e-05,
      "loss": 4.5636,
      "step": 5827
    },
    {
      "epoch": 0.5828,
      "grad_norm": 2.428561210632324,
      "learning_rate": 8.861031698659063e-05,
      "loss": 4.6504,
      "step": 5828
    },
    {
      "epoch": 0.5829,
      "grad_norm": 3.2698748111724854,
      "learning_rate": 8.857563824786596e-05,
      "loss": 4.4036,
      "step": 5829
    },
    {
      "epoch": 0.583,
      "grad_norm": 3.8956668376922607,
      "learning_rate": 8.854096090116508e-05,
      "loss": 5.5439,
      "step": 5830
    },
    {
      "epoch": 0.5831,
      "grad_norm": 3.6485400199890137,
      "learning_rate": 8.850628495071336e-05,
      "loss": 5.1522,
      "step": 5831
    },
    {
      "epoch": 0.5832,
      "grad_norm": 2.5850794315338135,
      "learning_rate": 8.847161040073594e-05,
      "loss": 4.3788,
      "step": 5832
    },
    {
      "epoch": 0.5833,
      "grad_norm": 2.478151559829712,
      "learning_rate": 8.843693725545786e-05,
      "loss": 4.6875,
      "step": 5833
    },
    {
      "epoch": 0.5834,
      "grad_norm": 2.663656234741211,
      "learning_rate": 8.840226551910387e-05,
      "loss": 4.531,
      "step": 5834
    },
    {
      "epoch": 0.5835,
      "grad_norm": 5.385288715362549,
      "learning_rate": 8.836759519589867e-05,
      "loss": 5.1244,
      "step": 5835
    },
    {
      "epoch": 0.5836,
      "grad_norm": 1.8924700021743774,
      "learning_rate": 8.833292629006668e-05,
      "loss": 4.5006,
      "step": 5836
    },
    {
      "epoch": 0.5837,
      "grad_norm": 4.528327941894531,
      "learning_rate": 8.829825880583226e-05,
      "loss": 5.5231,
      "step": 5837
    },
    {
      "epoch": 0.5838,
      "grad_norm": 3.398927688598633,
      "learning_rate": 8.826359274741953e-05,
      "loss": 4.7942,
      "step": 5838
    },
    {
      "epoch": 0.5839,
      "grad_norm": 1.9701274633407593,
      "learning_rate": 8.822892811905237e-05,
      "loss": 4.6807,
      "step": 5839
    },
    {
      "epoch": 0.584,
      "grad_norm": 2.7168354988098145,
      "learning_rate": 8.819426492495464e-05,
      "loss": 4.7666,
      "step": 5840
    },
    {
      "epoch": 0.5841,
      "grad_norm": 2.9048984050750732,
      "learning_rate": 8.81596031693499e-05,
      "loss": 4.8481,
      "step": 5841
    },
    {
      "epoch": 0.5842,
      "grad_norm": 3.0855140686035156,
      "learning_rate": 8.812494285646163e-05,
      "loss": 4.8509,
      "step": 5842
    },
    {
      "epoch": 0.5843,
      "grad_norm": 2.5303800106048584,
      "learning_rate": 8.809028399051302e-05,
      "loss": 4.8794,
      "step": 5843
    },
    {
      "epoch": 0.5844,
      "grad_norm": 2.790954828262329,
      "learning_rate": 8.805562657572723e-05,
      "loss": 4.9429,
      "step": 5844
    },
    {
      "epoch": 0.5845,
      "grad_norm": 2.684680461883545,
      "learning_rate": 8.802097061632705e-05,
      "loss": 4.642,
      "step": 5845
    },
    {
      "epoch": 0.5846,
      "grad_norm": 2.384114980697632,
      "learning_rate": 8.79863161165353e-05,
      "loss": 4.7861,
      "step": 5846
    },
    {
      "epoch": 0.5847,
      "grad_norm": 2.402505397796631,
      "learning_rate": 8.79516630805745e-05,
      "loss": 5.0022,
      "step": 5847
    },
    {
      "epoch": 0.5848,
      "grad_norm": 1.8748887777328491,
      "learning_rate": 8.791701151266696e-05,
      "loss": 4.5842,
      "step": 5848
    },
    {
      "epoch": 0.5849,
      "grad_norm": 3.811063766479492,
      "learning_rate": 8.788236141703498e-05,
      "loss": 4.6498,
      "step": 5849
    },
    {
      "epoch": 0.585,
      "grad_norm": 3.944868803024292,
      "learning_rate": 8.784771279790044e-05,
      "loss": 6.1665,
      "step": 5850
    },
    {
      "epoch": 0.5851,
      "grad_norm": 5.636005401611328,
      "learning_rate": 8.781306565948528e-05,
      "loss": 4.2308,
      "step": 5851
    },
    {
      "epoch": 0.5852,
      "grad_norm": 4.694180488586426,
      "learning_rate": 8.777842000601105e-05,
      "loss": 4.7253,
      "step": 5852
    },
    {
      "epoch": 0.5853,
      "grad_norm": 3.7290685176849365,
      "learning_rate": 8.774377584169933e-05,
      "loss": 4.6166,
      "step": 5853
    },
    {
      "epoch": 0.5854,
      "grad_norm": 2.6067535877227783,
      "learning_rate": 8.77091331707713e-05,
      "loss": 4.6906,
      "step": 5854
    },
    {
      "epoch": 0.5855,
      "grad_norm": 2.2246201038360596,
      "learning_rate": 8.767449199744814e-05,
      "loss": 4.3868,
      "step": 5855
    },
    {
      "epoch": 0.5856,
      "grad_norm": 4.03358268737793,
      "learning_rate": 8.763985232595075e-05,
      "loss": 5.4639,
      "step": 5856
    },
    {
      "epoch": 0.5857,
      "grad_norm": 2.2700252532958984,
      "learning_rate": 8.760521416049983e-05,
      "loss": 4.5678,
      "step": 5857
    },
    {
      "epoch": 0.5858,
      "grad_norm": 3.209782600402832,
      "learning_rate": 8.7570577505316e-05,
      "loss": 4.7676,
      "step": 5858
    },
    {
      "epoch": 0.5859,
      "grad_norm": 2.4438319206237793,
      "learning_rate": 8.753594236461957e-05,
      "loss": 4.88,
      "step": 5859
    },
    {
      "epoch": 0.586,
      "grad_norm": 4.40047025680542,
      "learning_rate": 8.750130874263077e-05,
      "loss": 4.7353,
      "step": 5860
    },
    {
      "epoch": 0.5861,
      "grad_norm": 3.3988280296325684,
      "learning_rate": 8.746667664356956e-05,
      "loss": 5.5363,
      "step": 5861
    },
    {
      "epoch": 0.5862,
      "grad_norm": 2.950096368789673,
      "learning_rate": 8.743204607165583e-05,
      "loss": 5.1642,
      "step": 5862
    },
    {
      "epoch": 0.5863,
      "grad_norm": 6.410384178161621,
      "learning_rate": 8.739741703110913e-05,
      "loss": 5.6011,
      "step": 5863
    },
    {
      "epoch": 0.5864,
      "grad_norm": 17.720970153808594,
      "learning_rate": 8.736278952614898e-05,
      "loss": 6.2707,
      "step": 5864
    },
    {
      "epoch": 0.5865,
      "grad_norm": 6.155170917510986,
      "learning_rate": 8.732816356099456e-05,
      "loss": 4.7339,
      "step": 5865
    },
    {
      "epoch": 0.5866,
      "grad_norm": 3.344444513320923,
      "learning_rate": 8.729353913986496e-05,
      "loss": 4.5542,
      "step": 5866
    },
    {
      "epoch": 0.5867,
      "grad_norm": 6.097910404205322,
      "learning_rate": 8.72589162669791e-05,
      "loss": 4.7185,
      "step": 5867
    },
    {
      "epoch": 0.5868,
      "grad_norm": 2.707233428955078,
      "learning_rate": 8.722429494655561e-05,
      "loss": 4.7694,
      "step": 5868
    },
    {
      "epoch": 0.5869,
      "grad_norm": 2.472458839416504,
      "learning_rate": 8.718967518281307e-05,
      "loss": 4.7473,
      "step": 5869
    },
    {
      "epoch": 0.587,
      "grad_norm": 2.217458486557007,
      "learning_rate": 8.715505697996971e-05,
      "loss": 5.1999,
      "step": 5870
    },
    {
      "epoch": 0.5871,
      "grad_norm": 2.1743006706237793,
      "learning_rate": 8.712044034224374e-05,
      "loss": 4.3371,
      "step": 5871
    },
    {
      "epoch": 0.5872,
      "grad_norm": 1.7569677829742432,
      "learning_rate": 8.708582527385301e-05,
      "loss": 4.4102,
      "step": 5872
    },
    {
      "epoch": 0.5873,
      "grad_norm": 3.4651291370391846,
      "learning_rate": 8.705121177901532e-05,
      "loss": 5.0331,
      "step": 5873
    },
    {
      "epoch": 0.5874,
      "grad_norm": 3.3023107051849365,
      "learning_rate": 8.70165998619482e-05,
      "loss": 5.6172,
      "step": 5874
    },
    {
      "epoch": 0.5875,
      "grad_norm": 2.286388874053955,
      "learning_rate": 8.698198952686896e-05,
      "loss": 4.9694,
      "step": 5875
    },
    {
      "epoch": 0.5876,
      "grad_norm": 4.126528263092041,
      "learning_rate": 8.694738077799488e-05,
      "loss": 5.0307,
      "step": 5876
    },
    {
      "epoch": 0.5877,
      "grad_norm": 2.903588056564331,
      "learning_rate": 8.691277361954279e-05,
      "loss": 5.0016,
      "step": 5877
    },
    {
      "epoch": 0.5878,
      "grad_norm": 4.102546215057373,
      "learning_rate": 8.687816805572956e-05,
      "loss": 5.3868,
      "step": 5878
    },
    {
      "epoch": 0.5879,
      "grad_norm": 2.0511927604675293,
      "learning_rate": 8.684356409077176e-05,
      "loss": 4.683,
      "step": 5879
    },
    {
      "epoch": 0.588,
      "grad_norm": 3.8065414428710938,
      "learning_rate": 8.680896172888576e-05,
      "loss": 4.7768,
      "step": 5880
    },
    {
      "epoch": 0.5881,
      "grad_norm": 2.688286781311035,
      "learning_rate": 8.677436097428775e-05,
      "loss": 5.1252,
      "step": 5881
    },
    {
      "epoch": 0.5882,
      "grad_norm": 2.6434707641601562,
      "learning_rate": 8.673976183119376e-05,
      "loss": 4.7296,
      "step": 5882
    },
    {
      "epoch": 0.5883,
      "grad_norm": 4.257566452026367,
      "learning_rate": 8.670516430381958e-05,
      "loss": 4.8122,
      "step": 5883
    },
    {
      "epoch": 0.5884,
      "grad_norm": 2.2576122283935547,
      "learning_rate": 8.667056839638075e-05,
      "loss": 5.0725,
      "step": 5884
    },
    {
      "epoch": 0.5885,
      "grad_norm": 2.346050500869751,
      "learning_rate": 8.663597411309279e-05,
      "loss": 4.8085,
      "step": 5885
    },
    {
      "epoch": 0.5886,
      "grad_norm": 5.626879692077637,
      "learning_rate": 8.660138145817079e-05,
      "loss": 4.7311,
      "step": 5886
    },
    {
      "epoch": 0.5887,
      "grad_norm": 2.0254733562469482,
      "learning_rate": 8.656679043582986e-05,
      "loss": 5.2596,
      "step": 5887
    },
    {
      "epoch": 0.5888,
      "grad_norm": 3.8574278354644775,
      "learning_rate": 8.653220105028474e-05,
      "loss": 4.7199,
      "step": 5888
    },
    {
      "epoch": 0.5889,
      "grad_norm": 3.5419304370880127,
      "learning_rate": 8.64976133057501e-05,
      "loss": 5.1111,
      "step": 5889
    },
    {
      "epoch": 0.589,
      "grad_norm": 3.4435248374938965,
      "learning_rate": 8.646302720644027e-05,
      "loss": 4.7874,
      "step": 5890
    },
    {
      "epoch": 0.5891,
      "grad_norm": 3.7451107501983643,
      "learning_rate": 8.642844275656957e-05,
      "loss": 5.0111,
      "step": 5891
    },
    {
      "epoch": 0.5892,
      "grad_norm": 10.21231746673584,
      "learning_rate": 8.639385996035194e-05,
      "loss": 5.329,
      "step": 5892
    },
    {
      "epoch": 0.5893,
      "grad_norm": 13.387052536010742,
      "learning_rate": 8.635927882200116e-05,
      "loss": 5.3711,
      "step": 5893
    },
    {
      "epoch": 0.5894,
      "grad_norm": 2.427917718887329,
      "learning_rate": 8.632469934573094e-05,
      "loss": 4.8195,
      "step": 5894
    },
    {
      "epoch": 0.5895,
      "grad_norm": 2.7691309452056885,
      "learning_rate": 8.629012153575458e-05,
      "loss": 5.5409,
      "step": 5895
    },
    {
      "epoch": 0.5896,
      "grad_norm": 2.7907841205596924,
      "learning_rate": 8.625554539628535e-05,
      "loss": 4.9327,
      "step": 5896
    },
    {
      "epoch": 0.5897,
      "grad_norm": 2.0589401721954346,
      "learning_rate": 8.62209709315362e-05,
      "loss": 4.6021,
      "step": 5897
    },
    {
      "epoch": 0.5898,
      "grad_norm": 3.848029851913452,
      "learning_rate": 8.618639814571996e-05,
      "loss": 4.933,
      "step": 5898
    },
    {
      "epoch": 0.5899,
      "grad_norm": 2.9761054515838623,
      "learning_rate": 8.615182704304918e-05,
      "loss": 4.7056,
      "step": 5899
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.350132703781128,
      "learning_rate": 8.611725762773631e-05,
      "loss": 4.5904,
      "step": 5900
    },
    {
      "epoch": 0.5901,
      "grad_norm": 2.462522506713867,
      "learning_rate": 8.608268990399349e-05,
      "loss": 4.5872,
      "step": 5901
    },
    {
      "epoch": 0.5902,
      "grad_norm": 4.710158824920654,
      "learning_rate": 8.604812387603265e-05,
      "loss": 4.7655,
      "step": 5902
    },
    {
      "epoch": 0.5903,
      "grad_norm": 4.275502681732178,
      "learning_rate": 8.601355954806561e-05,
      "loss": 4.2947,
      "step": 5903
    },
    {
      "epoch": 0.5904,
      "grad_norm": 3.1940150260925293,
      "learning_rate": 8.597899692430389e-05,
      "loss": 4.5996,
      "step": 5904
    },
    {
      "epoch": 0.5905,
      "grad_norm": 3.021012783050537,
      "learning_rate": 8.594443600895892e-05,
      "loss": 4.5051,
      "step": 5905
    },
    {
      "epoch": 0.5906,
      "grad_norm": 3.94128680229187,
      "learning_rate": 8.590987680624174e-05,
      "loss": 4.8949,
      "step": 5906
    },
    {
      "epoch": 0.5907,
      "grad_norm": 6.758218765258789,
      "learning_rate": 8.587531932036335e-05,
      "loss": 5.3219,
      "step": 5907
    },
    {
      "epoch": 0.5908,
      "grad_norm": 2.955117702484131,
      "learning_rate": 8.584076355553444e-05,
      "loss": 4.7326,
      "step": 5908
    },
    {
      "epoch": 0.5909,
      "grad_norm": 2.2485477924346924,
      "learning_rate": 8.580620951596557e-05,
      "loss": 4.7514,
      "step": 5909
    },
    {
      "epoch": 0.591,
      "grad_norm": 2.151196241378784,
      "learning_rate": 8.577165720586703e-05,
      "loss": 4.6279,
      "step": 5910
    },
    {
      "epoch": 0.5911,
      "grad_norm": 2.710581064224243,
      "learning_rate": 8.573710662944885e-05,
      "loss": 4.4584,
      "step": 5911
    },
    {
      "epoch": 0.5912,
      "grad_norm": 2.9923598766326904,
      "learning_rate": 8.570255779092098e-05,
      "loss": 5.0923,
      "step": 5912
    },
    {
      "epoch": 0.5913,
      "grad_norm": 2.5606110095977783,
      "learning_rate": 8.566801069449305e-05,
      "loss": 4.7858,
      "step": 5913
    },
    {
      "epoch": 0.5914,
      "grad_norm": 3.1194067001342773,
      "learning_rate": 8.56334653443746e-05,
      "loss": 5.3263,
      "step": 5914
    },
    {
      "epoch": 0.5915,
      "grad_norm": 3.1150436401367188,
      "learning_rate": 8.559892174477479e-05,
      "loss": 4.8942,
      "step": 5915
    },
    {
      "epoch": 0.5916,
      "grad_norm": 4.927653789520264,
      "learning_rate": 8.55643798999027e-05,
      "loss": 4.64,
      "step": 5916
    },
    {
      "epoch": 0.5917,
      "grad_norm": 2.804757595062256,
      "learning_rate": 8.55298398139671e-05,
      "loss": 5.4561,
      "step": 5917
    },
    {
      "epoch": 0.5918,
      "grad_norm": 4.709963321685791,
      "learning_rate": 8.549530149117664e-05,
      "loss": 4.4241,
      "step": 5918
    },
    {
      "epoch": 0.5919,
      "grad_norm": 3.9032628536224365,
      "learning_rate": 8.546076493573972e-05,
      "loss": 4.6303,
      "step": 5919
    },
    {
      "epoch": 0.592,
      "grad_norm": 3.246565341949463,
      "learning_rate": 8.542623015186445e-05,
      "loss": 5.1216,
      "step": 5920
    },
    {
      "epoch": 0.5921,
      "grad_norm": 3.022005081176758,
      "learning_rate": 8.539169714375885e-05,
      "loss": 4.8845,
      "step": 5921
    },
    {
      "epoch": 0.5922,
      "grad_norm": 4.715865612030029,
      "learning_rate": 8.535716591563062e-05,
      "loss": 4.3551,
      "step": 5922
    },
    {
      "epoch": 0.5923,
      "grad_norm": 3.1044342517852783,
      "learning_rate": 8.532263647168735e-05,
      "loss": 5.2408,
      "step": 5923
    },
    {
      "epoch": 0.5924,
      "grad_norm": 3.8912160396575928,
      "learning_rate": 8.528810881613626e-05,
      "loss": 4.8934,
      "step": 5924
    },
    {
      "epoch": 0.5925,
      "grad_norm": 2.9452359676361084,
      "learning_rate": 8.525358295318454e-05,
      "loss": 4.7441,
      "step": 5925
    },
    {
      "epoch": 0.5926,
      "grad_norm": 6.271330833435059,
      "learning_rate": 8.521905888703893e-05,
      "loss": 4.782,
      "step": 5926
    },
    {
      "epoch": 0.5927,
      "grad_norm": 3.394685745239258,
      "learning_rate": 8.51845366219062e-05,
      "loss": 5.3646,
      "step": 5927
    },
    {
      "epoch": 0.5928,
      "grad_norm": 5.3852858543396,
      "learning_rate": 8.515001616199279e-05,
      "loss": 4.822,
      "step": 5928
    },
    {
      "epoch": 0.5929,
      "grad_norm": 2.083998680114746,
      "learning_rate": 8.511549751150479e-05,
      "loss": 4.3682,
      "step": 5929
    },
    {
      "epoch": 0.593,
      "grad_norm": 2.4453330039978027,
      "learning_rate": 8.508098067464832e-05,
      "loss": 4.6662,
      "step": 5930
    },
    {
      "epoch": 0.5931,
      "grad_norm": 2.5372018814086914,
      "learning_rate": 8.504646565562906e-05,
      "loss": 4.766,
      "step": 5931
    },
    {
      "epoch": 0.5932,
      "grad_norm": 2.221017360687256,
      "learning_rate": 8.501195245865263e-05,
      "loss": 4.6534,
      "step": 5932
    },
    {
      "epoch": 0.5933,
      "grad_norm": 2.272307872772217,
      "learning_rate": 8.497744108792429e-05,
      "loss": 4.6078,
      "step": 5933
    },
    {
      "epoch": 0.5934,
      "grad_norm": 3.378390073776245,
      "learning_rate": 8.494293154764924e-05,
      "loss": 4.7746,
      "step": 5934
    },
    {
      "epoch": 0.5935,
      "grad_norm": 4.6569437980651855,
      "learning_rate": 8.490842384203225e-05,
      "loss": 5.1557,
      "step": 5935
    },
    {
      "epoch": 0.5936,
      "grad_norm": 3.259286642074585,
      "learning_rate": 8.487391797527808e-05,
      "loss": 5.3033,
      "step": 5936
    },
    {
      "epoch": 0.5937,
      "grad_norm": 2.711456298828125,
      "learning_rate": 8.483941395159114e-05,
      "loss": 5.1837,
      "step": 5937
    },
    {
      "epoch": 0.5938,
      "grad_norm": 2.650435209274292,
      "learning_rate": 8.480491177517557e-05,
      "loss": 4.6611,
      "step": 5938
    },
    {
      "epoch": 0.5939,
      "grad_norm": 4.374883651733398,
      "learning_rate": 8.477041145023546e-05,
      "loss": 4.7528,
      "step": 5939
    },
    {
      "epoch": 0.594,
      "grad_norm": 3.738973617553711,
      "learning_rate": 8.473591298097448e-05,
      "loss": 5.1814,
      "step": 5940
    },
    {
      "epoch": 0.5941,
      "grad_norm": 2.993853807449341,
      "learning_rate": 8.47014163715962e-05,
      "loss": 4.6536,
      "step": 5941
    },
    {
      "epoch": 0.5942,
      "grad_norm": 2.5986039638519287,
      "learning_rate": 8.466692162630392e-05,
      "loss": 4.738,
      "step": 5942
    },
    {
      "epoch": 0.5943,
      "grad_norm": 2.195176362991333,
      "learning_rate": 8.463242874930079e-05,
      "loss": 4.6148,
      "step": 5943
    },
    {
      "epoch": 0.5944,
      "grad_norm": 7.162106037139893,
      "learning_rate": 8.459793774478957e-05,
      "loss": 5.2595,
      "step": 5944
    },
    {
      "epoch": 0.5945,
      "grad_norm": 2.910146474838257,
      "learning_rate": 8.456344861697289e-05,
      "loss": 5.3986,
      "step": 5945
    },
    {
      "epoch": 0.5946,
      "grad_norm": 2.799480438232422,
      "learning_rate": 8.452896137005321e-05,
      "loss": 5.0835,
      "step": 5946
    },
    {
      "epoch": 0.5947,
      "grad_norm": 2.313985824584961,
      "learning_rate": 8.449447600823262e-05,
      "loss": 4.3918,
      "step": 5947
    },
    {
      "epoch": 0.5948,
      "grad_norm": 3.363571882247925,
      "learning_rate": 8.445999253571315e-05,
      "loss": 4.9979,
      "step": 5948
    },
    {
      "epoch": 0.5949,
      "grad_norm": 2.4080681800842285,
      "learning_rate": 8.442551095669639e-05,
      "loss": 5.1659,
      "step": 5949
    },
    {
      "epoch": 0.595,
      "grad_norm": 3.71187686920166,
      "learning_rate": 8.439103127538392e-05,
      "loss": 5.9762,
      "step": 5950
    },
    {
      "epoch": 0.5951,
      "grad_norm": 3.4272572994232178,
      "learning_rate": 8.435655349597689e-05,
      "loss": 5.0928,
      "step": 5951
    },
    {
      "epoch": 0.5952,
      "grad_norm": 3.276780366897583,
      "learning_rate": 8.432207762267644e-05,
      "loss": 4.5658,
      "step": 5952
    },
    {
      "epoch": 0.5953,
      "grad_norm": 3.7437314987182617,
      "learning_rate": 8.428760365968326e-05,
      "loss": 4.3639,
      "step": 5953
    },
    {
      "epoch": 0.5954,
      "grad_norm": 4.74917459487915,
      "learning_rate": 8.425313161119787e-05,
      "loss": 5.3667,
      "step": 5954
    },
    {
      "epoch": 0.5955,
      "grad_norm": 1.877126932144165,
      "learning_rate": 8.421866148142066e-05,
      "loss": 4.7854,
      "step": 5955
    },
    {
      "epoch": 0.5956,
      "grad_norm": 2.7228314876556396,
      "learning_rate": 8.418419327455164e-05,
      "loss": 5.0414,
      "step": 5956
    },
    {
      "epoch": 0.5957,
      "grad_norm": 3.702575922012329,
      "learning_rate": 8.414972699479075e-05,
      "loss": 5.4864,
      "step": 5957
    },
    {
      "epoch": 0.5958,
      "grad_norm": 2.369008779525757,
      "learning_rate": 8.41152626463375e-05,
      "loss": 4.8968,
      "step": 5958
    },
    {
      "epoch": 0.5959,
      "grad_norm": 3.974276542663574,
      "learning_rate": 8.408080023339133e-05,
      "loss": 4.8061,
      "step": 5959
    },
    {
      "epoch": 0.596,
      "grad_norm": 3.3229401111602783,
      "learning_rate": 8.404633976015134e-05,
      "loss": 4.8795,
      "step": 5960
    },
    {
      "epoch": 0.5961,
      "grad_norm": 2.340440273284912,
      "learning_rate": 8.401188123081653e-05,
      "loss": 4.6189,
      "step": 5961
    },
    {
      "epoch": 0.5962,
      "grad_norm": 4.720389366149902,
      "learning_rate": 8.397742464958547e-05,
      "loss": 5.3293,
      "step": 5962
    },
    {
      "epoch": 0.5963,
      "grad_norm": 2.0803613662719727,
      "learning_rate": 8.394297002065658e-05,
      "loss": 4.8442,
      "step": 5963
    },
    {
      "epoch": 0.5964,
      "grad_norm": 2.475350856781006,
      "learning_rate": 8.390851734822808e-05,
      "loss": 4.6848,
      "step": 5964
    },
    {
      "epoch": 0.5965,
      "grad_norm": 2.1117618083953857,
      "learning_rate": 8.387406663649795e-05,
      "loss": 4.9011,
      "step": 5965
    },
    {
      "epoch": 0.5966,
      "grad_norm": 3.3361258506774902,
      "learning_rate": 8.383961788966391e-05,
      "loss": 4.7719,
      "step": 5966
    },
    {
      "epoch": 0.5967,
      "grad_norm": 5.506298065185547,
      "learning_rate": 8.380517111192337e-05,
      "loss": 5.6093,
      "step": 5967
    },
    {
      "epoch": 0.5968,
      "grad_norm": 4.249314785003662,
      "learning_rate": 8.377072630747364e-05,
      "loss": 4.7379,
      "step": 5968
    },
    {
      "epoch": 0.5969,
      "grad_norm": 2.842495918273926,
      "learning_rate": 8.373628348051165e-05,
      "loss": 4.8133,
      "step": 5969
    },
    {
      "epoch": 0.597,
      "grad_norm": 3.0106747150421143,
      "learning_rate": 8.37018426352342e-05,
      "loss": 4.883,
      "step": 5970
    },
    {
      "epoch": 0.5971,
      "grad_norm": 2.4069252014160156,
      "learning_rate": 8.366740377583781e-05,
      "loss": 4.7185,
      "step": 5971
    },
    {
      "epoch": 0.5972,
      "grad_norm": 2.4635822772979736,
      "learning_rate": 8.363296690651868e-05,
      "loss": 4.7176,
      "step": 5972
    },
    {
      "epoch": 0.5973,
      "grad_norm": 2.1803791522979736,
      "learning_rate": 8.359853203147291e-05,
      "loss": 4.7833,
      "step": 5973
    },
    {
      "epoch": 0.5974,
      "grad_norm": 2.6739373207092285,
      "learning_rate": 8.356409915489625e-05,
      "loss": 4.714,
      "step": 5974
    },
    {
      "epoch": 0.5975,
      "grad_norm": 2.9635329246520996,
      "learning_rate": 8.352966828098428e-05,
      "loss": 4.8611,
      "step": 5975
    },
    {
      "epoch": 0.5976,
      "grad_norm": 3.0275683403015137,
      "learning_rate": 8.349523941393224e-05,
      "loss": 4.4739,
      "step": 5976
    },
    {
      "epoch": 0.5977,
      "grad_norm": 5.879847526550293,
      "learning_rate": 8.346081255793525e-05,
      "loss": 5.5408,
      "step": 5977
    },
    {
      "epoch": 0.5978,
      "grad_norm": 2.74707293510437,
      "learning_rate": 8.342638771718802e-05,
      "loss": 4.605,
      "step": 5978
    },
    {
      "epoch": 0.5979,
      "grad_norm": 2.2691259384155273,
      "learning_rate": 8.339196489588523e-05,
      "loss": 4.583,
      "step": 5979
    },
    {
      "epoch": 0.598,
      "grad_norm": 3.02341628074646,
      "learning_rate": 8.335754409822114e-05,
      "loss": 4.4818,
      "step": 5980
    },
    {
      "epoch": 0.5981,
      "grad_norm": 2.6564557552337646,
      "learning_rate": 8.332312532838978e-05,
      "loss": 4.7736,
      "step": 5981
    },
    {
      "epoch": 0.5982,
      "grad_norm": 5.776185035705566,
      "learning_rate": 8.328870859058506e-05,
      "loss": 5.3212,
      "step": 5982
    },
    {
      "epoch": 0.5983,
      "grad_norm": 4.05344295501709,
      "learning_rate": 8.325429388900046e-05,
      "loss": 4.8979,
      "step": 5983
    },
    {
      "epoch": 0.5984,
      "grad_norm": 2.7725651264190674,
      "learning_rate": 8.321988122782937e-05,
      "loss": 4.7017,
      "step": 5984
    },
    {
      "epoch": 0.5985,
      "grad_norm": 5.261909484863281,
      "learning_rate": 8.318547061126485e-05,
      "loss": 4.6129,
      "step": 5985
    },
    {
      "epoch": 0.5986,
      "grad_norm": 4.40648078918457,
      "learning_rate": 8.315106204349976e-05,
      "loss": 5.4805,
      "step": 5986
    },
    {
      "epoch": 0.5987,
      "grad_norm": 2.681908130645752,
      "learning_rate": 8.311665552872662e-05,
      "loss": 5.1418,
      "step": 5987
    },
    {
      "epoch": 0.5988,
      "grad_norm": 2.7893364429473877,
      "learning_rate": 8.30822510711378e-05,
      "loss": 4.7848,
      "step": 5988
    },
    {
      "epoch": 0.5989,
      "grad_norm": 3.3576998710632324,
      "learning_rate": 8.30478486749254e-05,
      "loss": 5.1112,
      "step": 5989
    },
    {
      "epoch": 0.599,
      "grad_norm": 2.1178743839263916,
      "learning_rate": 8.301344834428116e-05,
      "loss": 4.9786,
      "step": 5990
    },
    {
      "epoch": 0.5991,
      "grad_norm": 2.364778518676758,
      "learning_rate": 8.297905008339677e-05,
      "loss": 4.6566,
      "step": 5991
    },
    {
      "epoch": 0.5992,
      "grad_norm": 3.3462953567504883,
      "learning_rate": 8.294465389646345e-05,
      "loss": 4.8496,
      "step": 5992
    },
    {
      "epoch": 0.5993,
      "grad_norm": 2.6299054622650146,
      "learning_rate": 8.291025978767235e-05,
      "loss": 4.9874,
      "step": 5993
    },
    {
      "epoch": 0.5994,
      "grad_norm": 2.0550808906555176,
      "learning_rate": 8.287586776121423e-05,
      "loss": 4.5633,
      "step": 5994
    },
    {
      "epoch": 0.5995,
      "grad_norm": 2.8723244667053223,
      "learning_rate": 8.284147782127971e-05,
      "loss": 4.7934,
      "step": 5995
    },
    {
      "epoch": 0.5996,
      "grad_norm": 3.7339541912078857,
      "learning_rate": 8.280708997205904e-05,
      "loss": 5.8897,
      "step": 5996
    },
    {
      "epoch": 0.5997,
      "grad_norm": 3.1707911491394043,
      "learning_rate": 8.277270421774235e-05,
      "loss": 5.3423,
      "step": 5997
    },
    {
      "epoch": 0.5998,
      "grad_norm": 3.0164248943328857,
      "learning_rate": 8.273832056251937e-05,
      "loss": 4.9579,
      "step": 5998
    },
    {
      "epoch": 0.5999,
      "grad_norm": 1.849013328552246,
      "learning_rate": 8.270393901057964e-05,
      "loss": 4.4968,
      "step": 5999
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.734429359436035,
      "learning_rate": 8.266955956611253e-05,
      "loss": 5.1326,
      "step": 6000
    },
    {
      "epoch": 0.6001,
      "grad_norm": 2.619291067123413,
      "learning_rate": 8.263518223330697e-05,
      "loss": 4.7825,
      "step": 6001
    },
    {
      "epoch": 0.6002,
      "grad_norm": 3.202860116958618,
      "learning_rate": 8.26008070163518e-05,
      "loss": 4.4302,
      "step": 6002
    },
    {
      "epoch": 0.6003,
      "grad_norm": 4.227849960327148,
      "learning_rate": 8.256643391943551e-05,
      "loss": 5.0048,
      "step": 6003
    },
    {
      "epoch": 0.6004,
      "grad_norm": 2.1887154579162598,
      "learning_rate": 8.25320629467464e-05,
      "loss": 4.7402,
      "step": 6004
    },
    {
      "epoch": 0.6005,
      "grad_norm": 2.5022358894348145,
      "learning_rate": 8.249769410247239e-05,
      "loss": 5.0842,
      "step": 6005
    },
    {
      "epoch": 0.6006,
      "grad_norm": 7.583329677581787,
      "learning_rate": 8.24633273908013e-05,
      "loss": 5.6681,
      "step": 6006
    },
    {
      "epoch": 0.6007,
      "grad_norm": 3.0328378677368164,
      "learning_rate": 8.242896281592057e-05,
      "loss": 4.7987,
      "step": 6007
    },
    {
      "epoch": 0.6008,
      "grad_norm": 2.184514284133911,
      "learning_rate": 8.239460038201739e-05,
      "loss": 4.2254,
      "step": 6008
    },
    {
      "epoch": 0.6009,
      "grad_norm": 3.341282367706299,
      "learning_rate": 8.236024009327879e-05,
      "loss": 4.5687,
      "step": 6009
    },
    {
      "epoch": 0.601,
      "grad_norm": 5.207787036895752,
      "learning_rate": 8.23258819538914e-05,
      "loss": 5.1983,
      "step": 6010
    },
    {
      "epoch": 0.6011,
      "grad_norm": 3.4031941890716553,
      "learning_rate": 8.229152596804168e-05,
      "loss": 5.7035,
      "step": 6011
    },
    {
      "epoch": 0.6012,
      "grad_norm": 6.642244815826416,
      "learning_rate": 8.225717213991579e-05,
      "loss": 4.8635,
      "step": 6012
    },
    {
      "epoch": 0.6013,
      "grad_norm": 3.3394572734832764,
      "learning_rate": 8.222282047369971e-05,
      "loss": 4.5747,
      "step": 6013
    },
    {
      "epoch": 0.6014,
      "grad_norm": 9.04980182647705,
      "learning_rate": 8.218847097357898e-05,
      "loss": 5.243,
      "step": 6014
    },
    {
      "epoch": 0.6015,
      "grad_norm": 3.3485662937164307,
      "learning_rate": 8.215412364373907e-05,
      "loss": 4.6699,
      "step": 6015
    },
    {
      "epoch": 0.6016,
      "grad_norm": 5.000813961029053,
      "learning_rate": 8.211977848836506e-05,
      "loss": 4.5886,
      "step": 6016
    },
    {
      "epoch": 0.6017,
      "grad_norm": 6.920512676239014,
      "learning_rate": 8.208543551164178e-05,
      "loss": 4.6967,
      "step": 6017
    },
    {
      "epoch": 0.6018,
      "grad_norm": 3.1889517307281494,
      "learning_rate": 8.205109471775387e-05,
      "loss": 4.9354,
      "step": 6018
    },
    {
      "epoch": 0.6019,
      "grad_norm": 6.430717468261719,
      "learning_rate": 8.201675611088558e-05,
      "loss": 5.1134,
      "step": 6019
    },
    {
      "epoch": 0.602,
      "grad_norm": 3.850752592086792,
      "learning_rate": 8.198241969522107e-05,
      "loss": 5.3897,
      "step": 6020
    },
    {
      "epoch": 0.6021,
      "grad_norm": 3.1208248138427734,
      "learning_rate": 8.194808547494401e-05,
      "loss": 5.1653,
      "step": 6021
    },
    {
      "epoch": 0.6022,
      "grad_norm": 2.5546507835388184,
      "learning_rate": 8.191375345423799e-05,
      "loss": 5.0169,
      "step": 6022
    },
    {
      "epoch": 0.6023,
      "grad_norm": 2.250856399536133,
      "learning_rate": 8.187942363728625e-05,
      "loss": 4.6374,
      "step": 6023
    },
    {
      "epoch": 0.6024,
      "grad_norm": 2.6492257118225098,
      "learning_rate": 8.184509602827181e-05,
      "loss": 4.7144,
      "step": 6024
    },
    {
      "epoch": 0.6025,
      "grad_norm": 3.345231294631958,
      "learning_rate": 8.181077063137733e-05,
      "loss": 4.702,
      "step": 6025
    },
    {
      "epoch": 0.6026,
      "grad_norm": 2.4408531188964844,
      "learning_rate": 8.177644745078526e-05,
      "loss": 4.6804,
      "step": 6026
    },
    {
      "epoch": 0.6027,
      "grad_norm": 2.797683000564575,
      "learning_rate": 8.174212649067781e-05,
      "loss": 4.437,
      "step": 6027
    },
    {
      "epoch": 0.6028,
      "grad_norm": 3.501603364944458,
      "learning_rate": 8.170780775523684e-05,
      "loss": 5.268,
      "step": 6028
    },
    {
      "epoch": 0.6029,
      "grad_norm": 2.99406099319458,
      "learning_rate": 8.167349124864405e-05,
      "loss": 5.2019,
      "step": 6029
    },
    {
      "epoch": 0.603,
      "grad_norm": 1.9670592546463013,
      "learning_rate": 8.163917697508072e-05,
      "loss": 4.7427,
      "step": 6030
    },
    {
      "epoch": 0.6031,
      "grad_norm": 1.8723574876785278,
      "learning_rate": 8.160486493872798e-05,
      "loss": 4.995,
      "step": 6031
    },
    {
      "epoch": 0.6032,
      "grad_norm": 4.018259525299072,
      "learning_rate": 8.157055514376666e-05,
      "loss": 4.8669,
      "step": 6032
    },
    {
      "epoch": 0.6033,
      "grad_norm": 2.0495309829711914,
      "learning_rate": 8.153624759437732e-05,
      "loss": 4.4008,
      "step": 6033
    },
    {
      "epoch": 0.6034,
      "grad_norm": 2.2563836574554443,
      "learning_rate": 8.15019422947402e-05,
      "loss": 4.4746,
      "step": 6034
    },
    {
      "epoch": 0.6035,
      "grad_norm": 4.534444808959961,
      "learning_rate": 8.146763924903527e-05,
      "loss": 4.5443,
      "step": 6035
    },
    {
      "epoch": 0.6036,
      "grad_norm": 2.7409627437591553,
      "learning_rate": 8.14333384614423e-05,
      "loss": 5.1959,
      "step": 6036
    },
    {
      "epoch": 0.6037,
      "grad_norm": 6.1300458908081055,
      "learning_rate": 8.139903993614068e-05,
      "loss": 5.2132,
      "step": 6037
    },
    {
      "epoch": 0.6038,
      "grad_norm": 2.7449753284454346,
      "learning_rate": 8.136474367730969e-05,
      "loss": 4.3007,
      "step": 6038
    },
    {
      "epoch": 0.6039,
      "grad_norm": 4.0100016593933105,
      "learning_rate": 8.133044968912811e-05,
      "loss": 4.7347,
      "step": 6039
    },
    {
      "epoch": 0.604,
      "grad_norm": 2.7338621616363525,
      "learning_rate": 8.129615797577461e-05,
      "loss": 4.6907,
      "step": 6040
    },
    {
      "epoch": 0.6041,
      "grad_norm": 1.6686655282974243,
      "learning_rate": 8.126186854142752e-05,
      "loss": 4.6308,
      "step": 6041
    },
    {
      "epoch": 0.6042,
      "grad_norm": 2.1698601245880127,
      "learning_rate": 8.122758139026495e-05,
      "loss": 4.5803,
      "step": 6042
    },
    {
      "epoch": 0.6043,
      "grad_norm": 2.7327256202697754,
      "learning_rate": 8.119329652646463e-05,
      "loss": 4.655,
      "step": 6043
    },
    {
      "epoch": 0.6044,
      "grad_norm": 2.077510118484497,
      "learning_rate": 8.115901395420407e-05,
      "loss": 4.6439,
      "step": 6044
    },
    {
      "epoch": 0.6045,
      "grad_norm": 3.1080386638641357,
      "learning_rate": 8.11247336776605e-05,
      "loss": 4.8053,
      "step": 6045
    },
    {
      "epoch": 0.6046,
      "grad_norm": 4.294732570648193,
      "learning_rate": 8.109045570101086e-05,
      "loss": 4.7166,
      "step": 6046
    },
    {
      "epoch": 0.6047,
      "grad_norm": 3.4703891277313232,
      "learning_rate": 8.105618002843189e-05,
      "loss": 5.4449,
      "step": 6047
    },
    {
      "epoch": 0.6048,
      "grad_norm": 2.4390132427215576,
      "learning_rate": 8.102190666409987e-05,
      "loss": 4.6202,
      "step": 6048
    },
    {
      "epoch": 0.6049,
      "grad_norm": 4.364217281341553,
      "learning_rate": 8.0987635612191e-05,
      "loss": 5.0138,
      "step": 6049
    },
    {
      "epoch": 0.605,
      "grad_norm": 3.2460174560546875,
      "learning_rate": 8.095336687688102e-05,
      "loss": 5.3094,
      "step": 6050
    },
    {
      "epoch": 0.6051,
      "grad_norm": 5.1141486167907715,
      "learning_rate": 8.091910046234552e-05,
      "loss": 5.2067,
      "step": 6051
    },
    {
      "epoch": 0.6052,
      "grad_norm": 2.393165111541748,
      "learning_rate": 8.088483637275979e-05,
      "loss": 4.1374,
      "step": 6052
    },
    {
      "epoch": 0.6053,
      "grad_norm": 2.1886229515075684,
      "learning_rate": 8.085057461229872e-05,
      "loss": 4.736,
      "step": 6053
    },
    {
      "epoch": 0.6054,
      "grad_norm": 2.018340587615967,
      "learning_rate": 8.081631518513704e-05,
      "loss": 5.1262,
      "step": 6054
    },
    {
      "epoch": 0.6055,
      "grad_norm": 2.513336658477783,
      "learning_rate": 8.078205809544917e-05,
      "loss": 4.4991,
      "step": 6055
    },
    {
      "epoch": 0.6056,
      "grad_norm": 3.6070892810821533,
      "learning_rate": 8.074780334740928e-05,
      "loss": 5.122,
      "step": 6056
    },
    {
      "epoch": 0.6057,
      "grad_norm": 3.2908129692077637,
      "learning_rate": 8.071355094519109e-05,
      "loss": 5.0814,
      "step": 6057
    },
    {
      "epoch": 0.6058,
      "grad_norm": 3.780653238296509,
      "learning_rate": 8.067930089296827e-05,
      "loss": 5.0501,
      "step": 6058
    },
    {
      "epoch": 0.6059,
      "grad_norm": 3.306563377380371,
      "learning_rate": 8.064505319491398e-05,
      "loss": 4.4448,
      "step": 6059
    },
    {
      "epoch": 0.606,
      "grad_norm": 2.4433417320251465,
      "learning_rate": 8.061080785520126e-05,
      "loss": 5.0629,
      "step": 6060
    },
    {
      "epoch": 0.6061,
      "grad_norm": 5.828726291656494,
      "learning_rate": 8.057656487800282e-05,
      "loss": 4.5286,
      "step": 6061
    },
    {
      "epoch": 0.6062,
      "grad_norm": 2.5587997436523438,
      "learning_rate": 8.0542324267491e-05,
      "loss": 4.3576,
      "step": 6062
    },
    {
      "epoch": 0.6063,
      "grad_norm": 8.371116638183594,
      "learning_rate": 8.050808602783795e-05,
      "loss": 5.1756,
      "step": 6063
    },
    {
      "epoch": 0.6064,
      "grad_norm": 2.234090566635132,
      "learning_rate": 8.047385016321552e-05,
      "loss": 4.8621,
      "step": 6064
    },
    {
      "epoch": 0.6065,
      "grad_norm": 3.369337558746338,
      "learning_rate": 8.04396166777952e-05,
      "loss": 5.0046,
      "step": 6065
    },
    {
      "epoch": 0.6066,
      "grad_norm": 2.0262796878814697,
      "learning_rate": 8.040538557574822e-05,
      "loss": 4.5982,
      "step": 6066
    },
    {
      "epoch": 0.6067,
      "grad_norm": 6.77976655960083,
      "learning_rate": 8.037115686124564e-05,
      "loss": 4.6815,
      "step": 6067
    },
    {
      "epoch": 0.6068,
      "grad_norm": 2.014012336730957,
      "learning_rate": 8.033693053845801e-05,
      "loss": 4.8733,
      "step": 6068
    },
    {
      "epoch": 0.6069,
      "grad_norm": 2.4215517044067383,
      "learning_rate": 8.030270661155574e-05,
      "loss": 4.9767,
      "step": 6069
    },
    {
      "epoch": 0.607,
      "grad_norm": 2.9919793605804443,
      "learning_rate": 8.026848508470897e-05,
      "loss": 4.7789,
      "step": 6070
    },
    {
      "epoch": 0.6071,
      "grad_norm": 2.2420718669891357,
      "learning_rate": 8.023426596208739e-05,
      "loss": 4.76,
      "step": 6071
    },
    {
      "epoch": 0.6072,
      "grad_norm": 2.7339634895324707,
      "learning_rate": 8.020004924786059e-05,
      "loss": 4.491,
      "step": 6072
    },
    {
      "epoch": 0.6073,
      "grad_norm": 3.7877814769744873,
      "learning_rate": 8.016583494619769e-05,
      "loss": 5.617,
      "step": 6073
    },
    {
      "epoch": 0.6074,
      "grad_norm": 4.947831630706787,
      "learning_rate": 8.013162306126765e-05,
      "loss": 6.085,
      "step": 6074
    },
    {
      "epoch": 0.6075,
      "grad_norm": 4.091269493103027,
      "learning_rate": 8.009741359723906e-05,
      "loss": 5.1606,
      "step": 6075
    },
    {
      "epoch": 0.6076,
      "grad_norm": 2.8044893741607666,
      "learning_rate": 8.00632065582803e-05,
      "loss": 4.2961,
      "step": 6076
    },
    {
      "epoch": 0.6077,
      "grad_norm": 2.861511707305908,
      "learning_rate": 8.002900194855932e-05,
      "loss": 4.6913,
      "step": 6077
    },
    {
      "epoch": 0.6078,
      "grad_norm": 3.0967235565185547,
      "learning_rate": 7.999479977224384e-05,
      "loss": 4.4756,
      "step": 6078
    },
    {
      "epoch": 0.6079,
      "grad_norm": 2.2313690185546875,
      "learning_rate": 7.996060003350139e-05,
      "loss": 4.9278,
      "step": 6079
    },
    {
      "epoch": 0.608,
      "grad_norm": 2.7323412895202637,
      "learning_rate": 7.992640273649898e-05,
      "loss": 4.4564,
      "step": 6080
    },
    {
      "epoch": 0.6081,
      "grad_norm": 3.90022611618042,
      "learning_rate": 7.989220788540355e-05,
      "loss": 4.9071,
      "step": 6081
    },
    {
      "epoch": 0.6082,
      "grad_norm": 2.1617109775543213,
      "learning_rate": 7.985801548438157e-05,
      "loss": 4.5843,
      "step": 6082
    },
    {
      "epoch": 0.6083,
      "grad_norm": 2.1188628673553467,
      "learning_rate": 7.982382553759931e-05,
      "loss": 4.2885,
      "step": 6083
    },
    {
      "epoch": 0.6084,
      "grad_norm": 2.76750111579895,
      "learning_rate": 7.97896380492227e-05,
      "loss": 5.2685,
      "step": 6084
    },
    {
      "epoch": 0.6085,
      "grad_norm": 2.401951551437378,
      "learning_rate": 7.975545302341743e-05,
      "loss": 4.9603,
      "step": 6085
    },
    {
      "epoch": 0.6086,
      "grad_norm": 2.2924089431762695,
      "learning_rate": 7.972127046434878e-05,
      "loss": 4.5975,
      "step": 6086
    },
    {
      "epoch": 0.6087,
      "grad_norm": 3.980842351913452,
      "learning_rate": 7.96870903761818e-05,
      "loss": 4.5412,
      "step": 6087
    },
    {
      "epoch": 0.6088,
      "grad_norm": 2.9758217334747314,
      "learning_rate": 7.965291276308124e-05,
      "loss": 5.3031,
      "step": 6088
    },
    {
      "epoch": 0.6089,
      "grad_norm": 2.750027894973755,
      "learning_rate": 7.961873762921153e-05,
      "loss": 4.7115,
      "step": 6089
    },
    {
      "epoch": 0.609,
      "grad_norm": 2.1231467723846436,
      "learning_rate": 7.958456497873685e-05,
      "loss": 4.3606,
      "step": 6090
    },
    {
      "epoch": 0.6091,
      "grad_norm": 4.205121994018555,
      "learning_rate": 7.955039481582097e-05,
      "loss": 4.9814,
      "step": 6091
    },
    {
      "epoch": 0.6092,
      "grad_norm": 2.888401746749878,
      "learning_rate": 7.951622714462746e-05,
      "loss": 4.7687,
      "step": 6092
    },
    {
      "epoch": 0.6093,
      "grad_norm": 2.6875534057617188,
      "learning_rate": 7.948206196931954e-05,
      "loss": 4.6539,
      "step": 6093
    },
    {
      "epoch": 0.6094,
      "grad_norm": 1.7904789447784424,
      "learning_rate": 7.944789929406016e-05,
      "loss": 4.8092,
      "step": 6094
    },
    {
      "epoch": 0.6095,
      "grad_norm": 2.6260712146759033,
      "learning_rate": 7.941373912301189e-05,
      "loss": 4.6266,
      "step": 6095
    },
    {
      "epoch": 0.6096,
      "grad_norm": 3.4907281398773193,
      "learning_rate": 7.937958146033705e-05,
      "loss": 5.7927,
      "step": 6096
    },
    {
      "epoch": 0.6097,
      "grad_norm": 2.044506311416626,
      "learning_rate": 7.934542631019768e-05,
      "loss": 4.7861,
      "step": 6097
    },
    {
      "epoch": 0.6098,
      "grad_norm": 7.942724227905273,
      "learning_rate": 7.931127367675543e-05,
      "loss": 4.9437,
      "step": 6098
    },
    {
      "epoch": 0.6099,
      "grad_norm": 3.76773738861084,
      "learning_rate": 7.927712356417176e-05,
      "loss": 4.6884,
      "step": 6099
    },
    {
      "epoch": 0.61,
      "grad_norm": 2.523064374923706,
      "learning_rate": 7.92429759766077e-05,
      "loss": 4.6998,
      "step": 6100
    },
    {
      "epoch": 0.6101,
      "grad_norm": 2.4878458976745605,
      "learning_rate": 7.920883091822408e-05,
      "loss": 5.2197,
      "step": 6101
    },
    {
      "epoch": 0.6102,
      "grad_norm": 4.298656463623047,
      "learning_rate": 7.917468839318132e-05,
      "loss": 4.6654,
      "step": 6102
    },
    {
      "epoch": 0.6103,
      "grad_norm": 3.817013740539551,
      "learning_rate": 7.914054840563963e-05,
      "loss": 5.4938,
      "step": 6103
    },
    {
      "epoch": 0.6104,
      "grad_norm": 2.6510207653045654,
      "learning_rate": 7.910641095975886e-05,
      "loss": 4.8652,
      "step": 6104
    },
    {
      "epoch": 0.6105,
      "grad_norm": 2.5121607780456543,
      "learning_rate": 7.907227605969849e-05,
      "loss": 4.8475,
      "step": 6105
    },
    {
      "epoch": 0.6106,
      "grad_norm": 3.8875796794891357,
      "learning_rate": 7.903814370961784e-05,
      "loss": 4.7992,
      "step": 6106
    },
    {
      "epoch": 0.6107,
      "grad_norm": 3.35085129737854,
      "learning_rate": 7.900401391367576e-05,
      "loss": 5.3501,
      "step": 6107
    },
    {
      "epoch": 0.6108,
      "grad_norm": 2.9499077796936035,
      "learning_rate": 7.896988667603093e-05,
      "loss": 4.5775,
      "step": 6108
    },
    {
      "epoch": 0.6109,
      "grad_norm": 3.154494524002075,
      "learning_rate": 7.893576200084159e-05,
      "loss": 5.1154,
      "step": 6109
    },
    {
      "epoch": 0.611,
      "grad_norm": 4.938778877258301,
      "learning_rate": 7.89016398922658e-05,
      "loss": 5.1026,
      "step": 6110
    },
    {
      "epoch": 0.6111,
      "grad_norm": 2.4552900791168213,
      "learning_rate": 7.886752035446114e-05,
      "loss": 4.713,
      "step": 6111
    },
    {
      "epoch": 0.6112,
      "grad_norm": 2.9554691314697266,
      "learning_rate": 7.883340339158505e-05,
      "loss": 5.2054,
      "step": 6112
    },
    {
      "epoch": 0.6113,
      "grad_norm": 3.287768840789795,
      "learning_rate": 7.879928900779456e-05,
      "loss": 5.0488,
      "step": 6113
    },
    {
      "epoch": 0.6114,
      "grad_norm": 4.176562309265137,
      "learning_rate": 7.876517720724636e-05,
      "loss": 4.6134,
      "step": 6114
    },
    {
      "epoch": 0.6115,
      "grad_norm": 2.696925401687622,
      "learning_rate": 7.873106799409695e-05,
      "loss": 4.653,
      "step": 6115
    },
    {
      "epoch": 0.6116,
      "grad_norm": 5.483151435852051,
      "learning_rate": 7.869696137250235e-05,
      "loss": 4.6992,
      "step": 6116
    },
    {
      "epoch": 0.6117,
      "grad_norm": 3.7219579219818115,
      "learning_rate": 7.866285734661841e-05,
      "loss": 5.0945,
      "step": 6117
    },
    {
      "epoch": 0.6118,
      "grad_norm": 5.566192626953125,
      "learning_rate": 7.862875592060056e-05,
      "loss": 5.615,
      "step": 6118
    },
    {
      "epoch": 0.6119,
      "grad_norm": 9.56105899810791,
      "learning_rate": 7.8594657098604e-05,
      "loss": 4.7985,
      "step": 6119
    },
    {
      "epoch": 0.612,
      "grad_norm": 3.2784783840179443,
      "learning_rate": 7.856056088478352e-05,
      "loss": 4.6736,
      "step": 6120
    },
    {
      "epoch": 0.6121,
      "grad_norm": 5.821741580963135,
      "learning_rate": 7.852646728329368e-05,
      "loss": 4.4698,
      "step": 6121
    },
    {
      "epoch": 0.6122,
      "grad_norm": 3.3178393840789795,
      "learning_rate": 7.849237629828869e-05,
      "loss": 4.9757,
      "step": 6122
    },
    {
      "epoch": 0.6123,
      "grad_norm": 5.322384834289551,
      "learning_rate": 7.845828793392236e-05,
      "loss": 4.9469,
      "step": 6123
    },
    {
      "epoch": 0.6124,
      "grad_norm": 3.050485849380493,
      "learning_rate": 7.842420219434833e-05,
      "loss": 5.04,
      "step": 6124
    },
    {
      "epoch": 0.6125,
      "grad_norm": 4.958656311035156,
      "learning_rate": 7.83901190837198e-05,
      "loss": 5.037,
      "step": 6125
    },
    {
      "epoch": 0.6126,
      "grad_norm": 2.6735167503356934,
      "learning_rate": 7.835603860618972e-05,
      "loss": 5.1838,
      "step": 6126
    },
    {
      "epoch": 0.6127,
      "grad_norm": 2.0056467056274414,
      "learning_rate": 7.832196076591067e-05,
      "loss": 4.6151,
      "step": 6127
    },
    {
      "epoch": 0.6128,
      "grad_norm": 2.7567336559295654,
      "learning_rate": 7.828788556703498e-05,
      "loss": 4.9263,
      "step": 6128
    },
    {
      "epoch": 0.6129,
      "grad_norm": 2.0395667552948,
      "learning_rate": 7.825381301371452e-05,
      "loss": 4.5849,
      "step": 6129
    },
    {
      "epoch": 0.613,
      "grad_norm": 2.1816694736480713,
      "learning_rate": 7.821974311010102e-05,
      "loss": 4.497,
      "step": 6130
    },
    {
      "epoch": 0.6131,
      "grad_norm": 2.606822967529297,
      "learning_rate": 7.818567586034577e-05,
      "loss": 4.4347,
      "step": 6131
    },
    {
      "epoch": 0.6132,
      "grad_norm": 2.54903244972229,
      "learning_rate": 7.81516112685997e-05,
      "loss": 4.524,
      "step": 6132
    },
    {
      "epoch": 0.6133,
      "grad_norm": 1.8903318643569946,
      "learning_rate": 7.811754933901358e-05,
      "loss": 4.7081,
      "step": 6133
    },
    {
      "epoch": 0.6134,
      "grad_norm": 2.849365472793579,
      "learning_rate": 7.808349007573763e-05,
      "loss": 5.0154,
      "step": 6134
    },
    {
      "epoch": 0.6135,
      "grad_norm": 2.910702705383301,
      "learning_rate": 7.804943348292197e-05,
      "loss": 4.9255,
      "step": 6135
    },
    {
      "epoch": 0.6136,
      "grad_norm": 4.946155548095703,
      "learning_rate": 7.801537956471624e-05,
      "loss": 5.3249,
      "step": 6136
    },
    {
      "epoch": 0.6137,
      "grad_norm": 2.1343252658843994,
      "learning_rate": 7.798132832526986e-05,
      "loss": 4.8454,
      "step": 6137
    },
    {
      "epoch": 0.6138,
      "grad_norm": 4.264427661895752,
      "learning_rate": 7.79472797687318e-05,
      "loss": 4.6292,
      "step": 6138
    },
    {
      "epoch": 0.6139,
      "grad_norm": 3.189730644226074,
      "learning_rate": 7.791323389925084e-05,
      "loss": 4.9387,
      "step": 6139
    },
    {
      "epoch": 0.614,
      "grad_norm": 3.0797317028045654,
      "learning_rate": 7.787919072097531e-05,
      "loss": 4.8363,
      "step": 6140
    },
    {
      "epoch": 0.6141,
      "grad_norm": 3.3292579650878906,
      "learning_rate": 7.784515023805328e-05,
      "loss": 4.7737,
      "step": 6141
    },
    {
      "epoch": 0.6142,
      "grad_norm": 4.697385311126709,
      "learning_rate": 7.781111245463252e-05,
      "loss": 5.3606,
      "step": 6142
    },
    {
      "epoch": 0.6143,
      "grad_norm": 2.955953598022461,
      "learning_rate": 7.777707737486037e-05,
      "loss": 5.4124,
      "step": 6143
    },
    {
      "epoch": 0.6144,
      "grad_norm": 2.6560637950897217,
      "learning_rate": 7.774304500288394e-05,
      "loss": 5.0901,
      "step": 6144
    },
    {
      "epoch": 0.6145,
      "grad_norm": 3.0309107303619385,
      "learning_rate": 7.770901534284995e-05,
      "loss": 4.755,
      "step": 6145
    },
    {
      "epoch": 0.6146,
      "grad_norm": 2.156398057937622,
      "learning_rate": 7.767498839890488e-05,
      "loss": 5.271,
      "step": 6146
    },
    {
      "epoch": 0.6147,
      "grad_norm": 3.4797325134277344,
      "learning_rate": 7.76409641751947e-05,
      "loss": 4.4656,
      "step": 6147
    },
    {
      "epoch": 0.6148,
      "grad_norm": 1.9047383069992065,
      "learning_rate": 7.760694267586525e-05,
      "loss": 4.837,
      "step": 6148
    },
    {
      "epoch": 0.6149,
      "grad_norm": 2.493025302886963,
      "learning_rate": 7.75729239050619e-05,
      "loss": 4.6267,
      "step": 6149
    },
    {
      "epoch": 0.615,
      "grad_norm": 2.7014472484588623,
      "learning_rate": 7.753890786692972e-05,
      "loss": 4.7545,
      "step": 6150
    },
    {
      "epoch": 0.6151,
      "grad_norm": 2.375248670578003,
      "learning_rate": 7.750489456561352e-05,
      "loss": 4.5945,
      "step": 6151
    },
    {
      "epoch": 0.6152,
      "grad_norm": 2.753122568130493,
      "learning_rate": 7.747088400525766e-05,
      "loss": 4.6885,
      "step": 6152
    },
    {
      "epoch": 0.6153,
      "grad_norm": 1.903119444847107,
      "learning_rate": 7.743687619000626e-05,
      "loss": 5.7505,
      "step": 6153
    },
    {
      "epoch": 0.6154,
      "grad_norm": 3.345238447189331,
      "learning_rate": 7.740287112400303e-05,
      "loss": 4.7236,
      "step": 6154
    },
    {
      "epoch": 0.6155,
      "grad_norm": 2.04917049407959,
      "learning_rate": 7.736886881139142e-05,
      "loss": 4.5203,
      "step": 6155
    },
    {
      "epoch": 0.6156,
      "grad_norm": 2.7831592559814453,
      "learning_rate": 7.733486925631447e-05,
      "loss": 4.4831,
      "step": 6156
    },
    {
      "epoch": 0.6157,
      "grad_norm": 2.7849531173706055,
      "learning_rate": 7.730087246291502e-05,
      "loss": 4.9659,
      "step": 6157
    },
    {
      "epoch": 0.6158,
      "grad_norm": 3.386141061782837,
      "learning_rate": 7.726687843533538e-05,
      "loss": 4.5632,
      "step": 6158
    },
    {
      "epoch": 0.6159,
      "grad_norm": 2.3420169353485107,
      "learning_rate": 7.723288717771761e-05,
      "loss": 4.8176,
      "step": 6159
    },
    {
      "epoch": 0.616,
      "grad_norm": 2.5477023124694824,
      "learning_rate": 7.719889869420353e-05,
      "loss": 4.9483,
      "step": 6160
    },
    {
      "epoch": 0.6161,
      "grad_norm": 2.5646004676818848,
      "learning_rate": 7.716491298893442e-05,
      "loss": 5.0204,
      "step": 6161
    },
    {
      "epoch": 0.6162,
      "grad_norm": 3.0380871295928955,
      "learning_rate": 7.713093006605145e-05,
      "loss": 4.8978,
      "step": 6162
    },
    {
      "epoch": 0.6163,
      "grad_norm": 2.4960174560546875,
      "learning_rate": 7.709694992969526e-05,
      "loss": 5.076,
      "step": 6163
    },
    {
      "epoch": 0.6164,
      "grad_norm": 2.5632152557373047,
      "learning_rate": 7.706297258400624e-05,
      "loss": 5.1086,
      "step": 6164
    },
    {
      "epoch": 0.6165,
      "grad_norm": 15.457747459411621,
      "learning_rate": 7.702899803312443e-05,
      "loss": 5.7924,
      "step": 6165
    },
    {
      "epoch": 0.6166,
      "grad_norm": 3.8483059406280518,
      "learning_rate": 7.699502628118958e-05,
      "loss": 4.4909,
      "step": 6166
    },
    {
      "epoch": 0.6167,
      "grad_norm": 3.951277494430542,
      "learning_rate": 7.696105733234098e-05,
      "loss": 5.003,
      "step": 6167
    },
    {
      "epoch": 0.6168,
      "grad_norm": 2.972939968109131,
      "learning_rate": 7.692709119071762e-05,
      "loss": 4.4152,
      "step": 6168
    },
    {
      "epoch": 0.6169,
      "grad_norm": 4.070160865783691,
      "learning_rate": 7.689312786045823e-05,
      "loss": 4.7054,
      "step": 6169
    },
    {
      "epoch": 0.617,
      "grad_norm": 2.840028762817383,
      "learning_rate": 7.685916734570112e-05,
      "loss": 4.1241,
      "step": 6170
    },
    {
      "epoch": 0.6171,
      "grad_norm": 2.0085930824279785,
      "learning_rate": 7.682520965058428e-05,
      "loss": 4.5543,
      "step": 6171
    },
    {
      "epoch": 0.6172,
      "grad_norm": 2.7264978885650635,
      "learning_rate": 7.679125477924534e-05,
      "loss": 5.3281,
      "step": 6172
    },
    {
      "epoch": 0.6173,
      "grad_norm": 2.6376118659973145,
      "learning_rate": 7.67573027358216e-05,
      "loss": 4.8159,
      "step": 6173
    },
    {
      "epoch": 0.6174,
      "grad_norm": 2.318401336669922,
      "learning_rate": 7.672335352445002e-05,
      "loss": 5.23,
      "step": 6174
    },
    {
      "epoch": 0.6175,
      "grad_norm": 2.2357070446014404,
      "learning_rate": 7.668940714926725e-05,
      "loss": 4.7661,
      "step": 6175
    },
    {
      "epoch": 0.6176,
      "grad_norm": 3.370105743408203,
      "learning_rate": 7.66554636144095e-05,
      "loss": 4.8545,
      "step": 6176
    },
    {
      "epoch": 0.6177,
      "grad_norm": 2.907435655593872,
      "learning_rate": 7.662152292401264e-05,
      "loss": 5.2261,
      "step": 6177
    },
    {
      "epoch": 0.6178,
      "grad_norm": 2.1373233795166016,
      "learning_rate": 7.658758508221234e-05,
      "loss": 4.8204,
      "step": 6178
    },
    {
      "epoch": 0.6179,
      "grad_norm": 2.184804677963257,
      "learning_rate": 7.655365009314374e-05,
      "loss": 4.5558,
      "step": 6179
    },
    {
      "epoch": 0.618,
      "grad_norm": 5.1249613761901855,
      "learning_rate": 7.651971796094183e-05,
      "loss": 4.5984,
      "step": 6180
    },
    {
      "epoch": 0.6181,
      "grad_norm": 2.087038040161133,
      "learning_rate": 7.6485788689741e-05,
      "loss": 4.5847,
      "step": 6181
    },
    {
      "epoch": 0.6182,
      "grad_norm": 2.823840856552124,
      "learning_rate": 7.645186228367554e-05,
      "loss": 4.8234,
      "step": 6182
    },
    {
      "epoch": 0.6183,
      "grad_norm": 2.4413704872131348,
      "learning_rate": 7.641793874687918e-05,
      "loss": 5.0378,
      "step": 6183
    },
    {
      "epoch": 0.6184,
      "grad_norm": 1.9586607217788696,
      "learning_rate": 7.638401808348548e-05,
      "loss": 4.9743,
      "step": 6184
    },
    {
      "epoch": 0.6185,
      "grad_norm": 5.681273460388184,
      "learning_rate": 7.635010029762756e-05,
      "loss": 5.353,
      "step": 6185
    },
    {
      "epoch": 0.6186,
      "grad_norm": 3.3988828659057617,
      "learning_rate": 7.631618539343814e-05,
      "loss": 4.795,
      "step": 6186
    },
    {
      "epoch": 0.6187,
      "grad_norm": 4.2785725593566895,
      "learning_rate": 7.628227337504972e-05,
      "loss": 5.2382,
      "step": 6187
    },
    {
      "epoch": 0.6188,
      "grad_norm": 2.0152223110198975,
      "learning_rate": 7.62483642465943e-05,
      "loss": 4.5187,
      "step": 6188
    },
    {
      "epoch": 0.6189,
      "grad_norm": 5.446355819702148,
      "learning_rate": 7.621445801220371e-05,
      "loss": 4.8184,
      "step": 6189
    },
    {
      "epoch": 0.619,
      "grad_norm": 2.311471700668335,
      "learning_rate": 7.618055467600922e-05,
      "loss": 4.1803,
      "step": 6190
    },
    {
      "epoch": 0.6191,
      "grad_norm": 2.889647960662842,
      "learning_rate": 7.614665424214193e-05,
      "loss": 5.1721,
      "step": 6191
    },
    {
      "epoch": 0.6192,
      "grad_norm": 5.1931867599487305,
      "learning_rate": 7.611275671473245e-05,
      "loss": 4.91,
      "step": 6192
    },
    {
      "epoch": 0.6193,
      "grad_norm": 2.3053183555603027,
      "learning_rate": 7.607886209791107e-05,
      "loss": 4.7784,
      "step": 6193
    },
    {
      "epoch": 0.6194,
      "grad_norm": 1.7651475667953491,
      "learning_rate": 7.604497039580785e-05,
      "loss": 4.9657,
      "step": 6194
    },
    {
      "epoch": 0.6195,
      "grad_norm": 3.3554129600524902,
      "learning_rate": 7.601108161255226e-05,
      "loss": 5.4272,
      "step": 6195
    },
    {
      "epoch": 0.6196,
      "grad_norm": 2.2605085372924805,
      "learning_rate": 7.597719575227364e-05,
      "loss": 4.6688,
      "step": 6196
    },
    {
      "epoch": 0.6197,
      "grad_norm": 3.2898507118225098,
      "learning_rate": 7.594331281910082e-05,
      "loss": 4.7942,
      "step": 6197
    },
    {
      "epoch": 0.6198,
      "grad_norm": 2.6927032470703125,
      "learning_rate": 7.590943281716241e-05,
      "loss": 5.1532,
      "step": 6198
    },
    {
      "epoch": 0.6199,
      "grad_norm": 2.7784347534179688,
      "learning_rate": 7.587555575058649e-05,
      "loss": 4.0455,
      "step": 6199
    },
    {
      "epoch": 0.62,
      "grad_norm": 6.905799388885498,
      "learning_rate": 7.584168162350098e-05,
      "loss": 5.7678,
      "step": 6200
    },
    {
      "epoch": 0.6201,
      "grad_norm": 2.56076979637146,
      "learning_rate": 7.580781044003324e-05,
      "loss": 5.1207,
      "step": 6201
    },
    {
      "epoch": 0.6202,
      "grad_norm": 7.874569416046143,
      "learning_rate": 7.577394220431042e-05,
      "loss": 4.6476,
      "step": 6202
    },
    {
      "epoch": 0.6203,
      "grad_norm": 3.233734130859375,
      "learning_rate": 7.574007692045928e-05,
      "loss": 4.7386,
      "step": 6203
    },
    {
      "epoch": 0.6204,
      "grad_norm": 3.4051625728607178,
      "learning_rate": 7.570621459260615e-05,
      "loss": 4.9058,
      "step": 6204
    },
    {
      "epoch": 0.6205,
      "grad_norm": 2.302103042602539,
      "learning_rate": 7.567235522487712e-05,
      "loss": 4.792,
      "step": 6205
    },
    {
      "epoch": 0.6206,
      "grad_norm": 4.102400302886963,
      "learning_rate": 7.563849882139776e-05,
      "loss": 4.1181,
      "step": 6206
    },
    {
      "epoch": 0.6207,
      "grad_norm": 2.6068289279937744,
      "learning_rate": 7.560464538629344e-05,
      "loss": 4.4731,
      "step": 6207
    },
    {
      "epoch": 0.6208,
      "grad_norm": 2.1590378284454346,
      "learning_rate": 7.557079492368909e-05,
      "loss": 4.6534,
      "step": 6208
    },
    {
      "epoch": 0.6209,
      "grad_norm": 3.388484477996826,
      "learning_rate": 7.553694743770928e-05,
      "loss": 4.8623,
      "step": 6209
    },
    {
      "epoch": 0.621,
      "grad_norm": 2.4423866271972656,
      "learning_rate": 7.550310293247823e-05,
      "loss": 4.1468,
      "step": 6210
    },
    {
      "epoch": 0.6211,
      "grad_norm": 3.0565731525421143,
      "learning_rate": 7.546926141211974e-05,
      "loss": 4.6505,
      "step": 6211
    },
    {
      "epoch": 0.6212,
      "grad_norm": 6.096969127655029,
      "learning_rate": 7.543542288075739e-05,
      "loss": 4.8658,
      "step": 6212
    },
    {
      "epoch": 0.6213,
      "grad_norm": 4.194102764129639,
      "learning_rate": 7.54015873425142e-05,
      "loss": 5.6823,
      "step": 6213
    },
    {
      "epoch": 0.6214,
      "grad_norm": 3.3851511478424072,
      "learning_rate": 7.536775480151303e-05,
      "loss": 5.922,
      "step": 6214
    },
    {
      "epoch": 0.6215,
      "grad_norm": 3.3670175075531006,
      "learning_rate": 7.533392526187617e-05,
      "loss": 4.874,
      "step": 6215
    },
    {
      "epoch": 0.6216,
      "grad_norm": 1.8480710983276367,
      "learning_rate": 7.530009872772572e-05,
      "loss": 4.689,
      "step": 6216
    },
    {
      "epoch": 0.6217,
      "grad_norm": 2.7868402004241943,
      "learning_rate": 7.526627520318329e-05,
      "loss": 4.35,
      "step": 6217
    },
    {
      "epoch": 0.6218,
      "grad_norm": 2.4947171211242676,
      "learning_rate": 7.523245469237026e-05,
      "loss": 4.3954,
      "step": 6218
    },
    {
      "epoch": 0.6219,
      "grad_norm": 2.720874071121216,
      "learning_rate": 7.519863719940748e-05,
      "loss": 4.6734,
      "step": 6219
    },
    {
      "epoch": 0.622,
      "grad_norm": 4.473278522491455,
      "learning_rate": 7.516482272841549e-05,
      "loss": 4.3267,
      "step": 6220
    },
    {
      "epoch": 0.6221,
      "grad_norm": 5.496091365814209,
      "learning_rate": 7.513101128351454e-05,
      "loss": 5.4077,
      "step": 6221
    },
    {
      "epoch": 0.6222,
      "grad_norm": 3.070068120956421,
      "learning_rate": 7.50972028688244e-05,
      "loss": 4.6771,
      "step": 6222
    },
    {
      "epoch": 0.6223,
      "grad_norm": 2.931265354156494,
      "learning_rate": 7.506339748846461e-05,
      "loss": 4.6746,
      "step": 6223
    },
    {
      "epoch": 0.6224,
      "grad_norm": 2.1528570652008057,
      "learning_rate": 7.502959514655414e-05,
      "loss": 4.6572,
      "step": 6224
    },
    {
      "epoch": 0.6225,
      "grad_norm": 7.190796852111816,
      "learning_rate": 7.499579584721179e-05,
      "loss": 4.586,
      "step": 6225
    },
    {
      "epoch": 0.6226,
      "grad_norm": 2.972111463546753,
      "learning_rate": 7.496199959455584e-05,
      "loss": 4.5354,
      "step": 6226
    },
    {
      "epoch": 0.6227,
      "grad_norm": 12.403257369995117,
      "learning_rate": 7.492820639270434e-05,
      "loss": 5.1509,
      "step": 6227
    },
    {
      "epoch": 0.6228,
      "grad_norm": 3.370009183883667,
      "learning_rate": 7.489441624577485e-05,
      "loss": 4.7585,
      "step": 6228
    },
    {
      "epoch": 0.6229,
      "grad_norm": 12.078509330749512,
      "learning_rate": 7.486062915788452e-05,
      "loss": 5.0669,
      "step": 6229
    },
    {
      "epoch": 0.623,
      "grad_norm": 3.2156341075897217,
      "learning_rate": 7.48268451331503e-05,
      "loss": 4.9473,
      "step": 6230
    },
    {
      "epoch": 0.6231,
      "grad_norm": 1.9456443786621094,
      "learning_rate": 7.479306417568864e-05,
      "loss": 4.8379,
      "step": 6231
    },
    {
      "epoch": 0.6232,
      "grad_norm": 2.2426605224609375,
      "learning_rate": 7.475928628961566e-05,
      "loss": 4.8444,
      "step": 6232
    },
    {
      "epoch": 0.6233,
      "grad_norm": 2.0724949836730957,
      "learning_rate": 7.472551147904708e-05,
      "loss": 4.5781,
      "step": 6233
    },
    {
      "epoch": 0.6234,
      "grad_norm": 2.3937454223632812,
      "learning_rate": 7.469173974809826e-05,
      "loss": 4.6085,
      "step": 6234
    },
    {
      "epoch": 0.6235,
      "grad_norm": 2.480710744857788,
      "learning_rate": 7.465797110088417e-05,
      "loss": 4.6337,
      "step": 6235
    },
    {
      "epoch": 0.6236,
      "grad_norm": 3.9071829319000244,
      "learning_rate": 7.462420554151944e-05,
      "loss": 5.0072,
      "step": 6236
    },
    {
      "epoch": 0.6237,
      "grad_norm": 2.0865354537963867,
      "learning_rate": 7.459044307411832e-05,
      "loss": 3.998,
      "step": 6237
    },
    {
      "epoch": 0.6238,
      "grad_norm": 2.8415229320526123,
      "learning_rate": 7.45566837027946e-05,
      "loss": 4.8362,
      "step": 6238
    },
    {
      "epoch": 0.6239,
      "grad_norm": 4.788355350494385,
      "learning_rate": 7.45229274316618e-05,
      "loss": 4.52,
      "step": 6239
    },
    {
      "epoch": 0.624,
      "grad_norm": 2.737729072570801,
      "learning_rate": 7.448917426483299e-05,
      "loss": 5.0341,
      "step": 6240
    },
    {
      "epoch": 0.6241,
      "grad_norm": 2.703380823135376,
      "learning_rate": 7.445542420642097e-05,
      "loss": 4.6479,
      "step": 6241
    },
    {
      "epoch": 0.6242,
      "grad_norm": 3.1902811527252197,
      "learning_rate": 7.442167726053797e-05,
      "loss": 4.8111,
      "step": 6242
    },
    {
      "epoch": 0.6243,
      "grad_norm": 3.064455509185791,
      "learning_rate": 7.438793343129605e-05,
      "loss": 4.759,
      "step": 6243
    },
    {
      "epoch": 0.6244,
      "grad_norm": 3.052607297897339,
      "learning_rate": 7.435419272280672e-05,
      "loss": 5.0874,
      "step": 6244
    },
    {
      "epoch": 0.6245,
      "grad_norm": 3.32673978805542,
      "learning_rate": 7.432045513918122e-05,
      "loss": 4.9447,
      "step": 6245
    },
    {
      "epoch": 0.6246,
      "grad_norm": 3.8588497638702393,
      "learning_rate": 7.42867206845304e-05,
      "loss": 4.9141,
      "step": 6246
    },
    {
      "epoch": 0.6247,
      "grad_norm": 2.3375065326690674,
      "learning_rate": 7.425298936296463e-05,
      "loss": 4.8756,
      "step": 6247
    },
    {
      "epoch": 0.6248,
      "grad_norm": 2.2075653076171875,
      "learning_rate": 7.421926117859403e-05,
      "loss": 4.787,
      "step": 6248
    },
    {
      "epoch": 0.6249,
      "grad_norm": 2.757065773010254,
      "learning_rate": 7.418553613552825e-05,
      "loss": 4.3984,
      "step": 6249
    },
    {
      "epoch": 0.625,
      "grad_norm": 3.031680107116699,
      "learning_rate": 7.415181423787659e-05,
      "loss": 4.5363,
      "step": 6250
    },
    {
      "epoch": 0.6251,
      "grad_norm": 3.668482542037964,
      "learning_rate": 7.411809548974792e-05,
      "loss": 4.7591,
      "step": 6251
    },
    {
      "epoch": 0.6252,
      "grad_norm": 2.4354076385498047,
      "learning_rate": 7.408437989525085e-05,
      "loss": 4.7918,
      "step": 6252
    },
    {
      "epoch": 0.6253,
      "grad_norm": 2.519451141357422,
      "learning_rate": 7.405066745849346e-05,
      "loss": 4.4712,
      "step": 6253
    },
    {
      "epoch": 0.6254,
      "grad_norm": 2.6811537742614746,
      "learning_rate": 7.401695818358353e-05,
      "loss": 4.7951,
      "step": 6254
    },
    {
      "epoch": 0.6255,
      "grad_norm": 3.564635753631592,
      "learning_rate": 7.398325207462846e-05,
      "loss": 5.0953,
      "step": 6255
    },
    {
      "epoch": 0.6256,
      "grad_norm": 3.356534004211426,
      "learning_rate": 7.394954913573517e-05,
      "loss": 4.9808,
      "step": 6256
    },
    {
      "epoch": 0.6257,
      "grad_norm": 2.620450258255005,
      "learning_rate": 7.391584937101033e-05,
      "loss": 5.2681,
      "step": 6257
    },
    {
      "epoch": 0.6258,
      "grad_norm": 2.6394412517547607,
      "learning_rate": 7.38821527845601e-05,
      "loss": 4.5814,
      "step": 6258
    },
    {
      "epoch": 0.6259,
      "grad_norm": 4.423412799835205,
      "learning_rate": 7.384845938049031e-05,
      "loss": 5.1261,
      "step": 6259
    },
    {
      "epoch": 0.626,
      "grad_norm": 2.6278791427612305,
      "learning_rate": 7.381476916290644e-05,
      "loss": 4.8231,
      "step": 6260
    },
    {
      "epoch": 0.6261,
      "grad_norm": 4.490521430969238,
      "learning_rate": 7.378108213591355e-05,
      "loss": 5.0166,
      "step": 6261
    },
    {
      "epoch": 0.6262,
      "grad_norm": 2.62923264503479,
      "learning_rate": 7.374739830361621e-05,
      "loss": 4.3797,
      "step": 6262
    },
    {
      "epoch": 0.6263,
      "grad_norm": 4.278876781463623,
      "learning_rate": 7.37137176701188e-05,
      "loss": 5.7341,
      "step": 6263
    },
    {
      "epoch": 0.6264,
      "grad_norm": 6.087192535400391,
      "learning_rate": 7.368004023952517e-05,
      "loss": 4.9599,
      "step": 6264
    },
    {
      "epoch": 0.6265,
      "grad_norm": 3.8407540321350098,
      "learning_rate": 7.364636601593875e-05,
      "loss": 4.9634,
      "step": 6265
    },
    {
      "epoch": 0.6266,
      "grad_norm": 2.6375789642333984,
      "learning_rate": 7.361269500346274e-05,
      "loss": 4.5715,
      "step": 6266
    },
    {
      "epoch": 0.6267,
      "grad_norm": 2.3046836853027344,
      "learning_rate": 7.357902720619976e-05,
      "loss": 4.7902,
      "step": 6267
    },
    {
      "epoch": 0.6268,
      "grad_norm": 2.826462984085083,
      "learning_rate": 7.354536262825219e-05,
      "loss": 5.351,
      "step": 6268
    },
    {
      "epoch": 0.6269,
      "grad_norm": 3.023369789123535,
      "learning_rate": 7.351170127372191e-05,
      "loss": 4.6185,
      "step": 6269
    },
    {
      "epoch": 0.627,
      "grad_norm": 2.0984976291656494,
      "learning_rate": 7.347804314671055e-05,
      "loss": 4.4379,
      "step": 6270
    },
    {
      "epoch": 0.6271,
      "grad_norm": 2.632688522338867,
      "learning_rate": 7.344438825131911e-05,
      "loss": 4.7122,
      "step": 6271
    },
    {
      "epoch": 0.6272,
      "grad_norm": 2.370797872543335,
      "learning_rate": 7.341073659164848e-05,
      "loss": 4.5917,
      "step": 6272
    },
    {
      "epoch": 0.6273,
      "grad_norm": 2.7666687965393066,
      "learning_rate": 7.33770881717989e-05,
      "loss": 4.7361,
      "step": 6273
    },
    {
      "epoch": 0.6274,
      "grad_norm": 3.129645586013794,
      "learning_rate": 7.334344299587035e-05,
      "loss": 4.9245,
      "step": 6274
    },
    {
      "epoch": 0.6275,
      "grad_norm": 2.1002445220947266,
      "learning_rate": 7.330980106796246e-05,
      "loss": 4.5671,
      "step": 6275
    },
    {
      "epoch": 0.6276,
      "grad_norm": 2.163435697555542,
      "learning_rate": 7.327616239217431e-05,
      "loss": 4.7651,
      "step": 6276
    },
    {
      "epoch": 0.6277,
      "grad_norm": 3.5309245586395264,
      "learning_rate": 7.324252697260474e-05,
      "loss": 5.4295,
      "step": 6277
    },
    {
      "epoch": 0.6278,
      "grad_norm": 2.2017948627471924,
      "learning_rate": 7.320889481335207e-05,
      "loss": 4.801,
      "step": 6278
    },
    {
      "epoch": 0.6279,
      "grad_norm": 3.2965009212493896,
      "learning_rate": 7.317526591851433e-05,
      "loss": 5.0421,
      "step": 6279
    },
    {
      "epoch": 0.628,
      "grad_norm": 2.3067970275878906,
      "learning_rate": 7.314164029218904e-05,
      "loss": 4.8026,
      "step": 6280
    },
    {
      "epoch": 0.6281,
      "grad_norm": 4.807198524475098,
      "learning_rate": 7.310801793847344e-05,
      "loss": 4.7518,
      "step": 6281
    },
    {
      "epoch": 0.6282,
      "grad_norm": 4.009991645812988,
      "learning_rate": 7.307439886146428e-05,
      "loss": 4.789,
      "step": 6282
    },
    {
      "epoch": 0.6283,
      "grad_norm": 7.740306377410889,
      "learning_rate": 7.30407830652579e-05,
      "loss": 4.8471,
      "step": 6283
    },
    {
      "epoch": 0.6284,
      "grad_norm": 6.411609649658203,
      "learning_rate": 7.300717055395039e-05,
      "loss": 4.506,
      "step": 6284
    },
    {
      "epoch": 0.6285,
      "grad_norm": 1.8641574382781982,
      "learning_rate": 7.297356133163721e-05,
      "loss": 4.5238,
      "step": 6285
    },
    {
      "epoch": 0.6286,
      "grad_norm": 2.0086159706115723,
      "learning_rate": 7.293995540241366e-05,
      "loss": 4.7463,
      "step": 6286
    },
    {
      "epoch": 0.6287,
      "grad_norm": 2.533282995223999,
      "learning_rate": 7.290635277037442e-05,
      "loss": 4.8939,
      "step": 6287
    },
    {
      "epoch": 0.6288,
      "grad_norm": 2.4732961654663086,
      "learning_rate": 7.287275343961392e-05,
      "loss": 4.3843,
      "step": 6288
    },
    {
      "epoch": 0.6289,
      "grad_norm": 4.470192909240723,
      "learning_rate": 7.283915741422612e-05,
      "loss": 5.3789,
      "step": 6289
    },
    {
      "epoch": 0.629,
      "grad_norm": 3.0726513862609863,
      "learning_rate": 7.280556469830464e-05,
      "loss": 4.584,
      "step": 6290
    },
    {
      "epoch": 0.6291,
      "grad_norm": 7.2342915534973145,
      "learning_rate": 7.277197529594257e-05,
      "loss": 4.9788,
      "step": 6291
    },
    {
      "epoch": 0.6292,
      "grad_norm": 3.420624017715454,
      "learning_rate": 7.273838921123272e-05,
      "loss": 5.1132,
      "step": 6292
    },
    {
      "epoch": 0.6293,
      "grad_norm": 2.312993288040161,
      "learning_rate": 7.270480644826749e-05,
      "loss": 4.5994,
      "step": 6293
    },
    {
      "epoch": 0.6294,
      "grad_norm": 8.366740226745605,
      "learning_rate": 7.267122701113876e-05,
      "loss": 4.6965,
      "step": 6294
    },
    {
      "epoch": 0.6295,
      "grad_norm": 3.5007565021514893,
      "learning_rate": 7.263765090393817e-05,
      "loss": 4.5904,
      "step": 6295
    },
    {
      "epoch": 0.6296,
      "grad_norm": 1.7737981081008911,
      "learning_rate": 7.260407813075676e-05,
      "loss": 4.5768,
      "step": 6296
    },
    {
      "epoch": 0.6297,
      "grad_norm": 1.6475087404251099,
      "learning_rate": 7.257050869568535e-05,
      "loss": 4.3336,
      "step": 6297
    },
    {
      "epoch": 0.6298,
      "grad_norm": 1.8653929233551025,
      "learning_rate": 7.253694260281425e-05,
      "loss": 4.4621,
      "step": 6298
    },
    {
      "epoch": 0.6299,
      "grad_norm": 2.049273729324341,
      "learning_rate": 7.250337985623342e-05,
      "loss": 4.6958,
      "step": 6299
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.866360187530518,
      "learning_rate": 7.246982046003234e-05,
      "loss": 5.211,
      "step": 6300
    },
    {
      "epoch": 0.6301,
      "grad_norm": 2.656468629837036,
      "learning_rate": 7.243626441830009e-05,
      "loss": 4.6468,
      "step": 6301
    },
    {
      "epoch": 0.6302,
      "grad_norm": 2.09565806388855,
      "learning_rate": 7.240271173512546e-05,
      "loss": 4.5599,
      "step": 6302
    },
    {
      "epoch": 0.6303,
      "grad_norm": 2.508007287979126,
      "learning_rate": 7.236916241459663e-05,
      "loss": 5.2071,
      "step": 6303
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.9853707551956177,
      "learning_rate": 7.233561646080161e-05,
      "loss": 4.6878,
      "step": 6304
    },
    {
      "epoch": 0.6305,
      "grad_norm": 2.2505693435668945,
      "learning_rate": 7.230207387782776e-05,
      "loss": 4.5313,
      "step": 6305
    },
    {
      "epoch": 0.6306,
      "grad_norm": 3.7872231006622314,
      "learning_rate": 7.226853466976222e-05,
      "loss": 4.9154,
      "step": 6306
    },
    {
      "epoch": 0.6307,
      "grad_norm": 2.4175164699554443,
      "learning_rate": 7.22349988406916e-05,
      "loss": 4.7907,
      "step": 6307
    },
    {
      "epoch": 0.6308,
      "grad_norm": 4.152900695800781,
      "learning_rate": 7.220146639470218e-05,
      "loss": 4.8697,
      "step": 6308
    },
    {
      "epoch": 0.6309,
      "grad_norm": 3.4742400646209717,
      "learning_rate": 7.216793733587976e-05,
      "loss": 4.7728,
      "step": 6309
    },
    {
      "epoch": 0.631,
      "grad_norm": 4.478064060211182,
      "learning_rate": 7.21344116683097e-05,
      "loss": 5.1668,
      "step": 6310
    },
    {
      "epoch": 0.6311,
      "grad_norm": 2.6844542026519775,
      "learning_rate": 7.210088939607708e-05,
      "loss": 4.7471,
      "step": 6311
    },
    {
      "epoch": 0.6312,
      "grad_norm": 5.124731063842773,
      "learning_rate": 7.206737052326645e-05,
      "loss": 6.2767,
      "step": 6312
    },
    {
      "epoch": 0.6313,
      "grad_norm": 2.565000534057617,
      "learning_rate": 7.203385505396203e-05,
      "loss": 5.0031,
      "step": 6313
    },
    {
      "epoch": 0.6314,
      "grad_norm": 1.8571466207504272,
      "learning_rate": 7.20003429922475e-05,
      "loss": 4.2476,
      "step": 6314
    },
    {
      "epoch": 0.6315,
      "grad_norm": 2.327760934829712,
      "learning_rate": 7.196683434220625e-05,
      "loss": 5.2751,
      "step": 6315
    },
    {
      "epoch": 0.6316,
      "grad_norm": 1.904092788696289,
      "learning_rate": 7.193332910792124e-05,
      "loss": 4.545,
      "step": 6316
    },
    {
      "epoch": 0.6317,
      "grad_norm": 6.701279640197754,
      "learning_rate": 7.18998272934749e-05,
      "loss": 4.808,
      "step": 6317
    },
    {
      "epoch": 0.6318,
      "grad_norm": 2.079406261444092,
      "learning_rate": 7.186632890294941e-05,
      "loss": 4.74,
      "step": 6318
    },
    {
      "epoch": 0.6319,
      "grad_norm": 2.9246723651885986,
      "learning_rate": 7.183283394042634e-05,
      "loss": 4.885,
      "step": 6319
    },
    {
      "epoch": 0.632,
      "grad_norm": 3.25117564201355,
      "learning_rate": 7.179934240998706e-05,
      "loss": 4.376,
      "step": 6320
    },
    {
      "epoch": 0.6321,
      "grad_norm": 2.6860432624816895,
      "learning_rate": 7.176585431571235e-05,
      "loss": 4.4852,
      "step": 6321
    },
    {
      "epoch": 0.6322,
      "grad_norm": 2.3424596786499023,
      "learning_rate": 7.173236966168268e-05,
      "loss": 4.8974,
      "step": 6322
    },
    {
      "epoch": 0.6323,
      "grad_norm": 2.1588196754455566,
      "learning_rate": 7.169888845197798e-05,
      "loss": 4.8099,
      "step": 6323
    },
    {
      "epoch": 0.6324,
      "grad_norm": 3.3702869415283203,
      "learning_rate": 7.166541069067792e-05,
      "loss": 4.5595,
      "step": 6324
    },
    {
      "epoch": 0.6325,
      "grad_norm": 3.0001723766326904,
      "learning_rate": 7.163193638186158e-05,
      "loss": 5.0498,
      "step": 6325
    },
    {
      "epoch": 0.6326,
      "grad_norm": 1.921415090560913,
      "learning_rate": 7.159846552960774e-05,
      "loss": 4.8359,
      "step": 6326
    },
    {
      "epoch": 0.6327,
      "grad_norm": 2.975275754928589,
      "learning_rate": 7.156499813799476e-05,
      "loss": 4.7996,
      "step": 6327
    },
    {
      "epoch": 0.6328,
      "grad_norm": 2.71130108833313,
      "learning_rate": 7.153153421110048e-05,
      "loss": 5.0304,
      "step": 6328
    },
    {
      "epoch": 0.6329,
      "grad_norm": 2.6169629096984863,
      "learning_rate": 7.149807375300239e-05,
      "loss": 4.8667,
      "step": 6329
    },
    {
      "epoch": 0.633,
      "grad_norm": 1.9582200050354004,
      "learning_rate": 7.146461676777756e-05,
      "loss": 4.6738,
      "step": 6330
    },
    {
      "epoch": 0.6331,
      "grad_norm": 7.274276256561279,
      "learning_rate": 7.143116325950265e-05,
      "loss": 5.2848,
      "step": 6331
    },
    {
      "epoch": 0.6332,
      "grad_norm": 2.9885568618774414,
      "learning_rate": 7.139771323225381e-05,
      "loss": 4.6887,
      "step": 6332
    },
    {
      "epoch": 0.6333,
      "grad_norm": 3.4093823432922363,
      "learning_rate": 7.136426669010689e-05,
      "loss": 5.1031,
      "step": 6333
    },
    {
      "epoch": 0.6334,
      "grad_norm": 3.0938892364501953,
      "learning_rate": 7.13308236371372e-05,
      "loss": 4.6515,
      "step": 6334
    },
    {
      "epoch": 0.6335,
      "grad_norm": 3.633236885070801,
      "learning_rate": 7.129738407741964e-05,
      "loss": 4.6145,
      "step": 6335
    },
    {
      "epoch": 0.6336,
      "grad_norm": 2.3306853771209717,
      "learning_rate": 7.126394801502882e-05,
      "loss": 4.5098,
      "step": 6336
    },
    {
      "epoch": 0.6337,
      "grad_norm": 2.109853982925415,
      "learning_rate": 7.123051545403874e-05,
      "loss": 4.585,
      "step": 6337
    },
    {
      "epoch": 0.6338,
      "grad_norm": 2.845921754837036,
      "learning_rate": 7.119708639852312e-05,
      "loss": 4.4686,
      "step": 6338
    },
    {
      "epoch": 0.6339,
      "grad_norm": 2.517775774002075,
      "learning_rate": 7.11636608525551e-05,
      "loss": 4.4273,
      "step": 6339
    },
    {
      "epoch": 0.634,
      "grad_norm": 4.1325483322143555,
      "learning_rate": 7.113023882020757e-05,
      "loss": 5.2008,
      "step": 6340
    },
    {
      "epoch": 0.6341,
      "grad_norm": 3.254737138748169,
      "learning_rate": 7.109682030555283e-05,
      "loss": 4.9604,
      "step": 6341
    },
    {
      "epoch": 0.6342,
      "grad_norm": 2.2060468196868896,
      "learning_rate": 7.106340531266292e-05,
      "loss": 4.4365,
      "step": 6342
    },
    {
      "epoch": 0.6343,
      "grad_norm": 3.5167644023895264,
      "learning_rate": 7.102999384560927e-05,
      "loss": 4.4988,
      "step": 6343
    },
    {
      "epoch": 0.6344,
      "grad_norm": 2.5595178604125977,
      "learning_rate": 7.099658590846299e-05,
      "loss": 4.565,
      "step": 6344
    },
    {
      "epoch": 0.6345,
      "grad_norm": 2.51057767868042,
      "learning_rate": 7.096318150529477e-05,
      "loss": 4.5598,
      "step": 6345
    },
    {
      "epoch": 0.6346,
      "grad_norm": 1.978434681892395,
      "learning_rate": 7.092978064017475e-05,
      "loss": 5.2305,
      "step": 6346
    },
    {
      "epoch": 0.6347,
      "grad_norm": 2.694180965423584,
      "learning_rate": 7.089638331717284e-05,
      "loss": 5.2323,
      "step": 6347
    },
    {
      "epoch": 0.6348,
      "grad_norm": 4.983065605163574,
      "learning_rate": 7.08629895403583e-05,
      "loss": 4.8829,
      "step": 6348
    },
    {
      "epoch": 0.6349,
      "grad_norm": 2.9218013286590576,
      "learning_rate": 7.082959931380011e-05,
      "loss": 4.2598,
      "step": 6349
    },
    {
      "epoch": 0.635,
      "grad_norm": 4.015475749969482,
      "learning_rate": 7.079621264156675e-05,
      "loss": 4.756,
      "step": 6350
    },
    {
      "epoch": 0.6351,
      "grad_norm": 2.6050145626068115,
      "learning_rate": 7.076282952772633e-05,
      "loss": 4.6826,
      "step": 6351
    },
    {
      "epoch": 0.6352,
      "grad_norm": 2.501554489135742,
      "learning_rate": 7.072944997634646e-05,
      "loss": 4.4387,
      "step": 6352
    },
    {
      "epoch": 0.6353,
      "grad_norm": 3.024350166320801,
      "learning_rate": 7.069607399149428e-05,
      "loss": 5.2499,
      "step": 6353
    },
    {
      "epoch": 0.6354,
      "grad_norm": 2.1982226371765137,
      "learning_rate": 7.06627015772366e-05,
      "loss": 4.4472,
      "step": 6354
    },
    {
      "epoch": 0.6355,
      "grad_norm": 3.4929065704345703,
      "learning_rate": 7.062933273763975e-05,
      "loss": 4.6527,
      "step": 6355
    },
    {
      "epoch": 0.6356,
      "grad_norm": 6.048494815826416,
      "learning_rate": 7.059596747676962e-05,
      "loss": 4.6913,
      "step": 6356
    },
    {
      "epoch": 0.6357,
      "grad_norm": 2.651292085647583,
      "learning_rate": 7.056260579869165e-05,
      "loss": 4.7362,
      "step": 6357
    },
    {
      "epoch": 0.6358,
      "grad_norm": 2.710735559463501,
      "learning_rate": 7.052924770747087e-05,
      "loss": 4.7363,
      "step": 6358
    },
    {
      "epoch": 0.6359,
      "grad_norm": 3.7024552822113037,
      "learning_rate": 7.049589320717186e-05,
      "loss": 4.8352,
      "step": 6359
    },
    {
      "epoch": 0.636,
      "grad_norm": 2.2737741470336914,
      "learning_rate": 7.04625423018588e-05,
      "loss": 4.6018,
      "step": 6360
    },
    {
      "epoch": 0.6361,
      "grad_norm": 2.3920416831970215,
      "learning_rate": 7.042919499559537e-05,
      "loss": 4.8637,
      "step": 6361
    },
    {
      "epoch": 0.6362,
      "grad_norm": 2.8916404247283936,
      "learning_rate": 7.039585129244477e-05,
      "loss": 4.692,
      "step": 6362
    },
    {
      "epoch": 0.6363,
      "grad_norm": 2.952259063720703,
      "learning_rate": 7.036251119646992e-05,
      "loss": 5.2957,
      "step": 6363
    },
    {
      "epoch": 0.6364,
      "grad_norm": 2.9874274730682373,
      "learning_rate": 7.032917471173318e-05,
      "loss": 4.761,
      "step": 6364
    },
    {
      "epoch": 0.6365,
      "grad_norm": 5.012946605682373,
      "learning_rate": 7.029584184229653e-05,
      "loss": 5.6001,
      "step": 6365
    },
    {
      "epoch": 0.6366,
      "grad_norm": 4.345258712768555,
      "learning_rate": 7.026251259222141e-05,
      "loss": 5.0342,
      "step": 6366
    },
    {
      "epoch": 0.6367,
      "grad_norm": 2.6086902618408203,
      "learning_rate": 7.022918696556896e-05,
      "loss": 4.8748,
      "step": 6367
    },
    {
      "epoch": 0.6368,
      "grad_norm": 2.681215524673462,
      "learning_rate": 7.019586496639974e-05,
      "loss": 5.2718,
      "step": 6368
    },
    {
      "epoch": 0.6369,
      "grad_norm": 2.1439151763916016,
      "learning_rate": 7.016254659877398e-05,
      "loss": 4.7116,
      "step": 6369
    },
    {
      "epoch": 0.637,
      "grad_norm": 1.9769372940063477,
      "learning_rate": 7.012923186675144e-05,
      "loss": 4.6462,
      "step": 6370
    },
    {
      "epoch": 0.6371,
      "grad_norm": 5.539254665374756,
      "learning_rate": 7.009592077439134e-05,
      "loss": 4.504,
      "step": 6371
    },
    {
      "epoch": 0.6372,
      "grad_norm": 1.9705007076263428,
      "learning_rate": 7.00626133257526e-05,
      "loss": 4.4691,
      "step": 6372
    },
    {
      "epoch": 0.6373,
      "grad_norm": 2.1762938499450684,
      "learning_rate": 7.002930952489362e-05,
      "loss": 4.4345,
      "step": 6373
    },
    {
      "epoch": 0.6374,
      "grad_norm": 12.466574668884277,
      "learning_rate": 6.999600937587239e-05,
      "loss": 4.9427,
      "step": 6374
    },
    {
      "epoch": 0.6375,
      "grad_norm": 3.136801242828369,
      "learning_rate": 6.996271288274636e-05,
      "loss": 4.947,
      "step": 6375
    },
    {
      "epoch": 0.6376,
      "grad_norm": 2.5035507678985596,
      "learning_rate": 6.992942004957271e-05,
      "loss": 4.7947,
      "step": 6376
    },
    {
      "epoch": 0.6377,
      "grad_norm": 4.724710941314697,
      "learning_rate": 6.989613088040796e-05,
      "loss": 4.7464,
      "step": 6377
    },
    {
      "epoch": 0.6378,
      "grad_norm": 2.4508020877838135,
      "learning_rate": 6.986284537930838e-05,
      "loss": 4.3518,
      "step": 6378
    },
    {
      "epoch": 0.6379,
      "grad_norm": 12.113822937011719,
      "learning_rate": 6.982956355032968e-05,
      "loss": 5.8346,
      "step": 6379
    },
    {
      "epoch": 0.638,
      "grad_norm": 5.1807169914245605,
      "learning_rate": 6.979628539752711e-05,
      "loss": 5.1374,
      "step": 6380
    },
    {
      "epoch": 0.6381,
      "grad_norm": 8.158174514770508,
      "learning_rate": 6.976301092495556e-05,
      "loss": 6.2867,
      "step": 6381
    },
    {
      "epoch": 0.6382,
      "grad_norm": 2.771693468093872,
      "learning_rate": 6.972974013666942e-05,
      "loss": 5.0092,
      "step": 6382
    },
    {
      "epoch": 0.6383,
      "grad_norm": 2.4947433471679688,
      "learning_rate": 6.969647303672262e-05,
      "loss": 4.6447,
      "step": 6383
    },
    {
      "epoch": 0.6384,
      "grad_norm": 2.1372828483581543,
      "learning_rate": 6.966320962916864e-05,
      "loss": 4.7206,
      "step": 6384
    },
    {
      "epoch": 0.6385,
      "grad_norm": 4.157040119171143,
      "learning_rate": 6.962994991806059e-05,
      "loss": 5.5965,
      "step": 6385
    },
    {
      "epoch": 0.6386,
      "grad_norm": 3.343569278717041,
      "learning_rate": 6.959669390745097e-05,
      "loss": 5.1194,
      "step": 6386
    },
    {
      "epoch": 0.6387,
      "grad_norm": 3.939640760421753,
      "learning_rate": 6.956344160139201e-05,
      "loss": 4.8654,
      "step": 6387
    },
    {
      "epoch": 0.6388,
      "grad_norm": 4.724053859710693,
      "learning_rate": 6.953019300393538e-05,
      "loss": 5.1793,
      "step": 6388
    },
    {
      "epoch": 0.6389,
      "grad_norm": 2.1112911701202393,
      "learning_rate": 6.949694811913225e-05,
      "loss": 4.5635,
      "step": 6389
    },
    {
      "epoch": 0.639,
      "grad_norm": 3.7595248222351074,
      "learning_rate": 6.946370695103353e-05,
      "loss": 4.6723,
      "step": 6390
    },
    {
      "epoch": 0.6391,
      "grad_norm": 7.885161876678467,
      "learning_rate": 6.943046950368944e-05,
      "loss": 4.446,
      "step": 6391
    },
    {
      "epoch": 0.6392,
      "grad_norm": 2.993622064590454,
      "learning_rate": 6.939723578114993e-05,
      "loss": 4.8491,
      "step": 6392
    },
    {
      "epoch": 0.6393,
      "grad_norm": 4.459775447845459,
      "learning_rate": 6.93640057874644e-05,
      "loss": 4.5831,
      "step": 6393
    },
    {
      "epoch": 0.6394,
      "grad_norm": 3.0175745487213135,
      "learning_rate": 6.93307795266819e-05,
      "loss": 4.9277,
      "step": 6394
    },
    {
      "epoch": 0.6395,
      "grad_norm": 2.93483829498291,
      "learning_rate": 6.929755700285081e-05,
      "loss": 4.5356,
      "step": 6395
    },
    {
      "epoch": 0.6396,
      "grad_norm": 6.496286869049072,
      "learning_rate": 6.92643382200193e-05,
      "loss": 5.076,
      "step": 6396
    },
    {
      "epoch": 0.6397,
      "grad_norm": 3.1899237632751465,
      "learning_rate": 6.923112318223496e-05,
      "loss": 4.7178,
      "step": 6397
    },
    {
      "epoch": 0.6398,
      "grad_norm": 3.10856556892395,
      "learning_rate": 6.91979118935449e-05,
      "loss": 4.5946,
      "step": 6398
    },
    {
      "epoch": 0.6399,
      "grad_norm": 3.0957071781158447,
      "learning_rate": 6.916470435799587e-05,
      "loss": 4.5873,
      "step": 6399
    },
    {
      "epoch": 0.64,
      "grad_norm": 2.785719394683838,
      "learning_rate": 6.913150057963404e-05,
      "loss": 4.73,
      "step": 6400
    },
    {
      "epoch": 0.6401,
      "grad_norm": 2.411360740661621,
      "learning_rate": 6.909830056250527e-05,
      "loss": 4.8245,
      "step": 6401
    },
    {
      "epoch": 0.6402,
      "grad_norm": 4.3473711013793945,
      "learning_rate": 6.90651043106548e-05,
      "loss": 4.5453,
      "step": 6402
    },
    {
      "epoch": 0.6403,
      "grad_norm": 5.073803901672363,
      "learning_rate": 6.90319118281276e-05,
      "loss": 4.9258,
      "step": 6403
    },
    {
      "epoch": 0.6404,
      "grad_norm": 7.180359363555908,
      "learning_rate": 6.899872311896795e-05,
      "loss": 6.0736,
      "step": 6404
    },
    {
      "epoch": 0.6405,
      "grad_norm": 2.8183717727661133,
      "learning_rate": 6.896553818721989e-05,
      "loss": 4.5622,
      "step": 6405
    },
    {
      "epoch": 0.6406,
      "grad_norm": 2.523195505142212,
      "learning_rate": 6.893235703692685e-05,
      "loss": 4.6509,
      "step": 6406
    },
    {
      "epoch": 0.6407,
      "grad_norm": 2.4693031311035156,
      "learning_rate": 6.889917967213185e-05,
      "loss": 4.4503,
      "step": 6407
    },
    {
      "epoch": 0.6408,
      "grad_norm": 2.8809149265289307,
      "learning_rate": 6.88660060968775e-05,
      "loss": 5.181,
      "step": 6408
    },
    {
      "epoch": 0.6409,
      "grad_norm": 2.943193197250366,
      "learning_rate": 6.883283631520582e-05,
      "loss": 4.6169,
      "step": 6409
    },
    {
      "epoch": 0.641,
      "grad_norm": 2.07342267036438,
      "learning_rate": 6.879967033115853e-05,
      "loss": 5.3169,
      "step": 6410
    },
    {
      "epoch": 0.6411,
      "grad_norm": 2.516876459121704,
      "learning_rate": 6.876650814877674e-05,
      "loss": 5.3808,
      "step": 6411
    },
    {
      "epoch": 0.6412,
      "grad_norm": 3.5156798362731934,
      "learning_rate": 6.873334977210122e-05,
      "loss": 5.0477,
      "step": 6412
    },
    {
      "epoch": 0.6413,
      "grad_norm": 2.306854486465454,
      "learning_rate": 6.870019520517217e-05,
      "loss": 5.1716,
      "step": 6413
    },
    {
      "epoch": 0.6414,
      "grad_norm": 2.391256332397461,
      "learning_rate": 6.866704445202943e-05,
      "loss": 4.3888,
      "step": 6414
    },
    {
      "epoch": 0.6415,
      "grad_norm": 2.2195680141448975,
      "learning_rate": 6.863389751671225e-05,
      "loss": 4.4979,
      "step": 6415
    },
    {
      "epoch": 0.6416,
      "grad_norm": 3.5830280780792236,
      "learning_rate": 6.860075440325951e-05,
      "loss": 5.1611,
      "step": 6416
    },
    {
      "epoch": 0.6417,
      "grad_norm": 5.443673610687256,
      "learning_rate": 6.856761511570963e-05,
      "loss": 4.2663,
      "step": 6417
    },
    {
      "epoch": 0.6418,
      "grad_norm": 4.801024913787842,
      "learning_rate": 6.853447965810046e-05,
      "loss": 4.4588,
      "step": 6418
    },
    {
      "epoch": 0.6419,
      "grad_norm": 3.7212140560150146,
      "learning_rate": 6.850134803446954e-05,
      "loss": 4.6726,
      "step": 6419
    },
    {
      "epoch": 0.642,
      "grad_norm": 2.082162618637085,
      "learning_rate": 6.846822024885379e-05,
      "loss": 4.7156,
      "step": 6420
    },
    {
      "epoch": 0.6421,
      "grad_norm": 7.279184818267822,
      "learning_rate": 6.843509630528977e-05,
      "loss": 5.0928,
      "step": 6421
    },
    {
      "epoch": 0.6422,
      "grad_norm": 2.5886569023132324,
      "learning_rate": 6.840197620781349e-05,
      "loss": 4.4469,
      "step": 6422
    },
    {
      "epoch": 0.6423,
      "grad_norm": 3.874415397644043,
      "learning_rate": 6.836885996046061e-05,
      "loss": 5.2997,
      "step": 6423
    },
    {
      "epoch": 0.6424,
      "grad_norm": 1.961138129234314,
      "learning_rate": 6.833574756726618e-05,
      "loss": 4.7145,
      "step": 6424
    },
    {
      "epoch": 0.6425,
      "grad_norm": 4.65975284576416,
      "learning_rate": 6.830263903226483e-05,
      "loss": 5.5878,
      "step": 6425
    },
    {
      "epoch": 0.6426,
      "grad_norm": 3.23386549949646,
      "learning_rate": 6.826953435949081e-05,
      "loss": 4.9145,
      "step": 6426
    },
    {
      "epoch": 0.6427,
      "grad_norm": 3.8384532928466797,
      "learning_rate": 6.823643355297773e-05,
      "loss": 4.5931,
      "step": 6427
    },
    {
      "epoch": 0.6428,
      "grad_norm": 2.352092981338501,
      "learning_rate": 6.820333661675893e-05,
      "loss": 4.8724,
      "step": 6428
    },
    {
      "epoch": 0.6429,
      "grad_norm": 3.5900685787200928,
      "learning_rate": 6.817024355486706e-05,
      "loss": 4.598,
      "step": 6429
    },
    {
      "epoch": 0.643,
      "grad_norm": 2.612413167953491,
      "learning_rate": 6.81371543713345e-05,
      "loss": 4.8929,
      "step": 6430
    },
    {
      "epoch": 0.6431,
      "grad_norm": 3.2009403705596924,
      "learning_rate": 6.8104069070193e-05,
      "loss": 4.8538,
      "step": 6431
    },
    {
      "epoch": 0.6432,
      "grad_norm": 3.903242826461792,
      "learning_rate": 6.807098765547397e-05,
      "loss": 4.8809,
      "step": 6432
    },
    {
      "epoch": 0.6433,
      "grad_norm": 5.650837421417236,
      "learning_rate": 6.803791013120822e-05,
      "loss": 5.2946,
      "step": 6433
    },
    {
      "epoch": 0.6434,
      "grad_norm": 5.392282485961914,
      "learning_rate": 6.800483650142617e-05,
      "loss": 4.7648,
      "step": 6434
    },
    {
      "epoch": 0.6435,
      "grad_norm": 2.1924667358398438,
      "learning_rate": 6.797176677015775e-05,
      "loss": 4.9784,
      "step": 6435
    },
    {
      "epoch": 0.6436,
      "grad_norm": 2.627382516860962,
      "learning_rate": 6.793870094143238e-05,
      "loss": 4.7682,
      "step": 6436
    },
    {
      "epoch": 0.6437,
      "grad_norm": 1.9848096370697021,
      "learning_rate": 6.790563901927907e-05,
      "loss": 4.7766,
      "step": 6437
    },
    {
      "epoch": 0.6438,
      "grad_norm": 3.4778637886047363,
      "learning_rate": 6.787258100772627e-05,
      "loss": 4.5939,
      "step": 6438
    },
    {
      "epoch": 0.6439,
      "grad_norm": 3.0456759929656982,
      "learning_rate": 6.783952691080203e-05,
      "loss": 4.9759,
      "step": 6439
    },
    {
      "epoch": 0.644,
      "grad_norm": 3.0269110202789307,
      "learning_rate": 6.780647673253391e-05,
      "loss": 4.6252,
      "step": 6440
    },
    {
      "epoch": 0.6441,
      "grad_norm": 2.174715995788574,
      "learning_rate": 6.77734304769489e-05,
      "loss": 4.7133,
      "step": 6441
    },
    {
      "epoch": 0.6442,
      "grad_norm": 2.501497983932495,
      "learning_rate": 6.774038814807369e-05,
      "loss": 4.7267,
      "step": 6442
    },
    {
      "epoch": 0.6443,
      "grad_norm": 3.632874011993408,
      "learning_rate": 6.770734974993426e-05,
      "loss": 4.7538,
      "step": 6443
    },
    {
      "epoch": 0.6444,
      "grad_norm": 3.9601099491119385,
      "learning_rate": 6.767431528655635e-05,
      "loss": 5.1858,
      "step": 6444
    },
    {
      "epoch": 0.6445,
      "grad_norm": 2.400667428970337,
      "learning_rate": 6.764128476196505e-05,
      "loss": 4.8705,
      "step": 6445
    },
    {
      "epoch": 0.6446,
      "grad_norm": 4.705976963043213,
      "learning_rate": 6.760825818018508e-05,
      "loss": 5.0774,
      "step": 6446
    },
    {
      "epoch": 0.6447,
      "grad_norm": 2.2248568534851074,
      "learning_rate": 6.757523554524056e-05,
      "loss": 4.3045,
      "step": 6447
    },
    {
      "epoch": 0.6448,
      "grad_norm": 3.48330020904541,
      "learning_rate": 6.754221686115525e-05,
      "loss": 4.922,
      "step": 6448
    },
    {
      "epoch": 0.6449,
      "grad_norm": 2.0520241260528564,
      "learning_rate": 6.750920213195238e-05,
      "loss": 4.6254,
      "step": 6449
    },
    {
      "epoch": 0.645,
      "grad_norm": 3.646080493927002,
      "learning_rate": 6.747619136165463e-05,
      "loss": 4.7327,
      "step": 6450
    },
    {
      "epoch": 0.6451,
      "grad_norm": 2.9352097511291504,
      "learning_rate": 6.744318455428436e-05,
      "loss": 4.1627,
      "step": 6451
    },
    {
      "epoch": 0.6452,
      "grad_norm": 2.320873260498047,
      "learning_rate": 6.741018171386326e-05,
      "loss": 4.7994,
      "step": 6452
    },
    {
      "epoch": 0.6453,
      "grad_norm": 5.698821067810059,
      "learning_rate": 6.737718284441267e-05,
      "loss": 4.2566,
      "step": 6453
    },
    {
      "epoch": 0.6454,
      "grad_norm": 2.6182291507720947,
      "learning_rate": 6.734418794995337e-05,
      "loss": 4.9799,
      "step": 6454
    },
    {
      "epoch": 0.6455,
      "grad_norm": 2.9251067638397217,
      "learning_rate": 6.731119703450577e-05,
      "loss": 5.0924,
      "step": 6455
    },
    {
      "epoch": 0.6456,
      "grad_norm": 2.9507932662963867,
      "learning_rate": 6.727821010208962e-05,
      "loss": 5.2451,
      "step": 6456
    },
    {
      "epoch": 0.6457,
      "grad_norm": 2.933641195297241,
      "learning_rate": 6.724522715672432e-05,
      "loss": 4.564,
      "step": 6457
    },
    {
      "epoch": 0.6458,
      "grad_norm": 2.4691433906555176,
      "learning_rate": 6.721224820242875e-05,
      "loss": 4.3214,
      "step": 6458
    },
    {
      "epoch": 0.6459,
      "grad_norm": 3.5240726470947266,
      "learning_rate": 6.717927324322124e-05,
      "loss": 5.5497,
      "step": 6459
    },
    {
      "epoch": 0.646,
      "grad_norm": 5.221595764160156,
      "learning_rate": 6.714630228311978e-05,
      "loss": 4.8837,
      "step": 6460
    },
    {
      "epoch": 0.6461,
      "grad_norm": 2.27754282951355,
      "learning_rate": 6.711333532614168e-05,
      "loss": 4.4097,
      "step": 6461
    },
    {
      "epoch": 0.6462,
      "grad_norm": 2.1701204776763916,
      "learning_rate": 6.708037237630395e-05,
      "loss": 4.8891,
      "step": 6462
    },
    {
      "epoch": 0.6463,
      "grad_norm": 3.683351993560791,
      "learning_rate": 6.704741343762295e-05,
      "loss": 4.9054,
      "step": 6463
    },
    {
      "epoch": 0.6464,
      "grad_norm": 2.7007641792297363,
      "learning_rate": 6.701445851411472e-05,
      "loss": 5.8241,
      "step": 6464
    },
    {
      "epoch": 0.6465,
      "grad_norm": 3.9811131954193115,
      "learning_rate": 6.698150760979463e-05,
      "loss": 4.9092,
      "step": 6465
    },
    {
      "epoch": 0.6466,
      "grad_norm": 2.2803797721862793,
      "learning_rate": 6.69485607286777e-05,
      "loss": 4.7379,
      "step": 6466
    },
    {
      "epoch": 0.6467,
      "grad_norm": 2.907571315765381,
      "learning_rate": 6.69156178747784e-05,
      "loss": 5.2085,
      "step": 6467
    },
    {
      "epoch": 0.6468,
      "grad_norm": 3.452497959136963,
      "learning_rate": 6.688267905211067e-05,
      "loss": 4.5198,
      "step": 6468
    },
    {
      "epoch": 0.6469,
      "grad_norm": 2.6536519527435303,
      "learning_rate": 6.684974426468808e-05,
      "loss": 4.9154,
      "step": 6469
    },
    {
      "epoch": 0.647,
      "grad_norm": 4.202033042907715,
      "learning_rate": 6.681681351652356e-05,
      "loss": 4.772,
      "step": 6470
    },
    {
      "epoch": 0.6471,
      "grad_norm": 3.2180540561676025,
      "learning_rate": 6.67838868116297e-05,
      "loss": 5.1528,
      "step": 6471
    },
    {
      "epoch": 0.6472,
      "grad_norm": 3.1170074939727783,
      "learning_rate": 6.675096415401842e-05,
      "loss": 4.1348,
      "step": 6472
    },
    {
      "epoch": 0.6473,
      "grad_norm": 2.503807306289673,
      "learning_rate": 6.671804554770135e-05,
      "loss": 4.6232,
      "step": 6473
    },
    {
      "epoch": 0.6474,
      "grad_norm": 5.473371505737305,
      "learning_rate": 6.668513099668944e-05,
      "loss": 5.2756,
      "step": 6474
    },
    {
      "epoch": 0.6475,
      "grad_norm": 3.2897064685821533,
      "learning_rate": 6.66522205049933e-05,
      "loss": 4.9569,
      "step": 6475
    },
    {
      "epoch": 0.6476,
      "grad_norm": 2.3644633293151855,
      "learning_rate": 6.661931407662292e-05,
      "loss": 5.0132,
      "step": 6476
    },
    {
      "epoch": 0.6477,
      "grad_norm": 2.9611446857452393,
      "learning_rate": 6.658641171558785e-05,
      "loss": 4.8622,
      "step": 6477
    },
    {
      "epoch": 0.6478,
      "grad_norm": 2.0347883701324463,
      "learning_rate": 6.65535134258972e-05,
      "loss": 5.0706,
      "step": 6478
    },
    {
      "epoch": 0.6479,
      "grad_norm": 2.2280962467193604,
      "learning_rate": 6.652061921155943e-05,
      "loss": 4.4201,
      "step": 6479
    },
    {
      "epoch": 0.648,
      "grad_norm": 2.7920079231262207,
      "learning_rate": 6.648772907658272e-05,
      "loss": 4.4181,
      "step": 6480
    },
    {
      "epoch": 0.6481,
      "grad_norm": 2.8880162239074707,
      "learning_rate": 6.64548430249745e-05,
      "loss": 5.2655,
      "step": 6481
    },
    {
      "epoch": 0.6482,
      "grad_norm": 2.362459659576416,
      "learning_rate": 6.642196106074194e-05,
      "loss": 4.7745,
      "step": 6482
    },
    {
      "epoch": 0.6483,
      "grad_norm": 2.0626606941223145,
      "learning_rate": 6.638908318789156e-05,
      "loss": 4.666,
      "step": 6483
    },
    {
      "epoch": 0.6484,
      "grad_norm": 1.7069272994995117,
      "learning_rate": 6.635620941042945e-05,
      "loss": 4.9286,
      "step": 6484
    },
    {
      "epoch": 0.6485,
      "grad_norm": 3.1149590015411377,
      "learning_rate": 6.63233397323612e-05,
      "loss": 4.6228,
      "step": 6485
    },
    {
      "epoch": 0.6486,
      "grad_norm": 2.837143659591675,
      "learning_rate": 6.62904741576918e-05,
      "loss": 4.8289,
      "step": 6486
    },
    {
      "epoch": 0.6487,
      "grad_norm": 2.029904365539551,
      "learning_rate": 6.62576126904259e-05,
      "loss": 4.6212,
      "step": 6487
    },
    {
      "epoch": 0.6488,
      "grad_norm": 8.48405647277832,
      "learning_rate": 6.622475533456751e-05,
      "loss": 5.5275,
      "step": 6488
    },
    {
      "epoch": 0.6489,
      "grad_norm": 3.484537363052368,
      "learning_rate": 6.61919020941203e-05,
      "loss": 5.6563,
      "step": 6489
    },
    {
      "epoch": 0.649,
      "grad_norm": 3.473278284072876,
      "learning_rate": 6.61590529730872e-05,
      "loss": 5.2385,
      "step": 6490
    },
    {
      "epoch": 0.6491,
      "grad_norm": 4.066305637359619,
      "learning_rate": 6.612620797547087e-05,
      "loss": 5.433,
      "step": 6491
    },
    {
      "epoch": 0.6492,
      "grad_norm": 5.890203475952148,
      "learning_rate": 6.609336710527332e-05,
      "loss": 5.4073,
      "step": 6492
    },
    {
      "epoch": 0.6493,
      "grad_norm": 3.016288995742798,
      "learning_rate": 6.606053036649619e-05,
      "loss": 4.614,
      "step": 6493
    },
    {
      "epoch": 0.6494,
      "grad_norm": 2.596649169921875,
      "learning_rate": 6.602769776314049e-05,
      "loss": 4.1023,
      "step": 6494
    },
    {
      "epoch": 0.6495,
      "grad_norm": 2.3513758182525635,
      "learning_rate": 6.599486929920673e-05,
      "loss": 4.6345,
      "step": 6495
    },
    {
      "epoch": 0.6496,
      "grad_norm": 2.5235910415649414,
      "learning_rate": 6.5962044978695e-05,
      "loss": 4.2677,
      "step": 6496
    },
    {
      "epoch": 0.6497,
      "grad_norm": 4.581762790679932,
      "learning_rate": 6.592922480560483e-05,
      "loss": 5.3506,
      "step": 6497
    },
    {
      "epoch": 0.6498,
      "grad_norm": 4.442892551422119,
      "learning_rate": 6.589640878393531e-05,
      "loss": 5.8638,
      "step": 6498
    },
    {
      "epoch": 0.6499,
      "grad_norm": 3.3184561729431152,
      "learning_rate": 6.58635969176849e-05,
      "loss": 5.2803,
      "step": 6499
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.3881170749664307,
      "learning_rate": 6.583078921085167e-05,
      "loss": 4.7612,
      "step": 6500
    },
    {
      "epoch": 0.6501,
      "grad_norm": 2.2057316303253174,
      "learning_rate": 6.579798566743314e-05,
      "loss": 4.6829,
      "step": 6501
    },
    {
      "epoch": 0.6502,
      "grad_norm": 2.4529685974121094,
      "learning_rate": 6.57651862914263e-05,
      "loss": 4.7283,
      "step": 6502
    },
    {
      "epoch": 0.6503,
      "grad_norm": 6.080167293548584,
      "learning_rate": 6.573239108682768e-05,
      "loss": 4.6261,
      "step": 6503
    },
    {
      "epoch": 0.6504,
      "grad_norm": 2.5301156044006348,
      "learning_rate": 6.569960005763323e-05,
      "loss": 4.0557,
      "step": 6504
    },
    {
      "epoch": 0.6505,
      "grad_norm": 3.355543613433838,
      "learning_rate": 6.566681320783849e-05,
      "loss": 4.7184,
      "step": 6505
    },
    {
      "epoch": 0.6506,
      "grad_norm": 4.885681629180908,
      "learning_rate": 6.56340305414384e-05,
      "loss": 5.0273,
      "step": 6506
    },
    {
      "epoch": 0.6507,
      "grad_norm": 4.20935583114624,
      "learning_rate": 6.560125206242746e-05,
      "loss": 5.259,
      "step": 6507
    },
    {
      "epoch": 0.6508,
      "grad_norm": 4.369637489318848,
      "learning_rate": 6.55684777747996e-05,
      "loss": 5.037,
      "step": 6508
    },
    {
      "epoch": 0.6509,
      "grad_norm": 2.8199424743652344,
      "learning_rate": 6.55357076825483e-05,
      "loss": 4.9884,
      "step": 6509
    },
    {
      "epoch": 0.651,
      "grad_norm": 5.208587646484375,
      "learning_rate": 6.550294178966647e-05,
      "loss": 4.772,
      "step": 6510
    },
    {
      "epoch": 0.6511,
      "grad_norm": 2.8438525199890137,
      "learning_rate": 6.547018010014654e-05,
      "loss": 4.522,
      "step": 6511
    },
    {
      "epoch": 0.6512,
      "grad_norm": 2.147761344909668,
      "learning_rate": 6.543742261798045e-05,
      "loss": 5.261,
      "step": 6512
    },
    {
      "epoch": 0.6513,
      "grad_norm": 2.879274368286133,
      "learning_rate": 6.540466934715953e-05,
      "loss": 4.8693,
      "step": 6513
    },
    {
      "epoch": 0.6514,
      "grad_norm": 5.0659685134887695,
      "learning_rate": 6.537192029167474e-05,
      "loss": 4.9974,
      "step": 6514
    },
    {
      "epoch": 0.6515,
      "grad_norm": 3.0138838291168213,
      "learning_rate": 6.53391754555164e-05,
      "loss": 4.6263,
      "step": 6515
    },
    {
      "epoch": 0.6516,
      "grad_norm": 3.3754796981811523,
      "learning_rate": 6.530643484267443e-05,
      "loss": 4.8143,
      "step": 6516
    },
    {
      "epoch": 0.6517,
      "grad_norm": 2.6730358600616455,
      "learning_rate": 6.52736984571381e-05,
      "loss": 4.5644,
      "step": 6517
    },
    {
      "epoch": 0.6518,
      "grad_norm": 2.338937759399414,
      "learning_rate": 6.52409663028963e-05,
      "loss": 4.9069,
      "step": 6518
    },
    {
      "epoch": 0.6519,
      "grad_norm": 3.2276744842529297,
      "learning_rate": 6.520823838393731e-05,
      "loss": 5.0803,
      "step": 6519
    },
    {
      "epoch": 0.652,
      "grad_norm": 3.1610124111175537,
      "learning_rate": 6.517551470424895e-05,
      "loss": 5.02,
      "step": 6520
    },
    {
      "epoch": 0.6521,
      "grad_norm": 2.2030932903289795,
      "learning_rate": 6.51427952678185e-05,
      "loss": 4.6221,
      "step": 6521
    },
    {
      "epoch": 0.6522,
      "grad_norm": 3.719759702682495,
      "learning_rate": 6.511008007863268e-05,
      "loss": 4.2451,
      "step": 6522
    },
    {
      "epoch": 0.6523,
      "grad_norm": 2.053889036178589,
      "learning_rate": 6.507736914067781e-05,
      "loss": 4.7466,
      "step": 6523
    },
    {
      "epoch": 0.6524,
      "grad_norm": 3.179501533508301,
      "learning_rate": 6.504466245793955e-05,
      "loss": 4.7256,
      "step": 6524
    },
    {
      "epoch": 0.6525,
      "grad_norm": 2.1678712368011475,
      "learning_rate": 6.501196003440314e-05,
      "loss": 4.145,
      "step": 6525
    },
    {
      "epoch": 0.6526,
      "grad_norm": 2.2742679119110107,
      "learning_rate": 6.497926187405326e-05,
      "loss": 4.9628,
      "step": 6526
    },
    {
      "epoch": 0.6527,
      "grad_norm": 3.5380499362945557,
      "learning_rate": 6.494656798087412e-05,
      "loss": 4.4173,
      "step": 6527
    },
    {
      "epoch": 0.6528,
      "grad_norm": 2.071281909942627,
      "learning_rate": 6.49138783588493e-05,
      "loss": 4.5032,
      "step": 6528
    },
    {
      "epoch": 0.6529,
      "grad_norm": 2.2778656482696533,
      "learning_rate": 6.488119301196201e-05,
      "loss": 4.7479,
      "step": 6529
    },
    {
      "epoch": 0.653,
      "grad_norm": 2.4715614318847656,
      "learning_rate": 6.484851194419484e-05,
      "loss": 5.1473,
      "step": 6530
    },
    {
      "epoch": 0.6531,
      "grad_norm": 2.202962875366211,
      "learning_rate": 6.481583515952983e-05,
      "loss": 5.1067,
      "step": 6531
    },
    {
      "epoch": 0.6532,
      "grad_norm": 2.564614772796631,
      "learning_rate": 6.478316266194861e-05,
      "loss": 4.3921,
      "step": 6532
    },
    {
      "epoch": 0.6533,
      "grad_norm": 2.346012830734253,
      "learning_rate": 6.475049445543215e-05,
      "loss": 4.5278,
      "step": 6533
    },
    {
      "epoch": 0.6534,
      "grad_norm": 2.2895829677581787,
      "learning_rate": 6.471783054396105e-05,
      "loss": 5.037,
      "step": 6534
    },
    {
      "epoch": 0.6535,
      "grad_norm": 4.605841159820557,
      "learning_rate": 6.468517093151525e-05,
      "loss": 4.7374,
      "step": 6535
    },
    {
      "epoch": 0.6536,
      "grad_norm": 2.460458755493164,
      "learning_rate": 6.465251562207431e-05,
      "loss": 4.8186,
      "step": 6536
    },
    {
      "epoch": 0.6537,
      "grad_norm": 3.581894636154175,
      "learning_rate": 6.461986461961706e-05,
      "loss": 4.3678,
      "step": 6537
    },
    {
      "epoch": 0.6538,
      "grad_norm": 3.2136623859405518,
      "learning_rate": 6.458721792812204e-05,
      "loss": 4.8287,
      "step": 6538
    },
    {
      "epoch": 0.6539,
      "grad_norm": 2.897613763809204,
      "learning_rate": 6.455457555156705e-05,
      "loss": 4.5014,
      "step": 6539
    },
    {
      "epoch": 0.654,
      "grad_norm": 2.8079123497009277,
      "learning_rate": 6.452193749392952e-05,
      "loss": 4.6401,
      "step": 6540
    },
    {
      "epoch": 0.6541,
      "grad_norm": 2.51064133644104,
      "learning_rate": 6.448930375918631e-05,
      "loss": 4.552,
      "step": 6541
    },
    {
      "epoch": 0.6542,
      "grad_norm": 2.0984489917755127,
      "learning_rate": 6.44566743513137e-05,
      "loss": 5.0143,
      "step": 6542
    },
    {
      "epoch": 0.6543,
      "grad_norm": 2.89330792427063,
      "learning_rate": 6.442404927428751e-05,
      "loss": 5.1426,
      "step": 6543
    },
    {
      "epoch": 0.6544,
      "grad_norm": 3.498283624649048,
      "learning_rate": 6.4391428532083e-05,
      "loss": 4.9656,
      "step": 6544
    },
    {
      "epoch": 0.6545,
      "grad_norm": 3.4336600303649902,
      "learning_rate": 6.435881212867493e-05,
      "loss": 5.2395,
      "step": 6545
    },
    {
      "epoch": 0.6546,
      "grad_norm": 2.5972440242767334,
      "learning_rate": 6.432620006803746e-05,
      "loss": 4.7604,
      "step": 6546
    },
    {
      "epoch": 0.6547,
      "grad_norm": 2.9604427814483643,
      "learning_rate": 6.429359235414436e-05,
      "loss": 4.3775,
      "step": 6547
    },
    {
      "epoch": 0.6548,
      "grad_norm": 3.994678020477295,
      "learning_rate": 6.426098899096868e-05,
      "loss": 5.8169,
      "step": 6548
    },
    {
      "epoch": 0.6549,
      "grad_norm": 2.1263253688812256,
      "learning_rate": 6.422838998248307e-05,
      "loss": 4.8282,
      "step": 6549
    },
    {
      "epoch": 0.655,
      "grad_norm": 4.224761486053467,
      "learning_rate": 6.419579533265968e-05,
      "loss": 4.8467,
      "step": 6550
    },
    {
      "epoch": 0.6551,
      "grad_norm": 2.273690938949585,
      "learning_rate": 6.416320504546997e-05,
      "loss": 4.8025,
      "step": 6551
    },
    {
      "epoch": 0.6552,
      "grad_norm": 2.3039565086364746,
      "learning_rate": 6.413061912488508e-05,
      "loss": 4.6889,
      "step": 6552
    },
    {
      "epoch": 0.6553,
      "grad_norm": 3.147681951522827,
      "learning_rate": 6.409803757487538e-05,
      "loss": 5.0996,
      "step": 6553
    },
    {
      "epoch": 0.6554,
      "grad_norm": 2.6141300201416016,
      "learning_rate": 6.406546039941094e-05,
      "loss": 4.5802,
      "step": 6554
    },
    {
      "epoch": 0.6555,
      "grad_norm": 2.473703622817993,
      "learning_rate": 6.403288760246112e-05,
      "loss": 4.2019,
      "step": 6555
    },
    {
      "epoch": 0.6556,
      "grad_norm": 2.443788766860962,
      "learning_rate": 6.40003191879949e-05,
      "loss": 4.9323,
      "step": 6556
    },
    {
      "epoch": 0.6557,
      "grad_norm": 2.798102855682373,
      "learning_rate": 6.396775515998055e-05,
      "loss": 4.8144,
      "step": 6557
    },
    {
      "epoch": 0.6558,
      "grad_norm": 2.398477077484131,
      "learning_rate": 6.393519552238591e-05,
      "loss": 4.3256,
      "step": 6558
    },
    {
      "epoch": 0.6559,
      "grad_norm": 2.8075814247131348,
      "learning_rate": 6.390264027917836e-05,
      "loss": 4.6298,
      "step": 6559
    },
    {
      "epoch": 0.656,
      "grad_norm": 2.2271840572357178,
      "learning_rate": 6.387008943432455e-05,
      "loss": 4.4061,
      "step": 6560
    },
    {
      "epoch": 0.6561,
      "grad_norm": 2.944754123687744,
      "learning_rate": 6.383754299179079e-05,
      "loss": 4.7678,
      "step": 6561
    },
    {
      "epoch": 0.6562,
      "grad_norm": 1.9114500284194946,
      "learning_rate": 6.380500095554268e-05,
      "loss": 4.7029,
      "step": 6562
    },
    {
      "epoch": 0.6563,
      "grad_norm": 2.581695795059204,
      "learning_rate": 6.377246332954544e-05,
      "loss": 4.3411,
      "step": 6563
    },
    {
      "epoch": 0.6564,
      "grad_norm": 2.8499531745910645,
      "learning_rate": 6.373993011776367e-05,
      "loss": 5.2819,
      "step": 6564
    },
    {
      "epoch": 0.6565,
      "grad_norm": 2.3932578563690186,
      "learning_rate": 6.370740132416138e-05,
      "loss": 4.6225,
      "step": 6565
    },
    {
      "epoch": 0.6566,
      "grad_norm": 3.4358034133911133,
      "learning_rate": 6.367487695270217e-05,
      "loss": 4.3391,
      "step": 6566
    },
    {
      "epoch": 0.6567,
      "grad_norm": 2.477341651916504,
      "learning_rate": 6.364235700734903e-05,
      "loss": 4.7018,
      "step": 6567
    },
    {
      "epoch": 0.6568,
      "grad_norm": 3.328251361846924,
      "learning_rate": 6.360984149206439e-05,
      "loss": 4.3111,
      "step": 6568
    },
    {
      "epoch": 0.6569,
      "grad_norm": 4.123172760009766,
      "learning_rate": 6.357733041081018e-05,
      "loss": 5.1617,
      "step": 6569
    },
    {
      "epoch": 0.657,
      "grad_norm": 4.42378044128418,
      "learning_rate": 6.35448237675478e-05,
      "loss": 5.4658,
      "step": 6570
    },
    {
      "epoch": 0.6571,
      "grad_norm": 2.5106513500213623,
      "learning_rate": 6.351232156623803e-05,
      "loss": 4.5449,
      "step": 6571
    },
    {
      "epoch": 0.6572,
      "grad_norm": 3.573621988296509,
      "learning_rate": 6.347982381084123e-05,
      "loss": 5.2262,
      "step": 6572
    },
    {
      "epoch": 0.6573,
      "grad_norm": 2.2074480056762695,
      "learning_rate": 6.344733050531713e-05,
      "loss": 4.6612,
      "step": 6573
    },
    {
      "epoch": 0.6574,
      "grad_norm": 2.0092458724975586,
      "learning_rate": 6.341484165362487e-05,
      "loss": 4.8422,
      "step": 6574
    },
    {
      "epoch": 0.6575,
      "grad_norm": 2.506293535232544,
      "learning_rate": 6.338235725972325e-05,
      "loss": 4.3956,
      "step": 6575
    },
    {
      "epoch": 0.6576,
      "grad_norm": 3.2847628593444824,
      "learning_rate": 6.334987732757029e-05,
      "loss": 4.7787,
      "step": 6576
    },
    {
      "epoch": 0.6577,
      "grad_norm": 5.31828498840332,
      "learning_rate": 6.33174018611236e-05,
      "loss": 4.7096,
      "step": 6577
    },
    {
      "epoch": 0.6578,
      "grad_norm": 5.242201328277588,
      "learning_rate": 6.328493086434023e-05,
      "loss": 5.1698,
      "step": 6578
    },
    {
      "epoch": 0.6579,
      "grad_norm": 2.719313383102417,
      "learning_rate": 6.325246434117668e-05,
      "loss": 5.0352,
      "step": 6579
    },
    {
      "epoch": 0.658,
      "grad_norm": 2.183668851852417,
      "learning_rate": 6.322000229558887e-05,
      "loss": 5.0484,
      "step": 6580
    },
    {
      "epoch": 0.6581,
      "grad_norm": 3.3864779472351074,
      "learning_rate": 6.318754473153221e-05,
      "loss": 4.9692,
      "step": 6581
    },
    {
      "epoch": 0.6582,
      "grad_norm": 3.537602424621582,
      "learning_rate": 6.315509165296159e-05,
      "loss": 4.9567,
      "step": 6582
    },
    {
      "epoch": 0.6583,
      "grad_norm": 2.806657552719116,
      "learning_rate": 6.312264306383124e-05,
      "loss": 4.9602,
      "step": 6583
    },
    {
      "epoch": 0.6584,
      "grad_norm": 2.3373677730560303,
      "learning_rate": 6.309019896809503e-05,
      "loss": 4.838,
      "step": 6584
    },
    {
      "epoch": 0.6585,
      "grad_norm": 2.728254556655884,
      "learning_rate": 6.305775936970606e-05,
      "loss": 4.8114,
      "step": 6585
    },
    {
      "epoch": 0.6586,
      "grad_norm": 4.084047794342041,
      "learning_rate": 6.302532427261709e-05,
      "loss": 4.8023,
      "step": 6586
    },
    {
      "epoch": 0.6587,
      "grad_norm": 2.861610174179077,
      "learning_rate": 6.299289368078016e-05,
      "loss": 4.7801,
      "step": 6587
    },
    {
      "epoch": 0.6588,
      "grad_norm": 2.997673749923706,
      "learning_rate": 6.296046759814694e-05,
      "loss": 4.5679,
      "step": 6588
    },
    {
      "epoch": 0.6589,
      "grad_norm": 4.128951549530029,
      "learning_rate": 6.292804602866834e-05,
      "loss": 4.6542,
      "step": 6589
    },
    {
      "epoch": 0.659,
      "grad_norm": 3.8585867881774902,
      "learning_rate": 6.289562897629492e-05,
      "loss": 5.1507,
      "step": 6590
    },
    {
      "epoch": 0.6591,
      "grad_norm": 2.2798755168914795,
      "learning_rate": 6.286321644497655e-05,
      "loss": 5.088,
      "step": 6591
    },
    {
      "epoch": 0.6592,
      "grad_norm": 3.3860771656036377,
      "learning_rate": 6.283080843866256e-05,
      "loss": 4.6394,
      "step": 6592
    },
    {
      "epoch": 0.6593,
      "grad_norm": 4.430370330810547,
      "learning_rate": 6.27984049613019e-05,
      "loss": 5.1312,
      "step": 6593
    },
    {
      "epoch": 0.6594,
      "grad_norm": 3.7278599739074707,
      "learning_rate": 6.276600601684267e-05,
      "loss": 5.0766,
      "step": 6594
    },
    {
      "epoch": 0.6595,
      "grad_norm": 2.8325750827789307,
      "learning_rate": 6.27336116092327e-05,
      "loss": 4.6732,
      "step": 6595
    },
    {
      "epoch": 0.6596,
      "grad_norm": 3.243521213531494,
      "learning_rate": 6.27012217424191e-05,
      "loss": 4.982,
      "step": 6596
    },
    {
      "epoch": 0.6597,
      "grad_norm": 2.18959379196167,
      "learning_rate": 6.266883642034853e-05,
      "loss": 4.5756,
      "step": 6597
    },
    {
      "epoch": 0.6598,
      "grad_norm": 2.150588035583496,
      "learning_rate": 6.263645564696696e-05,
      "loss": 4.7448,
      "step": 6598
    },
    {
      "epoch": 0.6599,
      "grad_norm": 3.464353322982788,
      "learning_rate": 6.260407942621998e-05,
      "loss": 5.269,
      "step": 6599
    },
    {
      "epoch": 0.66,
      "grad_norm": 4.516343116760254,
      "learning_rate": 6.257170776205245e-05,
      "loss": 5.3542,
      "step": 6600
    },
    {
      "epoch": 0.6601,
      "grad_norm": 2.594989061355591,
      "learning_rate": 6.25393406584088e-05,
      "loss": 4.5666,
      "step": 6601
    },
    {
      "epoch": 0.6602,
      "grad_norm": 3.3048501014709473,
      "learning_rate": 6.25069781192329e-05,
      "loss": 5.7794,
      "step": 6602
    },
    {
      "epoch": 0.6603,
      "grad_norm": 3.467116355895996,
      "learning_rate": 6.247462014846792e-05,
      "loss": 4.5923,
      "step": 6603
    },
    {
      "epoch": 0.6604,
      "grad_norm": 3.8128745555877686,
      "learning_rate": 6.24422667500567e-05,
      "loss": 4.6406,
      "step": 6604
    },
    {
      "epoch": 0.6605,
      "grad_norm": 2.061635732650757,
      "learning_rate": 6.240991792794133e-05,
      "loss": 4.8568,
      "step": 6605
    },
    {
      "epoch": 0.6606,
      "grad_norm": 2.272500991821289,
      "learning_rate": 6.237757368606345e-05,
      "loss": 5.0022,
      "step": 6606
    },
    {
      "epoch": 0.6607,
      "grad_norm": 2.1731998920440674,
      "learning_rate": 6.234523402836407e-05,
      "loss": 4.9453,
      "step": 6607
    },
    {
      "epoch": 0.6608,
      "grad_norm": 3.793837070465088,
      "learning_rate": 6.231289895878375e-05,
      "loss": 5.0192,
      "step": 6608
    },
    {
      "epoch": 0.6609,
      "grad_norm": 4.924670219421387,
      "learning_rate": 6.228056848126236e-05,
      "loss": 4.4685,
      "step": 6609
    },
    {
      "epoch": 0.661,
      "grad_norm": 2.6252448558807373,
      "learning_rate": 6.224824259973925e-05,
      "loss": 5.1885,
      "step": 6610
    },
    {
      "epoch": 0.6611,
      "grad_norm": 11.40977668762207,
      "learning_rate": 6.22159213181533e-05,
      "loss": 5.0924,
      "step": 6611
    },
    {
      "epoch": 0.6612,
      "grad_norm": 10.023305892944336,
      "learning_rate": 6.21836046404427e-05,
      "loss": 5.6772,
      "step": 6612
    },
    {
      "epoch": 0.6613,
      "grad_norm": 2.002937078475952,
      "learning_rate": 6.215129257054522e-05,
      "loss": 4.9617,
      "step": 6613
    },
    {
      "epoch": 0.6614,
      "grad_norm": 3.643296003341675,
      "learning_rate": 6.211898511239787e-05,
      "loss": 4.8663,
      "step": 6614
    },
    {
      "epoch": 0.6615,
      "grad_norm": 3.395705461502075,
      "learning_rate": 6.20866822699373e-05,
      "loss": 4.6783,
      "step": 6615
    },
    {
      "epoch": 0.6616,
      "grad_norm": 3.5407962799072266,
      "learning_rate": 6.205438404709947e-05,
      "loss": 5.0268,
      "step": 6616
    },
    {
      "epoch": 0.6617,
      "grad_norm": 3.3066720962524414,
      "learning_rate": 6.20220904478199e-05,
      "loss": 5.0189,
      "step": 6617
    },
    {
      "epoch": 0.6618,
      "grad_norm": 2.909424304962158,
      "learning_rate": 6.198980147603339e-05,
      "loss": 4.7347,
      "step": 6618
    },
    {
      "epoch": 0.6619,
      "grad_norm": 11.601919174194336,
      "learning_rate": 6.195751713567427e-05,
      "loss": 4.526,
      "step": 6619
    },
    {
      "epoch": 0.662,
      "grad_norm": 1.9812384843826294,
      "learning_rate": 6.192523743067626e-05,
      "loss": 4.666,
      "step": 6620
    },
    {
      "epoch": 0.6621,
      "grad_norm": 3.8560993671417236,
      "learning_rate": 6.18929623649726e-05,
      "loss": 5.2225,
      "step": 6621
    },
    {
      "epoch": 0.6622,
      "grad_norm": 2.0577926635742188,
      "learning_rate": 6.18606919424959e-05,
      "loss": 4.5391,
      "step": 6622
    },
    {
      "epoch": 0.6623,
      "grad_norm": 2.339890718460083,
      "learning_rate": 6.182842616717817e-05,
      "loss": 4.5263,
      "step": 6623
    },
    {
      "epoch": 0.6624,
      "grad_norm": 2.725858211517334,
      "learning_rate": 6.179616504295092e-05,
      "loss": 4.3136,
      "step": 6624
    },
    {
      "epoch": 0.6625,
      "grad_norm": 4.371527194976807,
      "learning_rate": 6.176390857374507e-05,
      "loss": 4.8185,
      "step": 6625
    },
    {
      "epoch": 0.6626,
      "grad_norm": 2.9174234867095947,
      "learning_rate": 6.173165676349103e-05,
      "loss": 5.2799,
      "step": 6626
    },
    {
      "epoch": 0.6627,
      "grad_norm": 2.4148130416870117,
      "learning_rate": 6.169940961611854e-05,
      "loss": 4.535,
      "step": 6627
    },
    {
      "epoch": 0.6628,
      "grad_norm": 4.041593074798584,
      "learning_rate": 6.166716713555674e-05,
      "loss": 5.391,
      "step": 6628
    },
    {
      "epoch": 0.6629,
      "grad_norm": 3.5079545974731445,
      "learning_rate": 6.163492932573438e-05,
      "loss": 5.2277,
      "step": 6629
    },
    {
      "epoch": 0.663,
      "grad_norm": 2.41165828704834,
      "learning_rate": 6.160269619057951e-05,
      "loss": 4.5446,
      "step": 6630
    },
    {
      "epoch": 0.6631,
      "grad_norm": 4.170302867889404,
      "learning_rate": 6.157046773401964e-05,
      "loss": 5.3375,
      "step": 6631
    },
    {
      "epoch": 0.6632,
      "grad_norm": 5.230420112609863,
      "learning_rate": 6.153824395998168e-05,
      "loss": 4.828,
      "step": 6632
    },
    {
      "epoch": 0.6633,
      "grad_norm": 3.4771640300750732,
      "learning_rate": 6.150602487239206e-05,
      "loss": 4.724,
      "step": 6633
    },
    {
      "epoch": 0.6634,
      "grad_norm": 2.4454848766326904,
      "learning_rate": 6.147381047517655e-05,
      "loss": 4.5176,
      "step": 6634
    },
    {
      "epoch": 0.6635,
      "grad_norm": 5.450838565826416,
      "learning_rate": 6.144160077226036e-05,
      "loss": 4.5869,
      "step": 6635
    },
    {
      "epoch": 0.6636,
      "grad_norm": 4.776638507843018,
      "learning_rate": 6.140939576756817e-05,
      "loss": 5.0358,
      "step": 6636
    },
    {
      "epoch": 0.6637,
      "grad_norm": 2.1021623611450195,
      "learning_rate": 6.1377195465024e-05,
      "loss": 4.3018,
      "step": 6637
    },
    {
      "epoch": 0.6638,
      "grad_norm": 2.2933082580566406,
      "learning_rate": 6.134499986855144e-05,
      "loss": 4.577,
      "step": 6638
    },
    {
      "epoch": 0.6639,
      "grad_norm": 3.5626680850982666,
      "learning_rate": 6.131280898207339e-05,
      "loss": 4.7788,
      "step": 6639
    },
    {
      "epoch": 0.664,
      "grad_norm": 2.179405450820923,
      "learning_rate": 6.128062280951224e-05,
      "loss": 4.5201,
      "step": 6640
    },
    {
      "epoch": 0.6641,
      "grad_norm": 2.160891532897949,
      "learning_rate": 6.12484413547897e-05,
      "loss": 4.1461,
      "step": 6641
    },
    {
      "epoch": 0.6642,
      "grad_norm": 4.166542053222656,
      "learning_rate": 6.121626462182707e-05,
      "loss": 5.7921,
      "step": 6642
    },
    {
      "epoch": 0.6643,
      "grad_norm": 3.0123484134674072,
      "learning_rate": 6.118409261454494e-05,
      "loss": 4.9684,
      "step": 6643
    },
    {
      "epoch": 0.6644,
      "grad_norm": 2.186772108078003,
      "learning_rate": 6.11519253368634e-05,
      "loss": 4.547,
      "step": 6644
    },
    {
      "epoch": 0.6645,
      "grad_norm": 2.6823058128356934,
      "learning_rate": 6.111976279270192e-05,
      "loss": 4.7167,
      "step": 6645
    },
    {
      "epoch": 0.6646,
      "grad_norm": 2.2124929428100586,
      "learning_rate": 6.108760498597938e-05,
      "loss": 4.6802,
      "step": 6646
    },
    {
      "epoch": 0.6647,
      "grad_norm": 4.898305892944336,
      "learning_rate": 6.105545192061416e-05,
      "loss": 4.8582,
      "step": 6647
    },
    {
      "epoch": 0.6648,
      "grad_norm": 3.911170244216919,
      "learning_rate": 6.1023303600523975e-05,
      "loss": 4.6774,
      "step": 6648
    },
    {
      "epoch": 0.6649,
      "grad_norm": 2.7268412113189697,
      "learning_rate": 6.099116002962604e-05,
      "loss": 5.1419,
      "step": 6649
    },
    {
      "epoch": 0.665,
      "grad_norm": 2.751300573348999,
      "learning_rate": 6.09590212118369e-05,
      "loss": 4.4746,
      "step": 6650
    },
    {
      "epoch": 0.6651,
      "grad_norm": 2.9869225025177,
      "learning_rate": 6.092688715107264e-05,
      "loss": 5.3321,
      "step": 6651
    },
    {
      "epoch": 0.6652,
      "grad_norm": 1.6612420082092285,
      "learning_rate": 6.089475785124863e-05,
      "loss": 4.6441,
      "step": 6652
    },
    {
      "epoch": 0.6653,
      "grad_norm": 2.1390514373779297,
      "learning_rate": 6.086263331627976e-05,
      "loss": 4.898,
      "step": 6653
    },
    {
      "epoch": 0.6654,
      "grad_norm": 1.8714607954025269,
      "learning_rate": 6.0830513550080335e-05,
      "loss": 4.8399,
      "step": 6654
    },
    {
      "epoch": 0.6655,
      "grad_norm": 3.0362892150878906,
      "learning_rate": 6.0798398556563976e-05,
      "loss": 4.5916,
      "step": 6655
    },
    {
      "epoch": 0.6656,
      "grad_norm": 1.8305473327636719,
      "learning_rate": 6.076628833964388e-05,
      "loss": 4.6293,
      "step": 6656
    },
    {
      "epoch": 0.6657,
      "grad_norm": 2.123568534851074,
      "learning_rate": 6.073418290323251e-05,
      "loss": 4.5117,
      "step": 6657
    },
    {
      "epoch": 0.6658,
      "grad_norm": 1.9095451831817627,
      "learning_rate": 6.070208225124185e-05,
      "loss": 4.682,
      "step": 6658
    },
    {
      "epoch": 0.6659,
      "grad_norm": 2.913485527038574,
      "learning_rate": 6.066998638758326e-05,
      "loss": 4.9857,
      "step": 6659
    },
    {
      "epoch": 0.666,
      "grad_norm": 2.352752685546875,
      "learning_rate": 6.063789531616757e-05,
      "loss": 4.724,
      "step": 6660
    },
    {
      "epoch": 0.6661,
      "grad_norm": 2.0146567821502686,
      "learning_rate": 6.0605809040904894e-05,
      "loss": 4.4654,
      "step": 6661
    },
    {
      "epoch": 0.6662,
      "grad_norm": 10.755301475524902,
      "learning_rate": 6.05737275657049e-05,
      "loss": 5.8669,
      "step": 6662
    },
    {
      "epoch": 0.6663,
      "grad_norm": 5.567539215087891,
      "learning_rate": 6.054165089447663e-05,
      "loss": 5.2683,
      "step": 6663
    },
    {
      "epoch": 0.6664,
      "grad_norm": 5.74578857421875,
      "learning_rate": 6.0509579031128485e-05,
      "loss": 5.3746,
      "step": 6664
    },
    {
      "epoch": 0.6665,
      "grad_norm": 8.411175727844238,
      "learning_rate": 6.047751197956838e-05,
      "loss": 5.067,
      "step": 6665
    },
    {
      "epoch": 0.6666,
      "grad_norm": 2.1288673877716064,
      "learning_rate": 6.0445449743703516e-05,
      "loss": 4.6312,
      "step": 6666
    },
    {
      "epoch": 0.6667,
      "grad_norm": 8.33927059173584,
      "learning_rate": 6.0413392327440635e-05,
      "loss": 5.2637,
      "step": 6667
    },
    {
      "epoch": 0.6668,
      "grad_norm": 3.7518310546875,
      "learning_rate": 6.03813397346858e-05,
      "loss": 4.777,
      "step": 6668
    },
    {
      "epoch": 0.6669,
      "grad_norm": 3.079970359802246,
      "learning_rate": 6.034929196934459e-05,
      "loss": 4.6651,
      "step": 6669
    },
    {
      "epoch": 0.667,
      "grad_norm": 3.4413001537323,
      "learning_rate": 6.031724903532183e-05,
      "loss": 5.0561,
      "step": 6670
    },
    {
      "epoch": 0.6671,
      "grad_norm": 2.9802165031433105,
      "learning_rate": 6.0285210936521955e-05,
      "loss": 4.4514,
      "step": 6671
    },
    {
      "epoch": 0.6672,
      "grad_norm": 2.1879425048828125,
      "learning_rate": 6.025317767684864e-05,
      "loss": 4.7665,
      "step": 6672
    },
    {
      "epoch": 0.6673,
      "grad_norm": 2.0859744548797607,
      "learning_rate": 6.022114926020504e-05,
      "loss": 4.6773,
      "step": 6673
    },
    {
      "epoch": 0.6674,
      "grad_norm": 2.2625815868377686,
      "learning_rate": 6.018912569049376e-05,
      "loss": 4.5293,
      "step": 6674
    },
    {
      "epoch": 0.6675,
      "grad_norm": 6.380954265594482,
      "learning_rate": 6.015710697161674e-05,
      "loss": 4.7642,
      "step": 6675
    },
    {
      "epoch": 0.6676,
      "grad_norm": 3.4787886142730713,
      "learning_rate": 6.012509310747538e-05,
      "loss": 4.9472,
      "step": 6676
    },
    {
      "epoch": 0.6677,
      "grad_norm": 3.305877923965454,
      "learning_rate": 6.009308410197047e-05,
      "loss": 4.5413,
      "step": 6677
    },
    {
      "epoch": 0.6678,
      "grad_norm": 2.8985259532928467,
      "learning_rate": 6.006107995900224e-05,
      "loss": 4.7113,
      "step": 6678
    },
    {
      "epoch": 0.6679,
      "grad_norm": 3.2577857971191406,
      "learning_rate": 6.002908068247024e-05,
      "loss": 4.4619,
      "step": 6679
    },
    {
      "epoch": 0.668,
      "grad_norm": 3.8312571048736572,
      "learning_rate": 5.999708627627354e-05,
      "loss": 4.9877,
      "step": 6680
    },
    {
      "epoch": 0.6681,
      "grad_norm": 2.528709650039673,
      "learning_rate": 5.9965096744310526e-05,
      "loss": 4.9806,
      "step": 6681
    },
    {
      "epoch": 0.6682,
      "grad_norm": 3.3686516284942627,
      "learning_rate": 5.9933112090479006e-05,
      "loss": 5.118,
      "step": 6682
    },
    {
      "epoch": 0.6683,
      "grad_norm": 2.730006694793701,
      "learning_rate": 5.990113231867629e-05,
      "loss": 4.993,
      "step": 6683
    },
    {
      "epoch": 0.6684,
      "grad_norm": 3.915388584136963,
      "learning_rate": 5.9869157432798926e-05,
      "loss": 5.1133,
      "step": 6684
    },
    {
      "epoch": 0.6685,
      "grad_norm": 2.3262250423431396,
      "learning_rate": 5.9837187436743016e-05,
      "loss": 4.8173,
      "step": 6685
    },
    {
      "epoch": 0.6686,
      "grad_norm": 2.2568228244781494,
      "learning_rate": 5.9805222334404e-05,
      "loss": 4.9068,
      "step": 6686
    },
    {
      "epoch": 0.6687,
      "grad_norm": 3.4435782432556152,
      "learning_rate": 5.977326212967671e-05,
      "loss": 5.6956,
      "step": 6687
    },
    {
      "epoch": 0.6688,
      "grad_norm": 2.177375555038452,
      "learning_rate": 5.974130682645538e-05,
      "loss": 4.7534,
      "step": 6688
    },
    {
      "epoch": 0.6689,
      "grad_norm": 14.098068237304688,
      "learning_rate": 5.9709356428633746e-05,
      "loss": 6.0607,
      "step": 6689
    },
    {
      "epoch": 0.669,
      "grad_norm": 4.557988166809082,
      "learning_rate": 5.967741094010479e-05,
      "loss": 4.4657,
      "step": 6690
    },
    {
      "epoch": 0.6691,
      "grad_norm": 3.595097303390503,
      "learning_rate": 5.964547036476099e-05,
      "loss": 4.5746,
      "step": 6691
    },
    {
      "epoch": 0.6692,
      "grad_norm": 3.398721933364868,
      "learning_rate": 5.961353470649426e-05,
      "loss": 4.5497,
      "step": 6692
    },
    {
      "epoch": 0.6693,
      "grad_norm": 2.463338613510132,
      "learning_rate": 5.9581603969195766e-05,
      "loss": 4.5951,
      "step": 6693
    },
    {
      "epoch": 0.6694,
      "grad_norm": 4.024319171905518,
      "learning_rate": 5.954967815675627e-05,
      "loss": 5.023,
      "step": 6694
    },
    {
      "epoch": 0.6695,
      "grad_norm": 2.636017322540283,
      "learning_rate": 5.951775727306577e-05,
      "loss": 4.4517,
      "step": 6695
    },
    {
      "epoch": 0.6696,
      "grad_norm": 2.098031997680664,
      "learning_rate": 5.9485841322013755e-05,
      "loss": 4.4518,
      "step": 6696
    },
    {
      "epoch": 0.6697,
      "grad_norm": 2.431384801864624,
      "learning_rate": 5.94539303074891e-05,
      "loss": 4.1433,
      "step": 6697
    },
    {
      "epoch": 0.6698,
      "grad_norm": 2.198275327682495,
      "learning_rate": 5.9422024233380013e-05,
      "loss": 4.8287,
      "step": 6698
    },
    {
      "epoch": 0.6699,
      "grad_norm": 2.2690255641937256,
      "learning_rate": 5.9390123103574215e-05,
      "loss": 4.8001,
      "step": 6699
    },
    {
      "epoch": 0.67,
      "grad_norm": 3.0039589405059814,
      "learning_rate": 5.935822692195869e-05,
      "loss": 4.9865,
      "step": 6700
    },
    {
      "epoch": 0.6701,
      "grad_norm": 4.862521171569824,
      "learning_rate": 5.9326335692419995e-05,
      "loss": 5.5188,
      "step": 6701
    },
    {
      "epoch": 0.6702,
      "grad_norm": 2.293165445327759,
      "learning_rate": 5.929444941884388e-05,
      "loss": 4.6939,
      "step": 6702
    },
    {
      "epoch": 0.6703,
      "grad_norm": 2.2087697982788086,
      "learning_rate": 5.9262568105115654e-05,
      "loss": 4.3039,
      "step": 6703
    },
    {
      "epoch": 0.6704,
      "grad_norm": 4.969148635864258,
      "learning_rate": 5.92306917551199e-05,
      "loss": 5.1279,
      "step": 6704
    },
    {
      "epoch": 0.6705,
      "grad_norm": 2.4357900619506836,
      "learning_rate": 5.9198820372740726e-05,
      "loss": 4.2263,
      "step": 6705
    },
    {
      "epoch": 0.6706,
      "grad_norm": 2.093052387237549,
      "learning_rate": 5.916695396186154e-05,
      "loss": 4.5035,
      "step": 6706
    },
    {
      "epoch": 0.6707,
      "grad_norm": 5.436161518096924,
      "learning_rate": 5.913509252636511e-05,
      "loss": 4.5243,
      "step": 6707
    },
    {
      "epoch": 0.6708,
      "grad_norm": 2.4788379669189453,
      "learning_rate": 5.910323607013373e-05,
      "loss": 4.8485,
      "step": 6708
    },
    {
      "epoch": 0.6709,
      "grad_norm": 7.72588586807251,
      "learning_rate": 5.907138459704895e-05,
      "loss": 4.9128,
      "step": 6709
    },
    {
      "epoch": 0.671,
      "grad_norm": 2.078542470932007,
      "learning_rate": 5.903953811099183e-05,
      "loss": 4.6525,
      "step": 6710
    },
    {
      "epoch": 0.6711,
      "grad_norm": 5.908456802368164,
      "learning_rate": 5.900769661584272e-05,
      "loss": 5.2429,
      "step": 6711
    },
    {
      "epoch": 0.6712,
      "grad_norm": 3.2256293296813965,
      "learning_rate": 5.8975860115481487e-05,
      "loss": 4.4794,
      "step": 6712
    },
    {
      "epoch": 0.6713,
      "grad_norm": 3.1692419052124023,
      "learning_rate": 5.8944028613787206e-05,
      "loss": 5.4217,
      "step": 6713
    },
    {
      "epoch": 0.6714,
      "grad_norm": 2.6109120845794678,
      "learning_rate": 5.8912202114638524e-05,
      "loss": 4.6873,
      "step": 6714
    },
    {
      "epoch": 0.6715,
      "grad_norm": 2.4151451587677,
      "learning_rate": 5.88803806219134e-05,
      "loss": 4.3664,
      "step": 6715
    },
    {
      "epoch": 0.6716,
      "grad_norm": 2.4584708213806152,
      "learning_rate": 5.884856413948913e-05,
      "loss": 5.2123,
      "step": 6716
    },
    {
      "epoch": 0.6717,
      "grad_norm": 2.1827826499938965,
      "learning_rate": 5.881675267124254e-05,
      "loss": 4.7173,
      "step": 6717
    },
    {
      "epoch": 0.6718,
      "grad_norm": 2.36975359916687,
      "learning_rate": 5.8784946221049666e-05,
      "loss": 5.0726,
      "step": 6718
    },
    {
      "epoch": 0.6719,
      "grad_norm": 2.5922021865844727,
      "learning_rate": 5.8753144792786096e-05,
      "loss": 4.6777,
      "step": 6719
    },
    {
      "epoch": 0.672,
      "grad_norm": 3.4922235012054443,
      "learning_rate": 5.87213483903267e-05,
      "loss": 4.3157,
      "step": 6720
    },
    {
      "epoch": 0.6721,
      "grad_norm": 2.830590009689331,
      "learning_rate": 5.868955701754584e-05,
      "loss": 4.8629,
      "step": 6721
    },
    {
      "epoch": 0.6722,
      "grad_norm": 4.5467305183410645,
      "learning_rate": 5.86577706783171e-05,
      "loss": 5.1321,
      "step": 6722
    },
    {
      "epoch": 0.6723,
      "grad_norm": 3.228850841522217,
      "learning_rate": 5.862598937651365e-05,
      "loss": 5.1255,
      "step": 6723
    },
    {
      "epoch": 0.6724,
      "grad_norm": 2.8000850677490234,
      "learning_rate": 5.8594213116007855e-05,
      "loss": 5.4343,
      "step": 6724
    },
    {
      "epoch": 0.6725,
      "grad_norm": 2.3027126789093018,
      "learning_rate": 5.856244190067159e-05,
      "loss": 4.4909,
      "step": 6725
    },
    {
      "epoch": 0.6726,
      "grad_norm": 2.9004666805267334,
      "learning_rate": 5.853067573437612e-05,
      "loss": 4.6026,
      "step": 6726
    },
    {
      "epoch": 0.6727,
      "grad_norm": 2.844609260559082,
      "learning_rate": 5.849891462099198e-05,
      "loss": 4.4624,
      "step": 6727
    },
    {
      "epoch": 0.6728,
      "grad_norm": 2.0500564575195312,
      "learning_rate": 5.8467158564389234e-05,
      "loss": 5.0901,
      "step": 6728
    },
    {
      "epoch": 0.6729,
      "grad_norm": 2.81227970123291,
      "learning_rate": 5.8435407568437216e-05,
      "loss": 5.1258,
      "step": 6729
    },
    {
      "epoch": 0.673,
      "grad_norm": 3.311185121536255,
      "learning_rate": 5.8403661637004736e-05,
      "loss": 4.664,
      "step": 6730
    },
    {
      "epoch": 0.6731,
      "grad_norm": 3.5893850326538086,
      "learning_rate": 5.83719207739599e-05,
      "loss": 5.9186,
      "step": 6731
    },
    {
      "epoch": 0.6732,
      "grad_norm": 5.212854385375977,
      "learning_rate": 5.834018498317024e-05,
      "loss": 5.6933,
      "step": 6732
    },
    {
      "epoch": 0.6733,
      "grad_norm": 2.413203477859497,
      "learning_rate": 5.8308454268502675e-05,
      "loss": 4.901,
      "step": 6733
    },
    {
      "epoch": 0.6734,
      "grad_norm": 7.316823959350586,
      "learning_rate": 5.82767286338235e-05,
      "loss": 5.2525,
      "step": 6734
    },
    {
      "epoch": 0.6735,
      "grad_norm": 3.3964951038360596,
      "learning_rate": 5.8245008082998364e-05,
      "loss": 4.5935,
      "step": 6735
    },
    {
      "epoch": 0.6736,
      "grad_norm": 3.750383138656616,
      "learning_rate": 5.82132926198923e-05,
      "loss": 4.3097,
      "step": 6736
    },
    {
      "epoch": 0.6737,
      "grad_norm": 3.9666476249694824,
      "learning_rate": 5.818158224836987e-05,
      "loss": 5.7596,
      "step": 6737
    },
    {
      "epoch": 0.6738,
      "grad_norm": 3.325615644454956,
      "learning_rate": 5.814987697229471e-05,
      "loss": 4.9188,
      "step": 6738
    },
    {
      "epoch": 0.6739,
      "grad_norm": 2.785184383392334,
      "learning_rate": 5.8118176795530176e-05,
      "loss": 5.2416,
      "step": 6739
    },
    {
      "epoch": 0.674,
      "grad_norm": 4.633118629455566,
      "learning_rate": 5.808648172193868e-05,
      "loss": 4.6287,
      "step": 6740
    },
    {
      "epoch": 0.6741,
      "grad_norm": 2.681774854660034,
      "learning_rate": 5.805479175538229e-05,
      "loss": 4.6093,
      "step": 6741
    },
    {
      "epoch": 0.6742,
      "grad_norm": 7.034741401672363,
      "learning_rate": 5.802310689972233e-05,
      "loss": 5.728,
      "step": 6742
    },
    {
      "epoch": 0.6743,
      "grad_norm": 6.2685346603393555,
      "learning_rate": 5.799142715881938e-05,
      "loss": 5.4134,
      "step": 6743
    },
    {
      "epoch": 0.6744,
      "grad_norm": 1.8447237014770508,
      "learning_rate": 5.795975253653364e-05,
      "loss": 4.7017,
      "step": 6744
    },
    {
      "epoch": 0.6745,
      "grad_norm": 2.674394130706787,
      "learning_rate": 5.792808303672454e-05,
      "loss": 4.4956,
      "step": 6745
    },
    {
      "epoch": 0.6746,
      "grad_norm": 2.943096876144409,
      "learning_rate": 5.789641866325091e-05,
      "loss": 5.2128,
      "step": 6746
    },
    {
      "epoch": 0.6747,
      "grad_norm": 4.323859691619873,
      "learning_rate": 5.786475941997094e-05,
      "loss": 4.811,
      "step": 6747
    },
    {
      "epoch": 0.6748,
      "grad_norm": 2.3204681873321533,
      "learning_rate": 5.783310531074223e-05,
      "loss": 4.3809,
      "step": 6748
    },
    {
      "epoch": 0.6749,
      "grad_norm": 2.22916841506958,
      "learning_rate": 5.780145633942173e-05,
      "loss": 4.3337,
      "step": 6749
    },
    {
      "epoch": 0.675,
      "grad_norm": 2.3131744861602783,
      "learning_rate": 5.7769812509865773e-05,
      "loss": 5.0187,
      "step": 6750
    },
    {
      "epoch": 0.6751,
      "grad_norm": 13.890704154968262,
      "learning_rate": 5.773817382593008e-05,
      "loss": 6.3461,
      "step": 6751
    },
    {
      "epoch": 0.6752,
      "grad_norm": 2.306605577468872,
      "learning_rate": 5.770654029146969e-05,
      "loss": 4.7093,
      "step": 6752
    },
    {
      "epoch": 0.6753,
      "grad_norm": 3.6456243991851807,
      "learning_rate": 5.7674911910339094e-05,
      "loss": 4.5567,
      "step": 6753
    },
    {
      "epoch": 0.6754,
      "grad_norm": 2.3413898944854736,
      "learning_rate": 5.764328868639208e-05,
      "loss": 4.8217,
      "step": 6754
    },
    {
      "epoch": 0.6755,
      "grad_norm": 2.1638412475585938,
      "learning_rate": 5.7611670623481864e-05,
      "loss": 4.423,
      "step": 6755
    },
    {
      "epoch": 0.6756,
      "grad_norm": 2.8932440280914307,
      "learning_rate": 5.758005772546097e-05,
      "loss": 4.9228,
      "step": 6756
    },
    {
      "epoch": 0.6757,
      "grad_norm": 3.5785481929779053,
      "learning_rate": 5.754844999618144e-05,
      "loss": 4.6316,
      "step": 6757
    },
    {
      "epoch": 0.6758,
      "grad_norm": 3.5617711544036865,
      "learning_rate": 5.7516847439494435e-05,
      "loss": 5.2366,
      "step": 6758
    },
    {
      "epoch": 0.6759,
      "grad_norm": 3.2713146209716797,
      "learning_rate": 5.748525005925074e-05,
      "loss": 4.9103,
      "step": 6759
    },
    {
      "epoch": 0.676,
      "grad_norm": 2.00877046585083,
      "learning_rate": 5.7453657859300415e-05,
      "loss": 4.5779,
      "step": 6760
    },
    {
      "epoch": 0.6761,
      "grad_norm": 2.814478874206543,
      "learning_rate": 5.7422070843492734e-05,
      "loss": 4.9061,
      "step": 6761
    },
    {
      "epoch": 0.6762,
      "grad_norm": 3.8007943630218506,
      "learning_rate": 5.739048901567665e-05,
      "loss": 5.5649,
      "step": 6762
    },
    {
      "epoch": 0.6763,
      "grad_norm": 5.349628448486328,
      "learning_rate": 5.735891237970015e-05,
      "loss": 4.8794,
      "step": 6763
    },
    {
      "epoch": 0.6764,
      "grad_norm": 14.162971496582031,
      "learning_rate": 5.732734093941087e-05,
      "loss": 5.2969,
      "step": 6764
    },
    {
      "epoch": 0.6765,
      "grad_norm": 2.2174181938171387,
      "learning_rate": 5.7295774698655655e-05,
      "loss": 4.8403,
      "step": 6765
    },
    {
      "epoch": 0.6766,
      "grad_norm": 3.491833209991455,
      "learning_rate": 5.7264213661280765e-05,
      "loss": 4.9648,
      "step": 6766
    },
    {
      "epoch": 0.6767,
      "grad_norm": 3.0176310539245605,
      "learning_rate": 5.723265783113181e-05,
      "loss": 5.3788,
      "step": 6767
    },
    {
      "epoch": 0.6768,
      "grad_norm": 3.0257699489593506,
      "learning_rate": 5.720110721205376e-05,
      "loss": 4.6294,
      "step": 6768
    },
    {
      "epoch": 0.6769,
      "grad_norm": 10.336944580078125,
      "learning_rate": 5.716956180789098e-05,
      "loss": 4.9503,
      "step": 6769
    },
    {
      "epoch": 0.677,
      "grad_norm": 3.1716105937957764,
      "learning_rate": 5.713802162248718e-05,
      "loss": 4.8512,
      "step": 6770
    },
    {
      "epoch": 0.6771,
      "grad_norm": 2.4155521392822266,
      "learning_rate": 5.710648665968543e-05,
      "loss": 4.3686,
      "step": 6771
    },
    {
      "epoch": 0.6772,
      "grad_norm": 2.5509302616119385,
      "learning_rate": 5.707495692332816e-05,
      "loss": 4.7619,
      "step": 6772
    },
    {
      "epoch": 0.6773,
      "grad_norm": 4.523050785064697,
      "learning_rate": 5.704343241725719e-05,
      "loss": 4.7479,
      "step": 6773
    },
    {
      "epoch": 0.6774,
      "grad_norm": 2.218621015548706,
      "learning_rate": 5.701191314531364e-05,
      "loss": 4.8669,
      "step": 6774
    },
    {
      "epoch": 0.6775,
      "grad_norm": 4.509858131408691,
      "learning_rate": 5.6980399111338156e-05,
      "loss": 4.9515,
      "step": 6775
    },
    {
      "epoch": 0.6776,
      "grad_norm": 3.2419745922088623,
      "learning_rate": 5.694889031917047e-05,
      "loss": 5.757,
      "step": 6776
    },
    {
      "epoch": 0.6777,
      "grad_norm": 2.9195547103881836,
      "learning_rate": 5.691738677265e-05,
      "loss": 4.4095,
      "step": 6777
    },
    {
      "epoch": 0.6778,
      "grad_norm": 3.5676066875457764,
      "learning_rate": 5.6885888475615204e-05,
      "loss": 4.7937,
      "step": 6778
    },
    {
      "epoch": 0.6779,
      "grad_norm": 3.06833815574646,
      "learning_rate": 5.6854395431904094e-05,
      "loss": 4.4721,
      "step": 6779
    },
    {
      "epoch": 0.678,
      "grad_norm": 2.3984861373901367,
      "learning_rate": 5.68229076453541e-05,
      "loss": 4.3697,
      "step": 6780
    },
    {
      "epoch": 0.6781,
      "grad_norm": 3.2263145446777344,
      "learning_rate": 5.679142511980175e-05,
      "loss": 4.9883,
      "step": 6781
    },
    {
      "epoch": 0.6782,
      "grad_norm": 3.05383563041687,
      "learning_rate": 5.675994785908329e-05,
      "loss": 4.3241,
      "step": 6782
    },
    {
      "epoch": 0.6783,
      "grad_norm": 2.3170151710510254,
      "learning_rate": 5.6728475867033925e-05,
      "loss": 4.474,
      "step": 6783
    },
    {
      "epoch": 0.6784,
      "grad_norm": 2.6482601165771484,
      "learning_rate": 5.669700914748857e-05,
      "loss": 4.798,
      "step": 6784
    },
    {
      "epoch": 0.6785,
      "grad_norm": 2.268268346786499,
      "learning_rate": 5.666554770428129e-05,
      "loss": 4.96,
      "step": 6785
    },
    {
      "epoch": 0.6786,
      "grad_norm": 2.3894238471984863,
      "learning_rate": 5.663409154124557e-05,
      "loss": 4.8183,
      "step": 6786
    },
    {
      "epoch": 0.6787,
      "grad_norm": 2.5361833572387695,
      "learning_rate": 5.6602640662214256e-05,
      "loss": 4.3782,
      "step": 6787
    },
    {
      "epoch": 0.6788,
      "grad_norm": 6.046512126922607,
      "learning_rate": 5.657119507101954e-05,
      "loss": 5.6311,
      "step": 6788
    },
    {
      "epoch": 0.6789,
      "grad_norm": 7.639554977416992,
      "learning_rate": 5.653975477149298e-05,
      "loss": 4.6202,
      "step": 6789
    },
    {
      "epoch": 0.679,
      "grad_norm": 2.569551467895508,
      "learning_rate": 5.6508319767465465e-05,
      "loss": 5.1323,
      "step": 6790
    },
    {
      "epoch": 0.6791,
      "grad_norm": 2.9141974449157715,
      "learning_rate": 5.647689006276726e-05,
      "loss": 4.631,
      "step": 6791
    },
    {
      "epoch": 0.6792,
      "grad_norm": 2.276676893234253,
      "learning_rate": 5.6445465661227994e-05,
      "loss": 4.9817,
      "step": 6792
    },
    {
      "epoch": 0.6793,
      "grad_norm": 7.96079683303833,
      "learning_rate": 5.641404656667661e-05,
      "loss": 4.7059,
      "step": 6793
    },
    {
      "epoch": 0.6794,
      "grad_norm": 5.042358875274658,
      "learning_rate": 5.6382632782941405e-05,
      "loss": 4.7625,
      "step": 6794
    },
    {
      "epoch": 0.6795,
      "grad_norm": 4.694360256195068,
      "learning_rate": 5.635122431385016e-05,
      "loss": 4.0756,
      "step": 6795
    },
    {
      "epoch": 0.6796,
      "grad_norm": 3.178088426589966,
      "learning_rate": 5.63198211632298e-05,
      "loss": 4.5738,
      "step": 6796
    },
    {
      "epoch": 0.6797,
      "grad_norm": 3.453777551651001,
      "learning_rate": 5.6288423334906735e-05,
      "loss": 4.5321,
      "step": 6797
    },
    {
      "epoch": 0.6798,
      "grad_norm": 2.8418612480163574,
      "learning_rate": 5.6257030832706695e-05,
      "loss": 4.6925,
      "step": 6798
    },
    {
      "epoch": 0.6799,
      "grad_norm": 3.031054973602295,
      "learning_rate": 5.622564366045472e-05,
      "loss": 4.4197,
      "step": 6799
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.6352524757385254,
      "learning_rate": 5.619426182197536e-05,
      "loss": 4.5066,
      "step": 6800
    },
    {
      "epoch": 0.6801,
      "grad_norm": 2.5866732597351074,
      "learning_rate": 5.616288532109225e-05,
      "loss": 4.6579,
      "step": 6801
    },
    {
      "epoch": 0.6802,
      "grad_norm": 2.1487603187561035,
      "learning_rate": 5.6131514161628626e-05,
      "loss": 4.7643,
      "step": 6802
    },
    {
      "epoch": 0.6803,
      "grad_norm": 7.783400058746338,
      "learning_rate": 5.610014834740693e-05,
      "loss": 5.4006,
      "step": 6803
    },
    {
      "epoch": 0.6804,
      "grad_norm": 3.570895195007324,
      "learning_rate": 5.6068787882249005e-05,
      "loss": 4.658,
      "step": 6804
    },
    {
      "epoch": 0.6805,
      "grad_norm": 2.0780608654022217,
      "learning_rate": 5.6037432769976064e-05,
      "loss": 4.721,
      "step": 6805
    },
    {
      "epoch": 0.6806,
      "grad_norm": 3.958571434020996,
      "learning_rate": 5.6006083014408484e-05,
      "loss": 5.4028,
      "step": 6806
    },
    {
      "epoch": 0.6807,
      "grad_norm": 2.355330467224121,
      "learning_rate": 5.5974738619366295e-05,
      "loss": 4.5549,
      "step": 6807
    },
    {
      "epoch": 0.6808,
      "grad_norm": 2.8339555263519287,
      "learning_rate": 5.594339958866867e-05,
      "loss": 4.9679,
      "step": 6808
    },
    {
      "epoch": 0.6809,
      "grad_norm": 2.6633074283599854,
      "learning_rate": 5.591206592613416e-05,
      "loss": 4.887,
      "step": 6809
    },
    {
      "epoch": 0.681,
      "grad_norm": 3.208632707595825,
      "learning_rate": 5.588073763558068e-05,
      "loss": 4.8962,
      "step": 6810
    },
    {
      "epoch": 0.6811,
      "grad_norm": 2.4172682762145996,
      "learning_rate": 5.584941472082549e-05,
      "loss": 4.8659,
      "step": 6811
    },
    {
      "epoch": 0.6812,
      "grad_norm": 3.4283251762390137,
      "learning_rate": 5.5818097185685206e-05,
      "loss": 4.5993,
      "step": 6812
    },
    {
      "epoch": 0.6813,
      "grad_norm": 2.0890822410583496,
      "learning_rate": 5.578678503397574e-05,
      "loss": 4.6786,
      "step": 6813
    },
    {
      "epoch": 0.6814,
      "grad_norm": 3.315290689468384,
      "learning_rate": 5.575547826951242e-05,
      "loss": 5.4332,
      "step": 6814
    },
    {
      "epoch": 0.6815,
      "grad_norm": 2.078021764755249,
      "learning_rate": 5.572417689610987e-05,
      "loss": 4.5273,
      "step": 6815
    },
    {
      "epoch": 0.6816,
      "grad_norm": 2.5526726245880127,
      "learning_rate": 5.5692880917582046e-05,
      "loss": 5.1197,
      "step": 6816
    },
    {
      "epoch": 0.6817,
      "grad_norm": 3.0213701725006104,
      "learning_rate": 5.566159033774225e-05,
      "loss": 4.9063,
      "step": 6817
    },
    {
      "epoch": 0.6818,
      "grad_norm": 5.232877731323242,
      "learning_rate": 5.5630305160403275e-05,
      "loss": 4.8184,
      "step": 6818
    },
    {
      "epoch": 0.6819,
      "grad_norm": 4.148453712463379,
      "learning_rate": 5.5599025389376935e-05,
      "loss": 4.3512,
      "step": 6819
    },
    {
      "epoch": 0.682,
      "grad_norm": 2.314596176147461,
      "learning_rate": 5.556775102847475e-05,
      "loss": 4.6751,
      "step": 6820
    },
    {
      "epoch": 0.6821,
      "grad_norm": 3.359067916870117,
      "learning_rate": 5.553648208150728e-05,
      "loss": 4.4232,
      "step": 6821
    },
    {
      "epoch": 0.6822,
      "grad_norm": 3.4088170528411865,
      "learning_rate": 5.5505218552284565e-05,
      "loss": 4.2927,
      "step": 6822
    },
    {
      "epoch": 0.6823,
      "grad_norm": 3.071648359298706,
      "learning_rate": 5.547396044461608e-05,
      "loss": 4.7975,
      "step": 6823
    },
    {
      "epoch": 0.6824,
      "grad_norm": 5.035787582397461,
      "learning_rate": 5.544270776231038e-05,
      "loss": 4.9093,
      "step": 6824
    },
    {
      "epoch": 0.6825,
      "grad_norm": 2.5066304206848145,
      "learning_rate": 5.541146050917561e-05,
      "loss": 5.0125,
      "step": 6825
    },
    {
      "epoch": 0.6826,
      "grad_norm": 2.5772626399993896,
      "learning_rate": 5.5380218689019125e-05,
      "loss": 5.0832,
      "step": 6826
    },
    {
      "epoch": 0.6827,
      "grad_norm": 3.2768712043762207,
      "learning_rate": 5.5348982305647643e-05,
      "loss": 4.9304,
      "step": 6827
    },
    {
      "epoch": 0.6828,
      "grad_norm": 2.472059965133667,
      "learning_rate": 5.5317751362867234e-05,
      "loss": 4.5019,
      "step": 6828
    },
    {
      "epoch": 0.6829,
      "grad_norm": 3.1183438301086426,
      "learning_rate": 5.5286525864483285e-05,
      "loss": 4.9473,
      "step": 6829
    },
    {
      "epoch": 0.683,
      "grad_norm": 3.0710840225219727,
      "learning_rate": 5.525530581430054e-05,
      "loss": 4.9772,
      "step": 6830
    },
    {
      "epoch": 0.6831,
      "grad_norm": 2.667464256286621,
      "learning_rate": 5.522409121612304e-05,
      "loss": 4.8141,
      "step": 6831
    },
    {
      "epoch": 0.6832,
      "grad_norm": 3.2617201805114746,
      "learning_rate": 5.519288207375422e-05,
      "loss": 4.5394,
      "step": 6832
    },
    {
      "epoch": 0.6833,
      "grad_norm": 1.9903017282485962,
      "learning_rate": 5.5161678390996796e-05,
      "loss": 4.4114,
      "step": 6833
    },
    {
      "epoch": 0.6834,
      "grad_norm": 2.351922035217285,
      "learning_rate": 5.513048017165284e-05,
      "loss": 4.3807,
      "step": 6834
    },
    {
      "epoch": 0.6835,
      "grad_norm": 1.783882975578308,
      "learning_rate": 5.509928741952379e-05,
      "loss": 4.4778,
      "step": 6835
    },
    {
      "epoch": 0.6836,
      "grad_norm": 2.769124984741211,
      "learning_rate": 5.506810013841036e-05,
      "loss": 4.831,
      "step": 6836
    },
    {
      "epoch": 0.6837,
      "grad_norm": 3.013302803039551,
      "learning_rate": 5.50369183321126e-05,
      "loss": 4.8834,
      "step": 6837
    },
    {
      "epoch": 0.6838,
      "grad_norm": 3.8068394660949707,
      "learning_rate": 5.5005742004430025e-05,
      "loss": 4.7593,
      "step": 6838
    },
    {
      "epoch": 0.6839,
      "grad_norm": 4.0046515464782715,
      "learning_rate": 5.497457115916127e-05,
      "loss": 4.2812,
      "step": 6839
    },
    {
      "epoch": 0.684,
      "grad_norm": 3.1844894886016846,
      "learning_rate": 5.494340580010441e-05,
      "loss": 4.4944,
      "step": 6840
    },
    {
      "epoch": 0.6841,
      "grad_norm": 2.1585440635681152,
      "learning_rate": 5.491224593105695e-05,
      "loss": 5.0285,
      "step": 6841
    },
    {
      "epoch": 0.6842,
      "grad_norm": 3.715412139892578,
      "learning_rate": 5.488109155581549e-05,
      "loss": 5.419,
      "step": 6842
    },
    {
      "epoch": 0.6843,
      "grad_norm": 4.0664753913879395,
      "learning_rate": 5.484994267817624e-05,
      "loss": 4.9283,
      "step": 6843
    },
    {
      "epoch": 0.6844,
      "grad_norm": 7.698872089385986,
      "learning_rate": 5.481879930193443e-05,
      "loss": 5.1173,
      "step": 6844
    },
    {
      "epoch": 0.6845,
      "grad_norm": 7.140251636505127,
      "learning_rate": 5.478766143088492e-05,
      "loss": 6.1122,
      "step": 6845
    },
    {
      "epoch": 0.6846,
      "grad_norm": 1.9562199115753174,
      "learning_rate": 5.475652906882173e-05,
      "loss": 4.4881,
      "step": 6846
    },
    {
      "epoch": 0.6847,
      "grad_norm": 2.1470885276794434,
      "learning_rate": 5.4725402219538236e-05,
      "loss": 4.4538,
      "step": 6847
    },
    {
      "epoch": 0.6848,
      "grad_norm": 3.5741615295410156,
      "learning_rate": 5.469428088682717e-05,
      "loss": 4.6924,
      "step": 6848
    },
    {
      "epoch": 0.6849,
      "grad_norm": 4.17387580871582,
      "learning_rate": 5.466316507448048e-05,
      "loss": 4.7437,
      "step": 6849
    },
    {
      "epoch": 0.685,
      "grad_norm": 2.4662089347839355,
      "learning_rate": 5.4632054786289656e-05,
      "loss": 5.5544,
      "step": 6850
    },
    {
      "epoch": 0.6851,
      "grad_norm": 8.220888137817383,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 4.9337,
      "step": 6851
    },
    {
      "epoch": 0.6852,
      "grad_norm": 6.326123237609863,
      "learning_rate": 5.4569850797537536e-05,
      "loss": 4.2682,
      "step": 6852
    },
    {
      "epoch": 0.6853,
      "grad_norm": 2.120378255844116,
      "learning_rate": 5.4538757104555615e-05,
      "loss": 4.346,
      "step": 6853
    },
    {
      "epoch": 0.6854,
      "grad_norm": 2.5522348880767822,
      "learning_rate": 5.4507668950888245e-05,
      "loss": 4.8075,
      "step": 6854
    },
    {
      "epoch": 0.6855,
      "grad_norm": 2.7023961544036865,
      "learning_rate": 5.447658634032338e-05,
      "loss": 5.0525,
      "step": 6855
    },
    {
      "epoch": 0.6856,
      "grad_norm": 4.511447906494141,
      "learning_rate": 5.4445509276648466e-05,
      "loss": 4.8006,
      "step": 6856
    },
    {
      "epoch": 0.6857,
      "grad_norm": 1.6755415201187134,
      "learning_rate": 5.441443776365003e-05,
      "loss": 4.4191,
      "step": 6857
    },
    {
      "epoch": 0.6858,
      "grad_norm": 3.9617855548858643,
      "learning_rate": 5.438337180511406e-05,
      "loss": 4.6283,
      "step": 6858
    },
    {
      "epoch": 0.6859,
      "grad_norm": 2.1709351539611816,
      "learning_rate": 5.435231140482587e-05,
      "loss": 4.5402,
      "step": 6859
    },
    {
      "epoch": 0.686,
      "grad_norm": 3.2609004974365234,
      "learning_rate": 5.4321256566570036e-05,
      "loss": 4.378,
      "step": 6860
    },
    {
      "epoch": 0.6861,
      "grad_norm": 1.9680538177490234,
      "learning_rate": 5.4290207294130615e-05,
      "loss": 4.6919,
      "step": 6861
    },
    {
      "epoch": 0.6862,
      "grad_norm": 2.030886173248291,
      "learning_rate": 5.42591635912907e-05,
      "loss": 4.3484,
      "step": 6862
    },
    {
      "epoch": 0.6863,
      "grad_norm": 1.979516625404358,
      "learning_rate": 5.4228125461833024e-05,
      "loss": 4.4631,
      "step": 6863
    },
    {
      "epoch": 0.6864,
      "grad_norm": 2.7769641876220703,
      "learning_rate": 5.419709290953936e-05,
      "loss": 4.5722,
      "step": 6864
    },
    {
      "epoch": 0.6865,
      "grad_norm": 3.2034080028533936,
      "learning_rate": 5.416606593819101e-05,
      "loss": 5.1508,
      "step": 6865
    },
    {
      "epoch": 0.6866,
      "grad_norm": 2.944554090499878,
      "learning_rate": 5.4135044551568546e-05,
      "loss": 4.1639,
      "step": 6866
    },
    {
      "epoch": 0.6867,
      "grad_norm": 2.433816909790039,
      "learning_rate": 5.4104028753451696e-05,
      "loss": 4.534,
      "step": 6867
    },
    {
      "epoch": 0.6868,
      "grad_norm": 2.8553943634033203,
      "learning_rate": 5.407301854761977e-05,
      "loss": 4.3342,
      "step": 6868
    },
    {
      "epoch": 0.6869,
      "grad_norm": 3.410515546798706,
      "learning_rate": 5.404201393785122e-05,
      "loss": 4.6197,
      "step": 6869
    },
    {
      "epoch": 0.687,
      "grad_norm": 3.5745511054992676,
      "learning_rate": 5.401101492792386e-05,
      "loss": 4.6548,
      "step": 6870
    },
    {
      "epoch": 0.6871,
      "grad_norm": 2.5563671588897705,
      "learning_rate": 5.398002152161484e-05,
      "loss": 4.9066,
      "step": 6871
    },
    {
      "epoch": 0.6872,
      "grad_norm": 2.7828145027160645,
      "learning_rate": 5.394903372270062e-05,
      "loss": 4.6458,
      "step": 6872
    },
    {
      "epoch": 0.6873,
      "grad_norm": 2.033081531524658,
      "learning_rate": 5.3918051534956926e-05,
      "loss": 4.2228,
      "step": 6873
    },
    {
      "epoch": 0.6874,
      "grad_norm": 6.285080909729004,
      "learning_rate": 5.388707496215888e-05,
      "loss": 4.7733,
      "step": 6874
    },
    {
      "epoch": 0.6875,
      "grad_norm": 2.5971462726593018,
      "learning_rate": 5.3856104008080876e-05,
      "loss": 4.7912,
      "step": 6875
    },
    {
      "epoch": 0.6876,
      "grad_norm": 4.872189521789551,
      "learning_rate": 5.382513867649663e-05,
      "loss": 4.5254,
      "step": 6876
    },
    {
      "epoch": 0.6877,
      "grad_norm": 2.885028839111328,
      "learning_rate": 5.3794178971179165e-05,
      "loss": 5.2542,
      "step": 6877
    },
    {
      "epoch": 0.6878,
      "grad_norm": 2.4131710529327393,
      "learning_rate": 5.3763224895900846e-05,
      "loss": 4.4064,
      "step": 6878
    },
    {
      "epoch": 0.6879,
      "grad_norm": 3.338831663131714,
      "learning_rate": 5.373227645443332e-05,
      "loss": 4.8736,
      "step": 6879
    },
    {
      "epoch": 0.688,
      "grad_norm": 5.038492202758789,
      "learning_rate": 5.3701333650547525e-05,
      "loss": 5.0914,
      "step": 6880
    },
    {
      "epoch": 0.6881,
      "grad_norm": 2.113940954208374,
      "learning_rate": 5.3670396488013854e-05,
      "loss": 4.8475,
      "step": 6881
    },
    {
      "epoch": 0.6882,
      "grad_norm": 8.682156562805176,
      "learning_rate": 5.3639464970601775e-05,
      "loss": 4.8028,
      "step": 6882
    },
    {
      "epoch": 0.6883,
      "grad_norm": 6.136019229888916,
      "learning_rate": 5.360853910208028e-05,
      "loss": 5.9488,
      "step": 6883
    },
    {
      "epoch": 0.6884,
      "grad_norm": 9.463114738464355,
      "learning_rate": 5.357761888621764e-05,
      "loss": 5.4921,
      "step": 6884
    },
    {
      "epoch": 0.6885,
      "grad_norm": 3.1979243755340576,
      "learning_rate": 5.3546704326781236e-05,
      "loss": 4.3848,
      "step": 6885
    },
    {
      "epoch": 0.6886,
      "grad_norm": 2.6248364448547363,
      "learning_rate": 5.351579542753807e-05,
      "loss": 5.4376,
      "step": 6886
    },
    {
      "epoch": 0.6887,
      "grad_norm": 3.238502025604248,
      "learning_rate": 5.348489219225416e-05,
      "loss": 4.5806,
      "step": 6887
    },
    {
      "epoch": 0.6888,
      "grad_norm": 3.0147483348846436,
      "learning_rate": 5.345399462469509e-05,
      "loss": 4.4555,
      "step": 6888
    },
    {
      "epoch": 0.6889,
      "grad_norm": 2.2456464767456055,
      "learning_rate": 5.3423102728625574e-05,
      "loss": 4.7193,
      "step": 6889
    },
    {
      "epoch": 0.689,
      "grad_norm": 5.412311553955078,
      "learning_rate": 5.3392216507809714e-05,
      "loss": 4.7201,
      "step": 6890
    },
    {
      "epoch": 0.6891,
      "grad_norm": 5.701878547668457,
      "learning_rate": 5.33613359660109e-05,
      "loss": 4.7753,
      "step": 6891
    },
    {
      "epoch": 0.6892,
      "grad_norm": 2.631620407104492,
      "learning_rate": 5.33304611069918e-05,
      "loss": 5.4455,
      "step": 6892
    },
    {
      "epoch": 0.6893,
      "grad_norm": 3.139115571975708,
      "learning_rate": 5.329959193451448e-05,
      "loss": 5.9558,
      "step": 6893
    },
    {
      "epoch": 0.6894,
      "grad_norm": 2.3950371742248535,
      "learning_rate": 5.326872845234021e-05,
      "loss": 4.3777,
      "step": 6894
    },
    {
      "epoch": 0.6895,
      "grad_norm": 2.61476731300354,
      "learning_rate": 5.3237870664229636e-05,
      "loss": 4.7347,
      "step": 6895
    },
    {
      "epoch": 0.6896,
      "grad_norm": 2.197323799133301,
      "learning_rate": 5.320701857394268e-05,
      "loss": 4.8547,
      "step": 6896
    },
    {
      "epoch": 0.6897,
      "grad_norm": 5.55684232711792,
      "learning_rate": 5.317617218523856e-05,
      "loss": 5.1569,
      "step": 6897
    },
    {
      "epoch": 0.6898,
      "grad_norm": 2.15158748626709,
      "learning_rate": 5.3145331501875796e-05,
      "loss": 4.783,
      "step": 6898
    },
    {
      "epoch": 0.6899,
      "grad_norm": 2.6230251789093018,
      "learning_rate": 5.311449652761235e-05,
      "loss": 4.5033,
      "step": 6899
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.477896690368652,
      "learning_rate": 5.3083667266205194e-05,
      "loss": 4.3744,
      "step": 6900
    },
    {
      "epoch": 0.6901,
      "grad_norm": 4.256472110748291,
      "learning_rate": 5.305284372141095e-05,
      "loss": 4.32,
      "step": 6901
    },
    {
      "epoch": 0.6902,
      "grad_norm": 2.790827751159668,
      "learning_rate": 5.302202589698525e-05,
      "loss": 4.8389,
      "step": 6902
    },
    {
      "epoch": 0.6903,
      "grad_norm": 3.294597864151001,
      "learning_rate": 5.299121379668316e-05,
      "loss": 5.4213,
      "step": 6903
    },
    {
      "epoch": 0.6904,
      "grad_norm": 2.4269161224365234,
      "learning_rate": 5.296040742425916e-05,
      "loss": 4.3794,
      "step": 6904
    },
    {
      "epoch": 0.6905,
      "grad_norm": 2.9900238513946533,
      "learning_rate": 5.292960678346675e-05,
      "loss": 5.1487,
      "step": 6905
    },
    {
      "epoch": 0.6906,
      "grad_norm": 2.8507611751556396,
      "learning_rate": 5.2898811878059e-05,
      "loss": 4.9903,
      "step": 6906
    },
    {
      "epoch": 0.6907,
      "grad_norm": 2.268202543258667,
      "learning_rate": 5.286802271178815e-05,
      "loss": 4.7002,
      "step": 6907
    },
    {
      "epoch": 0.6908,
      "grad_norm": 2.286252737045288,
      "learning_rate": 5.2837239288405784e-05,
      "loss": 4.7491,
      "step": 6908
    },
    {
      "epoch": 0.6909,
      "grad_norm": 8.43924331665039,
      "learning_rate": 5.2806461611662735e-05,
      "loss": 4.7492,
      "step": 6909
    },
    {
      "epoch": 0.691,
      "grad_norm": 2.1278507709503174,
      "learning_rate": 5.277568968530919e-05,
      "loss": 4.2928,
      "step": 6910
    },
    {
      "epoch": 0.6911,
      "grad_norm": 2.2413976192474365,
      "learning_rate": 5.274492351309461e-05,
      "loss": 4.7378,
      "step": 6911
    },
    {
      "epoch": 0.6912,
      "grad_norm": 2.881967067718506,
      "learning_rate": 5.271416309876776e-05,
      "loss": 5.2093,
      "step": 6912
    },
    {
      "epoch": 0.6913,
      "grad_norm": 3.4673712253570557,
      "learning_rate": 5.26834084460767e-05,
      "loss": 4.5997,
      "step": 6913
    },
    {
      "epoch": 0.6914,
      "grad_norm": 3.1435375213623047,
      "learning_rate": 5.265265955876879e-05,
      "loss": 4.6236,
      "step": 6914
    },
    {
      "epoch": 0.6915,
      "grad_norm": 2.0196635723114014,
      "learning_rate": 5.2621916440590715e-05,
      "loss": 5.1282,
      "step": 6915
    },
    {
      "epoch": 0.6916,
      "grad_norm": 3.5578746795654297,
      "learning_rate": 5.259117909528839e-05,
      "loss": 4.5587,
      "step": 6916
    },
    {
      "epoch": 0.6917,
      "grad_norm": 2.8659110069274902,
      "learning_rate": 5.256044752660709e-05,
      "loss": 4.6121,
      "step": 6917
    },
    {
      "epoch": 0.6918,
      "grad_norm": 3.871779441833496,
      "learning_rate": 5.2529721738291315e-05,
      "loss": 4.2006,
      "step": 6918
    },
    {
      "epoch": 0.6919,
      "grad_norm": 3.6059629917144775,
      "learning_rate": 5.2499001734085044e-05,
      "loss": 4.7511,
      "step": 6919
    },
    {
      "epoch": 0.692,
      "grad_norm": 4.240866184234619,
      "learning_rate": 5.2468287517731276e-05,
      "loss": 4.9884,
      "step": 6920
    },
    {
      "epoch": 0.6921,
      "grad_norm": 3.547863721847534,
      "learning_rate": 5.243757909297247e-05,
      "loss": 4.8471,
      "step": 6921
    },
    {
      "epoch": 0.6922,
      "grad_norm": 3.686605215072632,
      "learning_rate": 5.2406876463550445e-05,
      "loss": 4.6607,
      "step": 6922
    },
    {
      "epoch": 0.6923,
      "grad_norm": 1.948486089706421,
      "learning_rate": 5.237617963320608e-05,
      "loss": 4.5399,
      "step": 6923
    },
    {
      "epoch": 0.6924,
      "grad_norm": 2.718032121658325,
      "learning_rate": 5.234548860567985e-05,
      "loss": 4.9118,
      "step": 6924
    },
    {
      "epoch": 0.6925,
      "grad_norm": 3.3793258666992188,
      "learning_rate": 5.2314803384711195e-05,
      "loss": 4.8626,
      "step": 6925
    },
    {
      "epoch": 0.6926,
      "grad_norm": 2.8998236656188965,
      "learning_rate": 5.2284123974039154e-05,
      "loss": 4.7689,
      "step": 6926
    },
    {
      "epoch": 0.6927,
      "grad_norm": 3.503214120864868,
      "learning_rate": 5.225345037740186e-05,
      "loss": 4.7739,
      "step": 6927
    },
    {
      "epoch": 0.6928,
      "grad_norm": 3.4066855907440186,
      "learning_rate": 5.222278259853681e-05,
      "loss": 5.0942,
      "step": 6928
    },
    {
      "epoch": 0.6929,
      "grad_norm": 3.5564990043640137,
      "learning_rate": 5.2192120641180786e-05,
      "loss": 4.3356,
      "step": 6929
    },
    {
      "epoch": 0.693,
      "grad_norm": 2.6904685497283936,
      "learning_rate": 5.216146450906984e-05,
      "loss": 4.519,
      "step": 6930
    },
    {
      "epoch": 0.6931,
      "grad_norm": 2.637399911880493,
      "learning_rate": 5.213081420593933e-05,
      "loss": 5.0847,
      "step": 6931
    },
    {
      "epoch": 0.6932,
      "grad_norm": 3.7224557399749756,
      "learning_rate": 5.210016973552391e-05,
      "loss": 5.8714,
      "step": 6932
    },
    {
      "epoch": 0.6933,
      "grad_norm": 2.0076327323913574,
      "learning_rate": 5.20695311015575e-05,
      "loss": 4.5276,
      "step": 6933
    },
    {
      "epoch": 0.6934,
      "grad_norm": 3.1592206954956055,
      "learning_rate": 5.2038898307773354e-05,
      "loss": 4.6092,
      "step": 6934
    },
    {
      "epoch": 0.6935,
      "grad_norm": 2.742241382598877,
      "learning_rate": 5.200827135790396e-05,
      "loss": 4.7599,
      "step": 6935
    },
    {
      "epoch": 0.6936,
      "grad_norm": 2.9975969791412354,
      "learning_rate": 5.197765025568109e-05,
      "loss": 4.425,
      "step": 6936
    },
    {
      "epoch": 0.6937,
      "grad_norm": 3.983813762664795,
      "learning_rate": 5.194703500483593e-05,
      "loss": 4.6904,
      "step": 6937
    },
    {
      "epoch": 0.6938,
      "grad_norm": 4.1896796226501465,
      "learning_rate": 5.1916425609098775e-05,
      "loss": 4.9934,
      "step": 6938
    },
    {
      "epoch": 0.6939,
      "grad_norm": 3.3688340187072754,
      "learning_rate": 5.188582207219931e-05,
      "loss": 5.5728,
      "step": 6939
    },
    {
      "epoch": 0.694,
      "grad_norm": 3.404736042022705,
      "learning_rate": 5.1855224397866476e-05,
      "loss": 5.5371,
      "step": 6940
    },
    {
      "epoch": 0.6941,
      "grad_norm": 3.636185884475708,
      "learning_rate": 5.182463258982846e-05,
      "loss": 4.1195,
      "step": 6941
    },
    {
      "epoch": 0.6942,
      "grad_norm": 2.741702079772949,
      "learning_rate": 5.179404665181291e-05,
      "loss": 5.1998,
      "step": 6942
    },
    {
      "epoch": 0.6943,
      "grad_norm": 2.0900609493255615,
      "learning_rate": 5.1763466587546485e-05,
      "loss": 4.6817,
      "step": 6943
    },
    {
      "epoch": 0.6944,
      "grad_norm": 5.727206707000732,
      "learning_rate": 5.1732892400755376e-05,
      "loss": 5.2918,
      "step": 6944
    },
    {
      "epoch": 0.6945,
      "grad_norm": 2.3521931171417236,
      "learning_rate": 5.170232409516496e-05,
      "loss": 4.8886,
      "step": 6945
    },
    {
      "epoch": 0.6946,
      "grad_norm": 2.7066915035247803,
      "learning_rate": 5.1671761674499765e-05,
      "loss": 5.1866,
      "step": 6946
    },
    {
      "epoch": 0.6947,
      "grad_norm": 3.0488953590393066,
      "learning_rate": 5.1641205142483894e-05,
      "loss": 4.8226,
      "step": 6947
    },
    {
      "epoch": 0.6948,
      "grad_norm": 3.1074233055114746,
      "learning_rate": 5.16106545028404e-05,
      "loss": 4.8916,
      "step": 6948
    },
    {
      "epoch": 0.6949,
      "grad_norm": 3.825730323791504,
      "learning_rate": 5.158010975929193e-05,
      "loss": 3.9031,
      "step": 6949
    },
    {
      "epoch": 0.695,
      "grad_norm": 2.1033952236175537,
      "learning_rate": 5.1549570915560206e-05,
      "loss": 4.5732,
      "step": 6950
    },
    {
      "epoch": 0.6951,
      "grad_norm": 2.33772349357605,
      "learning_rate": 5.15190379753663e-05,
      "loss": 4.1584,
      "step": 6951
    },
    {
      "epoch": 0.6952,
      "grad_norm": 4.2389912605285645,
      "learning_rate": 5.148851094243057e-05,
      "loss": 6.0797,
      "step": 6952
    },
    {
      "epoch": 0.6953,
      "grad_norm": 9.426643371582031,
      "learning_rate": 5.145798982047261e-05,
      "loss": 4.6587,
      "step": 6953
    },
    {
      "epoch": 0.6954,
      "grad_norm": 2.6133697032928467,
      "learning_rate": 5.1427474613211356e-05,
      "loss": 4.6143,
      "step": 6954
    },
    {
      "epoch": 0.6955,
      "grad_norm": 3.037203550338745,
      "learning_rate": 5.1396965324364986e-05,
      "loss": 5.122,
      "step": 6955
    },
    {
      "epoch": 0.6956,
      "grad_norm": 2.5753722190856934,
      "learning_rate": 5.1366461957650954e-05,
      "loss": 4.6591,
      "step": 6956
    },
    {
      "epoch": 0.6957,
      "grad_norm": 1.947648525238037,
      "learning_rate": 5.133596451678603e-05,
      "loss": 4.8488,
      "step": 6957
    },
    {
      "epoch": 0.6958,
      "grad_norm": 2.9757518768310547,
      "learning_rate": 5.13054730054862e-05,
      "loss": 4.9532,
      "step": 6958
    },
    {
      "epoch": 0.6959,
      "grad_norm": 1.6897209882736206,
      "learning_rate": 5.127498742746675e-05,
      "loss": 3.9835,
      "step": 6959
    },
    {
      "epoch": 0.696,
      "grad_norm": 3.949573040008545,
      "learning_rate": 5.1244507786442356e-05,
      "loss": 4.4609,
      "step": 6960
    },
    {
      "epoch": 0.6961,
      "grad_norm": 3.141953229904175,
      "learning_rate": 5.121403408612672e-05,
      "loss": 5.047,
      "step": 6961
    },
    {
      "epoch": 0.6962,
      "grad_norm": 3.3557395935058594,
      "learning_rate": 5.1183566330233124e-05,
      "loss": 5.0003,
      "step": 6962
    },
    {
      "epoch": 0.6963,
      "grad_norm": 2.6668055057525635,
      "learning_rate": 5.115310452247386e-05,
      "loss": 4.6512,
      "step": 6963
    },
    {
      "epoch": 0.6964,
      "grad_norm": 4.118471622467041,
      "learning_rate": 5.112264866656059e-05,
      "loss": 5.3676,
      "step": 6964
    },
    {
      "epoch": 0.6965,
      "grad_norm": 2.16334867477417,
      "learning_rate": 5.1092198766204415e-05,
      "loss": 4.6926,
      "step": 6965
    },
    {
      "epoch": 0.6966,
      "grad_norm": 2.868039608001709,
      "learning_rate": 5.1061754825115374e-05,
      "loss": 4.6782,
      "step": 6966
    },
    {
      "epoch": 0.6967,
      "grad_norm": 3.0979456901550293,
      "learning_rate": 5.103131684700314e-05,
      "loss": 4.9791,
      "step": 6967
    },
    {
      "epoch": 0.6968,
      "grad_norm": 3.70255184173584,
      "learning_rate": 5.100088483557634e-05,
      "loss": 4.6625,
      "step": 6968
    },
    {
      "epoch": 0.6969,
      "grad_norm": 2.4728176593780518,
      "learning_rate": 5.097045879454313e-05,
      "loss": 4.6969,
      "step": 6969
    },
    {
      "epoch": 0.697,
      "grad_norm": 4.609367847442627,
      "learning_rate": 5.0940038727610796e-05,
      "loss": 5.5874,
      "step": 6970
    },
    {
      "epoch": 0.6971,
      "grad_norm": 2.8525774478912354,
      "learning_rate": 5.090962463848592e-05,
      "loss": 4.64,
      "step": 6971
    },
    {
      "epoch": 0.6972,
      "grad_norm": 2.4954254627227783,
      "learning_rate": 5.087921653087437e-05,
      "loss": 5.1865,
      "step": 6972
    },
    {
      "epoch": 0.6973,
      "grad_norm": 2.4485366344451904,
      "learning_rate": 5.0848814408481305e-05,
      "loss": 4.837,
      "step": 6973
    },
    {
      "epoch": 0.6974,
      "grad_norm": 2.5799317359924316,
      "learning_rate": 5.0818418275011104e-05,
      "loss": 4.8147,
      "step": 6974
    },
    {
      "epoch": 0.6975,
      "grad_norm": 2.2005326747894287,
      "learning_rate": 5.0788028134167456e-05,
      "loss": 4.8686,
      "step": 6975
    },
    {
      "epoch": 0.6976,
      "grad_norm": 2.329097270965576,
      "learning_rate": 5.07576439896533e-05,
      "loss": 4.2773,
      "step": 6976
    },
    {
      "epoch": 0.6977,
      "grad_norm": 2.3172812461853027,
      "learning_rate": 5.072726584517086e-05,
      "loss": 4.8746,
      "step": 6977
    },
    {
      "epoch": 0.6978,
      "grad_norm": 3.530318260192871,
      "learning_rate": 5.069689370442161e-05,
      "loss": 5.0458,
      "step": 6978
    },
    {
      "epoch": 0.6979,
      "grad_norm": 4.574029922485352,
      "learning_rate": 5.066652757110628e-05,
      "loss": 5.098,
      "step": 6979
    },
    {
      "epoch": 0.698,
      "grad_norm": 4.074447154998779,
      "learning_rate": 5.0636167448924987e-05,
      "loss": 4.4174,
      "step": 6980
    },
    {
      "epoch": 0.6981,
      "grad_norm": 2.1564862728118896,
      "learning_rate": 5.0605813341576924e-05,
      "loss": 4.6384,
      "step": 6981
    },
    {
      "epoch": 0.6982,
      "grad_norm": 2.0424814224243164,
      "learning_rate": 5.057546525276068e-05,
      "loss": 4.6462,
      "step": 6982
    },
    {
      "epoch": 0.6983,
      "grad_norm": 3.8496456146240234,
      "learning_rate": 5.054512318617406e-05,
      "loss": 5.4354,
      "step": 6983
    },
    {
      "epoch": 0.6984,
      "grad_norm": 2.694380521774292,
      "learning_rate": 5.051478714551414e-05,
      "loss": 4.7147,
      "step": 6984
    },
    {
      "epoch": 0.6985,
      "grad_norm": 3.195373296737671,
      "learning_rate": 5.048445713447738e-05,
      "loss": 5.0237,
      "step": 6985
    },
    {
      "epoch": 0.6986,
      "grad_norm": 2.279259204864502,
      "learning_rate": 5.045413315675924e-05,
      "loss": 5.1328,
      "step": 6986
    },
    {
      "epoch": 0.6987,
      "grad_norm": 3.258068084716797,
      "learning_rate": 5.0423815216054724e-05,
      "loss": 5.3038,
      "step": 6987
    },
    {
      "epoch": 0.6988,
      "grad_norm": 6.491770267486572,
      "learning_rate": 5.039350331605794e-05,
      "loss": 5.0196,
      "step": 6988
    },
    {
      "epoch": 0.6989,
      "grad_norm": 2.366718292236328,
      "learning_rate": 5.036319746046232e-05,
      "loss": 4.6865,
      "step": 6989
    },
    {
      "epoch": 0.699,
      "grad_norm": 2.188589096069336,
      "learning_rate": 5.033289765296054e-05,
      "loss": 4.4043,
      "step": 6990
    },
    {
      "epoch": 0.6991,
      "grad_norm": 2.6954338550567627,
      "learning_rate": 5.0302603897244474e-05,
      "loss": 4.7074,
      "step": 6991
    },
    {
      "epoch": 0.6992,
      "grad_norm": 2.1041791439056396,
      "learning_rate": 5.0272316197005396e-05,
      "loss": 4.3686,
      "step": 6992
    },
    {
      "epoch": 0.6993,
      "grad_norm": 3.4208924770355225,
      "learning_rate": 5.024203455593375e-05,
      "loss": 5.0581,
      "step": 6993
    },
    {
      "epoch": 0.6994,
      "grad_norm": 1.8507839441299438,
      "learning_rate": 5.021175897771927e-05,
      "loss": 4.4381,
      "step": 6994
    },
    {
      "epoch": 0.6995,
      "grad_norm": 4.233869552612305,
      "learning_rate": 5.018148946605092e-05,
      "loss": 4.4421,
      "step": 6995
    },
    {
      "epoch": 0.6996,
      "grad_norm": 2.249678611755371,
      "learning_rate": 5.015122602461698e-05,
      "loss": 4.5214,
      "step": 6996
    },
    {
      "epoch": 0.6997,
      "grad_norm": 4.237335681915283,
      "learning_rate": 5.012096865710494e-05,
      "loss": 5.2423,
      "step": 6997
    },
    {
      "epoch": 0.6998,
      "grad_norm": 2.9207050800323486,
      "learning_rate": 5.0090717367201554e-05,
      "loss": 4.5994,
      "step": 6998
    },
    {
      "epoch": 0.6999,
      "grad_norm": 1.8369994163513184,
      "learning_rate": 5.006047215859289e-05,
      "loss": 4.405,
      "step": 6999
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.2034189701080322,
      "learning_rate": 5.003023303496419e-05,
      "loss": 4.6495,
      "step": 7000
    },
    {
      "epoch": 0.7001,
      "grad_norm": 2.5887351036071777,
      "learning_rate": 5.000000000000002e-05,
      "loss": 4.3142,
      "step": 7001
    },
    {
      "epoch": 0.7002,
      "grad_norm": 2.562621593475342,
      "learning_rate": 4.996977305738415e-05,
      "loss": 4.5214,
      "step": 7002
    },
    {
      "epoch": 0.7003,
      "grad_norm": 4.030638217926025,
      "learning_rate": 4.9939552210799755e-05,
      "loss": 4.7251,
      "step": 7003
    },
    {
      "epoch": 0.7004,
      "grad_norm": 3.423764228820801,
      "learning_rate": 4.990933746392899e-05,
      "loss": 4.7222,
      "step": 7004
    },
    {
      "epoch": 0.7005,
      "grad_norm": 2.760394334793091,
      "learning_rate": 4.98791288204536e-05,
      "loss": 4.2225,
      "step": 7005
    },
    {
      "epoch": 0.7006,
      "grad_norm": 3.3122079372406006,
      "learning_rate": 4.9848926284054255e-05,
      "loss": 4.8933,
      "step": 7006
    },
    {
      "epoch": 0.7007,
      "grad_norm": 2.3856377601623535,
      "learning_rate": 4.981872985841115e-05,
      "loss": 4.3643,
      "step": 7007
    },
    {
      "epoch": 0.7008,
      "grad_norm": 4.27390718460083,
      "learning_rate": 4.978853954720364e-05,
      "loss": 4.8441,
      "step": 7008
    },
    {
      "epoch": 0.7009,
      "grad_norm": 3.2019331455230713,
      "learning_rate": 4.97583553541102e-05,
      "loss": 4.8055,
      "step": 7009
    },
    {
      "epoch": 0.701,
      "grad_norm": 2.1121604442596436,
      "learning_rate": 4.97281772828088e-05,
      "loss": 4.5868,
      "step": 7010
    },
    {
      "epoch": 0.7011,
      "grad_norm": 3.2307565212249756,
      "learning_rate": 4.969800533697649e-05,
      "loss": 5.0909,
      "step": 7011
    },
    {
      "epoch": 0.7012,
      "grad_norm": 3.987091064453125,
      "learning_rate": 4.966783952028967e-05,
      "loss": 4.2097,
      "step": 7012
    },
    {
      "epoch": 0.7013,
      "grad_norm": 2.297231435775757,
      "learning_rate": 4.9637679836423924e-05,
      "loss": 4.3208,
      "step": 7013
    },
    {
      "epoch": 0.7014,
      "grad_norm": 2.476736068725586,
      "learning_rate": 4.960752628905412e-05,
      "loss": 4.2824,
      "step": 7014
    },
    {
      "epoch": 0.7015,
      "grad_norm": 2.5577034950256348,
      "learning_rate": 4.957737888185439e-05,
      "loss": 4.7218,
      "step": 7015
    },
    {
      "epoch": 0.7016,
      "grad_norm": 3.3604347705841064,
      "learning_rate": 4.9547237618498085e-05,
      "loss": 4.9452,
      "step": 7016
    },
    {
      "epoch": 0.7017,
      "grad_norm": 1.9816597700119019,
      "learning_rate": 4.9517102502657845e-05,
      "loss": 4.0962,
      "step": 7017
    },
    {
      "epoch": 0.7018,
      "grad_norm": 2.3066465854644775,
      "learning_rate": 4.9486973538005535e-05,
      "loss": 4.3921,
      "step": 7018
    },
    {
      "epoch": 0.7019,
      "grad_norm": 2.559091091156006,
      "learning_rate": 4.945685072821227e-05,
      "loss": 4.5124,
      "step": 7019
    },
    {
      "epoch": 0.702,
      "grad_norm": 3.3552496433258057,
      "learning_rate": 4.9426734076948436e-05,
      "loss": 4.5294,
      "step": 7020
    },
    {
      "epoch": 0.7021,
      "grad_norm": 3.282959461212158,
      "learning_rate": 4.939662358788364e-05,
      "loss": 4.6627,
      "step": 7021
    },
    {
      "epoch": 0.7022,
      "grad_norm": 4.057324409484863,
      "learning_rate": 4.9366519264686725e-05,
      "loss": 4.9224,
      "step": 7022
    },
    {
      "epoch": 0.7023,
      "grad_norm": 4.165923595428467,
      "learning_rate": 4.933642111102594e-05,
      "loss": 4.5898,
      "step": 7023
    },
    {
      "epoch": 0.7024,
      "grad_norm": 4.0702948570251465,
      "learning_rate": 4.9306329130568474e-05,
      "loss": 4.8366,
      "step": 7024
    },
    {
      "epoch": 0.7025,
      "grad_norm": 2.181971549987793,
      "learning_rate": 4.927624332698109e-05,
      "loss": 4.658,
      "step": 7025
    },
    {
      "epoch": 0.7026,
      "grad_norm": 3.4189705848693848,
      "learning_rate": 4.924616370392961e-05,
      "loss": 4.5809,
      "step": 7026
    },
    {
      "epoch": 0.7027,
      "grad_norm": 1.8163182735443115,
      "learning_rate": 4.921609026507907e-05,
      "loss": 4.3784,
      "step": 7027
    },
    {
      "epoch": 0.7028,
      "grad_norm": 2.2237801551818848,
      "learning_rate": 4.918602301409395e-05,
      "loss": 4.7991,
      "step": 7028
    },
    {
      "epoch": 0.7029,
      "grad_norm": 2.3658952713012695,
      "learning_rate": 4.915596195463773e-05,
      "loss": 4.7401,
      "step": 7029
    },
    {
      "epoch": 0.703,
      "grad_norm": 5.023770809173584,
      "learning_rate": 4.912590709037335e-05,
      "loss": 5.3427,
      "step": 7030
    },
    {
      "epoch": 0.7031,
      "grad_norm": 3.0239923000335693,
      "learning_rate": 4.909585842496287e-05,
      "loss": 4.1608,
      "step": 7031
    },
    {
      "epoch": 0.7032,
      "grad_norm": 2.93023943901062,
      "learning_rate": 4.906581596206764e-05,
      "loss": 5.2765,
      "step": 7032
    },
    {
      "epoch": 0.7033,
      "grad_norm": 4.188028812408447,
      "learning_rate": 4.9035779705348226e-05,
      "loss": 5.4313,
      "step": 7033
    },
    {
      "epoch": 0.7034,
      "grad_norm": 3.666435956954956,
      "learning_rate": 4.900574965846447e-05,
      "loss": 5.1225,
      "step": 7034
    },
    {
      "epoch": 0.7035,
      "grad_norm": 2.7282588481903076,
      "learning_rate": 4.8975725825075435e-05,
      "loss": 4.2808,
      "step": 7035
    },
    {
      "epoch": 0.7036,
      "grad_norm": 2.541781425476074,
      "learning_rate": 4.894570820883944e-05,
      "loss": 4.5502,
      "step": 7036
    },
    {
      "epoch": 0.7037,
      "grad_norm": 2.7386457920074463,
      "learning_rate": 4.8915696813414026e-05,
      "loss": 4.5014,
      "step": 7037
    },
    {
      "epoch": 0.7038,
      "grad_norm": 2.8449175357818604,
      "learning_rate": 4.888569164245601e-05,
      "loss": 4.7116,
      "step": 7038
    },
    {
      "epoch": 0.7039,
      "grad_norm": 3.2985308170318604,
      "learning_rate": 4.885569269962142e-05,
      "loss": 4.1694,
      "step": 7039
    },
    {
      "epoch": 0.704,
      "grad_norm": 5.109399318695068,
      "learning_rate": 4.8825699988565485e-05,
      "loss": 5.1969,
      "step": 7040
    },
    {
      "epoch": 0.7041,
      "grad_norm": 2.661114454269409,
      "learning_rate": 4.8795713512942865e-05,
      "loss": 4.4627,
      "step": 7041
    },
    {
      "epoch": 0.7042,
      "grad_norm": 1.9199107885360718,
      "learning_rate": 4.8765733276407156e-05,
      "loss": 4.4408,
      "step": 7042
    },
    {
      "epoch": 0.7043,
      "grad_norm": 3.195777416229248,
      "learning_rate": 4.8735759282611516e-05,
      "loss": 4.6741,
      "step": 7043
    },
    {
      "epoch": 0.7044,
      "grad_norm": 4.229799270629883,
      "learning_rate": 4.870579153520807e-05,
      "loss": 4.5307,
      "step": 7044
    },
    {
      "epoch": 0.7045,
      "grad_norm": 4.803525924682617,
      "learning_rate": 4.867583003784829e-05,
      "loss": 5.5275,
      "step": 7045
    },
    {
      "epoch": 0.7046,
      "grad_norm": 2.7893126010894775,
      "learning_rate": 4.864587479418302e-05,
      "loss": 4.6055,
      "step": 7046
    },
    {
      "epoch": 0.7047,
      "grad_norm": 4.74437141418457,
      "learning_rate": 4.861592580786205e-05,
      "loss": 5.0081,
      "step": 7047
    },
    {
      "epoch": 0.7048,
      "grad_norm": 3.359997034072876,
      "learning_rate": 4.858598308253473e-05,
      "loss": 4.608,
      "step": 7048
    },
    {
      "epoch": 0.7049,
      "grad_norm": 2.3954286575317383,
      "learning_rate": 4.8556046621849346e-05,
      "loss": 4.5026,
      "step": 7049
    },
    {
      "epoch": 0.705,
      "grad_norm": 2.225719451904297,
      "learning_rate": 4.852611642945368e-05,
      "loss": 4.8762,
      "step": 7050
    },
    {
      "epoch": 0.7051,
      "grad_norm": 2.5966708660125732,
      "learning_rate": 4.8496192508994576e-05,
      "loss": 4.9903,
      "step": 7051
    },
    {
      "epoch": 0.7052,
      "grad_norm": 3.825082302093506,
      "learning_rate": 4.84662748641182e-05,
      "loss": 4.9165,
      "step": 7052
    },
    {
      "epoch": 0.7053,
      "grad_norm": 2.350186586380005,
      "learning_rate": 4.8436363498469906e-05,
      "loss": 4.4177,
      "step": 7053
    },
    {
      "epoch": 0.7054,
      "grad_norm": 6.211674690246582,
      "learning_rate": 4.840645841569431e-05,
      "loss": 5.6452,
      "step": 7054
    },
    {
      "epoch": 0.7055,
      "grad_norm": 7.302852153778076,
      "learning_rate": 4.837655961943526e-05,
      "loss": 4.8999,
      "step": 7055
    },
    {
      "epoch": 0.7056,
      "grad_norm": 6.648589134216309,
      "learning_rate": 4.834666711333582e-05,
      "loss": 4.5545,
      "step": 7056
    },
    {
      "epoch": 0.7057,
      "grad_norm": 3.605699300765991,
      "learning_rate": 4.8316780901038314e-05,
      "loss": 4.9026,
      "step": 7057
    },
    {
      "epoch": 0.7058,
      "grad_norm": 2.3186447620391846,
      "learning_rate": 4.828690098618429e-05,
      "loss": 4.5801,
      "step": 7058
    },
    {
      "epoch": 0.7059,
      "grad_norm": 2.3493220806121826,
      "learning_rate": 4.825702737241452e-05,
      "loss": 4.7585,
      "step": 7059
    },
    {
      "epoch": 0.706,
      "grad_norm": 2.6517066955566406,
      "learning_rate": 4.822716006336897e-05,
      "loss": 4.5975,
      "step": 7060
    },
    {
      "epoch": 0.7061,
      "grad_norm": 3.439476251602173,
      "learning_rate": 4.8197299062686995e-05,
      "loss": 4.7316,
      "step": 7061
    },
    {
      "epoch": 0.7062,
      "grad_norm": 4.115218162536621,
      "learning_rate": 4.816744437400697e-05,
      "loss": 4.7101,
      "step": 7062
    },
    {
      "epoch": 0.7063,
      "grad_norm": 2.6849491596221924,
      "learning_rate": 4.8137596000966614e-05,
      "loss": 4.7466,
      "step": 7063
    },
    {
      "epoch": 0.7064,
      "grad_norm": 2.3758342266082764,
      "learning_rate": 4.810775394720286e-05,
      "loss": 4.5009,
      "step": 7064
    },
    {
      "epoch": 0.7065,
      "grad_norm": 3.4812657833099365,
      "learning_rate": 4.807791821635186e-05,
      "loss": 4.8454,
      "step": 7065
    },
    {
      "epoch": 0.7066,
      "grad_norm": 2.10143780708313,
      "learning_rate": 4.8048088812049096e-05,
      "loss": 4.601,
      "step": 7066
    },
    {
      "epoch": 0.7067,
      "grad_norm": 2.1902968883514404,
      "learning_rate": 4.8018265737929044e-05,
      "loss": 4.8402,
      "step": 7067
    },
    {
      "epoch": 0.7068,
      "grad_norm": 3.1989846229553223,
      "learning_rate": 4.798844899762568e-05,
      "loss": 5.1219,
      "step": 7068
    },
    {
      "epoch": 0.7069,
      "grad_norm": 3.2137742042541504,
      "learning_rate": 4.7958638594772064e-05,
      "loss": 5.0071,
      "step": 7069
    },
    {
      "epoch": 0.707,
      "grad_norm": 2.199840784072876,
      "learning_rate": 4.792883453300042e-05,
      "loss": 5.0929,
      "step": 7070
    },
    {
      "epoch": 0.7071,
      "grad_norm": 2.191516160964966,
      "learning_rate": 4.78990368159424e-05,
      "loss": 5.1076,
      "step": 7071
    },
    {
      "epoch": 0.7072,
      "grad_norm": 2.0672943592071533,
      "learning_rate": 4.786924544722864e-05,
      "loss": 4.1681,
      "step": 7072
    },
    {
      "epoch": 0.7073,
      "grad_norm": 1.9881120920181274,
      "learning_rate": 4.783946043048923e-05,
      "loss": 4.4623,
      "step": 7073
    },
    {
      "epoch": 0.7074,
      "grad_norm": 2.2748029232025146,
      "learning_rate": 4.780968176935333e-05,
      "loss": 4.6548,
      "step": 7074
    },
    {
      "epoch": 0.7075,
      "grad_norm": 2.6353299617767334,
      "learning_rate": 4.7779909467449414e-05,
      "loss": 5.1241,
      "step": 7075
    },
    {
      "epoch": 0.7076,
      "grad_norm": 3.821218252182007,
      "learning_rate": 4.7750143528405126e-05,
      "loss": 4.7912,
      "step": 7076
    },
    {
      "epoch": 0.7077,
      "grad_norm": 2.5312938690185547,
      "learning_rate": 4.7720383955847345e-05,
      "loss": 4.3092,
      "step": 7077
    },
    {
      "epoch": 0.7078,
      "grad_norm": 3.1581029891967773,
      "learning_rate": 4.769063075340222e-05,
      "loss": 4.4702,
      "step": 7078
    },
    {
      "epoch": 0.7079,
      "grad_norm": 5.920083045959473,
      "learning_rate": 4.766088392469506e-05,
      "loss": 4.7333,
      "step": 7079
    },
    {
      "epoch": 0.708,
      "grad_norm": 3.3971188068389893,
      "learning_rate": 4.763114347335043e-05,
      "loss": 4.6542,
      "step": 7080
    },
    {
      "epoch": 0.7081,
      "grad_norm": 2.406665086746216,
      "learning_rate": 4.7601409402992106e-05,
      "loss": 4.51,
      "step": 7081
    },
    {
      "epoch": 0.7082,
      "grad_norm": 2.679673671722412,
      "learning_rate": 4.757168171724311e-05,
      "loss": 5.3538,
      "step": 7082
    },
    {
      "epoch": 0.7083,
      "grad_norm": 2.835456609725952,
      "learning_rate": 4.7541960419725626e-05,
      "loss": 4.7976,
      "step": 7083
    },
    {
      "epoch": 0.7084,
      "grad_norm": 2.1510744094848633,
      "learning_rate": 4.7512245514061225e-05,
      "loss": 5.0088,
      "step": 7084
    },
    {
      "epoch": 0.7085,
      "grad_norm": 3.347835063934326,
      "learning_rate": 4.748253700387042e-05,
      "loss": 4.4824,
      "step": 7085
    },
    {
      "epoch": 0.7086,
      "grad_norm": 4.498300075531006,
      "learning_rate": 4.745283489277325e-05,
      "loss": 4.8972,
      "step": 7086
    },
    {
      "epoch": 0.7087,
      "grad_norm": 3.3994076251983643,
      "learning_rate": 4.742313918438872e-05,
      "loss": 4.8853,
      "step": 7087
    },
    {
      "epoch": 0.7088,
      "grad_norm": 4.9217939376831055,
      "learning_rate": 4.739344988233516e-05,
      "loss": 5.0127,
      "step": 7088
    },
    {
      "epoch": 0.7089,
      "grad_norm": 4.108308792114258,
      "learning_rate": 4.736376699023023e-05,
      "loss": 5.1268,
      "step": 7089
    },
    {
      "epoch": 0.709,
      "grad_norm": 2.7860875129699707,
      "learning_rate": 4.7334090511690554e-05,
      "loss": 4.6577,
      "step": 7090
    },
    {
      "epoch": 0.7091,
      "grad_norm": 4.900301456451416,
      "learning_rate": 4.7304420450332244e-05,
      "loss": 5.249,
      "step": 7091
    },
    {
      "epoch": 0.7092,
      "grad_norm": 4.92851448059082,
      "learning_rate": 4.7274756809770446e-05,
      "loss": 5.2336,
      "step": 7092
    },
    {
      "epoch": 0.7093,
      "grad_norm": 2.3126285076141357,
      "learning_rate": 4.724509959361961e-05,
      "loss": 5.108,
      "step": 7093
    },
    {
      "epoch": 0.7094,
      "grad_norm": 2.1367146968841553,
      "learning_rate": 4.721544880549337e-05,
      "loss": 4.6808,
      "step": 7094
    },
    {
      "epoch": 0.7095,
      "grad_norm": 4.891669750213623,
      "learning_rate": 4.7185804449004565e-05,
      "loss": 4.2353,
      "step": 7095
    },
    {
      "epoch": 0.7096,
      "grad_norm": 2.5101876258850098,
      "learning_rate": 4.71561665277653e-05,
      "loss": 4.1519,
      "step": 7096
    },
    {
      "epoch": 0.7097,
      "grad_norm": 2.0028374195098877,
      "learning_rate": 4.712653504538683e-05,
      "loss": 4.7041,
      "step": 7097
    },
    {
      "epoch": 0.7098,
      "grad_norm": 2.4243197441101074,
      "learning_rate": 4.70969100054797e-05,
      "loss": 4.804,
      "step": 7098
    },
    {
      "epoch": 0.7099,
      "grad_norm": 6.830991744995117,
      "learning_rate": 4.706729141165361e-05,
      "loss": 4.9799,
      "step": 7099
    },
    {
      "epoch": 0.71,
      "grad_norm": 2.5310821533203125,
      "learning_rate": 4.7037679267517495e-05,
      "loss": 5.3398,
      "step": 7100
    },
    {
      "epoch": 0.7101,
      "grad_norm": 2.3014795780181885,
      "learning_rate": 4.700807357667952e-05,
      "loss": 4.486,
      "step": 7101
    },
    {
      "epoch": 0.7102,
      "grad_norm": 1.9788185358047485,
      "learning_rate": 4.697847434274704e-05,
      "loss": 4.9472,
      "step": 7102
    },
    {
      "epoch": 0.7103,
      "grad_norm": 2.863104820251465,
      "learning_rate": 4.694888156932658e-05,
      "loss": 4.9904,
      "step": 7103
    },
    {
      "epoch": 0.7104,
      "grad_norm": 3.9741604328155518,
      "learning_rate": 4.6919295260024054e-05,
      "loss": 4.4874,
      "step": 7104
    },
    {
      "epoch": 0.7105,
      "grad_norm": 1.8539834022521973,
      "learning_rate": 4.688971541844436e-05,
      "loss": 4.8439,
      "step": 7105
    },
    {
      "epoch": 0.7106,
      "grad_norm": 2.247258424758911,
      "learning_rate": 4.68601420481917e-05,
      "loss": 4.516,
      "step": 7106
    },
    {
      "epoch": 0.7107,
      "grad_norm": 3.054584264755249,
      "learning_rate": 4.6830575152869616e-05,
      "loss": 4.9487,
      "step": 7107
    },
    {
      "epoch": 0.7108,
      "grad_norm": 4.856776714324951,
      "learning_rate": 4.6801014736080596e-05,
      "loss": 4.8976,
      "step": 7108
    },
    {
      "epoch": 0.7109,
      "grad_norm": 2.2354068756103516,
      "learning_rate": 4.6771460801426635e-05,
      "loss": 4.3174,
      "step": 7109
    },
    {
      "epoch": 0.711,
      "grad_norm": 3.14019775390625,
      "learning_rate": 4.674191335250865e-05,
      "loss": 4.6235,
      "step": 7110
    },
    {
      "epoch": 0.7111,
      "grad_norm": 2.554713010787964,
      "learning_rate": 4.6712372392927e-05,
      "loss": 5.2236,
      "step": 7111
    },
    {
      "epoch": 0.7112,
      "grad_norm": 3.76009202003479,
      "learning_rate": 4.668283792628114e-05,
      "loss": 5.0117,
      "step": 7112
    },
    {
      "epoch": 0.7113,
      "grad_norm": 3.7903053760528564,
      "learning_rate": 4.665330995616974e-05,
      "loss": 4.787,
      "step": 7113
    },
    {
      "epoch": 0.7114,
      "grad_norm": 3.031409502029419,
      "learning_rate": 4.6623788486190725e-05,
      "loss": 4.542,
      "step": 7114
    },
    {
      "epoch": 0.7115,
      "grad_norm": 2.6321492195129395,
      "learning_rate": 4.6594273519941154e-05,
      "loss": 4.7462,
      "step": 7115
    },
    {
      "epoch": 0.7116,
      "grad_norm": 2.352717876434326,
      "learning_rate": 4.656476506101737e-05,
      "loss": 4.3315,
      "step": 7116
    },
    {
      "epoch": 0.7117,
      "grad_norm": 2.3247873783111572,
      "learning_rate": 4.653526311301488e-05,
      "loss": 4.7173,
      "step": 7117
    },
    {
      "epoch": 0.7118,
      "grad_norm": 4.71337890625,
      "learning_rate": 4.65057676795284e-05,
      "loss": 5.4754,
      "step": 7118
    },
    {
      "epoch": 0.7119,
      "grad_norm": 2.7830302715301514,
      "learning_rate": 4.647627876415186e-05,
      "loss": 4.4448,
      "step": 7119
    },
    {
      "epoch": 0.712,
      "grad_norm": 3.7259299755096436,
      "learning_rate": 4.6446796370478394e-05,
      "loss": 4.706,
      "step": 7120
    },
    {
      "epoch": 0.7121,
      "grad_norm": 3.2623956203460693,
      "learning_rate": 4.6417320502100316e-05,
      "loss": 4.3812,
      "step": 7121
    },
    {
      "epoch": 0.7122,
      "grad_norm": 3.6644837856292725,
      "learning_rate": 4.6387851162609275e-05,
      "loss": 5.81,
      "step": 7122
    },
    {
      "epoch": 0.7123,
      "grad_norm": 2.240621328353882,
      "learning_rate": 4.6358388355595904e-05,
      "loss": 4.6247,
      "step": 7123
    },
    {
      "epoch": 0.7124,
      "grad_norm": 2.2902891635894775,
      "learning_rate": 4.632893208465021e-05,
      "loss": 4.949,
      "step": 7124
    },
    {
      "epoch": 0.7125,
      "grad_norm": 2.3892598152160645,
      "learning_rate": 4.629948235336133e-05,
      "loss": 4.2735,
      "step": 7125
    },
    {
      "epoch": 0.7126,
      "grad_norm": 1.9467114210128784,
      "learning_rate": 4.6270039165317605e-05,
      "loss": 4.6727,
      "step": 7126
    },
    {
      "epoch": 0.7127,
      "grad_norm": 2.2907660007476807,
      "learning_rate": 4.62406025241067e-05,
      "loss": 4.4417,
      "step": 7127
    },
    {
      "epoch": 0.7128,
      "grad_norm": 2.4301862716674805,
      "learning_rate": 4.621117243331523e-05,
      "loss": 4.6091,
      "step": 7128
    },
    {
      "epoch": 0.7129,
      "grad_norm": 3.4357736110687256,
      "learning_rate": 4.6181748896529273e-05,
      "loss": 5.4556,
      "step": 7129
    },
    {
      "epoch": 0.713,
      "grad_norm": 2.1740386486053467,
      "learning_rate": 4.615233191733398e-05,
      "loss": 4.5914,
      "step": 7130
    },
    {
      "epoch": 0.7131,
      "grad_norm": 2.0663669109344482,
      "learning_rate": 4.612292149931369e-05,
      "loss": 4.118,
      "step": 7131
    },
    {
      "epoch": 0.7132,
      "grad_norm": 2.5788145065307617,
      "learning_rate": 4.6093517646052034e-05,
      "loss": 5.2709,
      "step": 7132
    },
    {
      "epoch": 0.7133,
      "grad_norm": 1.9644798040390015,
      "learning_rate": 4.6064120361131656e-05,
      "loss": 4.5969,
      "step": 7133
    },
    {
      "epoch": 0.7134,
      "grad_norm": 2.1103341579437256,
      "learning_rate": 4.603472964813466e-05,
      "loss": 4.8065,
      "step": 7134
    },
    {
      "epoch": 0.7135,
      "grad_norm": 2.2107901573181152,
      "learning_rate": 4.600534551064215e-05,
      "loss": 5.2289,
      "step": 7135
    },
    {
      "epoch": 0.7136,
      "grad_norm": 2.809485912322998,
      "learning_rate": 4.59759679522345e-05,
      "loss": 4.5679,
      "step": 7136
    },
    {
      "epoch": 0.7137,
      "grad_norm": 3.8613271713256836,
      "learning_rate": 4.5946596976491295e-05,
      "loss": 5.5941,
      "step": 7137
    },
    {
      "epoch": 0.7138,
      "grad_norm": 2.811947822570801,
      "learning_rate": 4.591723258699127e-05,
      "loss": 4.5461,
      "step": 7138
    },
    {
      "epoch": 0.7139,
      "grad_norm": 4.888671875,
      "learning_rate": 4.588787478731242e-05,
      "loss": 4.2233,
      "step": 7139
    },
    {
      "epoch": 0.714,
      "grad_norm": 3.0462992191314697,
      "learning_rate": 4.5858523581031884e-05,
      "loss": 4.6881,
      "step": 7140
    },
    {
      "epoch": 0.7141,
      "grad_norm": 2.6601672172546387,
      "learning_rate": 4.582917897172603e-05,
      "loss": 4.3042,
      "step": 7141
    },
    {
      "epoch": 0.7142,
      "grad_norm": 2.8544304370880127,
      "learning_rate": 4.579984096297038e-05,
      "loss": 5.0763,
      "step": 7142
    },
    {
      "epoch": 0.7143,
      "grad_norm": 2.7703609466552734,
      "learning_rate": 4.577050955833973e-05,
      "loss": 6.084,
      "step": 7143
    },
    {
      "epoch": 0.7144,
      "grad_norm": 3.1345114707946777,
      "learning_rate": 4.574118476140794e-05,
      "loss": 4.6,
      "step": 7144
    },
    {
      "epoch": 0.7145,
      "grad_norm": 2.116111993789673,
      "learning_rate": 4.5711866575748276e-05,
      "loss": 4.2145,
      "step": 7145
    },
    {
      "epoch": 0.7146,
      "grad_norm": 4.710395336151123,
      "learning_rate": 4.568255500493292e-05,
      "loss": 4.5733,
      "step": 7146
    },
    {
      "epoch": 0.7147,
      "grad_norm": 5.147561550140381,
      "learning_rate": 4.565325005253356e-05,
      "loss": 5.2436,
      "step": 7147
    },
    {
      "epoch": 0.7148,
      "grad_norm": 3.0913071632385254,
      "learning_rate": 4.5623951722120736e-05,
      "loss": 4.4619,
      "step": 7148
    },
    {
      "epoch": 0.7149,
      "grad_norm": 2.5033254623413086,
      "learning_rate": 4.559466001726451e-05,
      "loss": 4.9221,
      "step": 7149
    },
    {
      "epoch": 0.715,
      "grad_norm": 2.6004085540771484,
      "learning_rate": 4.5565374941533965e-05,
      "loss": 4.7799,
      "step": 7150
    },
    {
      "epoch": 0.7151,
      "grad_norm": 2.9073476791381836,
      "learning_rate": 4.5536096498497295e-05,
      "loss": 4.7705,
      "step": 7151
    },
    {
      "epoch": 0.7152,
      "grad_norm": 2.0190272331237793,
      "learning_rate": 4.5506824691722126e-05,
      "loss": 4.5107,
      "step": 7152
    },
    {
      "epoch": 0.7153,
      "grad_norm": 2.9287993907928467,
      "learning_rate": 4.5477559524774994e-05,
      "loss": 4.5115,
      "step": 7153
    },
    {
      "epoch": 0.7154,
      "grad_norm": 2.8038973808288574,
      "learning_rate": 4.5448301001221895e-05,
      "loss": 4.332,
      "step": 7154
    },
    {
      "epoch": 0.7155,
      "grad_norm": 1.859408974647522,
      "learning_rate": 4.541904912462784e-05,
      "loss": 4.6537,
      "step": 7155
    },
    {
      "epoch": 0.7156,
      "grad_norm": 6.267601490020752,
      "learning_rate": 4.5389803898557106e-05,
      "loss": 4.8063,
      "step": 7156
    },
    {
      "epoch": 0.7157,
      "grad_norm": 2.4960858821868896,
      "learning_rate": 4.5360565326573104e-05,
      "loss": 4.52,
      "step": 7157
    },
    {
      "epoch": 0.7158,
      "grad_norm": 3.6741418838500977,
      "learning_rate": 4.5331333412238475e-05,
      "loss": 5.9286,
      "step": 7158
    },
    {
      "epoch": 0.7159,
      "grad_norm": 2.3189377784729004,
      "learning_rate": 4.530210815911504e-05,
      "loss": 5.2246,
      "step": 7159
    },
    {
      "epoch": 0.716,
      "grad_norm": 2.512092113494873,
      "learning_rate": 4.527288957076382e-05,
      "loss": 4.9964,
      "step": 7160
    },
    {
      "epoch": 0.7161,
      "grad_norm": 2.14766001701355,
      "learning_rate": 4.524367765074499e-05,
      "loss": 4.384,
      "step": 7161
    },
    {
      "epoch": 0.7162,
      "grad_norm": 3.1889290809631348,
      "learning_rate": 4.5214472402617944e-05,
      "loss": 4.7222,
      "step": 7162
    },
    {
      "epoch": 0.7163,
      "grad_norm": 3.2679104804992676,
      "learning_rate": 4.518527382994127e-05,
      "loss": 5.0869,
      "step": 7163
    },
    {
      "epoch": 0.7164,
      "grad_norm": 1.8026405572891235,
      "learning_rate": 4.515608193627265e-05,
      "loss": 4.5241,
      "step": 7164
    },
    {
      "epoch": 0.7165,
      "grad_norm": 5.024219512939453,
      "learning_rate": 4.512689672516918e-05,
      "loss": 4.6286,
      "step": 7165
    },
    {
      "epoch": 0.7166,
      "grad_norm": 3.2432141304016113,
      "learning_rate": 4.5097718200186814e-05,
      "loss": 4.075,
      "step": 7166
    },
    {
      "epoch": 0.7167,
      "grad_norm": 3.3172450065612793,
      "learning_rate": 4.506854636488103e-05,
      "loss": 4.6398,
      "step": 7167
    },
    {
      "epoch": 0.7168,
      "grad_norm": 4.943554878234863,
      "learning_rate": 4.50393812228062e-05,
      "loss": 4.6445,
      "step": 7168
    },
    {
      "epoch": 0.7169,
      "grad_norm": 8.884875297546387,
      "learning_rate": 4.501022277751602e-05,
      "loss": 5.0897,
      "step": 7169
    },
    {
      "epoch": 0.717,
      "grad_norm": 3.041759490966797,
      "learning_rate": 4.498107103256346e-05,
      "loss": 5.2714,
      "step": 7170
    },
    {
      "epoch": 0.7171,
      "grad_norm": 2.437929391860962,
      "learning_rate": 4.495192599150044e-05,
      "loss": 4.7107,
      "step": 7171
    },
    {
      "epoch": 0.7172,
      "grad_norm": 2.1254971027374268,
      "learning_rate": 4.4922787657878294e-05,
      "loss": 4.5691,
      "step": 7172
    },
    {
      "epoch": 0.7173,
      "grad_norm": 3.437901258468628,
      "learning_rate": 4.48936560352474e-05,
      "loss": 5.1776,
      "step": 7173
    },
    {
      "epoch": 0.7174,
      "grad_norm": 5.365175724029541,
      "learning_rate": 4.4864531127157374e-05,
      "loss": 4.4387,
      "step": 7174
    },
    {
      "epoch": 0.7175,
      "grad_norm": 3.5647025108337402,
      "learning_rate": 4.483541293715698e-05,
      "loss": 5.3119,
      "step": 7175
    },
    {
      "epoch": 0.7176,
      "grad_norm": 1.9649620056152344,
      "learning_rate": 4.480630146879419e-05,
      "loss": 4.909,
      "step": 7176
    },
    {
      "epoch": 0.7177,
      "grad_norm": 2.913442611694336,
      "learning_rate": 4.4777196725616146e-05,
      "loss": 4.8044,
      "step": 7177
    },
    {
      "epoch": 0.7178,
      "grad_norm": 5.643854141235352,
      "learning_rate": 4.474809871116916e-05,
      "loss": 4.8513,
      "step": 7178
    },
    {
      "epoch": 0.7179,
      "grad_norm": 3.9386534690856934,
      "learning_rate": 4.471900742899876e-05,
      "loss": 4.9464,
      "step": 7179
    },
    {
      "epoch": 0.718,
      "grad_norm": 2.174081802368164,
      "learning_rate": 4.4689922882649626e-05,
      "loss": 4.6505,
      "step": 7180
    },
    {
      "epoch": 0.7181,
      "grad_norm": 2.853820323944092,
      "learning_rate": 4.46608450756656e-05,
      "loss": 5.0866,
      "step": 7181
    },
    {
      "epoch": 0.7182,
      "grad_norm": 2.7403435707092285,
      "learning_rate": 4.463177401158975e-05,
      "loss": 4.7506,
      "step": 7182
    },
    {
      "epoch": 0.7183,
      "grad_norm": 2.3219122886657715,
      "learning_rate": 4.460270969396429e-05,
      "loss": 4.4655,
      "step": 7183
    },
    {
      "epoch": 0.7184,
      "grad_norm": 3.7583773136138916,
      "learning_rate": 4.457365212633058e-05,
      "loss": 4.647,
      "step": 7184
    },
    {
      "epoch": 0.7185,
      "grad_norm": 1.840060830116272,
      "learning_rate": 4.45446013122293e-05,
      "loss": 5.0182,
      "step": 7185
    },
    {
      "epoch": 0.7186,
      "grad_norm": 2.988677740097046,
      "learning_rate": 4.451555725520009e-05,
      "loss": 5.0701,
      "step": 7186
    },
    {
      "epoch": 0.7187,
      "grad_norm": 2.335810422897339,
      "learning_rate": 4.44865199587819e-05,
      "loss": 4.8344,
      "step": 7187
    },
    {
      "epoch": 0.7188,
      "grad_norm": 4.45152473449707,
      "learning_rate": 4.4457489426512947e-05,
      "loss": 4.6565,
      "step": 7188
    },
    {
      "epoch": 0.7189,
      "grad_norm": 1.8892040252685547,
      "learning_rate": 4.4428465661930343e-05,
      "loss": 4.7133,
      "step": 7189
    },
    {
      "epoch": 0.719,
      "grad_norm": 6.086344242095947,
      "learning_rate": 4.43994486685707e-05,
      "loss": 5.9903,
      "step": 7190
    },
    {
      "epoch": 0.7191,
      "grad_norm": 2.477044105529785,
      "learning_rate": 4.437043844996952e-05,
      "loss": 4.8521,
      "step": 7191
    },
    {
      "epoch": 0.7192,
      "grad_norm": 3.8793540000915527,
      "learning_rate": 4.43414350096617e-05,
      "loss": 4.4435,
      "step": 7192
    },
    {
      "epoch": 0.7193,
      "grad_norm": 7.685539722442627,
      "learning_rate": 4.431243835118124e-05,
      "loss": 5.3251,
      "step": 7193
    },
    {
      "epoch": 0.7194,
      "grad_norm": 2.462982654571533,
      "learning_rate": 4.428344847806116e-05,
      "loss": 4.5475,
      "step": 7194
    },
    {
      "epoch": 0.7195,
      "grad_norm": 3.617973566055298,
      "learning_rate": 4.425446539383393e-05,
      "loss": 5.1904,
      "step": 7195
    },
    {
      "epoch": 0.7196,
      "grad_norm": 6.1269683837890625,
      "learning_rate": 4.4225489102030995e-05,
      "loss": 5.5462,
      "step": 7196
    },
    {
      "epoch": 0.7197,
      "grad_norm": 3.8012943267822266,
      "learning_rate": 4.419651960618302e-05,
      "loss": 4.3943,
      "step": 7197
    },
    {
      "epoch": 0.7198,
      "grad_norm": 2.25460147857666,
      "learning_rate": 4.4167556909819874e-05,
      "loss": 4.3109,
      "step": 7198
    },
    {
      "epoch": 0.7199,
      "grad_norm": 6.705667972564697,
      "learning_rate": 4.413860101647055e-05,
      "loss": 4.7631,
      "step": 7199
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.488297939300537,
      "learning_rate": 4.4109651929663256e-05,
      "loss": 4.582,
      "step": 7200
    },
    {
      "epoch": 0.7201,
      "grad_norm": 3.2529332637786865,
      "learning_rate": 4.4080709652925336e-05,
      "loss": 4.7856,
      "step": 7201
    },
    {
      "epoch": 0.7202,
      "grad_norm": 3.0883913040161133,
      "learning_rate": 4.4051774189783315e-05,
      "loss": 5.1823,
      "step": 7202
    },
    {
      "epoch": 0.7203,
      "grad_norm": 2.566272020339966,
      "learning_rate": 4.4022845543762915e-05,
      "loss": 4.7756,
      "step": 7203
    },
    {
      "epoch": 0.7204,
      "grad_norm": 3.0003607273101807,
      "learning_rate": 4.399392371838897e-05,
      "loss": 5.2443,
      "step": 7204
    },
    {
      "epoch": 0.7205,
      "grad_norm": 2.626581907272339,
      "learning_rate": 4.396500871718555e-05,
      "loss": 4.591,
      "step": 7205
    },
    {
      "epoch": 0.7206,
      "grad_norm": 5.974673748016357,
      "learning_rate": 4.393610054367585e-05,
      "loss": 5.8373,
      "step": 7206
    },
    {
      "epoch": 0.7207,
      "grad_norm": 2.882585287094116,
      "learning_rate": 4.39071992013822e-05,
      "loss": 4.6402,
      "step": 7207
    },
    {
      "epoch": 0.7208,
      "grad_norm": 2.2206966876983643,
      "learning_rate": 4.387830469382624e-05,
      "loss": 4.6545,
      "step": 7208
    },
    {
      "epoch": 0.7209,
      "grad_norm": 2.8111462593078613,
      "learning_rate": 4.3849417024528564e-05,
      "loss": 4.7592,
      "step": 7209
    },
    {
      "epoch": 0.721,
      "grad_norm": 6.406638145446777,
      "learning_rate": 4.382053619700912e-05,
      "loss": 5.5156,
      "step": 7210
    },
    {
      "epoch": 0.7211,
      "grad_norm": 3.4164605140686035,
      "learning_rate": 4.379166221478697e-05,
      "loss": 4.661,
      "step": 7211
    },
    {
      "epoch": 0.7212,
      "grad_norm": 2.7279324531555176,
      "learning_rate": 4.3762795081380215e-05,
      "loss": 4.3474,
      "step": 7212
    },
    {
      "epoch": 0.7213,
      "grad_norm": 2.6133182048797607,
      "learning_rate": 4.3733934800306366e-05,
      "loss": 5.0064,
      "step": 7213
    },
    {
      "epoch": 0.7214,
      "grad_norm": 3.0591909885406494,
      "learning_rate": 4.37050813750818e-05,
      "loss": 4.7687,
      "step": 7214
    },
    {
      "epoch": 0.7215,
      "grad_norm": 3.613898277282715,
      "learning_rate": 4.367623480922236e-05,
      "loss": 5.0628,
      "step": 7215
    },
    {
      "epoch": 0.7216,
      "grad_norm": 4.942820072174072,
      "learning_rate": 4.364739510624286e-05,
      "loss": 4.8941,
      "step": 7216
    },
    {
      "epoch": 0.7217,
      "grad_norm": 3.4892873764038086,
      "learning_rate": 4.361856226965733e-05,
      "loss": 4.7012,
      "step": 7217
    },
    {
      "epoch": 0.7218,
      "grad_norm": 4.268299579620361,
      "learning_rate": 4.3589736302978954e-05,
      "loss": 4.4857,
      "step": 7218
    },
    {
      "epoch": 0.7219,
      "grad_norm": 2.546658754348755,
      "learning_rate": 4.356091720972011e-05,
      "loss": 4.6295,
      "step": 7219
    },
    {
      "epoch": 0.722,
      "grad_norm": 7.162531852722168,
      "learning_rate": 4.3532104993392306e-05,
      "loss": 4.4688,
      "step": 7220
    },
    {
      "epoch": 0.7221,
      "grad_norm": 3.092148780822754,
      "learning_rate": 4.350329965750621e-05,
      "loss": 5.1391,
      "step": 7221
    },
    {
      "epoch": 0.7222,
      "grad_norm": 1.9817537069320679,
      "learning_rate": 4.347450120557169e-05,
      "loss": 4.9206,
      "step": 7222
    },
    {
      "epoch": 0.7223,
      "grad_norm": 1.9718631505966187,
      "learning_rate": 4.3445709641097745e-05,
      "loss": 4.9848,
      "step": 7223
    },
    {
      "epoch": 0.7224,
      "grad_norm": 2.251905918121338,
      "learning_rate": 4.341692496759252e-05,
      "loss": 4.6432,
      "step": 7224
    },
    {
      "epoch": 0.7225,
      "grad_norm": 3.075547218322754,
      "learning_rate": 4.3388147188563325e-05,
      "loss": 4.9771,
      "step": 7225
    },
    {
      "epoch": 0.7226,
      "grad_norm": 4.097997188568115,
      "learning_rate": 4.335937630751674e-05,
      "loss": 5.3167,
      "step": 7226
    },
    {
      "epoch": 0.7227,
      "grad_norm": 2.250424385070801,
      "learning_rate": 4.333061232795826e-05,
      "loss": 4.561,
      "step": 7227
    },
    {
      "epoch": 0.7228,
      "grad_norm": 2.676558017730713,
      "learning_rate": 4.3301855253392864e-05,
      "loss": 4.3297,
      "step": 7228
    },
    {
      "epoch": 0.7229,
      "grad_norm": 2.7476320266723633,
      "learning_rate": 4.327310508732437e-05,
      "loss": 5.2717,
      "step": 7229
    },
    {
      "epoch": 0.723,
      "grad_norm": 2.5332863330841064,
      "learning_rate": 4.324436183325593e-05,
      "loss": 5.2093,
      "step": 7230
    },
    {
      "epoch": 0.7231,
      "grad_norm": 2.5286598205566406,
      "learning_rate": 4.32156254946899e-05,
      "loss": 4.6925,
      "step": 7231
    },
    {
      "epoch": 0.7232,
      "grad_norm": 3.3321094512939453,
      "learning_rate": 4.3186896075127595e-05,
      "loss": 4.4606,
      "step": 7232
    },
    {
      "epoch": 0.7233,
      "grad_norm": 4.28605842590332,
      "learning_rate": 4.315817357806974e-05,
      "loss": 5.0474,
      "step": 7233
    },
    {
      "epoch": 0.7234,
      "grad_norm": 4.865937232971191,
      "learning_rate": 4.3129458007015946e-05,
      "loss": 4.678,
      "step": 7234
    },
    {
      "epoch": 0.7235,
      "grad_norm": 2.3234975337982178,
      "learning_rate": 4.310074936546521e-05,
      "loss": 4.6686,
      "step": 7235
    },
    {
      "epoch": 0.7236,
      "grad_norm": 1.694849967956543,
      "learning_rate": 4.307204765691558e-05,
      "loss": 4.6227,
      "step": 7236
    },
    {
      "epoch": 0.7237,
      "grad_norm": 4.4177470207214355,
      "learning_rate": 4.304335288486426e-05,
      "loss": 5.0318,
      "step": 7237
    },
    {
      "epoch": 0.7238,
      "grad_norm": 5.262324333190918,
      "learning_rate": 4.301466505280762e-05,
      "loss": 5.0115,
      "step": 7238
    },
    {
      "epoch": 0.7239,
      "grad_norm": 2.1321237087249756,
      "learning_rate": 4.29859841642412e-05,
      "loss": 4.4853,
      "step": 7239
    },
    {
      "epoch": 0.724,
      "grad_norm": 4.534389972686768,
      "learning_rate": 4.295731022265966e-05,
      "loss": 4.8004,
      "step": 7240
    },
    {
      "epoch": 0.7241,
      "grad_norm": 2.1002891063690186,
      "learning_rate": 4.2928643231556844e-05,
      "loss": 5.1044,
      "step": 7241
    },
    {
      "epoch": 0.7242,
      "grad_norm": 3.6943633556365967,
      "learning_rate": 4.289998319442573e-05,
      "loss": 4.5374,
      "step": 7242
    },
    {
      "epoch": 0.7243,
      "grad_norm": 3.4873392581939697,
      "learning_rate": 4.287133011475847e-05,
      "loss": 4.4666,
      "step": 7243
    },
    {
      "epoch": 0.7244,
      "grad_norm": 3.1708571910858154,
      "learning_rate": 4.2842683996046327e-05,
      "loss": 4.355,
      "step": 7244
    },
    {
      "epoch": 0.7245,
      "grad_norm": 3.134197950363159,
      "learning_rate": 4.2814044841779745e-05,
      "loss": 4.8509,
      "step": 7245
    },
    {
      "epoch": 0.7246,
      "grad_norm": 2.9624831676483154,
      "learning_rate": 4.27854126554484e-05,
      "loss": 4.9456,
      "step": 7246
    },
    {
      "epoch": 0.7247,
      "grad_norm": 2.215092420578003,
      "learning_rate": 4.2756787440540936e-05,
      "loss": 4.8911,
      "step": 7247
    },
    {
      "epoch": 0.7248,
      "grad_norm": 3.642496109008789,
      "learning_rate": 4.2728169200545286e-05,
      "loss": 4.3131,
      "step": 7248
    },
    {
      "epoch": 0.7249,
      "grad_norm": 7.303675174713135,
      "learning_rate": 4.26995579389485e-05,
      "loss": 4.434,
      "step": 7249
    },
    {
      "epoch": 0.725,
      "grad_norm": 6.627284526824951,
      "learning_rate": 4.267095365923672e-05,
      "loss": 6.4967,
      "step": 7250
    },
    {
      "epoch": 0.7251,
      "grad_norm": 2.3651368618011475,
      "learning_rate": 4.264235636489542e-05,
      "loss": 4.4166,
      "step": 7251
    },
    {
      "epoch": 0.7252,
      "grad_norm": 2.6848740577697754,
      "learning_rate": 4.261376605940894e-05,
      "loss": 4.0901,
      "step": 7252
    },
    {
      "epoch": 0.7253,
      "grad_norm": 2.6748530864715576,
      "learning_rate": 4.2585182746261035e-05,
      "loss": 4.7632,
      "step": 7253
    },
    {
      "epoch": 0.7254,
      "grad_norm": 2.584235906600952,
      "learning_rate": 4.2556606428934443e-05,
      "loss": 4.8828,
      "step": 7254
    },
    {
      "epoch": 0.7255,
      "grad_norm": 2.096761703491211,
      "learning_rate": 4.252803711091112e-05,
      "loss": 4.734,
      "step": 7255
    },
    {
      "epoch": 0.7256,
      "grad_norm": 2.5597805976867676,
      "learning_rate": 4.249947479567218e-05,
      "loss": 4.681,
      "step": 7256
    },
    {
      "epoch": 0.7257,
      "grad_norm": 5.474852085113525,
      "learning_rate": 4.2470919486697744e-05,
      "loss": 4.453,
      "step": 7257
    },
    {
      "epoch": 0.7258,
      "grad_norm": 1.7060297727584839,
      "learning_rate": 4.244237118746731e-05,
      "loss": 4.6728,
      "step": 7258
    },
    {
      "epoch": 0.7259,
      "grad_norm": 1.9227999448776245,
      "learning_rate": 4.2413829901459344e-05,
      "loss": 4.4516,
      "step": 7259
    },
    {
      "epoch": 0.726,
      "grad_norm": 3.0945868492126465,
      "learning_rate": 4.238529563215153e-05,
      "loss": 4.6032,
      "step": 7260
    },
    {
      "epoch": 0.7261,
      "grad_norm": 1.94703209400177,
      "learning_rate": 4.235676838302068e-05,
      "loss": 4.2913,
      "step": 7261
    },
    {
      "epoch": 0.7262,
      "grad_norm": 2.2158782482147217,
      "learning_rate": 4.232824815754276e-05,
      "loss": 4.8738,
      "step": 7262
    },
    {
      "epoch": 0.7263,
      "grad_norm": 2.7764687538146973,
      "learning_rate": 4.229973495919286e-05,
      "loss": 4.3681,
      "step": 7263
    },
    {
      "epoch": 0.7264,
      "grad_norm": 3.0325236320495605,
      "learning_rate": 4.227122879144523e-05,
      "loss": 4.9542,
      "step": 7264
    },
    {
      "epoch": 0.7265,
      "grad_norm": 2.5965967178344727,
      "learning_rate": 4.224272965777326e-05,
      "loss": 4.5043,
      "step": 7265
    },
    {
      "epoch": 0.7266,
      "grad_norm": 2.0044033527374268,
      "learning_rate": 4.221423756164948e-05,
      "loss": 4.7746,
      "step": 7266
    },
    {
      "epoch": 0.7267,
      "grad_norm": 5.479520320892334,
      "learning_rate": 4.2185752506545585e-05,
      "loss": 4.5944,
      "step": 7267
    },
    {
      "epoch": 0.7268,
      "grad_norm": 2.752326726913452,
      "learning_rate": 4.215727449593233e-05,
      "loss": 4.2867,
      "step": 7268
    },
    {
      "epoch": 0.7269,
      "grad_norm": 2.631657361984253,
      "learning_rate": 4.212880353327979e-05,
      "loss": 5.0573,
      "step": 7269
    },
    {
      "epoch": 0.727,
      "grad_norm": 3.6340487003326416,
      "learning_rate": 4.210033962205694e-05,
      "loss": 4.8363,
      "step": 7270
    },
    {
      "epoch": 0.7271,
      "grad_norm": 1.6969188451766968,
      "learning_rate": 4.207188276573214e-05,
      "loss": 4.5331,
      "step": 7271
    },
    {
      "epoch": 0.7272,
      "grad_norm": 2.4808480739593506,
      "learning_rate": 4.204343296777265e-05,
      "loss": 4.8733,
      "step": 7272
    },
    {
      "epoch": 0.7273,
      "grad_norm": 4.429815769195557,
      "learning_rate": 4.201499023164508e-05,
      "loss": 6.1276,
      "step": 7273
    },
    {
      "epoch": 0.7274,
      "grad_norm": 2.1912670135498047,
      "learning_rate": 4.1986554560815096e-05,
      "loss": 4.711,
      "step": 7274
    },
    {
      "epoch": 0.7275,
      "grad_norm": 4.197017192840576,
      "learning_rate": 4.195812595874739e-05,
      "loss": 5.086,
      "step": 7275
    },
    {
      "epoch": 0.7276,
      "grad_norm": 2.30334210395813,
      "learning_rate": 4.1929704428906026e-05,
      "loss": 4.5404,
      "step": 7276
    },
    {
      "epoch": 0.7277,
      "grad_norm": 1.9973835945129395,
      "learning_rate": 4.190128997475402e-05,
      "loss": 4.4653,
      "step": 7277
    },
    {
      "epoch": 0.7278,
      "grad_norm": 2.033689022064209,
      "learning_rate": 4.1872882599753605e-05,
      "loss": 4.6948,
      "step": 7278
    },
    {
      "epoch": 0.7279,
      "grad_norm": 2.2098846435546875,
      "learning_rate": 4.184448230736613e-05,
      "loss": 4.6047,
      "step": 7279
    },
    {
      "epoch": 0.728,
      "grad_norm": 2.1860265731811523,
      "learning_rate": 4.181608910105207e-05,
      "loss": 4.5175,
      "step": 7280
    },
    {
      "epoch": 0.7281,
      "grad_norm": 2.312894582748413,
      "learning_rate": 4.1787702984271074e-05,
      "loss": 4.5201,
      "step": 7281
    },
    {
      "epoch": 0.7282,
      "grad_norm": 3.8111441135406494,
      "learning_rate": 4.175932396048188e-05,
      "loss": 4.6464,
      "step": 7282
    },
    {
      "epoch": 0.7283,
      "grad_norm": 2.3976593017578125,
      "learning_rate": 4.173095203314241e-05,
      "loss": 4.7082,
      "step": 7283
    },
    {
      "epoch": 0.7284,
      "grad_norm": 2.9067769050598145,
      "learning_rate": 4.170258720570968e-05,
      "loss": 4.2265,
      "step": 7284
    },
    {
      "epoch": 0.7285,
      "grad_norm": 2.328565835952759,
      "learning_rate": 4.167422948163986e-05,
      "loss": 4.6822,
      "step": 7285
    },
    {
      "epoch": 0.7286,
      "grad_norm": 2.571366310119629,
      "learning_rate": 4.1645878864388266e-05,
      "loss": 4.4145,
      "step": 7286
    },
    {
      "epoch": 0.7287,
      "grad_norm": 3.1886096000671387,
      "learning_rate": 4.161753535740932e-05,
      "loss": 4.7821,
      "step": 7287
    },
    {
      "epoch": 0.7288,
      "grad_norm": 1.969876766204834,
      "learning_rate": 4.158919896415656e-05,
      "loss": 4.5344,
      "step": 7288
    },
    {
      "epoch": 0.7289,
      "grad_norm": 3.1434326171875,
      "learning_rate": 4.15608696880828e-05,
      "loss": 4.8893,
      "step": 7289
    },
    {
      "epoch": 0.729,
      "grad_norm": 2.8757808208465576,
      "learning_rate": 4.153254753263974e-05,
      "loss": 4.6363,
      "step": 7290
    },
    {
      "epoch": 0.7291,
      "grad_norm": 3.5336430072784424,
      "learning_rate": 4.150423250127845e-05,
      "loss": 4.6385,
      "step": 7291
    },
    {
      "epoch": 0.7292,
      "grad_norm": 4.334225177764893,
      "learning_rate": 4.1475924597449024e-05,
      "loss": 4.5153,
      "step": 7292
    },
    {
      "epoch": 0.7293,
      "grad_norm": 2.956031084060669,
      "learning_rate": 4.144762382460059e-05,
      "loss": 4.7106,
      "step": 7293
    },
    {
      "epoch": 0.7294,
      "grad_norm": 2.876479148864746,
      "learning_rate": 4.141933018618165e-05,
      "loss": 5.3191,
      "step": 7294
    },
    {
      "epoch": 0.7295,
      "grad_norm": 2.530306816101074,
      "learning_rate": 4.1391043685639576e-05,
      "loss": 5.2186,
      "step": 7295
    },
    {
      "epoch": 0.7296,
      "grad_norm": 4.605071067810059,
      "learning_rate": 4.1362764326421064e-05,
      "loss": 4.8057,
      "step": 7296
    },
    {
      "epoch": 0.7297,
      "grad_norm": 3.718815803527832,
      "learning_rate": 4.133449211197188e-05,
      "loss": 4.6161,
      "step": 7297
    },
    {
      "epoch": 0.7298,
      "grad_norm": 5.647158622741699,
      "learning_rate": 4.130622704573685e-05,
      "loss": 4.4523,
      "step": 7298
    },
    {
      "epoch": 0.7299,
      "grad_norm": 3.0595650672912598,
      "learning_rate": 4.1277969131160045e-05,
      "loss": 4.5778,
      "step": 7299
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.0720949172973633,
      "learning_rate": 4.1249718371684564e-05,
      "loss": 4.3766,
      "step": 7300
    },
    {
      "epoch": 0.7301,
      "grad_norm": 1.970886468887329,
      "learning_rate": 4.12214747707527e-05,
      "loss": 4.3769,
      "step": 7301
    },
    {
      "epoch": 0.7302,
      "grad_norm": 6.208566665649414,
      "learning_rate": 4.1193238331805826e-05,
      "loss": 4.5691,
      "step": 7302
    },
    {
      "epoch": 0.7303,
      "grad_norm": 3.337543249130249,
      "learning_rate": 4.11650090582845e-05,
      "loss": 4.7282,
      "step": 7303
    },
    {
      "epoch": 0.7304,
      "grad_norm": 3.0048279762268066,
      "learning_rate": 4.1136786953628334e-05,
      "loss": 4.7842,
      "step": 7304
    },
    {
      "epoch": 0.7305,
      "grad_norm": 3.4567596912384033,
      "learning_rate": 4.110857202127615e-05,
      "loss": 4.6958,
      "step": 7305
    },
    {
      "epoch": 0.7306,
      "grad_norm": 3.280150890350342,
      "learning_rate": 4.1080364264665774e-05,
      "loss": 4.5518,
      "step": 7306
    },
    {
      "epoch": 0.7307,
      "grad_norm": 1.740828275680542,
      "learning_rate": 4.1052163687234366e-05,
      "loss": 4.6056,
      "step": 7307
    },
    {
      "epoch": 0.7308,
      "grad_norm": 3.2386786937713623,
      "learning_rate": 4.1023970292417935e-05,
      "loss": 4.5383,
      "step": 7308
    },
    {
      "epoch": 0.7309,
      "grad_norm": 6.157960414886475,
      "learning_rate": 4.099578408365191e-05,
      "loss": 4.9059,
      "step": 7309
    },
    {
      "epoch": 0.731,
      "grad_norm": 2.2504332065582275,
      "learning_rate": 4.096760506437057e-05,
      "loss": 4.4259,
      "step": 7310
    },
    {
      "epoch": 0.7311,
      "grad_norm": 3.863981246948242,
      "learning_rate": 4.093943323800745e-05,
      "loss": 4.7577,
      "step": 7311
    },
    {
      "epoch": 0.7312,
      "grad_norm": 3.437872886657715,
      "learning_rate": 4.0911268607995325e-05,
      "loss": 4.8131,
      "step": 7312
    },
    {
      "epoch": 0.7313,
      "grad_norm": 3.2035839557647705,
      "learning_rate": 4.08831111777658e-05,
      "loss": 4.837,
      "step": 7313
    },
    {
      "epoch": 0.7314,
      "grad_norm": 3.350346326828003,
      "learning_rate": 4.08549609507499e-05,
      "loss": 4.3888,
      "step": 7314
    },
    {
      "epoch": 0.7315,
      "grad_norm": 2.439079523086548,
      "learning_rate": 4.08268179303776e-05,
      "loss": 4.565,
      "step": 7315
    },
    {
      "epoch": 0.7316,
      "grad_norm": 3.008789539337158,
      "learning_rate": 4.0798682120078044e-05,
      "loss": 4.4735,
      "step": 7316
    },
    {
      "epoch": 0.7317,
      "grad_norm": 2.71110200881958,
      "learning_rate": 4.077055352327953e-05,
      "loss": 4.606,
      "step": 7317
    },
    {
      "epoch": 0.7318,
      "grad_norm": 2.3638410568237305,
      "learning_rate": 4.074243214340934e-05,
      "loss": 4.8435,
      "step": 7318
    },
    {
      "epoch": 0.7319,
      "grad_norm": 2.3387033939361572,
      "learning_rate": 4.071431798389408e-05,
      "loss": 4.677,
      "step": 7319
    },
    {
      "epoch": 0.732,
      "grad_norm": 4.5505242347717285,
      "learning_rate": 4.068621104815934e-05,
      "loss": 4.5856,
      "step": 7320
    },
    {
      "epoch": 0.7321,
      "grad_norm": 2.0502054691314697,
      "learning_rate": 4.065811133962987e-05,
      "loss": 4.7447,
      "step": 7321
    },
    {
      "epoch": 0.7322,
      "grad_norm": 2.7310826778411865,
      "learning_rate": 4.063001886172952e-05,
      "loss": 4.642,
      "step": 7322
    },
    {
      "epoch": 0.7323,
      "grad_norm": 3.6287145614624023,
      "learning_rate": 4.0601933617881294e-05,
      "loss": 3.9203,
      "step": 7323
    },
    {
      "epoch": 0.7324,
      "grad_norm": 5.8635969161987305,
      "learning_rate": 4.057385561150727e-05,
      "loss": 5.8065,
      "step": 7324
    },
    {
      "epoch": 0.7325,
      "grad_norm": 4.632020950317383,
      "learning_rate": 4.05457848460287e-05,
      "loss": 5.1555,
      "step": 7325
    },
    {
      "epoch": 0.7326,
      "grad_norm": 3.2392959594726562,
      "learning_rate": 4.0517721324865884e-05,
      "loss": 5.0474,
      "step": 7326
    },
    {
      "epoch": 0.7327,
      "grad_norm": 4.354494094848633,
      "learning_rate": 4.048966505143831e-05,
      "loss": 4.9887,
      "step": 7327
    },
    {
      "epoch": 0.7328,
      "grad_norm": 2.574937105178833,
      "learning_rate": 4.0461616029164526e-05,
      "loss": 4.7633,
      "step": 7328
    },
    {
      "epoch": 0.7329,
      "grad_norm": 2.1488709449768066,
      "learning_rate": 4.0433574261462206e-05,
      "loss": 4.7281,
      "step": 7329
    },
    {
      "epoch": 0.733,
      "grad_norm": 5.161465167999268,
      "learning_rate": 4.040553975174823e-05,
      "loss": 5.461,
      "step": 7330
    },
    {
      "epoch": 0.7331,
      "grad_norm": 3.9845783710479736,
      "learning_rate": 4.037751250343841e-05,
      "loss": 4.1786,
      "step": 7331
    },
    {
      "epoch": 0.7332,
      "grad_norm": 2.2582366466522217,
      "learning_rate": 4.0349492519947904e-05,
      "loss": 5.1037,
      "step": 7332
    },
    {
      "epoch": 0.7333,
      "grad_norm": 6.110312461853027,
      "learning_rate": 4.032147980469072e-05,
      "loss": 6.6221,
      "step": 7333
    },
    {
      "epoch": 0.7334,
      "grad_norm": 2.935959815979004,
      "learning_rate": 4.0293474361080244e-05,
      "loss": 4.3187,
      "step": 7334
    },
    {
      "epoch": 0.7335,
      "grad_norm": 2.2674238681793213,
      "learning_rate": 4.026547619252883e-05,
      "loss": 3.84,
      "step": 7335
    },
    {
      "epoch": 0.7336,
      "grad_norm": 3.991142511367798,
      "learning_rate": 4.023748530244789e-05,
      "loss": 4.4875,
      "step": 7336
    },
    {
      "epoch": 0.7337,
      "grad_norm": 3.752913475036621,
      "learning_rate": 4.020950169424815e-05,
      "loss": 4.7649,
      "step": 7337
    },
    {
      "epoch": 0.7338,
      "grad_norm": 3.4702603816986084,
      "learning_rate": 4.018152537133919e-05,
      "loss": 4.1656,
      "step": 7338
    },
    {
      "epoch": 0.7339,
      "grad_norm": 2.301286458969116,
      "learning_rate": 4.015355633712996e-05,
      "loss": 4.8713,
      "step": 7339
    },
    {
      "epoch": 0.734,
      "grad_norm": 2.9398882389068604,
      "learning_rate": 4.012559459502835e-05,
      "loss": 5.0337,
      "step": 7340
    },
    {
      "epoch": 0.7341,
      "grad_norm": 5.174771785736084,
      "learning_rate": 4.009764014844143e-05,
      "loss": 4.3414,
      "step": 7341
    },
    {
      "epoch": 0.7342,
      "grad_norm": 3.4554998874664307,
      "learning_rate": 4.006969300077534e-05,
      "loss": 4.873,
      "step": 7342
    },
    {
      "epoch": 0.7343,
      "grad_norm": 4.359077453613281,
      "learning_rate": 4.004175315543538e-05,
      "loss": 5.0654,
      "step": 7343
    },
    {
      "epoch": 0.7344,
      "grad_norm": 2.250117778778076,
      "learning_rate": 4.001382061582593e-05,
      "loss": 4.5381,
      "step": 7344
    },
    {
      "epoch": 0.7345,
      "grad_norm": 8.893909454345703,
      "learning_rate": 3.9985895385350456e-05,
      "loss": 5.1362,
      "step": 7345
    },
    {
      "epoch": 0.7346,
      "grad_norm": 3.2431068420410156,
      "learning_rate": 3.9957977467411615e-05,
      "loss": 4.9655,
      "step": 7346
    },
    {
      "epoch": 0.7347,
      "grad_norm": 2.1653082370758057,
      "learning_rate": 3.9930066865411075e-05,
      "loss": 4.5795,
      "step": 7347
    },
    {
      "epoch": 0.7348,
      "grad_norm": 3.422745943069458,
      "learning_rate": 3.990216358274969e-05,
      "loss": 5.0131,
      "step": 7348
    },
    {
      "epoch": 0.7349,
      "grad_norm": 4.474403381347656,
      "learning_rate": 3.987426762282733e-05,
      "loss": 4.9945,
      "step": 7349
    },
    {
      "epoch": 0.735,
      "grad_norm": 6.979869842529297,
      "learning_rate": 3.9846378989043156e-05,
      "loss": 5.5773,
      "step": 7350
    },
    {
      "epoch": 0.7351,
      "grad_norm": 2.374408721923828,
      "learning_rate": 3.981849768479517e-05,
      "loss": 4.2084,
      "step": 7351
    },
    {
      "epoch": 0.7352,
      "grad_norm": 3.210376739501953,
      "learning_rate": 3.979062371348075e-05,
      "loss": 5.0413,
      "step": 7352
    },
    {
      "epoch": 0.7353,
      "grad_norm": 3.2334320545196533,
      "learning_rate": 3.976275707849616e-05,
      "loss": 5.1118,
      "step": 7353
    },
    {
      "epoch": 0.7354,
      "grad_norm": 2.6072068214416504,
      "learning_rate": 3.973489778323688e-05,
      "loss": 5.1533,
      "step": 7354
    },
    {
      "epoch": 0.7355,
      "grad_norm": 2.919268846511841,
      "learning_rate": 3.9707045831097555e-05,
      "loss": 4.8914,
      "step": 7355
    },
    {
      "epoch": 0.7356,
      "grad_norm": 2.991365671157837,
      "learning_rate": 3.967920122547175e-05,
      "loss": 5.1073,
      "step": 7356
    },
    {
      "epoch": 0.7357,
      "grad_norm": 3.0695533752441406,
      "learning_rate": 3.9651363969752344e-05,
      "loss": 4.4877,
      "step": 7357
    },
    {
      "epoch": 0.7358,
      "grad_norm": 2.326404094696045,
      "learning_rate": 3.962353406733117e-05,
      "loss": 5.3825,
      "step": 7358
    },
    {
      "epoch": 0.7359,
      "grad_norm": 3.7242343425750732,
      "learning_rate": 3.9595711521599224e-05,
      "loss": 4.7034,
      "step": 7359
    },
    {
      "epoch": 0.736,
      "grad_norm": 2.302422285079956,
      "learning_rate": 3.956789633594661e-05,
      "loss": 4.1117,
      "step": 7360
    },
    {
      "epoch": 0.7361,
      "grad_norm": 3.3237578868865967,
      "learning_rate": 3.954008851376252e-05,
      "loss": 4.8103,
      "step": 7361
    },
    {
      "epoch": 0.7362,
      "grad_norm": 2.18306827545166,
      "learning_rate": 3.9512288058435256e-05,
      "loss": 4.0667,
      "step": 7362
    },
    {
      "epoch": 0.7363,
      "grad_norm": 2.5736522674560547,
      "learning_rate": 3.948449497335219e-05,
      "loss": 4.6852,
      "step": 7363
    },
    {
      "epoch": 0.7364,
      "grad_norm": 2.1494483947753906,
      "learning_rate": 3.945670926189987e-05,
      "loss": 4.8737,
      "step": 7364
    },
    {
      "epoch": 0.7365,
      "grad_norm": 2.990034341812134,
      "learning_rate": 3.942893092746387e-05,
      "loss": 4.7911,
      "step": 7365
    },
    {
      "epoch": 0.7366,
      "grad_norm": 2.8296263217926025,
      "learning_rate": 3.940115997342891e-05,
      "loss": 4.8784,
      "step": 7366
    },
    {
      "epoch": 0.7367,
      "grad_norm": 3.6825599670410156,
      "learning_rate": 3.9373396403178786e-05,
      "loss": 5.9766,
      "step": 7367
    },
    {
      "epoch": 0.7368,
      "grad_norm": 4.343380928039551,
      "learning_rate": 3.9345640220096417e-05,
      "loss": 5.1244,
      "step": 7368
    },
    {
      "epoch": 0.7369,
      "grad_norm": 2.000394105911255,
      "learning_rate": 3.931789142756377e-05,
      "loss": 4.8158,
      "step": 7369
    },
    {
      "epoch": 0.737,
      "grad_norm": 1.8428661823272705,
      "learning_rate": 3.9290150028962044e-05,
      "loss": 4.4169,
      "step": 7370
    },
    {
      "epoch": 0.7371,
      "grad_norm": 6.037509441375732,
      "learning_rate": 3.9262416027671356e-05,
      "loss": 5.2637,
      "step": 7371
    },
    {
      "epoch": 0.7372,
      "grad_norm": 2.705439805984497,
      "learning_rate": 3.9234689427071006e-05,
      "loss": 5.1037,
      "step": 7372
    },
    {
      "epoch": 0.7373,
      "grad_norm": 2.7529449462890625,
      "learning_rate": 3.920697023053949e-05,
      "loss": 4.9594,
      "step": 7373
    },
    {
      "epoch": 0.7374,
      "grad_norm": 3.1639668941497803,
      "learning_rate": 3.917925844145418e-05,
      "loss": 4.3339,
      "step": 7374
    },
    {
      "epoch": 0.7375,
      "grad_norm": 2.2646968364715576,
      "learning_rate": 3.915155406319181e-05,
      "loss": 4.5267,
      "step": 7375
    },
    {
      "epoch": 0.7376,
      "grad_norm": 4.080491065979004,
      "learning_rate": 3.9123857099127936e-05,
      "loss": 5.1945,
      "step": 7376
    },
    {
      "epoch": 0.7377,
      "grad_norm": 5.037154197692871,
      "learning_rate": 3.9096167552637454e-05,
      "loss": 4.7783,
      "step": 7377
    },
    {
      "epoch": 0.7378,
      "grad_norm": 3.6237452030181885,
      "learning_rate": 3.9068485427094205e-05,
      "loss": 5.4634,
      "step": 7378
    },
    {
      "epoch": 0.7379,
      "grad_norm": 7.461340427398682,
      "learning_rate": 3.904081072587119e-05,
      "loss": 5.6414,
      "step": 7379
    },
    {
      "epoch": 0.738,
      "grad_norm": 5.487738609313965,
      "learning_rate": 3.9013143452340475e-05,
      "loss": 4.5558,
      "step": 7380
    },
    {
      "epoch": 0.7381,
      "grad_norm": 8.286746978759766,
      "learning_rate": 3.8985483609873244e-05,
      "loss": 4.246,
      "step": 7381
    },
    {
      "epoch": 0.7382,
      "grad_norm": 2.5099358558654785,
      "learning_rate": 3.895783120183976e-05,
      "loss": 4.8323,
      "step": 7382
    },
    {
      "epoch": 0.7383,
      "grad_norm": 2.8239667415618896,
      "learning_rate": 3.893018623160938e-05,
      "loss": 4.7612,
      "step": 7383
    },
    {
      "epoch": 0.7384,
      "grad_norm": 3.58721923828125,
      "learning_rate": 3.890254870255055e-05,
      "loss": 4.8463,
      "step": 7384
    },
    {
      "epoch": 0.7385,
      "grad_norm": 3.582440137863159,
      "learning_rate": 3.887491861803085e-05,
      "loss": 4.769,
      "step": 7385
    },
    {
      "epoch": 0.7386,
      "grad_norm": 10.537494659423828,
      "learning_rate": 3.8847295981416896e-05,
      "loss": 5.1256,
      "step": 7386
    },
    {
      "epoch": 0.7387,
      "grad_norm": 4.973750591278076,
      "learning_rate": 3.88196807960744e-05,
      "loss": 5.1155,
      "step": 7387
    },
    {
      "epoch": 0.7388,
      "grad_norm": 3.0256540775299072,
      "learning_rate": 3.879207306536829e-05,
      "loss": 4.4469,
      "step": 7388
    },
    {
      "epoch": 0.7389,
      "grad_norm": 2.192155122756958,
      "learning_rate": 3.876447279266238e-05,
      "loss": 4.5062,
      "step": 7389
    },
    {
      "epoch": 0.739,
      "grad_norm": 3.134042501449585,
      "learning_rate": 3.8736879981319695e-05,
      "loss": 4.4226,
      "step": 7390
    },
    {
      "epoch": 0.7391,
      "grad_norm": 3.8458545207977295,
      "learning_rate": 3.8709294634702376e-05,
      "loss": 4.8445,
      "step": 7391
    },
    {
      "epoch": 0.7392,
      "grad_norm": 2.274876832962036,
      "learning_rate": 3.868171675617155e-05,
      "loss": 4.9104,
      "step": 7392
    },
    {
      "epoch": 0.7393,
      "grad_norm": 3.180072069168091,
      "learning_rate": 3.8654146349087606e-05,
      "loss": 4.4729,
      "step": 7393
    },
    {
      "epoch": 0.7394,
      "grad_norm": 5.220745086669922,
      "learning_rate": 3.862658341680977e-05,
      "loss": 4.821,
      "step": 7394
    },
    {
      "epoch": 0.7395,
      "grad_norm": 2.6124682426452637,
      "learning_rate": 3.859902796269663e-05,
      "loss": 4.7673,
      "step": 7395
    },
    {
      "epoch": 0.7396,
      "grad_norm": 2.174304246902466,
      "learning_rate": 3.857147999010567e-05,
      "loss": 4.4856,
      "step": 7396
    },
    {
      "epoch": 0.7397,
      "grad_norm": 2.2426915168762207,
      "learning_rate": 3.854393950239355e-05,
      "loss": 5.4132,
      "step": 7397
    },
    {
      "epoch": 0.7398,
      "grad_norm": 6.247959613800049,
      "learning_rate": 3.851640650291603e-05,
      "loss": 5.8637,
      "step": 7398
    },
    {
      "epoch": 0.7399,
      "grad_norm": 2.1342103481292725,
      "learning_rate": 3.8488880995027786e-05,
      "loss": 4.3236,
      "step": 7399
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.7833120822906494,
      "learning_rate": 3.846136298208285e-05,
      "loss": 4.5262,
      "step": 7400
    },
    {
      "epoch": 0.7401,
      "grad_norm": 6.635382175445557,
      "learning_rate": 3.843385246743417e-05,
      "loss": 5.6829,
      "step": 7401
    },
    {
      "epoch": 0.7402,
      "grad_norm": 2.7718710899353027,
      "learning_rate": 3.840634945443382e-05,
      "loss": 5.193,
      "step": 7402
    },
    {
      "epoch": 0.7403,
      "grad_norm": 2.8045408725738525,
      "learning_rate": 3.837885394643296e-05,
      "loss": 4.4468,
      "step": 7403
    },
    {
      "epoch": 0.7404,
      "grad_norm": 1.910025954246521,
      "learning_rate": 3.835136594678183e-05,
      "loss": 4.732,
      "step": 7404
    },
    {
      "epoch": 0.7405,
      "grad_norm": 2.7052576541900635,
      "learning_rate": 3.832388545882976e-05,
      "loss": 4.4687,
      "step": 7405
    },
    {
      "epoch": 0.7406,
      "grad_norm": 2.2044858932495117,
      "learning_rate": 3.829641248592515e-05,
      "loss": 4.7001,
      "step": 7406
    },
    {
      "epoch": 0.7407,
      "grad_norm": 2.251638174057007,
      "learning_rate": 3.8268947031415514e-05,
      "loss": 4.8424,
      "step": 7407
    },
    {
      "epoch": 0.7408,
      "grad_norm": 2.5906291007995605,
      "learning_rate": 3.824148909864744e-05,
      "loss": 4.5157,
      "step": 7408
    },
    {
      "epoch": 0.7409,
      "grad_norm": 1.6362214088439941,
      "learning_rate": 3.821403869096658e-05,
      "loss": 4.2518,
      "step": 7409
    },
    {
      "epoch": 0.741,
      "grad_norm": 2.634201765060425,
      "learning_rate": 3.818659581171766e-05,
      "loss": 5.0666,
      "step": 7410
    },
    {
      "epoch": 0.7411,
      "grad_norm": 2.0914173126220703,
      "learning_rate": 3.8159160464244606e-05,
      "loss": 4.8468,
      "step": 7411
    },
    {
      "epoch": 0.7412,
      "grad_norm": 3.4045798778533936,
      "learning_rate": 3.81317326518902e-05,
      "loss": 4.731,
      "step": 7412
    },
    {
      "epoch": 0.7413,
      "grad_norm": 4.801126956939697,
      "learning_rate": 3.8104312377996564e-05,
      "loss": 5.5084,
      "step": 7413
    },
    {
      "epoch": 0.7414,
      "grad_norm": 4.013615131378174,
      "learning_rate": 3.807689964590466e-05,
      "loss": 5.4976,
      "step": 7414
    },
    {
      "epoch": 0.7415,
      "grad_norm": 2.954400062561035,
      "learning_rate": 3.8049494458954725e-05,
      "loss": 4.1713,
      "step": 7415
    },
    {
      "epoch": 0.7416,
      "grad_norm": 3.121324062347412,
      "learning_rate": 3.802209682048602e-05,
      "loss": 5.6745,
      "step": 7416
    },
    {
      "epoch": 0.7417,
      "grad_norm": 2.263406276702881,
      "learning_rate": 3.799470673383674e-05,
      "loss": 5.1229,
      "step": 7417
    },
    {
      "epoch": 0.7418,
      "grad_norm": 2.536004066467285,
      "learning_rate": 3.796732420234443e-05,
      "loss": 5.2567,
      "step": 7418
    },
    {
      "epoch": 0.7419,
      "grad_norm": 2.850757122039795,
      "learning_rate": 3.793994922934544e-05,
      "loss": 5.3836,
      "step": 7419
    },
    {
      "epoch": 0.742,
      "grad_norm": 2.9680144786834717,
      "learning_rate": 3.791258181817542e-05,
      "loss": 4.6458,
      "step": 7420
    },
    {
      "epoch": 0.7421,
      "grad_norm": 2.474722146987915,
      "learning_rate": 3.788522197216897e-05,
      "loss": 4.5792,
      "step": 7421
    },
    {
      "epoch": 0.7422,
      "grad_norm": 4.419800281524658,
      "learning_rate": 3.785786969465981e-05,
      "loss": 5.0224,
      "step": 7422
    },
    {
      "epoch": 0.7423,
      "grad_norm": 2.3568809032440186,
      "learning_rate": 3.783052498898073e-05,
      "loss": 3.9979,
      "step": 7423
    },
    {
      "epoch": 0.7424,
      "grad_norm": 3.0555810928344727,
      "learning_rate": 3.7803187858463605e-05,
      "loss": 4.6268,
      "step": 7424
    },
    {
      "epoch": 0.7425,
      "grad_norm": 2.5833568572998047,
      "learning_rate": 3.777585830643937e-05,
      "loss": 4.5643,
      "step": 7425
    },
    {
      "epoch": 0.7426,
      "grad_norm": 3.2232677936553955,
      "learning_rate": 3.774853633623806e-05,
      "loss": 4.8124,
      "step": 7426
    },
    {
      "epoch": 0.7427,
      "grad_norm": 2.1635968685150146,
      "learning_rate": 3.7721221951188765e-05,
      "loss": 4.8036,
      "step": 7427
    },
    {
      "epoch": 0.7428,
      "grad_norm": 5.873354911804199,
      "learning_rate": 3.769391515461966e-05,
      "loss": 6.1119,
      "step": 7428
    },
    {
      "epoch": 0.7429,
      "grad_norm": 2.137972593307495,
      "learning_rate": 3.766661594985801e-05,
      "loss": 4.7352,
      "step": 7429
    },
    {
      "epoch": 0.743,
      "grad_norm": 3.636322259902954,
      "learning_rate": 3.7639324340230085e-05,
      "loss": 4.3652,
      "step": 7430
    },
    {
      "epoch": 0.7431,
      "grad_norm": 2.8464059829711914,
      "learning_rate": 3.7612040329061405e-05,
      "loss": 4.4994,
      "step": 7431
    },
    {
      "epoch": 0.7432,
      "grad_norm": 5.476583957672119,
      "learning_rate": 3.7584763919676294e-05,
      "loss": 4.5829,
      "step": 7432
    },
    {
      "epoch": 0.7433,
      "grad_norm": 2.7436256408691406,
      "learning_rate": 3.755749511539845e-05,
      "loss": 4.4645,
      "step": 7433
    },
    {
      "epoch": 0.7434,
      "grad_norm": 9.432605743408203,
      "learning_rate": 3.753023391955037e-05,
      "loss": 5.0728,
      "step": 7434
    },
    {
      "epoch": 0.7435,
      "grad_norm": 3.242563486099243,
      "learning_rate": 3.7502980335453774e-05,
      "loss": 4.3151,
      "step": 7435
    },
    {
      "epoch": 0.7436,
      "grad_norm": 2.8658759593963623,
      "learning_rate": 3.747573436642951e-05,
      "loss": 4.5272,
      "step": 7436
    },
    {
      "epoch": 0.7437,
      "grad_norm": 2.571413040161133,
      "learning_rate": 3.7448496015797295e-05,
      "loss": 4.9856,
      "step": 7437
    },
    {
      "epoch": 0.7438,
      "grad_norm": 2.721071481704712,
      "learning_rate": 3.742126528687614e-05,
      "loss": 4.525,
      "step": 7438
    },
    {
      "epoch": 0.7439,
      "grad_norm": 2.6745874881744385,
      "learning_rate": 3.739404218298398e-05,
      "loss": 4.9796,
      "step": 7439
    },
    {
      "epoch": 0.744,
      "grad_norm": 2.787709951400757,
      "learning_rate": 3.736682670743787e-05,
      "loss": 5.0433,
      "step": 7440
    },
    {
      "epoch": 0.7441,
      "grad_norm": 4.215086460113525,
      "learning_rate": 3.733961886355398e-05,
      "loss": 4.3906,
      "step": 7441
    },
    {
      "epoch": 0.7442,
      "grad_norm": 2.3593664169311523,
      "learning_rate": 3.7312418654647406e-05,
      "loss": 4.7395,
      "step": 7442
    },
    {
      "epoch": 0.7443,
      "grad_norm": 2.9756569862365723,
      "learning_rate": 3.728522608403249e-05,
      "loss": 4.7087,
      "step": 7443
    },
    {
      "epoch": 0.7444,
      "grad_norm": 3.076789617538452,
      "learning_rate": 3.725804115502254e-05,
      "loss": 4.6106,
      "step": 7444
    },
    {
      "epoch": 0.7445,
      "grad_norm": 3.473828077316284,
      "learning_rate": 3.7230863870929964e-05,
      "loss": 4.6285,
      "step": 7445
    },
    {
      "epoch": 0.7446,
      "grad_norm": 5.4060564041137695,
      "learning_rate": 3.720369423506622e-05,
      "loss": 4.8601,
      "step": 7446
    },
    {
      "epoch": 0.7447,
      "grad_norm": 2.7085139751434326,
      "learning_rate": 3.717653225074186e-05,
      "loss": 4.8337,
      "step": 7447
    },
    {
      "epoch": 0.7448,
      "grad_norm": 2.4101219177246094,
      "learning_rate": 3.714937792126647e-05,
      "loss": 4.3061,
      "step": 7448
    },
    {
      "epoch": 0.7449,
      "grad_norm": 2.4212114810943604,
      "learning_rate": 3.712223124994875e-05,
      "loss": 4.7736,
      "step": 7449
    },
    {
      "epoch": 0.745,
      "grad_norm": 2.954648733139038,
      "learning_rate": 3.709509224009641e-05,
      "loss": 4.4644,
      "step": 7450
    },
    {
      "epoch": 0.7451,
      "grad_norm": 3.1662940979003906,
      "learning_rate": 3.7067960895016275e-05,
      "loss": 4.9053,
      "step": 7451
    },
    {
      "epoch": 0.7452,
      "grad_norm": 2.4566173553466797,
      "learning_rate": 3.704083721801421e-05,
      "loss": 4.9484,
      "step": 7452
    },
    {
      "epoch": 0.7453,
      "grad_norm": 2.03592586517334,
      "learning_rate": 3.701372121239512e-05,
      "loss": 4.7384,
      "step": 7453
    },
    {
      "epoch": 0.7454,
      "grad_norm": 2.7895381450653076,
      "learning_rate": 3.698661288146311e-05,
      "loss": 4.257,
      "step": 7454
    },
    {
      "epoch": 0.7455,
      "grad_norm": 3.310062885284424,
      "learning_rate": 3.695951222852112e-05,
      "loss": 4.3051,
      "step": 7455
    },
    {
      "epoch": 0.7456,
      "grad_norm": 4.034221649169922,
      "learning_rate": 3.6932419256871406e-05,
      "loss": 4.6986,
      "step": 7456
    },
    {
      "epoch": 0.7457,
      "grad_norm": 3.6039161682128906,
      "learning_rate": 3.690533396981504e-05,
      "loss": 5.3439,
      "step": 7457
    },
    {
      "epoch": 0.7458,
      "grad_norm": 2.771193504333496,
      "learning_rate": 3.687825637065236e-05,
      "loss": 4.5917,
      "step": 7458
    },
    {
      "epoch": 0.7459,
      "grad_norm": 3.494795560836792,
      "learning_rate": 3.685118646268272e-05,
      "loss": 4.6838,
      "step": 7459
    },
    {
      "epoch": 0.746,
      "grad_norm": 2.6555981636047363,
      "learning_rate": 3.682412424920438e-05,
      "loss": 4.4832,
      "step": 7460
    },
    {
      "epoch": 0.7461,
      "grad_norm": 3.131434917449951,
      "learning_rate": 3.679706973351491e-05,
      "loss": 4.7157,
      "step": 7461
    },
    {
      "epoch": 0.7462,
      "grad_norm": 3.2190182209014893,
      "learning_rate": 3.677002291891078e-05,
      "loss": 4.6469,
      "step": 7462
    },
    {
      "epoch": 0.7463,
      "grad_norm": 3.3064401149749756,
      "learning_rate": 3.674298380868756e-05,
      "loss": 4.8481,
      "step": 7463
    },
    {
      "epoch": 0.7464,
      "grad_norm": 2.4345879554748535,
      "learning_rate": 3.6715952406139885e-05,
      "loss": 4.4771,
      "step": 7464
    },
    {
      "epoch": 0.7465,
      "grad_norm": 2.0760743618011475,
      "learning_rate": 3.668892871456144e-05,
      "loss": 4.5657,
      "step": 7465
    },
    {
      "epoch": 0.7466,
      "grad_norm": 2.7389047145843506,
      "learning_rate": 3.6661912737245e-05,
      "loss": 4.6368,
      "step": 7466
    },
    {
      "epoch": 0.7467,
      "grad_norm": 2.6871817111968994,
      "learning_rate": 3.6634904477482354e-05,
      "loss": 4.8506,
      "step": 7467
    },
    {
      "epoch": 0.7468,
      "grad_norm": 2.3580281734466553,
      "learning_rate": 3.6607903938564405e-05,
      "loss": 4.6051,
      "step": 7468
    },
    {
      "epoch": 0.7469,
      "grad_norm": 4.374931335449219,
      "learning_rate": 3.6580911123781056e-05,
      "loss": 4.4015,
      "step": 7469
    },
    {
      "epoch": 0.747,
      "grad_norm": 2.313063383102417,
      "learning_rate": 3.6553926036421326e-05,
      "loss": 5.0083,
      "step": 7470
    },
    {
      "epoch": 0.7471,
      "grad_norm": 4.084475994110107,
      "learning_rate": 3.6526948679773257e-05,
      "loss": 5.321,
      "step": 7471
    },
    {
      "epoch": 0.7472,
      "grad_norm": 2.3246994018554688,
      "learning_rate": 3.649997905712396e-05,
      "loss": 4.5458,
      "step": 7472
    },
    {
      "epoch": 0.7473,
      "grad_norm": 2.181446075439453,
      "learning_rate": 3.647301717175956e-05,
      "loss": 4.6015,
      "step": 7473
    },
    {
      "epoch": 0.7474,
      "grad_norm": 2.806213617324829,
      "learning_rate": 3.6446063026965384e-05,
      "loss": 4.7021,
      "step": 7474
    },
    {
      "epoch": 0.7475,
      "grad_norm": 4.253084659576416,
      "learning_rate": 3.641911662602559e-05,
      "loss": 5.2429,
      "step": 7475
    },
    {
      "epoch": 0.7476,
      "grad_norm": 2.313260316848755,
      "learning_rate": 3.6392177972223594e-05,
      "loss": 4.6141,
      "step": 7476
    },
    {
      "epoch": 0.7477,
      "grad_norm": 1.9722349643707275,
      "learning_rate": 3.6365247068841814e-05,
      "loss": 4.7655,
      "step": 7477
    },
    {
      "epoch": 0.7478,
      "grad_norm": 2.5791287422180176,
      "learning_rate": 3.633832391916159e-05,
      "loss": 5.0371,
      "step": 7478
    },
    {
      "epoch": 0.7479,
      "grad_norm": 2.226858139038086,
      "learning_rate": 3.631140852646355e-05,
      "loss": 4.3188,
      "step": 7479
    },
    {
      "epoch": 0.748,
      "grad_norm": 3.831892967224121,
      "learning_rate": 3.628450089402713e-05,
      "loss": 4.8008,
      "step": 7480
    },
    {
      "epoch": 0.7481,
      "grad_norm": 1.978545069694519,
      "learning_rate": 3.6257601025131026e-05,
      "loss": 4.3936,
      "step": 7481
    },
    {
      "epoch": 0.7482,
      "grad_norm": 4.960198879241943,
      "learning_rate": 3.62307089230529e-05,
      "loss": 4.5332,
      "step": 7482
    },
    {
      "epoch": 0.7483,
      "grad_norm": 2.911310911178589,
      "learning_rate": 3.6203824591069456e-05,
      "loss": 5.1849,
      "step": 7483
    },
    {
      "epoch": 0.7484,
      "grad_norm": 4.685695648193359,
      "learning_rate": 3.6176948032456473e-05,
      "loss": 4.7103,
      "step": 7484
    },
    {
      "epoch": 0.7485,
      "grad_norm": 5.327178955078125,
      "learning_rate": 3.615007925048878e-05,
      "loss": 5.0232,
      "step": 7485
    },
    {
      "epoch": 0.7486,
      "grad_norm": 2.6546785831451416,
      "learning_rate": 3.612321824844024e-05,
      "loss": 4.4978,
      "step": 7486
    },
    {
      "epoch": 0.7487,
      "grad_norm": 3.9112653732299805,
      "learning_rate": 3.60963650295838e-05,
      "loss": 4.8618,
      "step": 7487
    },
    {
      "epoch": 0.7488,
      "grad_norm": 3.0247962474823,
      "learning_rate": 3.606951959719145e-05,
      "loss": 4.6185,
      "step": 7488
    },
    {
      "epoch": 0.7489,
      "grad_norm": 2.944598436355591,
      "learning_rate": 3.604268195453421e-05,
      "loss": 4.8634,
      "step": 7489
    },
    {
      "epoch": 0.749,
      "grad_norm": 4.193759918212891,
      "learning_rate": 3.601585210488218e-05,
      "loss": 4.9732,
      "step": 7490
    },
    {
      "epoch": 0.7491,
      "grad_norm": 2.671940326690674,
      "learning_rate": 3.5989030051504434e-05,
      "loss": 4.5977,
      "step": 7491
    },
    {
      "epoch": 0.7492,
      "grad_norm": 7.636031150817871,
      "learning_rate": 3.59622157976693e-05,
      "loss": 4.4738,
      "step": 7492
    },
    {
      "epoch": 0.7493,
      "grad_norm": 2.4795100688934326,
      "learning_rate": 3.5935409346643835e-05,
      "loss": 4.7557,
      "step": 7493
    },
    {
      "epoch": 0.7494,
      "grad_norm": 4.130136966705322,
      "learning_rate": 3.590861070169449e-05,
      "loss": 5.0744,
      "step": 7494
    },
    {
      "epoch": 0.7495,
      "grad_norm": 5.978124141693115,
      "learning_rate": 3.5881819866086484e-05,
      "loss": 5.0075,
      "step": 7495
    },
    {
      "epoch": 0.7496,
      "grad_norm": 1.7893519401550293,
      "learning_rate": 3.585503684308421e-05,
      "loss": 4.5842,
      "step": 7496
    },
    {
      "epoch": 0.7497,
      "grad_norm": 3.935670852661133,
      "learning_rate": 3.582826163595119e-05,
      "loss": 5.0223,
      "step": 7497
    },
    {
      "epoch": 0.7498,
      "grad_norm": 5.097084045410156,
      "learning_rate": 3.580149424794976e-05,
      "loss": 5.201,
      "step": 7498
    },
    {
      "epoch": 0.7499,
      "grad_norm": 3.1531500816345215,
      "learning_rate": 3.577473468234156e-05,
      "loss": 4.8125,
      "step": 7499
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.3333353996276855,
      "learning_rate": 3.574798294238713e-05,
      "loss": 4.7048,
      "step": 7500
    },
    {
      "epoch": 0.7501,
      "grad_norm": 5.05497407913208,
      "learning_rate": 3.5721239031346066e-05,
      "loss": 5.2318,
      "step": 7501
    },
    {
      "epoch": 0.7502,
      "grad_norm": 2.6313507556915283,
      "learning_rate": 3.569450295247706e-05,
      "loss": 5.0558,
      "step": 7502
    },
    {
      "epoch": 0.7503,
      "grad_norm": 3.1252827644348145,
      "learning_rate": 3.56677747090378e-05,
      "loss": 5.6749,
      "step": 7503
    },
    {
      "epoch": 0.7504,
      "grad_norm": 2.240551471710205,
      "learning_rate": 3.564105430428506e-05,
      "loss": 4.6784,
      "step": 7504
    },
    {
      "epoch": 0.7505,
      "grad_norm": 2.9373457431793213,
      "learning_rate": 3.561434174147463e-05,
      "loss": 4.5743,
      "step": 7505
    },
    {
      "epoch": 0.7506,
      "grad_norm": 2.2142040729522705,
      "learning_rate": 3.558763702386135e-05,
      "loss": 4.5734,
      "step": 7506
    },
    {
      "epoch": 0.7507,
      "grad_norm": 2.4223318099975586,
      "learning_rate": 3.556094015469913e-05,
      "loss": 5.013,
      "step": 7507
    },
    {
      "epoch": 0.7508,
      "grad_norm": 2.650515556335449,
      "learning_rate": 3.553425113724088e-05,
      "loss": 5.0617,
      "step": 7508
    },
    {
      "epoch": 0.7509,
      "grad_norm": 2.0497140884399414,
      "learning_rate": 3.5507569974738574e-05,
      "loss": 5.0605,
      "step": 7509
    },
    {
      "epoch": 0.751,
      "grad_norm": 4.139504432678223,
      "learning_rate": 3.548089667044325e-05,
      "loss": 4.246,
      "step": 7510
    },
    {
      "epoch": 0.7511,
      "grad_norm": 4.080862045288086,
      "learning_rate": 3.545423122760493e-05,
      "loss": 4.6684,
      "step": 7511
    },
    {
      "epoch": 0.7512,
      "grad_norm": 3.82224702835083,
      "learning_rate": 3.542757364947281e-05,
      "loss": 5.4213,
      "step": 7512
    },
    {
      "epoch": 0.7513,
      "grad_norm": 2.709181547164917,
      "learning_rate": 3.540092393929494e-05,
      "loss": 4.5074,
      "step": 7513
    },
    {
      "epoch": 0.7514,
      "grad_norm": 2.38828444480896,
      "learning_rate": 3.537428210031849e-05,
      "loss": 4.4073,
      "step": 7514
    },
    {
      "epoch": 0.7515,
      "grad_norm": 2.0582339763641357,
      "learning_rate": 3.534764813578982e-05,
      "loss": 4.8481,
      "step": 7515
    },
    {
      "epoch": 0.7516,
      "grad_norm": 4.702442646026611,
      "learning_rate": 3.5321022048954035e-05,
      "loss": 5.492,
      "step": 7516
    },
    {
      "epoch": 0.7517,
      "grad_norm": 3.7118210792541504,
      "learning_rate": 3.52944038430556e-05,
      "loss": 4.9675,
      "step": 7517
    },
    {
      "epoch": 0.7518,
      "grad_norm": 2.6032662391662598,
      "learning_rate": 3.52677935213377e-05,
      "loss": 4.3287,
      "step": 7518
    },
    {
      "epoch": 0.7519,
      "grad_norm": 2.129124879837036,
      "learning_rate": 3.524119108704286e-05,
      "loss": 4.4574,
      "step": 7519
    },
    {
      "epoch": 0.752,
      "grad_norm": 2.586277723312378,
      "learning_rate": 3.521459654341244e-05,
      "loss": 4.1437,
      "step": 7520
    },
    {
      "epoch": 0.7521,
      "grad_norm": 3.531826972961426,
      "learning_rate": 3.518800989368691e-05,
      "loss": 4.7556,
      "step": 7521
    },
    {
      "epoch": 0.7522,
      "grad_norm": 1.7962501049041748,
      "learning_rate": 3.516143114110582e-05,
      "loss": 4.7777,
      "step": 7522
    },
    {
      "epoch": 0.7523,
      "grad_norm": 2.786057710647583,
      "learning_rate": 3.51348602889076e-05,
      "loss": 4.3339,
      "step": 7523
    },
    {
      "epoch": 0.7524,
      "grad_norm": 4.344465732574463,
      "learning_rate": 3.510829734032993e-05,
      "loss": 4.8063,
      "step": 7524
    },
    {
      "epoch": 0.7525,
      "grad_norm": 2.448122501373291,
      "learning_rate": 3.50817422986094e-05,
      "loss": 4.3938,
      "step": 7525
    },
    {
      "epoch": 0.7526,
      "grad_norm": 5.004829406738281,
      "learning_rate": 3.5055195166981645e-05,
      "loss": 4.7087,
      "step": 7526
    },
    {
      "epoch": 0.7527,
      "grad_norm": 2.6380536556243896,
      "learning_rate": 3.5028655948681355e-05,
      "loss": 4.7848,
      "step": 7527
    },
    {
      "epoch": 0.7528,
      "grad_norm": 3.305361270904541,
      "learning_rate": 3.500212464694227e-05,
      "loss": 4.6687,
      "step": 7528
    },
    {
      "epoch": 0.7529,
      "grad_norm": 3.390648365020752,
      "learning_rate": 3.497560126499709e-05,
      "loss": 4.6511,
      "step": 7529
    },
    {
      "epoch": 0.753,
      "grad_norm": 2.441277027130127,
      "learning_rate": 3.494908580607774e-05,
      "loss": 4.687,
      "step": 7530
    },
    {
      "epoch": 0.7531,
      "grad_norm": 2.629943609237671,
      "learning_rate": 3.492257827341492e-05,
      "loss": 4.9557,
      "step": 7531
    },
    {
      "epoch": 0.7532,
      "grad_norm": 2.49126935005188,
      "learning_rate": 3.489607867023854e-05,
      "loss": 5.0667,
      "step": 7532
    },
    {
      "epoch": 0.7533,
      "grad_norm": 1.8979851007461548,
      "learning_rate": 3.4869586999777495e-05,
      "loss": 4.6925,
      "step": 7533
    },
    {
      "epoch": 0.7534,
      "grad_norm": 1.959442138671875,
      "learning_rate": 3.484310326525967e-05,
      "loss": 4.5171,
      "step": 7534
    },
    {
      "epoch": 0.7535,
      "grad_norm": 3.385106086730957,
      "learning_rate": 3.481662746991214e-05,
      "loss": 4.4908,
      "step": 7535
    },
    {
      "epoch": 0.7536,
      "grad_norm": 2.630618095397949,
      "learning_rate": 3.479015961696077e-05,
      "loss": 5.0185,
      "step": 7536
    },
    {
      "epoch": 0.7537,
      "grad_norm": 12.294042587280273,
      "learning_rate": 3.4763699709630716e-05,
      "loss": 4.8934,
      "step": 7537
    },
    {
      "epoch": 0.7538,
      "grad_norm": 7.232980728149414,
      "learning_rate": 3.4737247751145896e-05,
      "loss": 4.6762,
      "step": 7538
    },
    {
      "epoch": 0.7539,
      "grad_norm": 2.8409855365753174,
      "learning_rate": 3.4710803744729515e-05,
      "loss": 4.905,
      "step": 7539
    },
    {
      "epoch": 0.754,
      "grad_norm": 5.446911334991455,
      "learning_rate": 3.468436769360368e-05,
      "loss": 5.06,
      "step": 7540
    },
    {
      "epoch": 0.7541,
      "grad_norm": 2.052952527999878,
      "learning_rate": 3.465793960098945e-05,
      "loss": 4.8704,
      "step": 7541
    },
    {
      "epoch": 0.7542,
      "grad_norm": 3.6328256130218506,
      "learning_rate": 3.463151947010712e-05,
      "loss": 4.6281,
      "step": 7542
    },
    {
      "epoch": 0.7543,
      "grad_norm": 3.5574634075164795,
      "learning_rate": 3.460510730417585e-05,
      "loss": 4.0359,
      "step": 7543
    },
    {
      "epoch": 0.7544,
      "grad_norm": 3.5813586711883545,
      "learning_rate": 3.4578703106413904e-05,
      "loss": 4.4076,
      "step": 7544
    },
    {
      "epoch": 0.7545,
      "grad_norm": 2.065000057220459,
      "learning_rate": 3.455230688003852e-05,
      "loss": 5.001,
      "step": 7545
    },
    {
      "epoch": 0.7546,
      "grad_norm": 2.616549491882324,
      "learning_rate": 3.452591862826603e-05,
      "loss": 4.9525,
      "step": 7546
    },
    {
      "epoch": 0.7547,
      "grad_norm": 4.1192145347595215,
      "learning_rate": 3.4499538354311755e-05,
      "loss": 5.0097,
      "step": 7547
    },
    {
      "epoch": 0.7548,
      "grad_norm": 7.09456205368042,
      "learning_rate": 3.447316606139004e-05,
      "loss": 5.9253,
      "step": 7548
    },
    {
      "epoch": 0.7549,
      "grad_norm": 3.385072946548462,
      "learning_rate": 3.444680175271428e-05,
      "loss": 4.5901,
      "step": 7549
    },
    {
      "epoch": 0.755,
      "grad_norm": 2.930264472961426,
      "learning_rate": 3.442044543149688e-05,
      "loss": 4.5917,
      "step": 7550
    },
    {
      "epoch": 0.7551,
      "grad_norm": 3.1859066486358643,
      "learning_rate": 3.439409710094929e-05,
      "loss": 5.2039,
      "step": 7551
    },
    {
      "epoch": 0.7552,
      "grad_norm": 3.9558510780334473,
      "learning_rate": 3.4367756764281955e-05,
      "loss": 4.7465,
      "step": 7552
    },
    {
      "epoch": 0.7553,
      "grad_norm": 3.4159457683563232,
      "learning_rate": 3.4341424424704375e-05,
      "loss": 5.0373,
      "step": 7553
    },
    {
      "epoch": 0.7554,
      "grad_norm": 1.7315213680267334,
      "learning_rate": 3.4315100085425034e-05,
      "loss": 4.481,
      "step": 7554
    },
    {
      "epoch": 0.7555,
      "grad_norm": 2.3384029865264893,
      "learning_rate": 3.4288783749651564e-05,
      "loss": 4.6397,
      "step": 7555
    },
    {
      "epoch": 0.7556,
      "grad_norm": 2.314638137817383,
      "learning_rate": 3.426247542059041e-05,
      "loss": 4.9329,
      "step": 7556
    },
    {
      "epoch": 0.7557,
      "grad_norm": 2.384962797164917,
      "learning_rate": 3.423617510144727e-05,
      "loss": 4.6848,
      "step": 7557
    },
    {
      "epoch": 0.7558,
      "grad_norm": 3.754581928253174,
      "learning_rate": 3.4209882795426716e-05,
      "loss": 4.5942,
      "step": 7558
    },
    {
      "epoch": 0.7559,
      "grad_norm": 1.9828150272369385,
      "learning_rate": 3.418359850573234e-05,
      "loss": 4.6001,
      "step": 7559
    },
    {
      "epoch": 0.756,
      "grad_norm": 1.8891481161117554,
      "learning_rate": 3.415732223556689e-05,
      "loss": 4.8267,
      "step": 7560
    },
    {
      "epoch": 0.7561,
      "grad_norm": 2.7930924892425537,
      "learning_rate": 3.413105398813195e-05,
      "loss": 4.9467,
      "step": 7561
    },
    {
      "epoch": 0.7562,
      "grad_norm": 2.1640641689300537,
      "learning_rate": 3.4104793766628304e-05,
      "loss": 4.6045,
      "step": 7562
    },
    {
      "epoch": 0.7563,
      "grad_norm": 3.066267967224121,
      "learning_rate": 3.4078541574255664e-05,
      "loss": 4.9632,
      "step": 7563
    },
    {
      "epoch": 0.7564,
      "grad_norm": 2.2765111923217773,
      "learning_rate": 3.4052297414212777e-05,
      "loss": 4.6782,
      "step": 7564
    },
    {
      "epoch": 0.7565,
      "grad_norm": 3.560899019241333,
      "learning_rate": 3.40260612896974e-05,
      "loss": 4.4249,
      "step": 7565
    },
    {
      "epoch": 0.7566,
      "grad_norm": 3.066643238067627,
      "learning_rate": 3.3999833203906326e-05,
      "loss": 4.8315,
      "step": 7566
    },
    {
      "epoch": 0.7567,
      "grad_norm": 2.1972830295562744,
      "learning_rate": 3.397361316003539e-05,
      "loss": 4.1811,
      "step": 7567
    },
    {
      "epoch": 0.7568,
      "grad_norm": 3.282735586166382,
      "learning_rate": 3.394740116127941e-05,
      "loss": 4.5345,
      "step": 7568
    },
    {
      "epoch": 0.7569,
      "grad_norm": 3.2736012935638428,
      "learning_rate": 3.3921197210832235e-05,
      "loss": 4.4858,
      "step": 7569
    },
    {
      "epoch": 0.757,
      "grad_norm": 2.6608493328094482,
      "learning_rate": 3.389500131188674e-05,
      "loss": 4.8205,
      "step": 7570
    },
    {
      "epoch": 0.7571,
      "grad_norm": 3.0682225227355957,
      "learning_rate": 3.386881346763483e-05,
      "loss": 4.7851,
      "step": 7571
    },
    {
      "epoch": 0.7572,
      "grad_norm": 2.768428325653076,
      "learning_rate": 3.3842633681267356e-05,
      "loss": 4.5076,
      "step": 7572
    },
    {
      "epoch": 0.7573,
      "grad_norm": 15.498311042785645,
      "learning_rate": 3.3816461955974365e-05,
      "loss": 5.8002,
      "step": 7573
    },
    {
      "epoch": 0.7574,
      "grad_norm": 3.6833555698394775,
      "learning_rate": 3.379029829494469e-05,
      "loss": 4.6291,
      "step": 7574
    },
    {
      "epoch": 0.7575,
      "grad_norm": 2.177237033843994,
      "learning_rate": 3.376414270136633e-05,
      "loss": 4.56,
      "step": 7575
    },
    {
      "epoch": 0.7576,
      "grad_norm": 2.5786707401275635,
      "learning_rate": 3.373799517842627e-05,
      "loss": 4.8589,
      "step": 7576
    },
    {
      "epoch": 0.7577,
      "grad_norm": 1.844935417175293,
      "learning_rate": 3.371185572931048e-05,
      "loss": 4.4416,
      "step": 7577
    },
    {
      "epoch": 0.7578,
      "grad_norm": 2.295696496963501,
      "learning_rate": 3.3685724357204054e-05,
      "loss": 4.5515,
      "step": 7578
    },
    {
      "epoch": 0.7579,
      "grad_norm": 2.636928081512451,
      "learning_rate": 3.3659601065290893e-05,
      "loss": 5.0763,
      "step": 7579
    },
    {
      "epoch": 0.758,
      "grad_norm": 2.405372381210327,
      "learning_rate": 3.363348585675414e-05,
      "loss": 4.6725,
      "step": 7580
    },
    {
      "epoch": 0.7581,
      "grad_norm": 2.1601765155792236,
      "learning_rate": 3.360737873477584e-05,
      "loss": 4.7594,
      "step": 7581
    },
    {
      "epoch": 0.7582,
      "grad_norm": 3.169466257095337,
      "learning_rate": 3.358127970253704e-05,
      "loss": 4.6895,
      "step": 7582
    },
    {
      "epoch": 0.7583,
      "grad_norm": 3.2889819145202637,
      "learning_rate": 3.355518876321787e-05,
      "loss": 4.356,
      "step": 7583
    },
    {
      "epoch": 0.7584,
      "grad_norm": 2.542431354522705,
      "learning_rate": 3.352910591999734e-05,
      "loss": 4.3706,
      "step": 7584
    },
    {
      "epoch": 0.7585,
      "grad_norm": 5.138837814331055,
      "learning_rate": 3.3503031176053656e-05,
      "loss": 4.8312,
      "step": 7585
    },
    {
      "epoch": 0.7586,
      "grad_norm": 1.9731264114379883,
      "learning_rate": 3.347696453456393e-05,
      "loss": 4.4811,
      "step": 7586
    },
    {
      "epoch": 0.7587,
      "grad_norm": 5.765212535858154,
      "learning_rate": 3.3450905998704275e-05,
      "loss": 4.8223,
      "step": 7587
    },
    {
      "epoch": 0.7588,
      "grad_norm": 2.907089948654175,
      "learning_rate": 3.342485557164986e-05,
      "loss": 4.544,
      "step": 7588
    },
    {
      "epoch": 0.7589,
      "grad_norm": 2.7156972885131836,
      "learning_rate": 3.339881325657484e-05,
      "loss": 4.9258,
      "step": 7589
    },
    {
      "epoch": 0.759,
      "grad_norm": 6.517168045043945,
      "learning_rate": 3.3372779056652426e-05,
      "loss": 4.6385,
      "step": 7590
    },
    {
      "epoch": 0.7591,
      "grad_norm": 2.7199673652648926,
      "learning_rate": 3.334675297505476e-05,
      "loss": 4.1487,
      "step": 7591
    },
    {
      "epoch": 0.7592,
      "grad_norm": 6.5071845054626465,
      "learning_rate": 3.3320735014953076e-05,
      "loss": 4.7811,
      "step": 7592
    },
    {
      "epoch": 0.7593,
      "grad_norm": 5.107992172241211,
      "learning_rate": 3.3294725179517574e-05,
      "loss": 4.5868,
      "step": 7593
    },
    {
      "epoch": 0.7594,
      "grad_norm": 4.754256725311279,
      "learning_rate": 3.326872347191746e-05,
      "loss": 5.0592,
      "step": 7594
    },
    {
      "epoch": 0.7595,
      "grad_norm": 2.3733572959899902,
      "learning_rate": 3.3242729895320946e-05,
      "loss": 5.3019,
      "step": 7595
    },
    {
      "epoch": 0.7596,
      "grad_norm": 2.4973249435424805,
      "learning_rate": 3.3216744452895354e-05,
      "loss": 4.805,
      "step": 7596
    },
    {
      "epoch": 0.7597,
      "grad_norm": 4.66115140914917,
      "learning_rate": 3.319076714780682e-05,
      "loss": 5.2505,
      "step": 7597
    },
    {
      "epoch": 0.7598,
      "grad_norm": 3.937340021133423,
      "learning_rate": 3.316479798322072e-05,
      "loss": 4.6896,
      "step": 7598
    },
    {
      "epoch": 0.7599,
      "grad_norm": 2.0259222984313965,
      "learning_rate": 3.313883696230119e-05,
      "loss": 4.8553,
      "step": 7599
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.499481678009033,
      "learning_rate": 3.311288408821159e-05,
      "loss": 5.0269,
      "step": 7600
    },
    {
      "epoch": 0.7601,
      "grad_norm": 2.9080963134765625,
      "learning_rate": 3.308693936411421e-05,
      "loss": 4.5367,
      "step": 7601
    },
    {
      "epoch": 0.7602,
      "grad_norm": 1.9968624114990234,
      "learning_rate": 3.306100279317024e-05,
      "loss": 4.5955,
      "step": 7602
    },
    {
      "epoch": 0.7603,
      "grad_norm": 2.7361972332000732,
      "learning_rate": 3.303507437854009e-05,
      "loss": 4.5172,
      "step": 7603
    },
    {
      "epoch": 0.7604,
      "grad_norm": 3.613154172897339,
      "learning_rate": 3.3009154123382936e-05,
      "loss": 4.6885,
      "step": 7604
    },
    {
      "epoch": 0.7605,
      "grad_norm": 2.6691715717315674,
      "learning_rate": 3.2983242030857174e-05,
      "loss": 4.422,
      "step": 7605
    },
    {
      "epoch": 0.7606,
      "grad_norm": 5.961757659912109,
      "learning_rate": 3.2957338104120096e-05,
      "loss": 5.0246,
      "step": 7606
    },
    {
      "epoch": 0.7607,
      "grad_norm": 2.680222749710083,
      "learning_rate": 3.2931442346328004e-05,
      "loss": 4.5308,
      "step": 7607
    },
    {
      "epoch": 0.7608,
      "grad_norm": 2.758513927459717,
      "learning_rate": 3.290555476063622e-05,
      "loss": 4.4606,
      "step": 7608
    },
    {
      "epoch": 0.7609,
      "grad_norm": 3.7139623165130615,
      "learning_rate": 3.287967535019908e-05,
      "loss": 4.5833,
      "step": 7609
    },
    {
      "epoch": 0.761,
      "grad_norm": 2.4176266193389893,
      "learning_rate": 3.285380411816988e-05,
      "loss": 4.8949,
      "step": 7610
    },
    {
      "epoch": 0.7611,
      "grad_norm": 2.85579514503479,
      "learning_rate": 3.2827941067700996e-05,
      "loss": 4.3283,
      "step": 7611
    },
    {
      "epoch": 0.7612,
      "grad_norm": 4.3982415199279785,
      "learning_rate": 3.2802086201943724e-05,
      "loss": 4.6484,
      "step": 7612
    },
    {
      "epoch": 0.7613,
      "grad_norm": 3.3758938312530518,
      "learning_rate": 3.277623952404842e-05,
      "loss": 5.5494,
      "step": 7613
    },
    {
      "epoch": 0.7614,
      "grad_norm": 2.875221014022827,
      "learning_rate": 3.275040103716441e-05,
      "loss": 4.0367,
      "step": 7614
    },
    {
      "epoch": 0.7615,
      "grad_norm": 2.329847812652588,
      "learning_rate": 3.272457074444003e-05,
      "loss": 4.7666,
      "step": 7615
    },
    {
      "epoch": 0.7616,
      "grad_norm": 3.952314853668213,
      "learning_rate": 3.269874864902269e-05,
      "loss": 4.8491,
      "step": 7616
    },
    {
      "epoch": 0.7617,
      "grad_norm": 2.963167190551758,
      "learning_rate": 3.2672934754058616e-05,
      "loss": 4.8135,
      "step": 7617
    },
    {
      "epoch": 0.7618,
      "grad_norm": 3.333523750305176,
      "learning_rate": 3.264712906269328e-05,
      "loss": 4.2931,
      "step": 7618
    },
    {
      "epoch": 0.7619,
      "grad_norm": 2.5544841289520264,
      "learning_rate": 3.2621331578070934e-05,
      "loss": 4.4663,
      "step": 7619
    },
    {
      "epoch": 0.762,
      "grad_norm": 2.226855993270874,
      "learning_rate": 3.2595542303334924e-05,
      "loss": 4.7371,
      "step": 7620
    },
    {
      "epoch": 0.7621,
      "grad_norm": 3.1512107849121094,
      "learning_rate": 3.2569761241627696e-05,
      "loss": 4.3051,
      "step": 7621
    },
    {
      "epoch": 0.7622,
      "grad_norm": 2.9892208576202393,
      "learning_rate": 3.254398839609044e-05,
      "loss": 4.9733,
      "step": 7622
    },
    {
      "epoch": 0.7623,
      "grad_norm": 2.006334066390991,
      "learning_rate": 3.251822376986363e-05,
      "loss": 4.6216,
      "step": 7623
    },
    {
      "epoch": 0.7624,
      "grad_norm": 2.1495778560638428,
      "learning_rate": 3.249246736608655e-05,
      "loss": 4.8426,
      "step": 7624
    },
    {
      "epoch": 0.7625,
      "grad_norm": 2.7530460357666016,
      "learning_rate": 3.246671918789755e-05,
      "loss": 5.0123,
      "step": 7625
    },
    {
      "epoch": 0.7626,
      "grad_norm": 2.380890369415283,
      "learning_rate": 3.244097923843398e-05,
      "loss": 4.4837,
      "step": 7626
    },
    {
      "epoch": 0.7627,
      "grad_norm": 2.473878860473633,
      "learning_rate": 3.2415247520832146e-05,
      "loss": 4.6825,
      "step": 7627
    },
    {
      "epoch": 0.7628,
      "grad_norm": 2.7611708641052246,
      "learning_rate": 3.23895240382274e-05,
      "loss": 4.7534,
      "step": 7628
    },
    {
      "epoch": 0.7629,
      "grad_norm": 4.842649459838867,
      "learning_rate": 3.236380879375408e-05,
      "loss": 5.5378,
      "step": 7629
    },
    {
      "epoch": 0.763,
      "grad_norm": 1.8492295742034912,
      "learning_rate": 3.233810179054548e-05,
      "loss": 4.3032,
      "step": 7630
    },
    {
      "epoch": 0.7631,
      "grad_norm": 2.665764570236206,
      "learning_rate": 3.231240303173394e-05,
      "loss": 4.6164,
      "step": 7631
    },
    {
      "epoch": 0.7632,
      "grad_norm": 2.150242328643799,
      "learning_rate": 3.2286712520450765e-05,
      "loss": 4.5856,
      "step": 7632
    },
    {
      "epoch": 0.7633,
      "grad_norm": 2.6797373294830322,
      "learning_rate": 3.226103025982628e-05,
      "loss": 4.6602,
      "step": 7633
    },
    {
      "epoch": 0.7634,
      "grad_norm": 2.396627426147461,
      "learning_rate": 3.223535625298979e-05,
      "loss": 4.6961,
      "step": 7634
    },
    {
      "epoch": 0.7635,
      "grad_norm": 2.449373245239258,
      "learning_rate": 3.220969050306955e-05,
      "loss": 4.5802,
      "step": 7635
    },
    {
      "epoch": 0.7636,
      "grad_norm": 2.962839365005493,
      "learning_rate": 3.218403301319296e-05,
      "loss": 4.8903,
      "step": 7636
    },
    {
      "epoch": 0.7637,
      "grad_norm": 4.670380115509033,
      "learning_rate": 3.21583837864862e-05,
      "loss": 4.4983,
      "step": 7637
    },
    {
      "epoch": 0.7638,
      "grad_norm": 2.9884896278381348,
      "learning_rate": 3.213274282607457e-05,
      "loss": 5.2643,
      "step": 7638
    },
    {
      "epoch": 0.7639,
      "grad_norm": 2.4238438606262207,
      "learning_rate": 3.210711013508242e-05,
      "loss": 5.1484,
      "step": 7639
    },
    {
      "epoch": 0.764,
      "grad_norm": 2.400446653366089,
      "learning_rate": 3.208148571663289e-05,
      "loss": 4.6525,
      "step": 7640
    },
    {
      "epoch": 0.7641,
      "grad_norm": 3.9363608360290527,
      "learning_rate": 3.205586957384838e-05,
      "loss": 4.2624,
      "step": 7641
    },
    {
      "epoch": 0.7642,
      "grad_norm": 3.017850399017334,
      "learning_rate": 3.2030261709849996e-05,
      "loss": 4.8441,
      "step": 7642
    },
    {
      "epoch": 0.7643,
      "grad_norm": 3.2562243938446045,
      "learning_rate": 3.200466212775808e-05,
      "loss": 4.8006,
      "step": 7643
    },
    {
      "epoch": 0.7644,
      "grad_norm": 2.883364200592041,
      "learning_rate": 3.197907083069184e-05,
      "loss": 4.468,
      "step": 7644
    },
    {
      "epoch": 0.7645,
      "grad_norm": 2.7024848461151123,
      "learning_rate": 3.195348782176948e-05,
      "loss": 4.633,
      "step": 7645
    },
    {
      "epoch": 0.7646,
      "grad_norm": 8.351130485534668,
      "learning_rate": 3.192791310410822e-05,
      "loss": 4.414,
      "step": 7646
    },
    {
      "epoch": 0.7647,
      "grad_norm": 3.0759549140930176,
      "learning_rate": 3.190234668082427e-05,
      "loss": 4.5364,
      "step": 7647
    },
    {
      "epoch": 0.7648,
      "grad_norm": 2.894117593765259,
      "learning_rate": 3.187678855503282e-05,
      "loss": 4.1209,
      "step": 7648
    },
    {
      "epoch": 0.7649,
      "grad_norm": 3.457918882369995,
      "learning_rate": 3.1851238729848034e-05,
      "loss": 4.642,
      "step": 7649
    },
    {
      "epoch": 0.765,
      "grad_norm": 3.486084461212158,
      "learning_rate": 3.1825697208383096e-05,
      "loss": 4.7789,
      "step": 7650
    },
    {
      "epoch": 0.7651,
      "grad_norm": 2.2972846031188965,
      "learning_rate": 3.1800163993750166e-05,
      "loss": 4.362,
      "step": 7651
    },
    {
      "epoch": 0.7652,
      "grad_norm": 3.413405656814575,
      "learning_rate": 3.1774639089060363e-05,
      "loss": 5.4129,
      "step": 7652
    },
    {
      "epoch": 0.7653,
      "grad_norm": 4.906002044677734,
      "learning_rate": 3.174912249742382e-05,
      "loss": 5.791,
      "step": 7653
    },
    {
      "epoch": 0.7654,
      "grad_norm": 4.617363452911377,
      "learning_rate": 3.172361422194974e-05,
      "loss": 5.0049,
      "step": 7654
    },
    {
      "epoch": 0.7655,
      "grad_norm": 2.082799196243286,
      "learning_rate": 3.1698114265746124e-05,
      "loss": 4.588,
      "step": 7655
    },
    {
      "epoch": 0.7656,
      "grad_norm": 2.2837514877319336,
      "learning_rate": 3.16726226319201e-05,
      "loss": 4.7025,
      "step": 7656
    },
    {
      "epoch": 0.7657,
      "grad_norm": 2.3034114837646484,
      "learning_rate": 3.164713932357776e-05,
      "loss": 4.9086,
      "step": 7657
    },
    {
      "epoch": 0.7658,
      "grad_norm": 2.315129041671753,
      "learning_rate": 3.162166434382412e-05,
      "loss": 5.1291,
      "step": 7658
    },
    {
      "epoch": 0.7659,
      "grad_norm": 4.363732814788818,
      "learning_rate": 3.159619769576333e-05,
      "loss": 5.7341,
      "step": 7659
    },
    {
      "epoch": 0.766,
      "grad_norm": 3.793386697769165,
      "learning_rate": 3.157073938249829e-05,
      "loss": 4.1965,
      "step": 7660
    },
    {
      "epoch": 0.7661,
      "grad_norm": 4.073892116546631,
      "learning_rate": 3.154528940713113e-05,
      "loss": 4.007,
      "step": 7661
    },
    {
      "epoch": 0.7662,
      "grad_norm": 4.935578346252441,
      "learning_rate": 3.15198477727628e-05,
      "loss": 4.5365,
      "step": 7662
    },
    {
      "epoch": 0.7663,
      "grad_norm": 3.1932404041290283,
      "learning_rate": 3.1494414482493304e-05,
      "loss": 4.5656,
      "step": 7663
    },
    {
      "epoch": 0.7664,
      "grad_norm": 4.682004451751709,
      "learning_rate": 3.146898953942163e-05,
      "loss": 4.8781,
      "step": 7664
    },
    {
      "epoch": 0.7665,
      "grad_norm": 2.9931561946868896,
      "learning_rate": 3.144357294664565e-05,
      "loss": 4.8031,
      "step": 7665
    },
    {
      "epoch": 0.7666,
      "grad_norm": 2.763993263244629,
      "learning_rate": 3.141816470726238e-05,
      "loss": 4.7214,
      "step": 7666
    },
    {
      "epoch": 0.7667,
      "grad_norm": 2.213055372238159,
      "learning_rate": 3.1392764824367704e-05,
      "loss": 4.4197,
      "step": 7667
    },
    {
      "epoch": 0.7668,
      "grad_norm": 2.1472017765045166,
      "learning_rate": 3.1367373301056536e-05,
      "loss": 4.6997,
      "step": 7668
    },
    {
      "epoch": 0.7669,
      "grad_norm": 2.573781728744507,
      "learning_rate": 3.134199014042274e-05,
      "loss": 4.6023,
      "step": 7669
    },
    {
      "epoch": 0.767,
      "grad_norm": 2.936748504638672,
      "learning_rate": 3.1316615345559185e-05,
      "loss": 4.4618,
      "step": 7670
    },
    {
      "epoch": 0.7671,
      "grad_norm": 2.3879146575927734,
      "learning_rate": 3.129124891955771e-05,
      "loss": 4.3455,
      "step": 7671
    },
    {
      "epoch": 0.7672,
      "grad_norm": 2.899282932281494,
      "learning_rate": 3.126589086550914e-05,
      "loss": 4.4105,
      "step": 7672
    },
    {
      "epoch": 0.7673,
      "grad_norm": 2.6707849502563477,
      "learning_rate": 3.124054118650327e-05,
      "loss": 4.6155,
      "step": 7673
    },
    {
      "epoch": 0.7674,
      "grad_norm": 3.600923776626587,
      "learning_rate": 3.12151998856289e-05,
      "loss": 4.8617,
      "step": 7674
    },
    {
      "epoch": 0.7675,
      "grad_norm": 1.8965871334075928,
      "learning_rate": 3.1189866965973766e-05,
      "loss": 4.2193,
      "step": 7675
    },
    {
      "epoch": 0.7676,
      "grad_norm": 1.906079888343811,
      "learning_rate": 3.116454243062459e-05,
      "loss": 4.7655,
      "step": 7676
    },
    {
      "epoch": 0.7677,
      "grad_norm": 4.122951507568359,
      "learning_rate": 3.113922628266718e-05,
      "loss": 4.9078,
      "step": 7677
    },
    {
      "epoch": 0.7678,
      "grad_norm": 3.0992681980133057,
      "learning_rate": 3.111391852518611e-05,
      "loss": 4.7,
      "step": 7678
    },
    {
      "epoch": 0.7679,
      "grad_norm": 1.9485007524490356,
      "learning_rate": 3.108861916126518e-05,
      "loss": 4.4315,
      "step": 7679
    },
    {
      "epoch": 0.768,
      "grad_norm": 2.9152560234069824,
      "learning_rate": 3.1063328193986904e-05,
      "loss": 4.7306,
      "step": 7680
    },
    {
      "epoch": 0.7681,
      "grad_norm": 2.206789255142212,
      "learning_rate": 3.103804562643302e-05,
      "loss": 4.8503,
      "step": 7681
    },
    {
      "epoch": 0.7682,
      "grad_norm": 13.994789123535156,
      "learning_rate": 3.101277146168412e-05,
      "loss": 5.1104,
      "step": 7682
    },
    {
      "epoch": 0.7683,
      "grad_norm": 3.0920097827911377,
      "learning_rate": 3.098750570281969e-05,
      "loss": 4.7381,
      "step": 7683
    },
    {
      "epoch": 0.7684,
      "grad_norm": 4.1020379066467285,
      "learning_rate": 3.096224835291839e-05,
      "loss": 4.6636,
      "step": 7684
    },
    {
      "epoch": 0.7685,
      "grad_norm": 3.459425926208496,
      "learning_rate": 3.093699941505771e-05,
      "loss": 4.5186,
      "step": 7685
    },
    {
      "epoch": 0.7686,
      "grad_norm": 2.9112205505371094,
      "learning_rate": 3.0911758892314166e-05,
      "loss": 5.1758,
      "step": 7686
    },
    {
      "epoch": 0.7687,
      "grad_norm": 2.4224534034729004,
      "learning_rate": 3.0886526787763234e-05,
      "loss": 4.7092,
      "step": 7687
    },
    {
      "epoch": 0.7688,
      "grad_norm": 3.337216377258301,
      "learning_rate": 3.086130310447937e-05,
      "loss": 4.8299,
      "step": 7688
    },
    {
      "epoch": 0.7689,
      "grad_norm": 4.303531169891357,
      "learning_rate": 3.0836087845536e-05,
      "loss": 4.9764,
      "step": 7689
    },
    {
      "epoch": 0.769,
      "grad_norm": 2.9196903705596924,
      "learning_rate": 3.081088101400552e-05,
      "loss": 4.6006,
      "step": 7690
    },
    {
      "epoch": 0.7691,
      "grad_norm": 5.851292610168457,
      "learning_rate": 3.078568261295933e-05,
      "loss": 5.4603,
      "step": 7691
    },
    {
      "epoch": 0.7692,
      "grad_norm": 4.745611190795898,
      "learning_rate": 3.0760492645467765e-05,
      "loss": 5.1086,
      "step": 7692
    },
    {
      "epoch": 0.7693,
      "grad_norm": 2.542325496673584,
      "learning_rate": 3.073531111460013e-05,
      "loss": 4.4298,
      "step": 7693
    },
    {
      "epoch": 0.7694,
      "grad_norm": 4.394464492797852,
      "learning_rate": 3.071013802342475e-05,
      "loss": 4.4149,
      "step": 7694
    },
    {
      "epoch": 0.7695,
      "grad_norm": 2.9415745735168457,
      "learning_rate": 3.068497337500886e-05,
      "loss": 4.555,
      "step": 7695
    },
    {
      "epoch": 0.7696,
      "grad_norm": 3.4429593086242676,
      "learning_rate": 3.0659817172418693e-05,
      "loss": 5.1679,
      "step": 7696
    },
    {
      "epoch": 0.7697,
      "grad_norm": 2.130169153213501,
      "learning_rate": 3.063466941871952e-05,
      "loss": 4.7974,
      "step": 7697
    },
    {
      "epoch": 0.7698,
      "grad_norm": 2.6064419746398926,
      "learning_rate": 3.060953011697545e-05,
      "loss": 4.9565,
      "step": 7698
    },
    {
      "epoch": 0.7699,
      "grad_norm": 2.1705596446990967,
      "learning_rate": 3.058439927024962e-05,
      "loss": 4.5759,
      "step": 7699
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.550297975540161,
      "learning_rate": 3.0559276881604236e-05,
      "loss": 4.3366,
      "step": 7700
    },
    {
      "epoch": 0.7701,
      "grad_norm": 4.367553234100342,
      "learning_rate": 3.053416295410026e-05,
      "loss": 4.7648,
      "step": 7701
    },
    {
      "epoch": 0.7702,
      "grad_norm": 4.210424423217773,
      "learning_rate": 3.0509057490797888e-05,
      "loss": 5.257,
      "step": 7702
    },
    {
      "epoch": 0.7703,
      "grad_norm": 2.196640968322754,
      "learning_rate": 3.0483960494756016e-05,
      "loss": 4.8118,
      "step": 7703
    },
    {
      "epoch": 0.7704,
      "grad_norm": 2.7671241760253906,
      "learning_rate": 3.045887196903271e-05,
      "loss": 4.4636,
      "step": 7704
    },
    {
      "epoch": 0.7705,
      "grad_norm": 1.6965277194976807,
      "learning_rate": 3.0433791916684916e-05,
      "loss": 4.7978,
      "step": 7705
    },
    {
      "epoch": 0.7706,
      "grad_norm": 6.885673999786377,
      "learning_rate": 3.0408720340768572e-05,
      "loss": 4.5009,
      "step": 7706
    },
    {
      "epoch": 0.7707,
      "grad_norm": 4.967752933502197,
      "learning_rate": 3.038365724433858e-05,
      "loss": 4.7371,
      "step": 7707
    },
    {
      "epoch": 0.7708,
      "grad_norm": 2.4584827423095703,
      "learning_rate": 3.035860263044873e-05,
      "loss": 5.1302,
      "step": 7708
    },
    {
      "epoch": 0.7709,
      "grad_norm": 2.899493932723999,
      "learning_rate": 3.0333556502151926e-05,
      "loss": 5.307,
      "step": 7709
    },
    {
      "epoch": 0.771,
      "grad_norm": 2.086968183517456,
      "learning_rate": 3.0308518862499957e-05,
      "loss": 4.8651,
      "step": 7710
    },
    {
      "epoch": 0.7711,
      "grad_norm": 2.420409917831421,
      "learning_rate": 3.0283489714543556e-05,
      "loss": 4.3714,
      "step": 7711
    },
    {
      "epoch": 0.7712,
      "grad_norm": 2.3879730701446533,
      "learning_rate": 3.0258469061332463e-05,
      "loss": 5.0454,
      "step": 7712
    },
    {
      "epoch": 0.7713,
      "grad_norm": 2.4759740829467773,
      "learning_rate": 3.023345690591537e-05,
      "loss": 5.0113,
      "step": 7713
    },
    {
      "epoch": 0.7714,
      "grad_norm": 2.9768290519714355,
      "learning_rate": 3.0208453251339885e-05,
      "loss": 4.451,
      "step": 7714
    },
    {
      "epoch": 0.7715,
      "grad_norm": 3.012145519256592,
      "learning_rate": 3.018345810065275e-05,
      "loss": 5.4098,
      "step": 7715
    },
    {
      "epoch": 0.7716,
      "grad_norm": 2.8213024139404297,
      "learning_rate": 3.0158471456899428e-05,
      "loss": 4.6245,
      "step": 7716
    },
    {
      "epoch": 0.7717,
      "grad_norm": 11.101506233215332,
      "learning_rate": 3.0133493323124505e-05,
      "loss": 4.8133,
      "step": 7717
    },
    {
      "epoch": 0.7718,
      "grad_norm": 3.439927816390991,
      "learning_rate": 3.0108523702371505e-05,
      "loss": 4.891,
      "step": 7718
    },
    {
      "epoch": 0.7719,
      "grad_norm": 2.6772994995117188,
      "learning_rate": 3.0083562597682847e-05,
      "loss": 4.1098,
      "step": 7719
    },
    {
      "epoch": 0.772,
      "grad_norm": 4.255144119262695,
      "learning_rate": 3.0058610012100074e-05,
      "loss": 5.3752,
      "step": 7720
    },
    {
      "epoch": 0.7721,
      "grad_norm": 2.145113468170166,
      "learning_rate": 3.0033665948663448e-05,
      "loss": 4.7823,
      "step": 7721
    },
    {
      "epoch": 0.7722,
      "grad_norm": 2.3394405841827393,
      "learning_rate": 3.0008730410412466e-05,
      "loss": 5.0136,
      "step": 7722
    },
    {
      "epoch": 0.7723,
      "grad_norm": 3.5276105403900146,
      "learning_rate": 2.9983803400385312e-05,
      "loss": 4.5328,
      "step": 7723
    },
    {
      "epoch": 0.7724,
      "grad_norm": 2.6077537536621094,
      "learning_rate": 2.9958884921619367e-05,
      "loss": 4.694,
      "step": 7724
    },
    {
      "epoch": 0.7725,
      "grad_norm": 2.2622578144073486,
      "learning_rate": 2.993397497715086e-05,
      "loss": 4.778,
      "step": 7725
    },
    {
      "epoch": 0.7726,
      "grad_norm": 2.3612146377563477,
      "learning_rate": 2.9909073570014912e-05,
      "loss": 4.9091,
      "step": 7726
    },
    {
      "epoch": 0.7727,
      "grad_norm": 2.7243380546569824,
      "learning_rate": 2.9884180703245767e-05,
      "loss": 4.5325,
      "step": 7727
    },
    {
      "epoch": 0.7728,
      "grad_norm": 6.958564281463623,
      "learning_rate": 2.9859296379876523e-05,
      "loss": 5.5141,
      "step": 7728
    },
    {
      "epoch": 0.7729,
      "grad_norm": 5.35446834564209,
      "learning_rate": 2.9834420602939263e-05,
      "loss": 5.0554,
      "step": 7729
    },
    {
      "epoch": 0.773,
      "grad_norm": 2.35310697555542,
      "learning_rate": 2.9809553375465004e-05,
      "loss": 4.5783,
      "step": 7730
    },
    {
      "epoch": 0.7731,
      "grad_norm": 2.3722739219665527,
      "learning_rate": 2.9784694700483762e-05,
      "loss": 4.0075,
      "step": 7731
    },
    {
      "epoch": 0.7732,
      "grad_norm": 7.166333198547363,
      "learning_rate": 2.9759844581024486e-05,
      "loss": 5.0258,
      "step": 7732
    },
    {
      "epoch": 0.7733,
      "grad_norm": 3.8832433223724365,
      "learning_rate": 2.9735003020115092e-05,
      "loss": 5.6254,
      "step": 7733
    },
    {
      "epoch": 0.7734,
      "grad_norm": 2.023676633834839,
      "learning_rate": 2.9710170020782435e-05,
      "loss": 4.8427,
      "step": 7734
    },
    {
      "epoch": 0.7735,
      "grad_norm": 2.3003339767456055,
      "learning_rate": 2.968534558605236e-05,
      "loss": 4.7268,
      "step": 7735
    },
    {
      "epoch": 0.7736,
      "grad_norm": 3.3973000049591064,
      "learning_rate": 2.9660529718949627e-05,
      "loss": 5.3417,
      "step": 7736
    },
    {
      "epoch": 0.7737,
      "grad_norm": 2.5941641330718994,
      "learning_rate": 2.9635722422497993e-05,
      "loss": 4.8596,
      "step": 7737
    },
    {
      "epoch": 0.7738,
      "grad_norm": 6.797102928161621,
      "learning_rate": 2.961092369972014e-05,
      "loss": 5.3085,
      "step": 7738
    },
    {
      "epoch": 0.7739,
      "grad_norm": 3.409196138381958,
      "learning_rate": 2.9586133553637683e-05,
      "loss": 5.1365,
      "step": 7739
    },
    {
      "epoch": 0.774,
      "grad_norm": 3.456545352935791,
      "learning_rate": 2.9561351987271334e-05,
      "loss": 5.3523,
      "step": 7740
    },
    {
      "epoch": 0.7741,
      "grad_norm": 3.0127756595611572,
      "learning_rate": 2.953657900364053e-05,
      "loss": 4.7385,
      "step": 7741
    },
    {
      "epoch": 0.7742,
      "grad_norm": 3.16072154045105,
      "learning_rate": 2.9511814605763855e-05,
      "loss": 5.2943,
      "step": 7742
    },
    {
      "epoch": 0.7743,
      "grad_norm": 2.8057444095611572,
      "learning_rate": 2.9487058796658783e-05,
      "loss": 5.021,
      "step": 7743
    },
    {
      "epoch": 0.7744,
      "grad_norm": 2.5454442501068115,
      "learning_rate": 2.9462311579341663e-05,
      "loss": 4.54,
      "step": 7744
    },
    {
      "epoch": 0.7745,
      "grad_norm": 2.241765260696411,
      "learning_rate": 2.9437572956827964e-05,
      "loss": 4.7657,
      "step": 7745
    },
    {
      "epoch": 0.7746,
      "grad_norm": 1.9138015508651733,
      "learning_rate": 2.94128429321319e-05,
      "loss": 4.932,
      "step": 7746
    },
    {
      "epoch": 0.7747,
      "grad_norm": 3.0102908611297607,
      "learning_rate": 2.938812150826684e-05,
      "loss": 4.8333,
      "step": 7747
    },
    {
      "epoch": 0.7748,
      "grad_norm": 2.2673895359039307,
      "learning_rate": 2.9363408688245e-05,
      "loss": 4.6673,
      "step": 7748
    },
    {
      "epoch": 0.7749,
      "grad_norm": 2.17747163772583,
      "learning_rate": 2.933870447507753e-05,
      "loss": 4.4351,
      "step": 7749
    },
    {
      "epoch": 0.775,
      "grad_norm": 2.0996692180633545,
      "learning_rate": 2.931400887177459e-05,
      "loss": 4.4291,
      "step": 7750
    },
    {
      "epoch": 0.7751,
      "grad_norm": 3.5090930461883545,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 4.7445,
      "step": 7751
    },
    {
      "epoch": 0.7752,
      "grad_norm": 2.2578277587890625,
      "learning_rate": 2.926464350679756e-05,
      "loss": 4.6174,
      "step": 7752
    },
    {
      "epoch": 0.7753,
      "grad_norm": 3.0936949253082275,
      "learning_rate": 2.9239973751138495e-05,
      "loss": 4.1972,
      "step": 7753
    },
    {
      "epoch": 0.7754,
      "grad_norm": 3.004002332687378,
      "learning_rate": 2.921531261737398e-05,
      "loss": 4.707,
      "step": 7754
    },
    {
      "epoch": 0.7755,
      "grad_norm": 3.214609384536743,
      "learning_rate": 2.9190660108508917e-05,
      "loss": 4.1599,
      "step": 7755
    },
    {
      "epoch": 0.7756,
      "grad_norm": 2.2577409744262695,
      "learning_rate": 2.9166016227547133e-05,
      "loss": 4.6849,
      "step": 7756
    },
    {
      "epoch": 0.7757,
      "grad_norm": 3.199007749557495,
      "learning_rate": 2.9141380977491373e-05,
      "loss": 5.0798,
      "step": 7757
    },
    {
      "epoch": 0.7758,
      "grad_norm": 3.3973610401153564,
      "learning_rate": 2.911675436134347e-05,
      "loss": 5.5956,
      "step": 7758
    },
    {
      "epoch": 0.7759,
      "grad_norm": 3.084195375442505,
      "learning_rate": 2.9092136382103973e-05,
      "loss": 4.6549,
      "step": 7759
    },
    {
      "epoch": 0.776,
      "grad_norm": 3.1787216663360596,
      "learning_rate": 2.9067527042772636e-05,
      "loss": 5.0627,
      "step": 7760
    },
    {
      "epoch": 0.7761,
      "grad_norm": 2.3907439708709717,
      "learning_rate": 2.904292634634793e-05,
      "loss": 4.8975,
      "step": 7761
    },
    {
      "epoch": 0.7762,
      "grad_norm": 2.506600856781006,
      "learning_rate": 2.9018334295827388e-05,
      "loss": 4.2078,
      "step": 7762
    },
    {
      "epoch": 0.7763,
      "grad_norm": 3.1367764472961426,
      "learning_rate": 2.899375089420756e-05,
      "loss": 4.5723,
      "step": 7763
    },
    {
      "epoch": 0.7764,
      "grad_norm": 2.7409822940826416,
      "learning_rate": 2.8969176144483744e-05,
      "loss": 4.841,
      "step": 7764
    },
    {
      "epoch": 0.7765,
      "grad_norm": 2.807870626449585,
      "learning_rate": 2.894461004965038e-05,
      "loss": 4.629,
      "step": 7765
    },
    {
      "epoch": 0.7766,
      "grad_norm": 2.6885249614715576,
      "learning_rate": 2.8920052612700754e-05,
      "loss": 5.2352,
      "step": 7766
    },
    {
      "epoch": 0.7767,
      "grad_norm": 10.92166805267334,
      "learning_rate": 2.8895503836627103e-05,
      "loss": 4.7854,
      "step": 7767
    },
    {
      "epoch": 0.7768,
      "grad_norm": 1.9612269401550293,
      "learning_rate": 2.887096372442063e-05,
      "loss": 4.3068,
      "step": 7768
    },
    {
      "epoch": 0.7769,
      "grad_norm": 2.246487617492676,
      "learning_rate": 2.8846432279071467e-05,
      "loss": 4.7655,
      "step": 7769
    },
    {
      "epoch": 0.777,
      "grad_norm": 3.3648903369903564,
      "learning_rate": 2.88219095035687e-05,
      "loss": 4.6898,
      "step": 7770
    },
    {
      "epoch": 0.7771,
      "grad_norm": 4.172203540802002,
      "learning_rate": 2.879739540090036e-05,
      "loss": 4.8784,
      "step": 7771
    },
    {
      "epoch": 0.7772,
      "grad_norm": 3.9742562770843506,
      "learning_rate": 2.877288997405341e-05,
      "loss": 5.1551,
      "step": 7772
    },
    {
      "epoch": 0.7773,
      "grad_norm": 2.368929386138916,
      "learning_rate": 2.874839322601375e-05,
      "loss": 4.3537,
      "step": 7773
    },
    {
      "epoch": 0.7774,
      "grad_norm": 2.9751925468444824,
      "learning_rate": 2.872390515976625e-05,
      "loss": 4.6272,
      "step": 7774
    },
    {
      "epoch": 0.7775,
      "grad_norm": 1.7615219354629517,
      "learning_rate": 2.869942577829471e-05,
      "loss": 4.5557,
      "step": 7775
    },
    {
      "epoch": 0.7776,
      "grad_norm": 3.076413154602051,
      "learning_rate": 2.8674955084581857e-05,
      "loss": 5.0178,
      "step": 7776
    },
    {
      "epoch": 0.7777,
      "grad_norm": 4.318208694458008,
      "learning_rate": 2.865049308160934e-05,
      "loss": 5.1506,
      "step": 7777
    },
    {
      "epoch": 0.7778,
      "grad_norm": 3.972670555114746,
      "learning_rate": 2.8626039772357882e-05,
      "loss": 4.2815,
      "step": 7778
    },
    {
      "epoch": 0.7779,
      "grad_norm": 3.1744439601898193,
      "learning_rate": 2.860159515980695e-05,
      "loss": 4.7787,
      "step": 7779
    },
    {
      "epoch": 0.778,
      "grad_norm": 3.5712852478027344,
      "learning_rate": 2.8577159246935037e-05,
      "loss": 4.7529,
      "step": 7780
    },
    {
      "epoch": 0.7781,
      "grad_norm": 2.576824426651001,
      "learning_rate": 2.8552732036719687e-05,
      "loss": 4.6986,
      "step": 7781
    },
    {
      "epoch": 0.7782,
      "grad_norm": 1.8332459926605225,
      "learning_rate": 2.852831353213715e-05,
      "loss": 4.1503,
      "step": 7782
    },
    {
      "epoch": 0.7783,
      "grad_norm": 2.738271951675415,
      "learning_rate": 2.8503903736162875e-05,
      "loss": 5.1151,
      "step": 7783
    },
    {
      "epoch": 0.7784,
      "grad_norm": 5.868711471557617,
      "learning_rate": 2.8479502651770995e-05,
      "loss": 4.2995,
      "step": 7784
    },
    {
      "epoch": 0.7785,
      "grad_norm": 2.5979199409484863,
      "learning_rate": 2.8455110281934803e-05,
      "loss": 4.582,
      "step": 7785
    },
    {
      "epoch": 0.7786,
      "grad_norm": 2.548337459564209,
      "learning_rate": 2.8430726629626413e-05,
      "loss": 4.6399,
      "step": 7786
    },
    {
      "epoch": 0.7787,
      "grad_norm": 2.4103500843048096,
      "learning_rate": 2.840635169781688e-05,
      "loss": 4.76,
      "step": 7787
    },
    {
      "epoch": 0.7788,
      "grad_norm": 2.2593984603881836,
      "learning_rate": 2.838198548947627e-05,
      "loss": 4.2895,
      "step": 7788
    },
    {
      "epoch": 0.7789,
      "grad_norm": 2.4296557903289795,
      "learning_rate": 2.835762800757341e-05,
      "loss": 4.3806,
      "step": 7789
    },
    {
      "epoch": 0.779,
      "grad_norm": 2.639138698577881,
      "learning_rate": 2.8333279255076306e-05,
      "loss": 4.6798,
      "step": 7790
    },
    {
      "epoch": 0.7791,
      "grad_norm": 2.2101244926452637,
      "learning_rate": 2.8308939234951726e-05,
      "loss": 4.7442,
      "step": 7791
    },
    {
      "epoch": 0.7792,
      "grad_norm": 7.030245304107666,
      "learning_rate": 2.8284607950165442e-05,
      "loss": 5.4168,
      "step": 7792
    },
    {
      "epoch": 0.7793,
      "grad_norm": 5.49186372756958,
      "learning_rate": 2.826028540368215e-05,
      "loss": 5.2986,
      "step": 7793
    },
    {
      "epoch": 0.7794,
      "grad_norm": 1.9628527164459229,
      "learning_rate": 2.823597159846547e-05,
      "loss": 4.7244,
      "step": 7794
    },
    {
      "epoch": 0.7795,
      "grad_norm": 2.2986598014831543,
      "learning_rate": 2.8211666537477933e-05,
      "loss": 4.4939,
      "step": 7795
    },
    {
      "epoch": 0.7796,
      "grad_norm": 3.761861801147461,
      "learning_rate": 2.8187370223681132e-05,
      "loss": 5.0353,
      "step": 7796
    },
    {
      "epoch": 0.7797,
      "grad_norm": 6.489050388336182,
      "learning_rate": 2.816308266003541e-05,
      "loss": 4.379,
      "step": 7797
    },
    {
      "epoch": 0.7798,
      "grad_norm": 2.3663642406463623,
      "learning_rate": 2.813880384950016e-05,
      "loss": 5.0419,
      "step": 7798
    },
    {
      "epoch": 0.7799,
      "grad_norm": 3.1018357276916504,
      "learning_rate": 2.8114533795033683e-05,
      "loss": 4.1673,
      "step": 7799
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.250300884246826,
      "learning_rate": 2.8090272499593173e-05,
      "loss": 4.9638,
      "step": 7800
    },
    {
      "epoch": 0.7801,
      "grad_norm": 3.1466000080108643,
      "learning_rate": 2.8066019966134904e-05,
      "loss": 4.699,
      "step": 7801
    },
    {
      "epoch": 0.7802,
      "grad_norm": 2.9622180461883545,
      "learning_rate": 2.8041776197613844e-05,
      "loss": 4.6591,
      "step": 7802
    },
    {
      "epoch": 0.7803,
      "grad_norm": 2.4130470752716064,
      "learning_rate": 2.8017541196984142e-05,
      "loss": 4.157,
      "step": 7803
    },
    {
      "epoch": 0.7804,
      "grad_norm": 7.025087833404541,
      "learning_rate": 2.7993314967198635e-05,
      "loss": 4.0264,
      "step": 7804
    },
    {
      "epoch": 0.7805,
      "grad_norm": 5.318014144897461,
      "learning_rate": 2.7969097511209308e-05,
      "loss": 4.9933,
      "step": 7805
    },
    {
      "epoch": 0.7806,
      "grad_norm": 2.716134786605835,
      "learning_rate": 2.7944888831966987e-05,
      "loss": 4.4599,
      "step": 7806
    },
    {
      "epoch": 0.7807,
      "grad_norm": 1.8952680826187134,
      "learning_rate": 2.7920688932421335e-05,
      "loss": 4.6998,
      "step": 7807
    },
    {
      "epoch": 0.7808,
      "grad_norm": 2.642230272293091,
      "learning_rate": 2.7896497815521128e-05,
      "loss": 4.6663,
      "step": 7808
    },
    {
      "epoch": 0.7809,
      "grad_norm": 2.7880163192749023,
      "learning_rate": 2.7872315484213952e-05,
      "loss": 4.6582,
      "step": 7809
    },
    {
      "epoch": 0.781,
      "grad_norm": 2.090909004211426,
      "learning_rate": 2.7848141941446347e-05,
      "loss": 4.9677,
      "step": 7810
    },
    {
      "epoch": 0.7811,
      "grad_norm": 2.041571855545044,
      "learning_rate": 2.7823977190163786e-05,
      "loss": 4.5462,
      "step": 7811
    },
    {
      "epoch": 0.7812,
      "grad_norm": 3.3168728351593018,
      "learning_rate": 2.7799821233310674e-05,
      "loss": 4.673,
      "step": 7812
    },
    {
      "epoch": 0.7813,
      "grad_norm": 2.3642239570617676,
      "learning_rate": 2.7775674073830337e-05,
      "loss": 4.6065,
      "step": 7813
    },
    {
      "epoch": 0.7814,
      "grad_norm": 2.576277494430542,
      "learning_rate": 2.775153571466502e-05,
      "loss": 4.7177,
      "step": 7814
    },
    {
      "epoch": 0.7815,
      "grad_norm": 3.5087082386016846,
      "learning_rate": 2.772740615875594e-05,
      "loss": 5.8303,
      "step": 7815
    },
    {
      "epoch": 0.7816,
      "grad_norm": 3.2718276977539062,
      "learning_rate": 2.770328540904319e-05,
      "loss": 5.054,
      "step": 7816
    },
    {
      "epoch": 0.7817,
      "grad_norm": 1.686061143875122,
      "learning_rate": 2.7679173468465812e-05,
      "loss": 4.5393,
      "step": 7817
    },
    {
      "epoch": 0.7818,
      "grad_norm": 4.578314781188965,
      "learning_rate": 2.7655070339961776e-05,
      "loss": 4.9894,
      "step": 7818
    },
    {
      "epoch": 0.7819,
      "grad_norm": 5.084556579589844,
      "learning_rate": 2.7630976026467968e-05,
      "loss": 5.3273,
      "step": 7819
    },
    {
      "epoch": 0.782,
      "grad_norm": 2.054654121398926,
      "learning_rate": 2.7606890530920195e-05,
      "loss": 4.3659,
      "step": 7820
    },
    {
      "epoch": 0.7821,
      "grad_norm": 2.38702654838562,
      "learning_rate": 2.7582813856253275e-05,
      "loss": 4.2663,
      "step": 7821
    },
    {
      "epoch": 0.7822,
      "grad_norm": 2.4502789974212646,
      "learning_rate": 2.755874600540078e-05,
      "loss": 4.8683,
      "step": 7822
    },
    {
      "epoch": 0.7823,
      "grad_norm": 2.383751153945923,
      "learning_rate": 2.753468698129533e-05,
      "loss": 4.6822,
      "step": 7823
    },
    {
      "epoch": 0.7824,
      "grad_norm": 3.2372241020202637,
      "learning_rate": 2.7510636786868514e-05,
      "loss": 4.5075,
      "step": 7824
    },
    {
      "epoch": 0.7825,
      "grad_norm": 2.6895554065704346,
      "learning_rate": 2.7486595425050665e-05,
      "loss": 4.8483,
      "step": 7825
    },
    {
      "epoch": 0.7826,
      "grad_norm": 2.571627378463745,
      "learning_rate": 2.746256289877126e-05,
      "loss": 4.3844,
      "step": 7826
    },
    {
      "epoch": 0.7827,
      "grad_norm": 2.1909804344177246,
      "learning_rate": 2.743853921095848e-05,
      "loss": 4.739,
      "step": 7827
    },
    {
      "epoch": 0.7828,
      "grad_norm": 2.046905994415283,
      "learning_rate": 2.741452436453963e-05,
      "loss": 4.8558,
      "step": 7828
    },
    {
      "epoch": 0.7829,
      "grad_norm": 3.470899820327759,
      "learning_rate": 2.7390518362440808e-05,
      "loss": 4.8352,
      "step": 7829
    },
    {
      "epoch": 0.783,
      "grad_norm": 5.3773956298828125,
      "learning_rate": 2.736652120758708e-05,
      "loss": 5.4672,
      "step": 7830
    },
    {
      "epoch": 0.7831,
      "grad_norm": 2.7895798683166504,
      "learning_rate": 2.734253290290242e-05,
      "loss": 4.2425,
      "step": 7831
    },
    {
      "epoch": 0.7832,
      "grad_norm": 3.2366886138916016,
      "learning_rate": 2.7318553451309726e-05,
      "loss": 5.5564,
      "step": 7832
    },
    {
      "epoch": 0.7833,
      "grad_norm": 2.2544209957122803,
      "learning_rate": 2.7294582855730832e-05,
      "loss": 4.3428,
      "step": 7833
    },
    {
      "epoch": 0.7834,
      "grad_norm": 2.3950819969177246,
      "learning_rate": 2.727062111908647e-05,
      "loss": 4.4036,
      "step": 7834
    },
    {
      "epoch": 0.7835,
      "grad_norm": 2.871689796447754,
      "learning_rate": 2.7246668244296323e-05,
      "loss": 4.6121,
      "step": 7835
    },
    {
      "epoch": 0.7836,
      "grad_norm": 2.3664023876190186,
      "learning_rate": 2.722272423427896e-05,
      "loss": 4.5657,
      "step": 7836
    },
    {
      "epoch": 0.7837,
      "grad_norm": 2.9521725177764893,
      "learning_rate": 2.7198789091951902e-05,
      "loss": 4.8634,
      "step": 7837
    },
    {
      "epoch": 0.7838,
      "grad_norm": 4.4387054443359375,
      "learning_rate": 2.717486282023153e-05,
      "loss": 5.466,
      "step": 7838
    },
    {
      "epoch": 0.7839,
      "grad_norm": 2.4955079555511475,
      "learning_rate": 2.715094542203327e-05,
      "loss": 4.7988,
      "step": 7839
    },
    {
      "epoch": 0.784,
      "grad_norm": 2.6838624477386475,
      "learning_rate": 2.7127036900271317e-05,
      "loss": 4.4889,
      "step": 7840
    },
    {
      "epoch": 0.7841,
      "grad_norm": 2.874039888381958,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 4.363,
      "step": 7841
    },
    {
      "epoch": 0.7842,
      "grad_norm": 2.4990992546081543,
      "learning_rate": 2.707924649770802e-05,
      "loss": 4.6849,
      "step": 7842
    },
    {
      "epoch": 0.7843,
      "grad_norm": 4.253323554992676,
      "learning_rate": 2.7055364622729773e-05,
      "loss": 4.4516,
      "step": 7843
    },
    {
      "epoch": 0.7844,
      "grad_norm": 3.0367581844329834,
      "learning_rate": 2.7031491635834137e-05,
      "loss": 4.8768,
      "step": 7844
    },
    {
      "epoch": 0.7845,
      "grad_norm": 2.1031203269958496,
      "learning_rate": 2.700762753992985e-05,
      "loss": 4.8238,
      "step": 7845
    },
    {
      "epoch": 0.7846,
      "grad_norm": 5.2342376708984375,
      "learning_rate": 2.698377233792476e-05,
      "loss": 5.9975,
      "step": 7846
    },
    {
      "epoch": 0.7847,
      "grad_norm": 4.897358417510986,
      "learning_rate": 2.6959926032725535e-05,
      "loss": 5.1151,
      "step": 7847
    },
    {
      "epoch": 0.7848,
      "grad_norm": 2.98384690284729,
      "learning_rate": 2.6936088627237765e-05,
      "loss": 4.9579,
      "step": 7848
    },
    {
      "epoch": 0.7849,
      "grad_norm": 2.6612393856048584,
      "learning_rate": 2.6912260124366006e-05,
      "loss": 5.1394,
      "step": 7849
    },
    {
      "epoch": 0.785,
      "grad_norm": 2.6485843658447266,
      "learning_rate": 2.688844052701359e-05,
      "loss": 4.346,
      "step": 7850
    },
    {
      "epoch": 0.7851,
      "grad_norm": 2.57389760017395,
      "learning_rate": 2.6864629838082956e-05,
      "loss": 4.7853,
      "step": 7851
    },
    {
      "epoch": 0.7852,
      "grad_norm": 2.016345500946045,
      "learning_rate": 2.6840828060475332e-05,
      "loss": 4.4892,
      "step": 7852
    },
    {
      "epoch": 0.7853,
      "grad_norm": 2.5559985637664795,
      "learning_rate": 2.681703519709089e-05,
      "loss": 4.7263,
      "step": 7853
    },
    {
      "epoch": 0.7854,
      "grad_norm": 2.7908754348754883,
      "learning_rate": 2.679325125082872e-05,
      "loss": 4.9275,
      "step": 7854
    },
    {
      "epoch": 0.7855,
      "grad_norm": 4.534905910491943,
      "learning_rate": 2.676947622458683e-05,
      "loss": 4.3775,
      "step": 7855
    },
    {
      "epoch": 0.7856,
      "grad_norm": 3.7399086952209473,
      "learning_rate": 2.6745710121262136e-05,
      "loss": 5.0964,
      "step": 7856
    },
    {
      "epoch": 0.7857,
      "grad_norm": 2.0193212032318115,
      "learning_rate": 2.672195294375045e-05,
      "loss": 4.5145,
      "step": 7857
    },
    {
      "epoch": 0.7858,
      "grad_norm": 3.928802728652954,
      "learning_rate": 2.6698204694946527e-05,
      "loss": 5.2826,
      "step": 7858
    },
    {
      "epoch": 0.7859,
      "grad_norm": 3.289609670639038,
      "learning_rate": 2.6674465377744017e-05,
      "loss": 4.3919,
      "step": 7859
    },
    {
      "epoch": 0.786,
      "grad_norm": 2.7344157695770264,
      "learning_rate": 2.6650734995035477e-05,
      "loss": 4.7818,
      "step": 7860
    },
    {
      "epoch": 0.7861,
      "grad_norm": 2.3610661029815674,
      "learning_rate": 2.6627013549712355e-05,
      "loss": 4.7345,
      "step": 7861
    },
    {
      "epoch": 0.7862,
      "grad_norm": 3.364828586578369,
      "learning_rate": 2.660330104466513e-05,
      "loss": 4.9596,
      "step": 7862
    },
    {
      "epoch": 0.7863,
      "grad_norm": 3.084545612335205,
      "learning_rate": 2.657959748278297e-05,
      "loss": 4.4056,
      "step": 7863
    },
    {
      "epoch": 0.7864,
      "grad_norm": 2.6140034198760986,
      "learning_rate": 2.655590286695422e-05,
      "loss": 4.6118,
      "step": 7864
    },
    {
      "epoch": 0.7865,
      "grad_norm": 3.684454917907715,
      "learning_rate": 2.6532217200065858e-05,
      "loss": 4.7303,
      "step": 7865
    },
    {
      "epoch": 0.7866,
      "grad_norm": 2.2032341957092285,
      "learning_rate": 2.6508540485004006e-05,
      "loss": 4.7352,
      "step": 7866
    },
    {
      "epoch": 0.7867,
      "grad_norm": 2.0554733276367188,
      "learning_rate": 2.6484872724653608e-05,
      "loss": 4.5783,
      "step": 7867
    },
    {
      "epoch": 0.7868,
      "grad_norm": 2.7479329109191895,
      "learning_rate": 2.646121392189841e-05,
      "loss": 4.424,
      "step": 7868
    },
    {
      "epoch": 0.7869,
      "grad_norm": 3.0418763160705566,
      "learning_rate": 2.6437564079621267e-05,
      "loss": 4.4521,
      "step": 7869
    },
    {
      "epoch": 0.787,
      "grad_norm": 2.818743944168091,
      "learning_rate": 2.6413923200703794e-05,
      "loss": 4.4146,
      "step": 7870
    },
    {
      "epoch": 0.7871,
      "grad_norm": 3.619446039199829,
      "learning_rate": 2.639029128802657e-05,
      "loss": 5.0914,
      "step": 7871
    },
    {
      "epoch": 0.7872,
      "grad_norm": 6.173349380493164,
      "learning_rate": 2.636666834446907e-05,
      "loss": 4.8742,
      "step": 7872
    },
    {
      "epoch": 0.7873,
      "grad_norm": 2.587836742401123,
      "learning_rate": 2.634305437290968e-05,
      "loss": 4.8661,
      "step": 7873
    },
    {
      "epoch": 0.7874,
      "grad_norm": 3.446193218231201,
      "learning_rate": 2.631944937622569e-05,
      "loss": 4.9299,
      "step": 7874
    },
    {
      "epoch": 0.7875,
      "grad_norm": 2.446593761444092,
      "learning_rate": 2.6295853357293298e-05,
      "loss": 4.612,
      "step": 7875
    },
    {
      "epoch": 0.7876,
      "grad_norm": 2.5596299171447754,
      "learning_rate": 2.6272266318987603e-05,
      "loss": 4.9944,
      "step": 7876
    },
    {
      "epoch": 0.7877,
      "grad_norm": 3.5615992546081543,
      "learning_rate": 2.624868826418262e-05,
      "loss": 4.4367,
      "step": 7877
    },
    {
      "epoch": 0.7878,
      "grad_norm": 4.5302324295043945,
      "learning_rate": 2.6225119195751258e-05,
      "loss": 5.1529,
      "step": 7878
    },
    {
      "epoch": 0.7879,
      "grad_norm": 1.7242714166641235,
      "learning_rate": 2.6201559116565345e-05,
      "loss": 4.478,
      "step": 7879
    },
    {
      "epoch": 0.788,
      "grad_norm": 3.522368907928467,
      "learning_rate": 2.6178008029495592e-05,
      "loss": 4.2938,
      "step": 7880
    },
    {
      "epoch": 0.7881,
      "grad_norm": 3.293041944503784,
      "learning_rate": 2.615446593741161e-05,
      "loss": 4.8705,
      "step": 7881
    },
    {
      "epoch": 0.7882,
      "grad_norm": 2.171454668045044,
      "learning_rate": 2.613093284318201e-05,
      "loss": 4.5804,
      "step": 7882
    },
    {
      "epoch": 0.7883,
      "grad_norm": 2.246358633041382,
      "learning_rate": 2.6107408749674122e-05,
      "loss": 4.5534,
      "step": 7883
    },
    {
      "epoch": 0.7884,
      "grad_norm": 2.6767053604125977,
      "learning_rate": 2.6083893659754356e-05,
      "loss": 4.5104,
      "step": 7884
    },
    {
      "epoch": 0.7885,
      "grad_norm": 5.936943054199219,
      "learning_rate": 2.606038757628798e-05,
      "loss": 5.4967,
      "step": 7885
    },
    {
      "epoch": 0.7886,
      "grad_norm": 5.16802978515625,
      "learning_rate": 2.603689050213902e-05,
      "loss": 4.9286,
      "step": 7886
    },
    {
      "epoch": 0.7887,
      "grad_norm": 2.2711048126220703,
      "learning_rate": 2.6013402440170676e-05,
      "loss": 4.7858,
      "step": 7887
    },
    {
      "epoch": 0.7888,
      "grad_norm": 3.000715494155884,
      "learning_rate": 2.5989923393244742e-05,
      "loss": 4.9811,
      "step": 7888
    },
    {
      "epoch": 0.7889,
      "grad_norm": 3.02803111076355,
      "learning_rate": 2.5966453364222186e-05,
      "loss": 4.9176,
      "step": 7889
    },
    {
      "epoch": 0.789,
      "grad_norm": 3.4632208347320557,
      "learning_rate": 2.5942992355962727e-05,
      "loss": 4.4998,
      "step": 7890
    },
    {
      "epoch": 0.7891,
      "grad_norm": 2.459174633026123,
      "learning_rate": 2.5919540371325e-05,
      "loss": 4.9415,
      "step": 7891
    },
    {
      "epoch": 0.7892,
      "grad_norm": 12.29137897491455,
      "learning_rate": 2.5896097413166564e-05,
      "loss": 4.7445,
      "step": 7892
    },
    {
      "epoch": 0.7893,
      "grad_norm": 5.028990268707275,
      "learning_rate": 2.5872663484343884e-05,
      "loss": 4.1849,
      "step": 7893
    },
    {
      "epoch": 0.7894,
      "grad_norm": 2.006373167037964,
      "learning_rate": 2.584923858771231e-05,
      "loss": 4.7117,
      "step": 7894
    },
    {
      "epoch": 0.7895,
      "grad_norm": 2.584047555923462,
      "learning_rate": 2.582582272612609e-05,
      "loss": 4.6985,
      "step": 7895
    },
    {
      "epoch": 0.7896,
      "grad_norm": 2.056345224380493,
      "learning_rate": 2.580241590243837e-05,
      "loss": 4.1804,
      "step": 7896
    },
    {
      "epoch": 0.7897,
      "grad_norm": 2.1620020866394043,
      "learning_rate": 2.5779018119501208e-05,
      "loss": 4.8643,
      "step": 7897
    },
    {
      "epoch": 0.7898,
      "grad_norm": 4.352365016937256,
      "learning_rate": 2.575562938016556e-05,
      "loss": 5.5656,
      "step": 7898
    },
    {
      "epoch": 0.7899,
      "grad_norm": 2.3620877265930176,
      "learning_rate": 2.573224968728123e-05,
      "loss": 4.7532,
      "step": 7899
    },
    {
      "epoch": 0.79,
      "grad_norm": 3.976269245147705,
      "learning_rate": 2.5708879043697054e-05,
      "loss": 4.4623,
      "step": 7900
    },
    {
      "epoch": 0.7901,
      "grad_norm": 3.388272285461426,
      "learning_rate": 2.5685517452260567e-05,
      "loss": 4.4018,
      "step": 7901
    },
    {
      "epoch": 0.7902,
      "grad_norm": 2.5948352813720703,
      "learning_rate": 2.566216491581841e-05,
      "loss": 4.4297,
      "step": 7902
    },
    {
      "epoch": 0.7903,
      "grad_norm": 2.332882881164551,
      "learning_rate": 2.5638821437215944e-05,
      "loss": 4.6675,
      "step": 7903
    },
    {
      "epoch": 0.7904,
      "grad_norm": 5.515114784240723,
      "learning_rate": 2.561548701929749e-05,
      "loss": 4.6103,
      "step": 7904
    },
    {
      "epoch": 0.7905,
      "grad_norm": 3.579763412475586,
      "learning_rate": 2.5592161664906368e-05,
      "loss": 5.3806,
      "step": 7905
    },
    {
      "epoch": 0.7906,
      "grad_norm": 2.5458741188049316,
      "learning_rate": 2.5568845376884587e-05,
      "loss": 5.0774,
      "step": 7906
    },
    {
      "epoch": 0.7907,
      "grad_norm": 2.8660550117492676,
      "learning_rate": 2.554553815807328e-05,
      "loss": 4.5459,
      "step": 7907
    },
    {
      "epoch": 0.7908,
      "grad_norm": 4.0276970863342285,
      "learning_rate": 2.5522240011312247e-05,
      "loss": 5.1083,
      "step": 7908
    },
    {
      "epoch": 0.7909,
      "grad_norm": 2.1624913215637207,
      "learning_rate": 2.549895093944039e-05,
      "loss": 4.7919,
      "step": 7909
    },
    {
      "epoch": 0.791,
      "grad_norm": 1.8098511695861816,
      "learning_rate": 2.547567094529537e-05,
      "loss": 4.2669,
      "step": 7910
    },
    {
      "epoch": 0.7911,
      "grad_norm": 3.442654848098755,
      "learning_rate": 2.5452400031713785e-05,
      "loss": 4.9981,
      "step": 7911
    },
    {
      "epoch": 0.7912,
      "grad_norm": 1.8600319623947144,
      "learning_rate": 2.542913820153113e-05,
      "loss": 4.439,
      "step": 7912
    },
    {
      "epoch": 0.7913,
      "grad_norm": 1.9983973503112793,
      "learning_rate": 2.540588545758179e-05,
      "loss": 4.3704,
      "step": 7913
    },
    {
      "epoch": 0.7914,
      "grad_norm": 4.158737659454346,
      "learning_rate": 2.5382641802699035e-05,
      "loss": 5.3477,
      "step": 7914
    },
    {
      "epoch": 0.7915,
      "grad_norm": 2.2491424083709717,
      "learning_rate": 2.535940723971505e-05,
      "loss": 5.0143,
      "step": 7915
    },
    {
      "epoch": 0.7916,
      "grad_norm": 1.8892459869384766,
      "learning_rate": 2.5336181771460876e-05,
      "loss": 4.4491,
      "step": 7916
    },
    {
      "epoch": 0.7917,
      "grad_norm": 3.2542197704315186,
      "learning_rate": 2.5312965400766474e-05,
      "loss": 5.7716,
      "step": 7917
    },
    {
      "epoch": 0.7918,
      "grad_norm": 2.938157558441162,
      "learning_rate": 2.5289758130460683e-05,
      "loss": 5.3572,
      "step": 7918
    },
    {
      "epoch": 0.7919,
      "grad_norm": 3.198354959487915,
      "learning_rate": 2.5266559963371216e-05,
      "loss": 5.348,
      "step": 7919
    },
    {
      "epoch": 0.792,
      "grad_norm": 2.974135398864746,
      "learning_rate": 2.5243370902324792e-05,
      "loss": 5.2276,
      "step": 7920
    },
    {
      "epoch": 0.7921,
      "grad_norm": 2.5552074909210205,
      "learning_rate": 2.5220190950146827e-05,
      "loss": 4.6662,
      "step": 7921
    },
    {
      "epoch": 0.7922,
      "grad_norm": 2.503061056137085,
      "learning_rate": 2.5197020109661772e-05,
      "loss": 4.3607,
      "step": 7922
    },
    {
      "epoch": 0.7923,
      "grad_norm": 4.213957786560059,
      "learning_rate": 2.5173858383692906e-05,
      "loss": 4.852,
      "step": 7923
    },
    {
      "epoch": 0.7924,
      "grad_norm": 2.349419116973877,
      "learning_rate": 2.51507057750624e-05,
      "loss": 4.6087,
      "step": 7924
    },
    {
      "epoch": 0.7925,
      "grad_norm": 2.7591493129730225,
      "learning_rate": 2.512756228659141e-05,
      "loss": 4.8338,
      "step": 7925
    },
    {
      "epoch": 0.7926,
      "grad_norm": 4.730348110198975,
      "learning_rate": 2.5104427921099782e-05,
      "loss": 4.1108,
      "step": 7926
    },
    {
      "epoch": 0.7927,
      "grad_norm": 5.090383529663086,
      "learning_rate": 2.508130268140646e-05,
      "loss": 4.4701,
      "step": 7927
    },
    {
      "epoch": 0.7928,
      "grad_norm": 2.3492774963378906,
      "learning_rate": 2.5058186570329156e-05,
      "loss": 4.4815,
      "step": 7928
    },
    {
      "epoch": 0.7929,
      "grad_norm": 3.8558833599090576,
      "learning_rate": 2.5035079590684497e-05,
      "loss": 5.1115,
      "step": 7929
    },
    {
      "epoch": 0.793,
      "grad_norm": 2.525808811187744,
      "learning_rate": 2.5011981745288015e-05,
      "loss": 5.3895,
      "step": 7930
    },
    {
      "epoch": 0.7931,
      "grad_norm": 2.2723255157470703,
      "learning_rate": 2.4988893036954043e-05,
      "loss": 4.8394,
      "step": 7931
    },
    {
      "epoch": 0.7932,
      "grad_norm": 3.0181057453155518,
      "learning_rate": 2.496581346849596e-05,
      "loss": 4.7153,
      "step": 7932
    },
    {
      "epoch": 0.7933,
      "grad_norm": 2.242152452468872,
      "learning_rate": 2.4942743042725892e-05,
      "loss": 4.8186,
      "step": 7933
    },
    {
      "epoch": 0.7934,
      "grad_norm": 2.120868444442749,
      "learning_rate": 2.4919681762454918e-05,
      "loss": 4.6525,
      "step": 7934
    },
    {
      "epoch": 0.7935,
      "grad_norm": 2.4847192764282227,
      "learning_rate": 2.4896629630492973e-05,
      "loss": 4.7407,
      "step": 7935
    },
    {
      "epoch": 0.7936,
      "grad_norm": 3.9656362533569336,
      "learning_rate": 2.4873586649648894e-05,
      "loss": 4.6293,
      "step": 7936
    },
    {
      "epoch": 0.7937,
      "grad_norm": 3.4293856620788574,
      "learning_rate": 2.48505528227304e-05,
      "loss": 4.6106,
      "step": 7937
    },
    {
      "epoch": 0.7938,
      "grad_norm": 2.6792445182800293,
      "learning_rate": 2.48275281525441e-05,
      "loss": 4.7019,
      "step": 7938
    },
    {
      "epoch": 0.7939,
      "grad_norm": 3.012484073638916,
      "learning_rate": 2.480451264189546e-05,
      "loss": 4.5795,
      "step": 7939
    },
    {
      "epoch": 0.794,
      "grad_norm": 2.4639272689819336,
      "learning_rate": 2.4781506293588873e-05,
      "loss": 4.1135,
      "step": 7940
    },
    {
      "epoch": 0.7941,
      "grad_norm": 2.895838737487793,
      "learning_rate": 2.4758509110427575e-05,
      "loss": 4.4697,
      "step": 7941
    },
    {
      "epoch": 0.7942,
      "grad_norm": 2.382452964782715,
      "learning_rate": 2.4735521095213687e-05,
      "loss": 5.1538,
      "step": 7942
    },
    {
      "epoch": 0.7943,
      "grad_norm": 3.795281171798706,
      "learning_rate": 2.4712542250748304e-05,
      "loss": 4.8159,
      "step": 7943
    },
    {
      "epoch": 0.7944,
      "grad_norm": 3.478447198867798,
      "learning_rate": 2.4689572579831222e-05,
      "loss": 5.0409,
      "step": 7944
    },
    {
      "epoch": 0.7945,
      "grad_norm": 2.7204487323760986,
      "learning_rate": 2.4666612085261342e-05,
      "loss": 4.2337,
      "step": 7945
    },
    {
      "epoch": 0.7946,
      "grad_norm": 3.6047592163085938,
      "learning_rate": 2.464366076983623e-05,
      "loss": 4.9576,
      "step": 7946
    },
    {
      "epoch": 0.7947,
      "grad_norm": 4.055906772613525,
      "learning_rate": 2.4620718636352457e-05,
      "loss": 4.5341,
      "step": 7947
    },
    {
      "epoch": 0.7948,
      "grad_norm": 2.333341598510742,
      "learning_rate": 2.4597785687605513e-05,
      "loss": 4.7947,
      "step": 7948
    },
    {
      "epoch": 0.7949,
      "grad_norm": 3.7113027572631836,
      "learning_rate": 2.4574861926389615e-05,
      "loss": 5.0591,
      "step": 7949
    },
    {
      "epoch": 0.795,
      "grad_norm": 2.2922868728637695,
      "learning_rate": 2.4551947355498027e-05,
      "loss": 5.1101,
      "step": 7950
    },
    {
      "epoch": 0.7951,
      "grad_norm": 2.3673412799835205,
      "learning_rate": 2.45290419777228e-05,
      "loss": 4.4247,
      "step": 7951
    },
    {
      "epoch": 0.7952,
      "grad_norm": 4.637848377227783,
      "learning_rate": 2.4506145795854873e-05,
      "loss": 4.6976,
      "step": 7952
    },
    {
      "epoch": 0.7953,
      "grad_norm": 2.4847264289855957,
      "learning_rate": 2.4483258812684096e-05,
      "loss": 4.4466,
      "step": 7953
    },
    {
      "epoch": 0.7954,
      "grad_norm": 6.058928489685059,
      "learning_rate": 2.4460381030999158e-05,
      "loss": 5.3306,
      "step": 7954
    },
    {
      "epoch": 0.7955,
      "grad_norm": 2.6534922122955322,
      "learning_rate": 2.443751245358765e-05,
      "loss": 4.7594,
      "step": 7955
    },
    {
      "epoch": 0.7956,
      "grad_norm": 3.5591745376586914,
      "learning_rate": 2.441465308323605e-05,
      "loss": 4.5223,
      "step": 7956
    },
    {
      "epoch": 0.7957,
      "grad_norm": 2.9098825454711914,
      "learning_rate": 2.43918029227297e-05,
      "loss": 5.0748,
      "step": 7957
    },
    {
      "epoch": 0.7958,
      "grad_norm": 3.6014769077301025,
      "learning_rate": 2.436896197485282e-05,
      "loss": 4.9501,
      "step": 7958
    },
    {
      "epoch": 0.7959,
      "grad_norm": 3.0741894245147705,
      "learning_rate": 2.43461302423885e-05,
      "loss": 4.7418,
      "step": 7959
    },
    {
      "epoch": 0.796,
      "grad_norm": 9.07675838470459,
      "learning_rate": 2.4323307728118738e-05,
      "loss": 5.1935,
      "step": 7960
    },
    {
      "epoch": 0.7961,
      "grad_norm": 2.8046655654907227,
      "learning_rate": 2.4300494434824373e-05,
      "loss": 4.9687,
      "step": 7961
    },
    {
      "epoch": 0.7962,
      "grad_norm": 2.4475817680358887,
      "learning_rate": 2.4277690365285112e-05,
      "loss": 4.445,
      "step": 7962
    },
    {
      "epoch": 0.7963,
      "grad_norm": 1.9277777671813965,
      "learning_rate": 2.425489552227964e-05,
      "loss": 5.0223,
      "step": 7963
    },
    {
      "epoch": 0.7964,
      "grad_norm": 2.0721850395202637,
      "learning_rate": 2.4232109908585377e-05,
      "loss": 4.6993,
      "step": 7964
    },
    {
      "epoch": 0.7965,
      "grad_norm": 2.312730073928833,
      "learning_rate": 2.420933352697865e-05,
      "loss": 4.5419,
      "step": 7965
    },
    {
      "epoch": 0.7966,
      "grad_norm": 4.306256294250488,
      "learning_rate": 2.4186566380234798e-05,
      "loss": 4.929,
      "step": 7966
    },
    {
      "epoch": 0.7967,
      "grad_norm": 4.129608631134033,
      "learning_rate": 2.4163808471127812e-05,
      "loss": 4.9084,
      "step": 7967
    },
    {
      "epoch": 0.7968,
      "grad_norm": 2.060520648956299,
      "learning_rate": 2.4141059802430777e-05,
      "loss": 4.6757,
      "step": 7968
    },
    {
      "epoch": 0.7969,
      "grad_norm": 5.1675333976745605,
      "learning_rate": 2.411832037691545e-05,
      "loss": 5.0485,
      "step": 7969
    },
    {
      "epoch": 0.797,
      "grad_norm": 2.815091133117676,
      "learning_rate": 2.4095590197352635e-05,
      "loss": 4.7997,
      "step": 7970
    },
    {
      "epoch": 0.7971,
      "grad_norm": 2.444312334060669,
      "learning_rate": 2.407286926651192e-05,
      "loss": 4.3176,
      "step": 7971
    },
    {
      "epoch": 0.7972,
      "grad_norm": 3.5573692321777344,
      "learning_rate": 2.405015758716177e-05,
      "loss": 4.7035,
      "step": 7972
    },
    {
      "epoch": 0.7973,
      "grad_norm": 3.3959364891052246,
      "learning_rate": 2.4027455162069567e-05,
      "loss": 4.7547,
      "step": 7973
    },
    {
      "epoch": 0.7974,
      "grad_norm": 2.6235392093658447,
      "learning_rate": 2.4004761994001435e-05,
      "loss": 4.6837,
      "step": 7974
    },
    {
      "epoch": 0.7975,
      "grad_norm": 2.9914159774780273,
      "learning_rate": 2.3982078085722582e-05,
      "loss": 4.5066,
      "step": 7975
    },
    {
      "epoch": 0.7976,
      "grad_norm": 2.877686023712158,
      "learning_rate": 2.3959403439996907e-05,
      "loss": 4.4456,
      "step": 7976
    },
    {
      "epoch": 0.7977,
      "grad_norm": 2.39951229095459,
      "learning_rate": 2.3936738059587282e-05,
      "loss": 5.1535,
      "step": 7977
    },
    {
      "epoch": 0.7978,
      "grad_norm": 3.371448516845703,
      "learning_rate": 2.3914081947255397e-05,
      "loss": 4.8022,
      "step": 7978
    },
    {
      "epoch": 0.7979,
      "grad_norm": 2.9091763496398926,
      "learning_rate": 2.389143510576184e-05,
      "loss": 5.0728,
      "step": 7979
    },
    {
      "epoch": 0.798,
      "grad_norm": 2.272561550140381,
      "learning_rate": 2.3868797537866016e-05,
      "loss": 4.7949,
      "step": 7980
    },
    {
      "epoch": 0.7981,
      "grad_norm": 2.3972084522247314,
      "learning_rate": 2.3846169246326343e-05,
      "loss": 4.6618,
      "step": 7981
    },
    {
      "epoch": 0.7982,
      "grad_norm": 2.5610997676849365,
      "learning_rate": 2.3823550233899915e-05,
      "loss": 4.9782,
      "step": 7982
    },
    {
      "epoch": 0.7983,
      "grad_norm": 2.713954210281372,
      "learning_rate": 2.3800940503342828e-05,
      "loss": 4.116,
      "step": 7983
    },
    {
      "epoch": 0.7984,
      "grad_norm": 2.1075165271759033,
      "learning_rate": 2.3778340057409998e-05,
      "loss": 4.3762,
      "step": 7984
    },
    {
      "epoch": 0.7985,
      "grad_norm": 2.118223190307617,
      "learning_rate": 2.37557488988552e-05,
      "loss": 4.7697,
      "step": 7985
    },
    {
      "epoch": 0.7986,
      "grad_norm": 2.866859197616577,
      "learning_rate": 2.373316703043119e-05,
      "loss": 4.2875,
      "step": 7986
    },
    {
      "epoch": 0.7987,
      "grad_norm": 3.6573681831359863,
      "learning_rate": 2.3710594454889378e-05,
      "loss": 5.1984,
      "step": 7987
    },
    {
      "epoch": 0.7988,
      "grad_norm": 2.529506206512451,
      "learning_rate": 2.3688031174980275e-05,
      "loss": 4.4407,
      "step": 7988
    },
    {
      "epoch": 0.7989,
      "grad_norm": 2.9606761932373047,
      "learning_rate": 2.3665477193453034e-05,
      "loss": 4.9321,
      "step": 7989
    },
    {
      "epoch": 0.799,
      "grad_norm": 2.6233956813812256,
      "learning_rate": 2.3642932513055884e-05,
      "loss": 4.9164,
      "step": 7990
    },
    {
      "epoch": 0.7991,
      "grad_norm": 2.6657137870788574,
      "learning_rate": 2.362039713653581e-05,
      "loss": 4.6227,
      "step": 7991
    },
    {
      "epoch": 0.7992,
      "grad_norm": 2.2499265670776367,
      "learning_rate": 2.359787106663861e-05,
      "loss": 4.9279,
      "step": 7992
    },
    {
      "epoch": 0.7993,
      "grad_norm": 3.7281668186187744,
      "learning_rate": 2.35753543061091e-05,
      "loss": 4.8737,
      "step": 7993
    },
    {
      "epoch": 0.7994,
      "grad_norm": 3.1699342727661133,
      "learning_rate": 2.3552846857690846e-05,
      "loss": 4.8079,
      "step": 7994
    },
    {
      "epoch": 0.7995,
      "grad_norm": 3.1185007095336914,
      "learning_rate": 2.3530348724126307e-05,
      "loss": 4.4718,
      "step": 7995
    },
    {
      "epoch": 0.7996,
      "grad_norm": 4.799326419830322,
      "learning_rate": 2.3507859908156827e-05,
      "loss": 4.6161,
      "step": 7996
    },
    {
      "epoch": 0.7997,
      "grad_norm": 2.9995181560516357,
      "learning_rate": 2.3485380412522585e-05,
      "loss": 5.0816,
      "step": 7997
    },
    {
      "epoch": 0.7998,
      "grad_norm": 2.182861566543579,
      "learning_rate": 2.3462910239962654e-05,
      "loss": 4.6523,
      "step": 7998
    },
    {
      "epoch": 0.7999,
      "grad_norm": 2.3130247592926025,
      "learning_rate": 2.3440449393214948e-05,
      "loss": 4.8092,
      "step": 7999
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.290402412414551,
      "learning_rate": 2.341799787501625e-05,
      "loss": 4.6378,
      "step": 8000
    },
    {
      "epoch": 0.8001,
      "grad_norm": 2.7434208393096924,
      "learning_rate": 2.339555568810221e-05,
      "loss": 4.7871,
      "step": 8001
    },
    {
      "epoch": 0.8002,
      "grad_norm": 2.3146748542785645,
      "learning_rate": 2.337312283520735e-05,
      "loss": 4.3876,
      "step": 8002
    },
    {
      "epoch": 0.8003,
      "grad_norm": 2.8315248489379883,
      "learning_rate": 2.3350699319065026e-05,
      "loss": 5.1494,
      "step": 8003
    },
    {
      "epoch": 0.8004,
      "grad_norm": 5.687308311462402,
      "learning_rate": 2.3328285142407503e-05,
      "loss": 4.5795,
      "step": 8004
    },
    {
      "epoch": 0.8005,
      "grad_norm": 2.3403759002685547,
      "learning_rate": 2.3305880307965833e-05,
      "loss": 4.6924,
      "step": 8005
    },
    {
      "epoch": 0.8006,
      "grad_norm": 2.2935800552368164,
      "learning_rate": 2.328348481847006e-05,
      "loss": 4.4837,
      "step": 8006
    },
    {
      "epoch": 0.8007,
      "grad_norm": 2.1452510356903076,
      "learning_rate": 2.3261098676648907e-05,
      "loss": 4.5556,
      "step": 8007
    },
    {
      "epoch": 0.8008,
      "grad_norm": 2.799797296524048,
      "learning_rate": 2.323872188523013e-05,
      "loss": 4.7207,
      "step": 8008
    },
    {
      "epoch": 0.8009,
      "grad_norm": 2.929349184036255,
      "learning_rate": 2.321635444694028e-05,
      "loss": 4.6949,
      "step": 8009
    },
    {
      "epoch": 0.801,
      "grad_norm": 3.136687994003296,
      "learning_rate": 2.319399636450468e-05,
      "loss": 3.9686,
      "step": 8010
    },
    {
      "epoch": 0.8011,
      "grad_norm": 3.7100203037261963,
      "learning_rate": 2.3171647640647687e-05,
      "loss": 4.7825,
      "step": 8011
    },
    {
      "epoch": 0.8012,
      "grad_norm": 2.2142727375030518,
      "learning_rate": 2.3149308278092342e-05,
      "loss": 4.3259,
      "step": 8012
    },
    {
      "epoch": 0.8013,
      "grad_norm": 2.3582141399383545,
      "learning_rate": 2.3126978279560684e-05,
      "loss": 4.668,
      "step": 8013
    },
    {
      "epoch": 0.8014,
      "grad_norm": 3.6136176586151123,
      "learning_rate": 2.3104657647773554e-05,
      "loss": 4.7826,
      "step": 8014
    },
    {
      "epoch": 0.8015,
      "grad_norm": 6.972061634063721,
      "learning_rate": 2.3082346385450637e-05,
      "loss": 6.6387,
      "step": 8015
    },
    {
      "epoch": 0.8016,
      "grad_norm": 2.8325142860412598,
      "learning_rate": 2.3060044495310505e-05,
      "loss": 4.9207,
      "step": 8016
    },
    {
      "epoch": 0.8017,
      "grad_norm": 3.0464138984680176,
      "learning_rate": 2.3037751980070555e-05,
      "loss": 4.7604,
      "step": 8017
    },
    {
      "epoch": 0.8018,
      "grad_norm": 2.1690709590911865,
      "learning_rate": 2.3015468842447086e-05,
      "loss": 4.6576,
      "step": 8018
    },
    {
      "epoch": 0.8019,
      "grad_norm": 2.8691599369049072,
      "learning_rate": 2.2993195085155205e-05,
      "loss": 5.244,
      "step": 8019
    },
    {
      "epoch": 0.802,
      "grad_norm": 2.3306825160980225,
      "learning_rate": 2.2970930710908935e-05,
      "loss": 4.4097,
      "step": 8020
    },
    {
      "epoch": 0.8021,
      "grad_norm": 1.9080525636672974,
      "learning_rate": 2.2948675722421086e-05,
      "loss": 4.5618,
      "step": 8021
    },
    {
      "epoch": 0.8022,
      "grad_norm": 3.657074213027954,
      "learning_rate": 2.2926430122403386e-05,
      "loss": 5.0928,
      "step": 8022
    },
    {
      "epoch": 0.8023,
      "grad_norm": 2.3197412490844727,
      "learning_rate": 2.2904193913566363e-05,
      "loss": 4.2808,
      "step": 8023
    },
    {
      "epoch": 0.8024,
      "grad_norm": 3.0522375106811523,
      "learning_rate": 2.2881967098619506e-05,
      "loss": 4.7945,
      "step": 8024
    },
    {
      "epoch": 0.8025,
      "grad_norm": 4.143770694732666,
      "learning_rate": 2.2859749680270982e-05,
      "loss": 5.2486,
      "step": 8025
    },
    {
      "epoch": 0.8026,
      "grad_norm": 3.410865545272827,
      "learning_rate": 2.2837541661228025e-05,
      "loss": 5.2889,
      "step": 8026
    },
    {
      "epoch": 0.8027,
      "grad_norm": 3.7813990116119385,
      "learning_rate": 2.2815343044196525e-05,
      "loss": 5.0456,
      "step": 8027
    },
    {
      "epoch": 0.8028,
      "grad_norm": 2.738081693649292,
      "learning_rate": 2.279315383188132e-05,
      "loss": 4.8791,
      "step": 8028
    },
    {
      "epoch": 0.8029,
      "grad_norm": 1.9137247800827026,
      "learning_rate": 2.277097402698619e-05,
      "loss": 4.3596,
      "step": 8029
    },
    {
      "epoch": 0.803,
      "grad_norm": 4.924509048461914,
      "learning_rate": 2.2748803632213557e-05,
      "loss": 4.8302,
      "step": 8030
    },
    {
      "epoch": 0.8031,
      "grad_norm": 2.435649871826172,
      "learning_rate": 2.2726642650264895e-05,
      "loss": 4.6241,
      "step": 8031
    },
    {
      "epoch": 0.8032,
      "grad_norm": 3.141439914703369,
      "learning_rate": 2.270449108384044e-05,
      "loss": 4.8323,
      "step": 8032
    },
    {
      "epoch": 0.8033,
      "grad_norm": 2.9818315505981445,
      "learning_rate": 2.2682348935639274e-05,
      "loss": 4.3798,
      "step": 8033
    },
    {
      "epoch": 0.8034,
      "grad_norm": 2.736936330795288,
      "learning_rate": 2.2660216208359365e-05,
      "loss": 4.6908,
      "step": 8034
    },
    {
      "epoch": 0.8035,
      "grad_norm": 3.505295991897583,
      "learning_rate": 2.2638092904697518e-05,
      "loss": 5.3693,
      "step": 8035
    },
    {
      "epoch": 0.8036,
      "grad_norm": 3.33463454246521,
      "learning_rate": 2.2615979027349387e-05,
      "loss": 4.2335,
      "step": 8036
    },
    {
      "epoch": 0.8037,
      "grad_norm": 2.348550319671631,
      "learning_rate": 2.259387457900948e-05,
      "loss": 4.8238,
      "step": 8037
    },
    {
      "epoch": 0.8038,
      "grad_norm": 3.0589492321014404,
      "learning_rate": 2.2571779562371153e-05,
      "loss": 4.6978,
      "step": 8038
    },
    {
      "epoch": 0.8039,
      "grad_norm": 4.962616443634033,
      "learning_rate": 2.254969398012663e-05,
      "loss": 4.5555,
      "step": 8039
    },
    {
      "epoch": 0.804,
      "grad_norm": 2.5533688068389893,
      "learning_rate": 2.2527617834966954e-05,
      "loss": 4.8345,
      "step": 8040
    },
    {
      "epoch": 0.8041,
      "grad_norm": 4.3777289390563965,
      "learning_rate": 2.2505551129582047e-05,
      "loss": 4.2349,
      "step": 8041
    },
    {
      "epoch": 0.8042,
      "grad_norm": 2.9013564586639404,
      "learning_rate": 2.2483493866660676e-05,
      "loss": 4.7991,
      "step": 8042
    },
    {
      "epoch": 0.8043,
      "grad_norm": 3.8614535331726074,
      "learning_rate": 2.246144604889042e-05,
      "loss": 4.5594,
      "step": 8043
    },
    {
      "epoch": 0.8044,
      "grad_norm": 2.0890660285949707,
      "learning_rate": 2.2439407678957812e-05,
      "loss": 4.2959,
      "step": 8044
    },
    {
      "epoch": 0.8045,
      "grad_norm": 4.465095043182373,
      "learning_rate": 2.241737875954808e-05,
      "loss": 5.4379,
      "step": 8045
    },
    {
      "epoch": 0.8046,
      "grad_norm": 1.8204615116119385,
      "learning_rate": 2.2395359293345396e-05,
      "loss": 4.8995,
      "step": 8046
    },
    {
      "epoch": 0.8047,
      "grad_norm": 2.1314303874969482,
      "learning_rate": 2.237334928303283e-05,
      "loss": 4.4109,
      "step": 8047
    },
    {
      "epoch": 0.8048,
      "grad_norm": 2.897576332092285,
      "learning_rate": 2.235134873129213e-05,
      "loss": 5.017,
      "step": 8048
    },
    {
      "epoch": 0.8049,
      "grad_norm": 1.7612487077713013,
      "learning_rate": 2.2329357640804117e-05,
      "loss": 4.6651,
      "step": 8049
    },
    {
      "epoch": 0.805,
      "grad_norm": 2.901675224304199,
      "learning_rate": 2.2307376014248216e-05,
      "loss": 4.7668,
      "step": 8050
    },
    {
      "epoch": 0.8051,
      "grad_norm": 3.307384967803955,
      "learning_rate": 2.2285403854302912e-05,
      "loss": 4.2384,
      "step": 8051
    },
    {
      "epoch": 0.8052,
      "grad_norm": 2.649427890777588,
      "learning_rate": 2.2263441163645403e-05,
      "loss": 4.6441,
      "step": 8052
    },
    {
      "epoch": 0.8053,
      "grad_norm": 3.0534093379974365,
      "learning_rate": 2.22414879449518e-05,
      "loss": 5.7527,
      "step": 8053
    },
    {
      "epoch": 0.8054,
      "grad_norm": 2.043818473815918,
      "learning_rate": 2.2219544200897025e-05,
      "loss": 4.6249,
      "step": 8054
    },
    {
      "epoch": 0.8055,
      "grad_norm": 2.4989521503448486,
      "learning_rate": 2.219760993415485e-05,
      "loss": 4.538,
      "step": 8055
    },
    {
      "epoch": 0.8056,
      "grad_norm": 2.1336312294006348,
      "learning_rate": 2.2175685147397906e-05,
      "loss": 4.8076,
      "step": 8056
    },
    {
      "epoch": 0.8057,
      "grad_norm": 2.1215484142303467,
      "learning_rate": 2.2153769843297667e-05,
      "loss": 4.7597,
      "step": 8057
    },
    {
      "epoch": 0.8058,
      "grad_norm": 6.04477071762085,
      "learning_rate": 2.213186402452443e-05,
      "loss": 4.8872,
      "step": 8058
    },
    {
      "epoch": 0.8059,
      "grad_norm": 3.4038069248199463,
      "learning_rate": 2.210996769374737e-05,
      "loss": 4.4996,
      "step": 8059
    },
    {
      "epoch": 0.806,
      "grad_norm": 3.0249807834625244,
      "learning_rate": 2.2088080853634473e-05,
      "loss": 4.5086,
      "step": 8060
    },
    {
      "epoch": 0.8061,
      "grad_norm": 2.958819627761841,
      "learning_rate": 2.2066203506852566e-05,
      "loss": 4.4455,
      "step": 8061
    },
    {
      "epoch": 0.8062,
      "grad_norm": 4.382347106933594,
      "learning_rate": 2.204433565606743e-05,
      "loss": 5.8872,
      "step": 8062
    },
    {
      "epoch": 0.8063,
      "grad_norm": 3.3288064002990723,
      "learning_rate": 2.202247730394349e-05,
      "loss": 4.5249,
      "step": 8063
    },
    {
      "epoch": 0.8064,
      "grad_norm": 2.4791717529296875,
      "learning_rate": 2.200062845314417e-05,
      "loss": 4.7625,
      "step": 8064
    },
    {
      "epoch": 0.8065,
      "grad_norm": 2.6996073722839355,
      "learning_rate": 2.1978789106331665e-05,
      "loss": 4.9931,
      "step": 8065
    },
    {
      "epoch": 0.8066,
      "grad_norm": 2.735774517059326,
      "learning_rate": 2.195695926616702e-05,
      "loss": 4.9791,
      "step": 8066
    },
    {
      "epoch": 0.8067,
      "grad_norm": 2.145768642425537,
      "learning_rate": 2.1935138935310206e-05,
      "loss": 4.9773,
      "step": 8067
    },
    {
      "epoch": 0.8068,
      "grad_norm": 3.976475477218628,
      "learning_rate": 2.1913328116419873e-05,
      "loss": 5.0089,
      "step": 8068
    },
    {
      "epoch": 0.8069,
      "grad_norm": 2.457306146621704,
      "learning_rate": 2.1891526812153672e-05,
      "loss": 4.8568,
      "step": 8069
    },
    {
      "epoch": 0.807,
      "grad_norm": 2.6296563148498535,
      "learning_rate": 2.1869735025168026e-05,
      "loss": 5.4814,
      "step": 8070
    },
    {
      "epoch": 0.8071,
      "grad_norm": 4.843747615814209,
      "learning_rate": 2.1847952758118117e-05,
      "loss": 4.61,
      "step": 8071
    },
    {
      "epoch": 0.8072,
      "grad_norm": 2.4819650650024414,
      "learning_rate": 2.1826180013658172e-05,
      "loss": 4.4937,
      "step": 8072
    },
    {
      "epoch": 0.8073,
      "grad_norm": 3.5444746017456055,
      "learning_rate": 2.1804416794440995e-05,
      "loss": 5.2379,
      "step": 8073
    },
    {
      "epoch": 0.8074,
      "grad_norm": 2.400266170501709,
      "learning_rate": 2.178266310311847e-05,
      "loss": 5.0231,
      "step": 8074
    },
    {
      "epoch": 0.8075,
      "grad_norm": 2.176754951477051,
      "learning_rate": 2.1760918942341192e-05,
      "loss": 4.9748,
      "step": 8075
    },
    {
      "epoch": 0.8076,
      "grad_norm": 2.8607208728790283,
      "learning_rate": 2.173918431475861e-05,
      "loss": 4.1168,
      "step": 8076
    },
    {
      "epoch": 0.8077,
      "grad_norm": 2.5842928886413574,
      "learning_rate": 2.1717459223019033e-05,
      "loss": 4.663,
      "step": 8077
    },
    {
      "epoch": 0.8078,
      "grad_norm": 3.849823474884033,
      "learning_rate": 2.1695743669769596e-05,
      "loss": 4.8452,
      "step": 8078
    },
    {
      "epoch": 0.8079,
      "grad_norm": 2.3780887126922607,
      "learning_rate": 2.1674037657656266e-05,
      "loss": 4.715,
      "step": 8079
    },
    {
      "epoch": 0.808,
      "grad_norm": 3.379889488220215,
      "learning_rate": 2.1652341189323866e-05,
      "loss": 4.5215,
      "step": 8080
    },
    {
      "epoch": 0.8081,
      "grad_norm": 4.706051349639893,
      "learning_rate": 2.163065426741603e-05,
      "loss": 4.8684,
      "step": 8081
    },
    {
      "epoch": 0.8082,
      "grad_norm": 4.163350582122803,
      "learning_rate": 2.160897689457526e-05,
      "loss": 4.6089,
      "step": 8082
    },
    {
      "epoch": 0.8083,
      "grad_norm": 2.3966362476348877,
      "learning_rate": 2.1587309073442863e-05,
      "loss": 4.8318,
      "step": 8083
    },
    {
      "epoch": 0.8084,
      "grad_norm": 3.980822801589966,
      "learning_rate": 2.1565650806658975e-05,
      "loss": 4.3403,
      "step": 8084
    },
    {
      "epoch": 0.8085,
      "grad_norm": 4.359767436981201,
      "learning_rate": 2.154400209686268e-05,
      "loss": 5.8402,
      "step": 8085
    },
    {
      "epoch": 0.8086,
      "grad_norm": 2.315685272216797,
      "learning_rate": 2.1522362946691698e-05,
      "loss": 4.8936,
      "step": 8086
    },
    {
      "epoch": 0.8087,
      "grad_norm": 2.159010648727417,
      "learning_rate": 2.1500733358782786e-05,
      "loss": 4.47,
      "step": 8087
    },
    {
      "epoch": 0.8088,
      "grad_norm": 1.9812536239624023,
      "learning_rate": 2.1479113335771383e-05,
      "loss": 4.1944,
      "step": 8088
    },
    {
      "epoch": 0.8089,
      "grad_norm": 3.4751336574554443,
      "learning_rate": 2.1457502880291812e-05,
      "loss": 4.7095,
      "step": 8089
    },
    {
      "epoch": 0.809,
      "grad_norm": 3.6332740783691406,
      "learning_rate": 2.1435901994977326e-05,
      "loss": 4.8031,
      "step": 8090
    },
    {
      "epoch": 0.8091,
      "grad_norm": 2.2532904148101807,
      "learning_rate": 2.1414310682459802e-05,
      "loss": 4.6143,
      "step": 8091
    },
    {
      "epoch": 0.8092,
      "grad_norm": 2.5255377292633057,
      "learning_rate": 2.1392728945370222e-05,
      "loss": 4.7854,
      "step": 8092
    },
    {
      "epoch": 0.8093,
      "grad_norm": 5.012913227081299,
      "learning_rate": 2.137115678633811e-05,
      "loss": 4.5115,
      "step": 8093
    },
    {
      "epoch": 0.8094,
      "grad_norm": 2.969583749771118,
      "learning_rate": 2.1349594207992064e-05,
      "loss": 4.729,
      "step": 8094
    },
    {
      "epoch": 0.8095,
      "grad_norm": 3.791407585144043,
      "learning_rate": 2.1328041212959403e-05,
      "loss": 4.8643,
      "step": 8095
    },
    {
      "epoch": 0.8096,
      "grad_norm": 2.2843892574310303,
      "learning_rate": 2.1306497803866277e-05,
      "loss": 4.4642,
      "step": 8096
    },
    {
      "epoch": 0.8097,
      "grad_norm": 2.206742286682129,
      "learning_rate": 2.128496398333768e-05,
      "loss": 4.6449,
      "step": 8097
    },
    {
      "epoch": 0.8098,
      "grad_norm": 3.5441064834594727,
      "learning_rate": 2.126343975399747e-05,
      "loss": 5.213,
      "step": 8098
    },
    {
      "epoch": 0.8099,
      "grad_norm": 3.935055732727051,
      "learning_rate": 2.1241925118468287e-05,
      "loss": 5.073,
      "step": 8099
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.6420938968658447,
      "learning_rate": 2.1220420079371628e-05,
      "loss": 4.591,
      "step": 8100
    },
    {
      "epoch": 0.8101,
      "grad_norm": 5.050716876983643,
      "learning_rate": 2.119892463932781e-05,
      "loss": 4.802,
      "step": 8101
    },
    {
      "epoch": 0.8102,
      "grad_norm": 3.855355739593506,
      "learning_rate": 2.1177438800956007e-05,
      "loss": 5.1723,
      "step": 8102
    },
    {
      "epoch": 0.8103,
      "grad_norm": 2.155489444732666,
      "learning_rate": 2.115596256687419e-05,
      "loss": 4.7146,
      "step": 8103
    },
    {
      "epoch": 0.8104,
      "grad_norm": 5.877534866333008,
      "learning_rate": 2.113449593969915e-05,
      "loss": 4.2761,
      "step": 8104
    },
    {
      "epoch": 0.8105,
      "grad_norm": 3.3134429454803467,
      "learning_rate": 2.1113038922046602e-05,
      "loss": 4.5876,
      "step": 8105
    },
    {
      "epoch": 0.8106,
      "grad_norm": 4.289638996124268,
      "learning_rate": 2.1091591516530952e-05,
      "loss": 4.4829,
      "step": 8106
    },
    {
      "epoch": 0.8107,
      "grad_norm": 2.5427095890045166,
      "learning_rate": 2.107015372576552e-05,
      "loss": 4.507,
      "step": 8107
    },
    {
      "epoch": 0.8108,
      "grad_norm": 2.6233410835266113,
      "learning_rate": 2.1048725552362435e-05,
      "loss": 4.855,
      "step": 8108
    },
    {
      "epoch": 0.8109,
      "grad_norm": 2.048806667327881,
      "learning_rate": 2.1027306998932627e-05,
      "loss": 4.8459,
      "step": 8109
    },
    {
      "epoch": 0.811,
      "grad_norm": 5.365761756896973,
      "learning_rate": 2.100589806808597e-05,
      "loss": 4.7809,
      "step": 8110
    },
    {
      "epoch": 0.8111,
      "grad_norm": 4.909718990325928,
      "learning_rate": 2.098449876243096e-05,
      "loss": 5.1167,
      "step": 8111
    },
    {
      "epoch": 0.8112,
      "grad_norm": 7.018162250518799,
      "learning_rate": 2.096310908457513e-05,
      "loss": 5.4447,
      "step": 8112
    },
    {
      "epoch": 0.8113,
      "grad_norm": 5.724162578582764,
      "learning_rate": 2.09417290371247e-05,
      "loss": 5.4238,
      "step": 8113
    },
    {
      "epoch": 0.8114,
      "grad_norm": 3.439617156982422,
      "learning_rate": 2.0920358622684788e-05,
      "loss": 4.4921,
      "step": 8114
    },
    {
      "epoch": 0.8115,
      "grad_norm": 1.985856056213379,
      "learning_rate": 2.0898997843859334e-05,
      "loss": 4.6358,
      "step": 8115
    },
    {
      "epoch": 0.8116,
      "grad_norm": 1.9476251602172852,
      "learning_rate": 2.0877646703250996e-05,
      "loss": 4.7147,
      "step": 8116
    },
    {
      "epoch": 0.8117,
      "grad_norm": 2.7719533443450928,
      "learning_rate": 2.0856305203461436e-05,
      "loss": 4.4761,
      "step": 8117
    },
    {
      "epoch": 0.8118,
      "grad_norm": 2.6881797313690186,
      "learning_rate": 2.0834973347091014e-05,
      "loss": 4.1758,
      "step": 8118
    },
    {
      "epoch": 0.8119,
      "grad_norm": 2.3239688873291016,
      "learning_rate": 2.0813651136738954e-05,
      "loss": 4.9712,
      "step": 8119
    },
    {
      "epoch": 0.812,
      "grad_norm": 2.7869112491607666,
      "learning_rate": 2.07923385750033e-05,
      "loss": 4.7951,
      "step": 8120
    },
    {
      "epoch": 0.8121,
      "grad_norm": 2.401865243911743,
      "learning_rate": 2.0771035664480942e-05,
      "loss": 4.5549,
      "step": 8121
    },
    {
      "epoch": 0.8122,
      "grad_norm": 3.596365451812744,
      "learning_rate": 2.0749742407767546e-05,
      "loss": 4.8932,
      "step": 8122
    },
    {
      "epoch": 0.8123,
      "grad_norm": 3.1564061641693115,
      "learning_rate": 2.0728458807457662e-05,
      "loss": 4.4961,
      "step": 8123
    },
    {
      "epoch": 0.8124,
      "grad_norm": 2.9673099517822266,
      "learning_rate": 2.0707184866144603e-05,
      "loss": 5.4499,
      "step": 8124
    },
    {
      "epoch": 0.8125,
      "grad_norm": 2.578117609024048,
      "learning_rate": 2.068592058642056e-05,
      "loss": 4.5999,
      "step": 8125
    },
    {
      "epoch": 0.8126,
      "grad_norm": 2.7290446758270264,
      "learning_rate": 2.0664665970876496e-05,
      "loss": 4.3068,
      "step": 8126
    },
    {
      "epoch": 0.8127,
      "grad_norm": 8.10662841796875,
      "learning_rate": 2.0643421022102217e-05,
      "loss": 4.9055,
      "step": 8127
    },
    {
      "epoch": 0.8128,
      "grad_norm": 3.0821774005889893,
      "learning_rate": 2.0622185742686416e-05,
      "loss": 4.7658,
      "step": 8128
    },
    {
      "epoch": 0.8129,
      "grad_norm": 1.9806674718856812,
      "learning_rate": 2.0600960135216462e-05,
      "loss": 4.8663,
      "step": 8129
    },
    {
      "epoch": 0.813,
      "grad_norm": 3.229379177093506,
      "learning_rate": 2.0579744202278718e-05,
      "loss": 4.8486,
      "step": 8130
    },
    {
      "epoch": 0.8131,
      "grad_norm": 3.0395121574401855,
      "learning_rate": 2.0558537946458177e-05,
      "loss": 4.8727,
      "step": 8131
    },
    {
      "epoch": 0.8132,
      "grad_norm": 2.958653688430786,
      "learning_rate": 2.0537341370338857e-05,
      "loss": 4.9035,
      "step": 8132
    },
    {
      "epoch": 0.8133,
      "grad_norm": 3.0177271366119385,
      "learning_rate": 2.051615447650347e-05,
      "loss": 4.5833,
      "step": 8133
    },
    {
      "epoch": 0.8134,
      "grad_norm": 2.8899080753326416,
      "learning_rate": 2.049497726753351e-05,
      "loss": 5.2545,
      "step": 8134
    },
    {
      "epoch": 0.8135,
      "grad_norm": 1.960601806640625,
      "learning_rate": 2.0473809746009444e-05,
      "loss": 4.3242,
      "step": 8135
    },
    {
      "epoch": 0.8136,
      "grad_norm": 1.9072729349136353,
      "learning_rate": 2.045265191451041e-05,
      "loss": 4.4794,
      "step": 8136
    },
    {
      "epoch": 0.8137,
      "grad_norm": 2.8200550079345703,
      "learning_rate": 2.0431503775614457e-05,
      "loss": 4.7121,
      "step": 8137
    },
    {
      "epoch": 0.8138,
      "grad_norm": 2.4160492420196533,
      "learning_rate": 2.0410365331898416e-05,
      "loss": 4.8988,
      "step": 8138
    },
    {
      "epoch": 0.8139,
      "grad_norm": 2.17561936378479,
      "learning_rate": 2.0389236585937945e-05,
      "loss": 4.4021,
      "step": 8139
    },
    {
      "epoch": 0.814,
      "grad_norm": 3.5657384395599365,
      "learning_rate": 2.0368117540307496e-05,
      "loss": 4.5047,
      "step": 8140
    },
    {
      "epoch": 0.8141,
      "grad_norm": 2.8933489322662354,
      "learning_rate": 2.0347008197580374e-05,
      "loss": 4.5881,
      "step": 8141
    },
    {
      "epoch": 0.8142,
      "grad_norm": 3.8208751678466797,
      "learning_rate": 2.03259085603287e-05,
      "loss": 4.6383,
      "step": 8142
    },
    {
      "epoch": 0.8143,
      "grad_norm": 2.4745094776153564,
      "learning_rate": 2.030481863112339e-05,
      "loss": 4.9813,
      "step": 8143
    },
    {
      "epoch": 0.8144,
      "grad_norm": 3.969186782836914,
      "learning_rate": 2.028373841253419e-05,
      "loss": 5.3844,
      "step": 8144
    },
    {
      "epoch": 0.8145,
      "grad_norm": 4.29534912109375,
      "learning_rate": 2.026266790712965e-05,
      "loss": 5.0193,
      "step": 8145
    },
    {
      "epoch": 0.8146,
      "grad_norm": 3.978422164916992,
      "learning_rate": 2.024160711747717e-05,
      "loss": 4.8712,
      "step": 8146
    },
    {
      "epoch": 0.8147,
      "grad_norm": 2.673948287963867,
      "learning_rate": 2.0220556046142893e-05,
      "loss": 4.9154,
      "step": 8147
    },
    {
      "epoch": 0.8148,
      "grad_norm": 3.4999449253082275,
      "learning_rate": 2.019951469569191e-05,
      "loss": 4.5081,
      "step": 8148
    },
    {
      "epoch": 0.8149,
      "grad_norm": 2.4312355518341064,
      "learning_rate": 2.017848306868797e-05,
      "loss": 5.1483,
      "step": 8149
    },
    {
      "epoch": 0.815,
      "grad_norm": 4.318506240844727,
      "learning_rate": 2.0157461167693758e-05,
      "loss": 4.7756,
      "step": 8150
    },
    {
      "epoch": 0.8151,
      "grad_norm": 3.007946729660034,
      "learning_rate": 2.013644899527074e-05,
      "loss": 4.855,
      "step": 8151
    },
    {
      "epoch": 0.8152,
      "grad_norm": 3.2958340644836426,
      "learning_rate": 2.01154465539791e-05,
      "loss": 4.639,
      "step": 8152
    },
    {
      "epoch": 0.8153,
      "grad_norm": 2.760615587234497,
      "learning_rate": 2.009445384637805e-05,
      "loss": 4.8112,
      "step": 8153
    },
    {
      "epoch": 0.8154,
      "grad_norm": 2.1615653038024902,
      "learning_rate": 2.0073470875025358e-05,
      "loss": 4.5249,
      "step": 8154
    },
    {
      "epoch": 0.8155,
      "grad_norm": 3.049898862838745,
      "learning_rate": 2.005249764247783e-05,
      "loss": 4.477,
      "step": 8155
    },
    {
      "epoch": 0.8156,
      "grad_norm": 2.9754230976104736,
      "learning_rate": 2.0031534151290943e-05,
      "loss": 5.3676,
      "step": 8156
    },
    {
      "epoch": 0.8157,
      "grad_norm": 2.70991587638855,
      "learning_rate": 2.0010580404019062e-05,
      "loss": 5.0062,
      "step": 8157
    },
    {
      "epoch": 0.8158,
      "grad_norm": 2.1519625186920166,
      "learning_rate": 1.9989636403215328e-05,
      "loss": 4.5655,
      "step": 8158
    },
    {
      "epoch": 0.8159,
      "grad_norm": 1.8965320587158203,
      "learning_rate": 1.9968702151431696e-05,
      "loss": 4.7717,
      "step": 8159
    },
    {
      "epoch": 0.816,
      "grad_norm": 2.933422803878784,
      "learning_rate": 1.9947777651218946e-05,
      "loss": 4.2428,
      "step": 8160
    },
    {
      "epoch": 0.8161,
      "grad_norm": 1.9472681283950806,
      "learning_rate": 1.9926862905126665e-05,
      "loss": 4.4322,
      "step": 8161
    },
    {
      "epoch": 0.8162,
      "grad_norm": 2.504744529724121,
      "learning_rate": 1.9905957915703244e-05,
      "loss": 4.4848,
      "step": 8162
    },
    {
      "epoch": 0.8163,
      "grad_norm": 3.5566203594207764,
      "learning_rate": 1.9885062685495904e-05,
      "loss": 4.4256,
      "step": 8163
    },
    {
      "epoch": 0.8164,
      "grad_norm": 4.542937278747559,
      "learning_rate": 1.9864177217050674e-05,
      "loss": 5.0271,
      "step": 8164
    },
    {
      "epoch": 0.8165,
      "grad_norm": 3.5416297912597656,
      "learning_rate": 1.9843301512912327e-05,
      "loss": 4.0439,
      "step": 8165
    },
    {
      "epoch": 0.8166,
      "grad_norm": 2.9168875217437744,
      "learning_rate": 1.9822435575624608e-05,
      "loss": 4.5944,
      "step": 8166
    },
    {
      "epoch": 0.8167,
      "grad_norm": 4.975122451782227,
      "learning_rate": 1.9801579407729866e-05,
      "loss": 5.2757,
      "step": 8167
    },
    {
      "epoch": 0.8168,
      "grad_norm": 3.1146914958953857,
      "learning_rate": 1.9780733011769447e-05,
      "loss": 4.5149,
      "step": 8168
    },
    {
      "epoch": 0.8169,
      "grad_norm": 3.1793978214263916,
      "learning_rate": 1.9759896390283362e-05,
      "loss": 4.8064,
      "step": 8169
    },
    {
      "epoch": 0.817,
      "grad_norm": 2.3790063858032227,
      "learning_rate": 1.9739069545810485e-05,
      "loss": 5.2074,
      "step": 8170
    },
    {
      "epoch": 0.8171,
      "grad_norm": 2.709153890609741,
      "learning_rate": 1.9718252480888566e-05,
      "loss": 4.6515,
      "step": 8171
    },
    {
      "epoch": 0.8172,
      "grad_norm": 6.553076267242432,
      "learning_rate": 1.969744519805402e-05,
      "loss": 4.7985,
      "step": 8172
    },
    {
      "epoch": 0.8173,
      "grad_norm": 3.2681431770324707,
      "learning_rate": 1.9676647699842242e-05,
      "loss": 4.7968,
      "step": 8173
    },
    {
      "epoch": 0.8174,
      "grad_norm": 2.429187059402466,
      "learning_rate": 1.965585998878724e-05,
      "loss": 4.6987,
      "step": 8174
    },
    {
      "epoch": 0.8175,
      "grad_norm": 3.006871223449707,
      "learning_rate": 1.963508206742202e-05,
      "loss": 4.4301,
      "step": 8175
    },
    {
      "epoch": 0.8176,
      "grad_norm": 1.9208670854568481,
      "learning_rate": 1.9614313938278272e-05,
      "loss": 5.0444,
      "step": 8176
    },
    {
      "epoch": 0.8177,
      "grad_norm": 3.5974814891815186,
      "learning_rate": 1.9593555603886538e-05,
      "loss": 5.4089,
      "step": 8177
    },
    {
      "epoch": 0.8178,
      "grad_norm": 3.2078726291656494,
      "learning_rate": 1.9572807066776143e-05,
      "loss": 4.9405,
      "step": 8178
    },
    {
      "epoch": 0.8179,
      "grad_norm": 3.046823024749756,
      "learning_rate": 1.955206832947526e-05,
      "loss": 4.8544,
      "step": 8179
    },
    {
      "epoch": 0.818,
      "grad_norm": 2.331585645675659,
      "learning_rate": 1.9531339394510827e-05,
      "loss": 4.7138,
      "step": 8180
    },
    {
      "epoch": 0.8181,
      "grad_norm": 2.9424030780792236,
      "learning_rate": 1.9510620264408596e-05,
      "loss": 5.031,
      "step": 8181
    },
    {
      "epoch": 0.8182,
      "grad_norm": 2.713698625564575,
      "learning_rate": 1.9489910941693133e-05,
      "loss": 4.3722,
      "step": 8182
    },
    {
      "epoch": 0.8183,
      "grad_norm": 1.9690908193588257,
      "learning_rate": 1.946921142888781e-05,
      "loss": 4.3704,
      "step": 8183
    },
    {
      "epoch": 0.8184,
      "grad_norm": 2.3002092838287354,
      "learning_rate": 1.9448521728514802e-05,
      "loss": 4.5055,
      "step": 8184
    },
    {
      "epoch": 0.8185,
      "grad_norm": 3.492933750152588,
      "learning_rate": 1.9427841843095063e-05,
      "loss": 5.1587,
      "step": 8185
    },
    {
      "epoch": 0.8186,
      "grad_norm": 2.3009517192840576,
      "learning_rate": 1.9407171775148436e-05,
      "loss": 4.8087,
      "step": 8186
    },
    {
      "epoch": 0.8187,
      "grad_norm": 4.8885955810546875,
      "learning_rate": 1.938651152719344e-05,
      "loss": 6.4324,
      "step": 8187
    },
    {
      "epoch": 0.8188,
      "grad_norm": 2.0979692935943604,
      "learning_rate": 1.9365861101747485e-05,
      "loss": 4.5065,
      "step": 8188
    },
    {
      "epoch": 0.8189,
      "grad_norm": 2.0531766414642334,
      "learning_rate": 1.9345220501326777e-05,
      "loss": 4.9047,
      "step": 8189
    },
    {
      "epoch": 0.819,
      "grad_norm": 3.38999605178833,
      "learning_rate": 1.9324589728446262e-05,
      "loss": 4.5279,
      "step": 8190
    },
    {
      "epoch": 0.8191,
      "grad_norm": 2.69038987159729,
      "learning_rate": 1.930396878561983e-05,
      "loss": 4.9637,
      "step": 8191
    },
    {
      "epoch": 0.8192,
      "grad_norm": 3.837857484817505,
      "learning_rate": 1.928335767535997e-05,
      "loss": 4.7605,
      "step": 8192
    },
    {
      "epoch": 0.8193,
      "grad_norm": 2.5155506134033203,
      "learning_rate": 1.9262756400178162e-05,
      "loss": 4.6276,
      "step": 8193
    },
    {
      "epoch": 0.8194,
      "grad_norm": 2.3777689933776855,
      "learning_rate": 1.9242164962584618e-05,
      "loss": 4.6162,
      "step": 8194
    },
    {
      "epoch": 0.8195,
      "grad_norm": 2.0888774394989014,
      "learning_rate": 1.922158336508825e-05,
      "loss": 4.4998,
      "step": 8195
    },
    {
      "epoch": 0.8196,
      "grad_norm": 5.830031871795654,
      "learning_rate": 1.9201011610196973e-05,
      "loss": 5.5088,
      "step": 8196
    },
    {
      "epoch": 0.8197,
      "grad_norm": 2.132692813873291,
      "learning_rate": 1.918044970041729e-05,
      "loss": 4.2771,
      "step": 8197
    },
    {
      "epoch": 0.8198,
      "grad_norm": 3.646813154220581,
      "learning_rate": 1.91598976382547e-05,
      "loss": 5.403,
      "step": 8198
    },
    {
      "epoch": 0.8199,
      "grad_norm": 6.019252300262451,
      "learning_rate": 1.9139355426213347e-05,
      "loss": 5.6305,
      "step": 8199
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.78515625,
      "learning_rate": 1.9118823066796276e-05,
      "loss": 4.5231,
      "step": 8200
    },
    {
      "epoch": 0.8201,
      "grad_norm": 2.1479365825653076,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 4.2887,
      "step": 8201
    },
    {
      "epoch": 0.8202,
      "grad_norm": 2.1652235984802246,
      "learning_rate": 1.9077787915840927e-05,
      "loss": 4.4079,
      "step": 8202
    },
    {
      "epoch": 0.8203,
      "grad_norm": 4.416387557983398,
      "learning_rate": 1.9057285129302683e-05,
      "loss": 3.9272,
      "step": 8203
    },
    {
      "epoch": 0.8204,
      "grad_norm": 4.24630880355835,
      "learning_rate": 1.903679220538871e-05,
      "loss": 4.4252,
      "step": 8204
    },
    {
      "epoch": 0.8205,
      "grad_norm": 9.325063705444336,
      "learning_rate": 1.9016309146596023e-05,
      "loss": 5.0883,
      "step": 8205
    },
    {
      "epoch": 0.8206,
      "grad_norm": 2.417182445526123,
      "learning_rate": 1.8995835955420416e-05,
      "loss": 4.5593,
      "step": 8206
    },
    {
      "epoch": 0.8207,
      "grad_norm": 3.868600845336914,
      "learning_rate": 1.897537263435648e-05,
      "loss": 4.6371,
      "step": 8207
    },
    {
      "epoch": 0.8208,
      "grad_norm": 6.806074619293213,
      "learning_rate": 1.895491918589759e-05,
      "loss": 5.1776,
      "step": 8208
    },
    {
      "epoch": 0.8209,
      "grad_norm": 3.9865355491638184,
      "learning_rate": 1.8934475612536017e-05,
      "loss": 4.3188,
      "step": 8209
    },
    {
      "epoch": 0.821,
      "grad_norm": 2.6623194217681885,
      "learning_rate": 1.891404191676265e-05,
      "loss": 4.8101,
      "step": 8210
    },
    {
      "epoch": 0.8211,
      "grad_norm": 3.1274220943450928,
      "learning_rate": 1.8893618101067355e-05,
      "loss": 4.9336,
      "step": 8211
    },
    {
      "epoch": 0.8212,
      "grad_norm": 2.68304181098938,
      "learning_rate": 1.8873204167938653e-05,
      "loss": 4.8277,
      "step": 8212
    },
    {
      "epoch": 0.8213,
      "grad_norm": 7.940675258636475,
      "learning_rate": 1.885280011986391e-05,
      "loss": 4.8792,
      "step": 8213
    },
    {
      "epoch": 0.8214,
      "grad_norm": 4.8603034019470215,
      "learning_rate": 1.883240595932938e-05,
      "loss": 4.7783,
      "step": 8214
    },
    {
      "epoch": 0.8215,
      "grad_norm": 2.3079640865325928,
      "learning_rate": 1.8812021688819915e-05,
      "loss": 4.8152,
      "step": 8215
    },
    {
      "epoch": 0.8216,
      "grad_norm": 3.026264190673828,
      "learning_rate": 1.879164731081937e-05,
      "loss": 4.5128,
      "step": 8216
    },
    {
      "epoch": 0.8217,
      "grad_norm": 2.688722848892212,
      "learning_rate": 1.877128282781028e-05,
      "loss": 4.4833,
      "step": 8217
    },
    {
      "epoch": 0.8218,
      "grad_norm": 2.238301992416382,
      "learning_rate": 1.8750928242273968e-05,
      "loss": 4.5149,
      "step": 8218
    },
    {
      "epoch": 0.8219,
      "grad_norm": 2.024993658065796,
      "learning_rate": 1.8730583556690605e-05,
      "loss": 4.3784,
      "step": 8219
    },
    {
      "epoch": 0.822,
      "grad_norm": 3.12314772605896,
      "learning_rate": 1.8710248773539118e-05,
      "loss": 4.7319,
      "step": 8220
    },
    {
      "epoch": 0.8221,
      "grad_norm": 2.396855354309082,
      "learning_rate": 1.8689923895297245e-05,
      "loss": 4.6354,
      "step": 8221
    },
    {
      "epoch": 0.8222,
      "grad_norm": 3.0408236980438232,
      "learning_rate": 1.8669608924441496e-05,
      "loss": 4.9982,
      "step": 8222
    },
    {
      "epoch": 0.8223,
      "grad_norm": 2.3709917068481445,
      "learning_rate": 1.8649303863447198e-05,
      "loss": 4.8677,
      "step": 8223
    },
    {
      "epoch": 0.8224,
      "grad_norm": 3.0860648155212402,
      "learning_rate": 1.8629008714788464e-05,
      "loss": 4.931,
      "step": 8224
    },
    {
      "epoch": 0.8225,
      "grad_norm": 2.5329437255859375,
      "learning_rate": 1.8608723480938206e-05,
      "loss": 4.7932,
      "step": 8225
    },
    {
      "epoch": 0.8226,
      "grad_norm": 4.653234481811523,
      "learning_rate": 1.858844816436809e-05,
      "loss": 5.5151,
      "step": 8226
    },
    {
      "epoch": 0.8227,
      "grad_norm": 3.0877108573913574,
      "learning_rate": 1.8568182767548626e-05,
      "loss": 4.6939,
      "step": 8227
    },
    {
      "epoch": 0.8228,
      "grad_norm": 2.0997772216796875,
      "learning_rate": 1.854792729294905e-05,
      "loss": 4.3732,
      "step": 8228
    },
    {
      "epoch": 0.8229,
      "grad_norm": 3.482410430908203,
      "learning_rate": 1.852768174303752e-05,
      "loss": 4.1067,
      "step": 8229
    },
    {
      "epoch": 0.823,
      "grad_norm": 2.320124387741089,
      "learning_rate": 1.8507446120280814e-05,
      "loss": 4.5579,
      "step": 8230
    },
    {
      "epoch": 0.8231,
      "grad_norm": 2.973386764526367,
      "learning_rate": 1.848722042714457e-05,
      "loss": 4.7992,
      "step": 8231
    },
    {
      "epoch": 0.8232,
      "grad_norm": 3.099205732345581,
      "learning_rate": 1.8467004666093325e-05,
      "loss": 4.6166,
      "step": 8232
    },
    {
      "epoch": 0.8233,
      "grad_norm": 2.9469547271728516,
      "learning_rate": 1.8446798839590186e-05,
      "loss": 4.8207,
      "step": 8233
    },
    {
      "epoch": 0.8234,
      "grad_norm": 2.247553586959839,
      "learning_rate": 1.8426602950097283e-05,
      "loss": 4.589,
      "step": 8234
    },
    {
      "epoch": 0.8235,
      "grad_norm": 4.416903972625732,
      "learning_rate": 1.8406417000075325e-05,
      "loss": 4.5421,
      "step": 8235
    },
    {
      "epoch": 0.8236,
      "grad_norm": 3.163965940475464,
      "learning_rate": 1.838624099198397e-05,
      "loss": 4.659,
      "step": 8236
    },
    {
      "epoch": 0.8237,
      "grad_norm": 2.7330026626586914,
      "learning_rate": 1.8366074928281607e-05,
      "loss": 4.6595,
      "step": 8237
    },
    {
      "epoch": 0.8238,
      "grad_norm": 3.5846099853515625,
      "learning_rate": 1.834591881142538e-05,
      "loss": 4.7892,
      "step": 8238
    },
    {
      "epoch": 0.8239,
      "grad_norm": 2.730104446411133,
      "learning_rate": 1.8325772643871265e-05,
      "loss": 4.8683,
      "step": 8239
    },
    {
      "epoch": 0.824,
      "grad_norm": 2.86130690574646,
      "learning_rate": 1.8305636428074014e-05,
      "loss": 4.2571,
      "step": 8240
    },
    {
      "epoch": 0.8241,
      "grad_norm": 2.6172447204589844,
      "learning_rate": 1.8285510166487152e-05,
      "loss": 5.0152,
      "step": 8241
    },
    {
      "epoch": 0.8242,
      "grad_norm": 1.9560370445251465,
      "learning_rate": 1.826539386156302e-05,
      "loss": 4.1873,
      "step": 8242
    },
    {
      "epoch": 0.8243,
      "grad_norm": 2.493773937225342,
      "learning_rate": 1.8245287515752708e-05,
      "loss": 5.1643,
      "step": 8243
    },
    {
      "epoch": 0.8244,
      "grad_norm": 2.7644991874694824,
      "learning_rate": 1.8225191131506126e-05,
      "loss": 4.592,
      "step": 8244
    },
    {
      "epoch": 0.8245,
      "grad_norm": 2.281520366668701,
      "learning_rate": 1.8205104711271957e-05,
      "loss": 5.0988,
      "step": 8245
    },
    {
      "epoch": 0.8246,
      "grad_norm": 3.530271530151367,
      "learning_rate": 1.818502825749764e-05,
      "loss": 5.0739,
      "step": 8246
    },
    {
      "epoch": 0.8247,
      "grad_norm": 1.8412647247314453,
      "learning_rate": 1.816496177262952e-05,
      "loss": 4.2646,
      "step": 8247
    },
    {
      "epoch": 0.8248,
      "grad_norm": 3.226578950881958,
      "learning_rate": 1.8144905259112542e-05,
      "loss": 4.9212,
      "step": 8248
    },
    {
      "epoch": 0.8249,
      "grad_norm": 2.4646053314208984,
      "learning_rate": 1.812485871939056e-05,
      "loss": 4.81,
      "step": 8249
    },
    {
      "epoch": 0.825,
      "grad_norm": 3.270296812057495,
      "learning_rate": 1.81048221559062e-05,
      "loss": 4.6733,
      "step": 8250
    },
    {
      "epoch": 0.8251,
      "grad_norm": 2.2640187740325928,
      "learning_rate": 1.808479557110081e-05,
      "loss": 4.7572,
      "step": 8251
    },
    {
      "epoch": 0.8252,
      "grad_norm": 3.0268778800964355,
      "learning_rate": 1.8064778967414662e-05,
      "loss": 4.819,
      "step": 8252
    },
    {
      "epoch": 0.8253,
      "grad_norm": 3.222177743911743,
      "learning_rate": 1.804477234728661e-05,
      "loss": 4.7262,
      "step": 8253
    },
    {
      "epoch": 0.8254,
      "grad_norm": 2.2558934688568115,
      "learning_rate": 1.8024775713154473e-05,
      "loss": 4.5376,
      "step": 8254
    },
    {
      "epoch": 0.8255,
      "grad_norm": 4.133777141571045,
      "learning_rate": 1.8004789067454764e-05,
      "loss": 5.2302,
      "step": 8255
    },
    {
      "epoch": 0.8256,
      "grad_norm": 6.181481838226318,
      "learning_rate": 1.7984812412622787e-05,
      "loss": 5.2395,
      "step": 8256
    },
    {
      "epoch": 0.8257,
      "grad_norm": 7.38854455947876,
      "learning_rate": 1.7964845751092664e-05,
      "loss": 5.269,
      "step": 8257
    },
    {
      "epoch": 0.8258,
      "grad_norm": 2.4870057106018066,
      "learning_rate": 1.794488908529719e-05,
      "loss": 4.2725,
      "step": 8258
    },
    {
      "epoch": 0.8259,
      "grad_norm": 5.325161933898926,
      "learning_rate": 1.792494241766811e-05,
      "loss": 5.0597,
      "step": 8259
    },
    {
      "epoch": 0.826,
      "grad_norm": 2.6873137950897217,
      "learning_rate": 1.790500575063584e-05,
      "loss": 4.6638,
      "step": 8260
    },
    {
      "epoch": 0.8261,
      "grad_norm": 1.793818712234497,
      "learning_rate": 1.78850790866296e-05,
      "loss": 4.6593,
      "step": 8261
    },
    {
      "epoch": 0.8262,
      "grad_norm": 3.2518134117126465,
      "learning_rate": 1.7865162428077386e-05,
      "loss": 5.0411,
      "step": 8262
    },
    {
      "epoch": 0.8263,
      "grad_norm": 2.453263521194458,
      "learning_rate": 1.7845255777406e-05,
      "loss": 4.5165,
      "step": 8263
    },
    {
      "epoch": 0.8264,
      "grad_norm": 5.073136806488037,
      "learning_rate": 1.7825359137040988e-05,
      "loss": 5.0659,
      "step": 8264
    },
    {
      "epoch": 0.8265,
      "grad_norm": 9.441554069519043,
      "learning_rate": 1.7805472509406696e-05,
      "loss": 5.0096,
      "step": 8265
    },
    {
      "epoch": 0.8266,
      "grad_norm": 2.4946236610412598,
      "learning_rate": 1.7785595896926265e-05,
      "loss": 4.6843,
      "step": 8266
    },
    {
      "epoch": 0.8267,
      "grad_norm": 5.5865478515625,
      "learning_rate": 1.7765729302021596e-05,
      "loss": 4.4685,
      "step": 8267
    },
    {
      "epoch": 0.8268,
      "grad_norm": 2.6050920486450195,
      "learning_rate": 1.7745872727113356e-05,
      "loss": 4.7487,
      "step": 8268
    },
    {
      "epoch": 0.8269,
      "grad_norm": 2.3670618534088135,
      "learning_rate": 1.7726026174621003e-05,
      "loss": 4.5792,
      "step": 8269
    },
    {
      "epoch": 0.827,
      "grad_norm": 3.769702672958374,
      "learning_rate": 1.7706189646962847e-05,
      "loss": 4.5784,
      "step": 8270
    },
    {
      "epoch": 0.8271,
      "grad_norm": 4.419459819793701,
      "learning_rate": 1.7686363146555805e-05,
      "loss": 5.3739,
      "step": 8271
    },
    {
      "epoch": 0.8272,
      "grad_norm": 2.5005664825439453,
      "learning_rate": 1.7666546675815778e-05,
      "loss": 4.5898,
      "step": 8272
    },
    {
      "epoch": 0.8273,
      "grad_norm": 2.104174852371216,
      "learning_rate": 1.7646740237157256e-05,
      "loss": 4.9351,
      "step": 8273
    },
    {
      "epoch": 0.8274,
      "grad_norm": 1.957179069519043,
      "learning_rate": 1.7626943832993647e-05,
      "loss": 4.4306,
      "step": 8274
    },
    {
      "epoch": 0.8275,
      "grad_norm": 1.8974186182022095,
      "learning_rate": 1.760715746573709e-05,
      "loss": 4.8481,
      "step": 8275
    },
    {
      "epoch": 0.8276,
      "grad_norm": 2.7487504482269287,
      "learning_rate": 1.7587381137798432e-05,
      "loss": 4.7387,
      "step": 8276
    },
    {
      "epoch": 0.8277,
      "grad_norm": 2.3167998790740967,
      "learning_rate": 1.7567614851587443e-05,
      "loss": 4.8265,
      "step": 8277
    },
    {
      "epoch": 0.8278,
      "grad_norm": 2.7892723083496094,
      "learning_rate": 1.754785860951249e-05,
      "loss": 4.3233,
      "step": 8278
    },
    {
      "epoch": 0.8279,
      "grad_norm": 2.3880770206451416,
      "learning_rate": 1.752811241398089e-05,
      "loss": 5.3169,
      "step": 8279
    },
    {
      "epoch": 0.828,
      "grad_norm": 5.579233169555664,
      "learning_rate": 1.750837626739863e-05,
      "loss": 5.2885,
      "step": 8280
    },
    {
      "epoch": 0.8281,
      "grad_norm": 2.2881596088409424,
      "learning_rate": 1.7488650172170496e-05,
      "loss": 5.136,
      "step": 8281
    },
    {
      "epoch": 0.8282,
      "grad_norm": 3.180952310562134,
      "learning_rate": 1.7468934130700044e-05,
      "loss": 4.4786,
      "step": 8282
    },
    {
      "epoch": 0.8283,
      "grad_norm": 3.4376533031463623,
      "learning_rate": 1.744922814538964e-05,
      "loss": 5.1556,
      "step": 8283
    },
    {
      "epoch": 0.8284,
      "grad_norm": 2.949368476867676,
      "learning_rate": 1.7429532218640377e-05,
      "loss": 4.6543,
      "step": 8284
    },
    {
      "epoch": 0.8285,
      "grad_norm": 3.132915735244751,
      "learning_rate": 1.7409846352852143e-05,
      "loss": 4.7849,
      "step": 8285
    },
    {
      "epoch": 0.8286,
      "grad_norm": 2.9997682571411133,
      "learning_rate": 1.7390170550423625e-05,
      "loss": 4.8396,
      "step": 8286
    },
    {
      "epoch": 0.8287,
      "grad_norm": 7.520700931549072,
      "learning_rate": 1.737050481375223e-05,
      "loss": 6.9267,
      "step": 8287
    },
    {
      "epoch": 0.8288,
      "grad_norm": 3.543893337249756,
      "learning_rate": 1.7350849145234184e-05,
      "loss": 4.6943,
      "step": 8288
    },
    {
      "epoch": 0.8289,
      "grad_norm": 2.5301175117492676,
      "learning_rate": 1.7331203547264453e-05,
      "loss": 4.8095,
      "step": 8289
    },
    {
      "epoch": 0.829,
      "grad_norm": 4.141503810882568,
      "learning_rate": 1.7311568022236845e-05,
      "loss": 4.9724,
      "step": 8290
    },
    {
      "epoch": 0.8291,
      "grad_norm": 2.719101905822754,
      "learning_rate": 1.7291942572543807e-05,
      "loss": 4.7722,
      "step": 8291
    },
    {
      "epoch": 0.8292,
      "grad_norm": 2.9215264320373535,
      "learning_rate": 1.7272327200576742e-05,
      "loss": 5.2529,
      "step": 8292
    },
    {
      "epoch": 0.8293,
      "grad_norm": 2.3794264793395996,
      "learning_rate": 1.7252721908725632e-05,
      "loss": 4.7604,
      "step": 8293
    },
    {
      "epoch": 0.8294,
      "grad_norm": 2.867161273956299,
      "learning_rate": 1.7233126699379343e-05,
      "loss": 4.9087,
      "step": 8294
    },
    {
      "epoch": 0.8295,
      "grad_norm": 2.3504581451416016,
      "learning_rate": 1.721354157492555e-05,
      "loss": 4.8145,
      "step": 8295
    },
    {
      "epoch": 0.8296,
      "grad_norm": 4.067580223083496,
      "learning_rate": 1.719396653775056e-05,
      "loss": 4.762,
      "step": 8296
    },
    {
      "epoch": 0.8297,
      "grad_norm": 5.420118808746338,
      "learning_rate": 1.7174401590239587e-05,
      "loss": 5.2405,
      "step": 8297
    },
    {
      "epoch": 0.8298,
      "grad_norm": 2.0480000972747803,
      "learning_rate": 1.715484673477654e-05,
      "loss": 4.8179,
      "step": 8298
    },
    {
      "epoch": 0.8299,
      "grad_norm": 3.862480878829956,
      "learning_rate": 1.7135301973744124e-05,
      "loss": 5.1998,
      "step": 8299
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.3174545764923096,
      "learning_rate": 1.7115767309523812e-05,
      "loss": 4.5484,
      "step": 8300
    },
    {
      "epoch": 0.8301,
      "grad_norm": 2.4578161239624023,
      "learning_rate": 1.7096242744495837e-05,
      "loss": 4.6818,
      "step": 8301
    },
    {
      "epoch": 0.8302,
      "grad_norm": 2.1117496490478516,
      "learning_rate": 1.70767282810392e-05,
      "loss": 4.524,
      "step": 8302
    },
    {
      "epoch": 0.8303,
      "grad_norm": 2.2208144664764404,
      "learning_rate": 1.7057223921531707e-05,
      "loss": 5.0991,
      "step": 8303
    },
    {
      "epoch": 0.8304,
      "grad_norm": 1.70479154586792,
      "learning_rate": 1.7037729668349877e-05,
      "loss": 4.2723,
      "step": 8304
    },
    {
      "epoch": 0.8305,
      "grad_norm": 4.044407367706299,
      "learning_rate": 1.7018245523869036e-05,
      "loss": 4.2038,
      "step": 8305
    },
    {
      "epoch": 0.8306,
      "grad_norm": 4.37038516998291,
      "learning_rate": 1.6998771490463262e-05,
      "loss": 5.4768,
      "step": 8306
    },
    {
      "epoch": 0.8307,
      "grad_norm": 3.4458329677581787,
      "learning_rate": 1.697930757050542e-05,
      "loss": 4.9759,
      "step": 8307
    },
    {
      "epoch": 0.8308,
      "grad_norm": 3.813075065612793,
      "learning_rate": 1.6959853766367118e-05,
      "loss": 4.8002,
      "step": 8308
    },
    {
      "epoch": 0.8309,
      "grad_norm": 2.1652321815490723,
      "learning_rate": 1.6940410080418723e-05,
      "loss": 4.3412,
      "step": 8309
    },
    {
      "epoch": 0.831,
      "grad_norm": 2.2876217365264893,
      "learning_rate": 1.6920976515029463e-05,
      "loss": 4.3462,
      "step": 8310
    },
    {
      "epoch": 0.8311,
      "grad_norm": 2.751765012741089,
      "learning_rate": 1.690155307256719e-05,
      "loss": 4.8085,
      "step": 8311
    },
    {
      "epoch": 0.8312,
      "grad_norm": 3.100776433944702,
      "learning_rate": 1.6882139755398574e-05,
      "loss": 4.5001,
      "step": 8312
    },
    {
      "epoch": 0.8313,
      "grad_norm": 2.6970081329345703,
      "learning_rate": 1.686273656588917e-05,
      "loss": 4.9319,
      "step": 8313
    },
    {
      "epoch": 0.8314,
      "grad_norm": 2.671107053756714,
      "learning_rate": 1.6843343506403075e-05,
      "loss": 4.8766,
      "step": 8314
    },
    {
      "epoch": 0.8315,
      "grad_norm": 2.4603111743927,
      "learning_rate": 1.6823960579303377e-05,
      "loss": 4.7241,
      "step": 8315
    },
    {
      "epoch": 0.8316,
      "grad_norm": 5.4033308029174805,
      "learning_rate": 1.680458778695174e-05,
      "loss": 4.5297,
      "step": 8316
    },
    {
      "epoch": 0.8317,
      "grad_norm": 1.984423041343689,
      "learning_rate": 1.678522513170875e-05,
      "loss": 4.1945,
      "step": 8317
    },
    {
      "epoch": 0.8318,
      "grad_norm": 4.279323577880859,
      "learning_rate": 1.6765872615933677e-05,
      "loss": 5.1453,
      "step": 8318
    },
    {
      "epoch": 0.8319,
      "grad_norm": 2.2332332134246826,
      "learning_rate": 1.6746530241984504e-05,
      "loss": 4.6889,
      "step": 8319
    },
    {
      "epoch": 0.832,
      "grad_norm": 3.9556894302368164,
      "learning_rate": 1.6727198012218114e-05,
      "loss": 4.3576,
      "step": 8320
    },
    {
      "epoch": 0.8321,
      "grad_norm": 2.6696743965148926,
      "learning_rate": 1.6707875928990058e-05,
      "loss": 4.4001,
      "step": 8321
    },
    {
      "epoch": 0.8322,
      "grad_norm": 2.425154447555542,
      "learning_rate": 1.668856399465466e-05,
      "loss": 4.4505,
      "step": 8322
    },
    {
      "epoch": 0.8323,
      "grad_norm": 1.8491668701171875,
      "learning_rate": 1.6669262211565028e-05,
      "loss": 4.5433,
      "step": 8323
    },
    {
      "epoch": 0.8324,
      "grad_norm": 2.21916127204895,
      "learning_rate": 1.6649970582073028e-05,
      "loss": 4.5625,
      "step": 8324
    },
    {
      "epoch": 0.8325,
      "grad_norm": 2.414166212081909,
      "learning_rate": 1.6630689108529284e-05,
      "loss": 4.2214,
      "step": 8325
    },
    {
      "epoch": 0.8326,
      "grad_norm": 3.3529281616210938,
      "learning_rate": 1.661141779328319e-05,
      "loss": 4.7018,
      "step": 8326
    },
    {
      "epoch": 0.8327,
      "grad_norm": 3.0511839389801025,
      "learning_rate": 1.6592156638682886e-05,
      "loss": 4.6499,
      "step": 8327
    },
    {
      "epoch": 0.8328,
      "grad_norm": 2.1411778926849365,
      "learning_rate": 1.6572905647075298e-05,
      "loss": 4.3512,
      "step": 8328
    },
    {
      "epoch": 0.8329,
      "grad_norm": 2.509873867034912,
      "learning_rate": 1.6553664820806102e-05,
      "loss": 5.1102,
      "step": 8329
    },
    {
      "epoch": 0.833,
      "grad_norm": 2.539642810821533,
      "learning_rate": 1.6534434162219727e-05,
      "loss": 5.1439,
      "step": 8330
    },
    {
      "epoch": 0.8331,
      "grad_norm": 2.0563361644744873,
      "learning_rate": 1.6515213673659357e-05,
      "loss": 4.5256,
      "step": 8331
    },
    {
      "epoch": 0.8332,
      "grad_norm": 3.574312925338745,
      "learning_rate": 1.649600335746695e-05,
      "loss": 4.4283,
      "step": 8332
    },
    {
      "epoch": 0.8333,
      "grad_norm": 4.659883975982666,
      "learning_rate": 1.6476803215983294e-05,
      "loss": 4.7724,
      "step": 8333
    },
    {
      "epoch": 0.8334,
      "grad_norm": 2.664025068283081,
      "learning_rate": 1.6457613251547754e-05,
      "loss": 4.6867,
      "step": 8334
    },
    {
      "epoch": 0.8335,
      "grad_norm": 3.000577926635742,
      "learning_rate": 1.643843346649866e-05,
      "loss": 5.1923,
      "step": 8335
    },
    {
      "epoch": 0.8336,
      "grad_norm": 3.4110267162323,
      "learning_rate": 1.6419263863172997e-05,
      "loss": 4.4194,
      "step": 8336
    },
    {
      "epoch": 0.8337,
      "grad_norm": 4.433845043182373,
      "learning_rate": 1.640010444390646e-05,
      "loss": 4.6284,
      "step": 8337
    },
    {
      "epoch": 0.8338,
      "grad_norm": 4.381348133087158,
      "learning_rate": 1.6380955211033656e-05,
      "loss": 4.5202,
      "step": 8338
    },
    {
      "epoch": 0.8339,
      "grad_norm": 2.397902250289917,
      "learning_rate": 1.6361816166887768e-05,
      "loss": 5.0919,
      "step": 8339
    },
    {
      "epoch": 0.834,
      "grad_norm": 3.588691473007202,
      "learning_rate": 1.634268731380091e-05,
      "loss": 5.3431,
      "step": 8340
    },
    {
      "epoch": 0.8341,
      "grad_norm": 2.6990954875946045,
      "learning_rate": 1.632356865410384e-05,
      "loss": 5.0274,
      "step": 8341
    },
    {
      "epoch": 0.8342,
      "grad_norm": 2.478534460067749,
      "learning_rate": 1.6304460190126103e-05,
      "loss": 4.4793,
      "step": 8342
    },
    {
      "epoch": 0.8343,
      "grad_norm": 2.389993906021118,
      "learning_rate": 1.628536192419603e-05,
      "loss": 4.7114,
      "step": 8343
    },
    {
      "epoch": 0.8344,
      "grad_norm": 2.4070374965667725,
      "learning_rate": 1.6266273858640656e-05,
      "loss": 4.8908,
      "step": 8344
    },
    {
      "epoch": 0.8345,
      "grad_norm": 3.3367557525634766,
      "learning_rate": 1.6247195995785837e-05,
      "loss": 4.5976,
      "step": 8345
    },
    {
      "epoch": 0.8346,
      "grad_norm": 3.6356613636016846,
      "learning_rate": 1.622812833795613e-05,
      "loss": 4.6389,
      "step": 8346
    },
    {
      "epoch": 0.8347,
      "grad_norm": 3.9304018020629883,
      "learning_rate": 1.6209070887474876e-05,
      "loss": 5.1114,
      "step": 8347
    },
    {
      "epoch": 0.8348,
      "grad_norm": 2.330498218536377,
      "learning_rate": 1.6190023646664175e-05,
      "loss": 4.8766,
      "step": 8348
    },
    {
      "epoch": 0.8349,
      "grad_norm": 3.7238361835479736,
      "learning_rate": 1.6170986617844863e-05,
      "loss": 4.9319,
      "step": 8349
    },
    {
      "epoch": 0.835,
      "grad_norm": 2.584528923034668,
      "learning_rate": 1.6151959803336535e-05,
      "loss": 4.8582,
      "step": 8350
    },
    {
      "epoch": 0.8351,
      "grad_norm": 2.148691177368164,
      "learning_rate": 1.6132943205457606e-05,
      "loss": 4.2487,
      "step": 8351
    },
    {
      "epoch": 0.8352,
      "grad_norm": 3.3975648880004883,
      "learning_rate": 1.611393682652511e-05,
      "loss": 4.7609,
      "step": 8352
    },
    {
      "epoch": 0.8353,
      "grad_norm": 2.4862828254699707,
      "learning_rate": 1.6094940668855006e-05,
      "loss": 4.4282,
      "step": 8353
    },
    {
      "epoch": 0.8354,
      "grad_norm": 2.4736266136169434,
      "learning_rate": 1.6075954734761845e-05,
      "loss": 4.4691,
      "step": 8354
    },
    {
      "epoch": 0.8355,
      "grad_norm": 2.842794418334961,
      "learning_rate": 1.6056979026559004e-05,
      "loss": 3.968,
      "step": 8355
    },
    {
      "epoch": 0.8356,
      "grad_norm": 3.034294843673706,
      "learning_rate": 1.6038013546558696e-05,
      "loss": 4.6887,
      "step": 8356
    },
    {
      "epoch": 0.8357,
      "grad_norm": 4.481576919555664,
      "learning_rate": 1.601905829707171e-05,
      "loss": 5.0004,
      "step": 8357
    },
    {
      "epoch": 0.8358,
      "grad_norm": 1.9879233837127686,
      "learning_rate": 1.600011328040777e-05,
      "loss": 4.6695,
      "step": 8358
    },
    {
      "epoch": 0.8359,
      "grad_norm": 2.4367964267730713,
      "learning_rate": 1.598117849887518e-05,
      "loss": 4.5221,
      "step": 8359
    },
    {
      "epoch": 0.836,
      "grad_norm": 2.820342779159546,
      "learning_rate": 1.596225395478116e-05,
      "loss": 5.0508,
      "step": 8360
    },
    {
      "epoch": 0.8361,
      "grad_norm": 2.539299488067627,
      "learning_rate": 1.5943339650431576e-05,
      "loss": 4.4821,
      "step": 8361
    },
    {
      "epoch": 0.8362,
      "grad_norm": 2.931351900100708,
      "learning_rate": 1.592443558813109e-05,
      "loss": 4.6653,
      "step": 8362
    },
    {
      "epoch": 0.8363,
      "grad_norm": 3.05306339263916,
      "learning_rate": 1.5905541770183096e-05,
      "loss": 5.0522,
      "step": 8363
    },
    {
      "epoch": 0.8364,
      "grad_norm": 5.011849880218506,
      "learning_rate": 1.588665819888976e-05,
      "loss": 4.4409,
      "step": 8364
    },
    {
      "epoch": 0.8365,
      "grad_norm": 3.7281811237335205,
      "learning_rate": 1.586778487655197e-05,
      "loss": 4.5035,
      "step": 8365
    },
    {
      "epoch": 0.8366,
      "grad_norm": 2.320030450820923,
      "learning_rate": 1.5848921805469397e-05,
      "loss": 4.8831,
      "step": 8366
    },
    {
      "epoch": 0.8367,
      "grad_norm": 2.541588544845581,
      "learning_rate": 1.5830068987940438e-05,
      "loss": 4.5921,
      "step": 8367
    },
    {
      "epoch": 0.8368,
      "grad_norm": 2.3169679641723633,
      "learning_rate": 1.581122642626226e-05,
      "loss": 4.7523,
      "step": 8368
    },
    {
      "epoch": 0.8369,
      "grad_norm": 3.2706212997436523,
      "learning_rate": 1.579239412273078e-05,
      "loss": 5.0783,
      "step": 8369
    },
    {
      "epoch": 0.837,
      "grad_norm": 1.9223201274871826,
      "learning_rate": 1.577357207964062e-05,
      "loss": 4.2612,
      "step": 8370
    },
    {
      "epoch": 0.8371,
      "grad_norm": 4.612407207489014,
      "learning_rate": 1.5754760299285252e-05,
      "loss": 4.8877,
      "step": 8371
    },
    {
      "epoch": 0.8372,
      "grad_norm": 3.0301241874694824,
      "learning_rate": 1.5735958783956794e-05,
      "loss": 4.7942,
      "step": 8372
    },
    {
      "epoch": 0.8373,
      "grad_norm": 2.6140060424804688,
      "learning_rate": 1.5717167535946144e-05,
      "loss": 4.354,
      "step": 8373
    },
    {
      "epoch": 0.8374,
      "grad_norm": 3.197247266769409,
      "learning_rate": 1.569838655754298e-05,
      "loss": 5.1631,
      "step": 8374
    },
    {
      "epoch": 0.8375,
      "grad_norm": 2.5974998474121094,
      "learning_rate": 1.567961585103567e-05,
      "loss": 4.2448,
      "step": 8375
    },
    {
      "epoch": 0.8376,
      "grad_norm": 2.05110764503479,
      "learning_rate": 1.566085541871145e-05,
      "loss": 4.3757,
      "step": 8376
    },
    {
      "epoch": 0.8377,
      "grad_norm": 4.433166980743408,
      "learning_rate": 1.564210526285612e-05,
      "loss": 4.9825,
      "step": 8377
    },
    {
      "epoch": 0.8378,
      "grad_norm": 3.914884567260742,
      "learning_rate": 1.5623365385754408e-05,
      "loss": 4.8446,
      "step": 8378
    },
    {
      "epoch": 0.8379,
      "grad_norm": 3.939854621887207,
      "learning_rate": 1.560463578968967e-05,
      "loss": 4.8217,
      "step": 8379
    },
    {
      "epoch": 0.838,
      "grad_norm": 2.152724504470825,
      "learning_rate": 1.5585916476944073e-05,
      "loss": 4.8014,
      "step": 8380
    },
    {
      "epoch": 0.8381,
      "grad_norm": 2.9439680576324463,
      "learning_rate": 1.5567207449798515e-05,
      "loss": 4.5717,
      "step": 8381
    },
    {
      "epoch": 0.8382,
      "grad_norm": 2.7080535888671875,
      "learning_rate": 1.5548508710532572e-05,
      "loss": 4.495,
      "step": 8382
    },
    {
      "epoch": 0.8383,
      "grad_norm": 2.7765471935272217,
      "learning_rate": 1.5529820261424698e-05,
      "loss": 4.7941,
      "step": 8383
    },
    {
      "epoch": 0.8384,
      "grad_norm": 2.2377915382385254,
      "learning_rate": 1.551114210475201e-05,
      "loss": 4.6639,
      "step": 8384
    },
    {
      "epoch": 0.8385,
      "grad_norm": 2.0982115268707275,
      "learning_rate": 1.5492474242790366e-05,
      "loss": 4.6626,
      "step": 8385
    },
    {
      "epoch": 0.8386,
      "grad_norm": 2.416977882385254,
      "learning_rate": 1.547381667781439e-05,
      "loss": 4.4363,
      "step": 8386
    },
    {
      "epoch": 0.8387,
      "grad_norm": 5.0876946449279785,
      "learning_rate": 1.545516941209747e-05,
      "loss": 4.2236,
      "step": 8387
    },
    {
      "epoch": 0.8388,
      "grad_norm": 5.614290714263916,
      "learning_rate": 1.54365324479117e-05,
      "loss": 4.756,
      "step": 8388
    },
    {
      "epoch": 0.8389,
      "grad_norm": 3.3475213050842285,
      "learning_rate": 1.541790578752794e-05,
      "loss": 5.0024,
      "step": 8389
    },
    {
      "epoch": 0.839,
      "grad_norm": 2.5527381896972656,
      "learning_rate": 1.539928943321579e-05,
      "loss": 4.3744,
      "step": 8390
    },
    {
      "epoch": 0.8391,
      "grad_norm": 4.533365726470947,
      "learning_rate": 1.538068338724361e-05,
      "loss": 4.7457,
      "step": 8391
    },
    {
      "epoch": 0.8392,
      "grad_norm": 2.133216381072998,
      "learning_rate": 1.5362087651878475e-05,
      "loss": 4.3921,
      "step": 8392
    },
    {
      "epoch": 0.8393,
      "grad_norm": 2.544677495956421,
      "learning_rate": 1.5343502229386207e-05,
      "loss": 4.6006,
      "step": 8393
    },
    {
      "epoch": 0.8394,
      "grad_norm": 3.8756237030029297,
      "learning_rate": 1.532492712203145e-05,
      "loss": 4.2133,
      "step": 8394
    },
    {
      "epoch": 0.8395,
      "grad_norm": 1.9210290908813477,
      "learning_rate": 1.530636233207743e-05,
      "loss": 4.6661,
      "step": 8395
    },
    {
      "epoch": 0.8396,
      "grad_norm": 2.6174938678741455,
      "learning_rate": 1.528780786178631e-05,
      "loss": 4.4425,
      "step": 8396
    },
    {
      "epoch": 0.8397,
      "grad_norm": 3.1927268505096436,
      "learning_rate": 1.526926371341878e-05,
      "loss": 4.8463,
      "step": 8397
    },
    {
      "epoch": 0.8398,
      "grad_norm": 2.1200826168060303,
      "learning_rate": 1.5250729889234482e-05,
      "loss": 4.4853,
      "step": 8398
    },
    {
      "epoch": 0.8399,
      "grad_norm": 4.39580774307251,
      "learning_rate": 1.5232206391491699e-05,
      "loss": 5.0783,
      "step": 8399
    },
    {
      "epoch": 0.84,
      "grad_norm": 2.0401320457458496,
      "learning_rate": 1.521369322244739e-05,
      "loss": 4.1426,
      "step": 8400
    },
    {
      "epoch": 0.8401,
      "grad_norm": 2.624584197998047,
      "learning_rate": 1.5195190384357404e-05,
      "loss": 5.0444,
      "step": 8401
    },
    {
      "epoch": 0.8402,
      "grad_norm": 2.885201930999756,
      "learning_rate": 1.5176697879476232e-05,
      "loss": 4.671,
      "step": 8402
    },
    {
      "epoch": 0.8403,
      "grad_norm": 2.5527687072753906,
      "learning_rate": 1.5158215710057123e-05,
      "loss": 4.6968,
      "step": 8403
    },
    {
      "epoch": 0.8404,
      "grad_norm": 4.099034786224365,
      "learning_rate": 1.5139743878352075e-05,
      "loss": 5.6867,
      "step": 8404
    },
    {
      "epoch": 0.8405,
      "grad_norm": 2.6020140647888184,
      "learning_rate": 1.5121282386611824e-05,
      "loss": 4.8126,
      "step": 8405
    },
    {
      "epoch": 0.8406,
      "grad_norm": 2.0498909950256348,
      "learning_rate": 1.5102831237085857e-05,
      "loss": 4.7479,
      "step": 8406
    },
    {
      "epoch": 0.8407,
      "grad_norm": 1.9450671672821045,
      "learning_rate": 1.5084390432022377e-05,
      "loss": 4.3872,
      "step": 8407
    },
    {
      "epoch": 0.8408,
      "grad_norm": 2.175541877746582,
      "learning_rate": 1.5065959973668353e-05,
      "loss": 4.4806,
      "step": 8408
    },
    {
      "epoch": 0.8409,
      "grad_norm": 2.178286075592041,
      "learning_rate": 1.5047539864269478e-05,
      "loss": 4.387,
      "step": 8409
    },
    {
      "epoch": 0.841,
      "grad_norm": 2.0876309871673584,
      "learning_rate": 1.5029130106070167e-05,
      "loss": 4.6668,
      "step": 8410
    },
    {
      "epoch": 0.8411,
      "grad_norm": 2.9254512786865234,
      "learning_rate": 1.5010730701313625e-05,
      "loss": 4.6312,
      "step": 8411
    },
    {
      "epoch": 0.8412,
      "grad_norm": 3.253906488418579,
      "learning_rate": 1.4992341652241737e-05,
      "loss": 4.8915,
      "step": 8412
    },
    {
      "epoch": 0.8413,
      "grad_norm": 2.6741437911987305,
      "learning_rate": 1.4973962961095134e-05,
      "loss": 4.2712,
      "step": 8413
    },
    {
      "epoch": 0.8414,
      "grad_norm": 3.162914514541626,
      "learning_rate": 1.495559463011329e-05,
      "loss": 4.7449,
      "step": 8414
    },
    {
      "epoch": 0.8415,
      "grad_norm": 4.383924961090088,
      "learning_rate": 1.4937236661534226e-05,
      "loss": 5.7002,
      "step": 8415
    },
    {
      "epoch": 0.8416,
      "grad_norm": 2.530860185623169,
      "learning_rate": 1.4918889057594876e-05,
      "loss": 4.9195,
      "step": 8416
    },
    {
      "epoch": 0.8417,
      "grad_norm": 3.501871109008789,
      "learning_rate": 1.4900551820530828e-05,
      "loss": 4.6415,
      "step": 8417
    },
    {
      "epoch": 0.8418,
      "grad_norm": 2.9891762733459473,
      "learning_rate": 1.4882224952576373e-05,
      "loss": 4.8109,
      "step": 8418
    },
    {
      "epoch": 0.8419,
      "grad_norm": 2.2939772605895996,
      "learning_rate": 1.486390845596466e-05,
      "loss": 4.5597,
      "step": 8419
    },
    {
      "epoch": 0.842,
      "grad_norm": 2.480377674102783,
      "learning_rate": 1.484560233292741e-05,
      "loss": 5.0181,
      "step": 8420
    },
    {
      "epoch": 0.8421,
      "grad_norm": 3.126617670059204,
      "learning_rate": 1.4827306585695234e-05,
      "loss": 4.3799,
      "step": 8421
    },
    {
      "epoch": 0.8422,
      "grad_norm": 3.949697732925415,
      "learning_rate": 1.4809021216497399e-05,
      "loss": 4.546,
      "step": 8422
    },
    {
      "epoch": 0.8423,
      "grad_norm": 2.7392776012420654,
      "learning_rate": 1.4790746227561925e-05,
      "loss": 4.3563,
      "step": 8423
    },
    {
      "epoch": 0.8424,
      "grad_norm": 3.7241575717926025,
      "learning_rate": 1.4772481621115541e-05,
      "loss": 5.016,
      "step": 8424
    },
    {
      "epoch": 0.8425,
      "grad_norm": 2.6908905506134033,
      "learning_rate": 1.4754227399383757e-05,
      "loss": 4.5806,
      "step": 8425
    },
    {
      "epoch": 0.8426,
      "grad_norm": 3.6295464038848877,
      "learning_rate": 1.4735983564590783e-05,
      "loss": 4.8233,
      "step": 8426
    },
    {
      "epoch": 0.8427,
      "grad_norm": 2.575174570083618,
      "learning_rate": 1.4717750118959584e-05,
      "loss": 4.4204,
      "step": 8427
    },
    {
      "epoch": 0.8428,
      "grad_norm": 2.372087240219116,
      "learning_rate": 1.4699527064711838e-05,
      "loss": 4.5144,
      "step": 8428
    },
    {
      "epoch": 0.8429,
      "grad_norm": 2.8439273834228516,
      "learning_rate": 1.468131440406798e-05,
      "loss": 4.6412,
      "step": 8429
    },
    {
      "epoch": 0.843,
      "grad_norm": 4.8302531242370605,
      "learning_rate": 1.466311213924716e-05,
      "loss": 4.4849,
      "step": 8430
    },
    {
      "epoch": 0.8431,
      "grad_norm": 2.0850470066070557,
      "learning_rate": 1.4644920272467244e-05,
      "loss": 4.4811,
      "step": 8431
    },
    {
      "epoch": 0.8432,
      "grad_norm": 3.0821125507354736,
      "learning_rate": 1.462673880594494e-05,
      "loss": 5.1093,
      "step": 8432
    },
    {
      "epoch": 0.8433,
      "grad_norm": 3.10746431350708,
      "learning_rate": 1.4608567741895495e-05,
      "loss": 4.975,
      "step": 8433
    },
    {
      "epoch": 0.8434,
      "grad_norm": 3.2377891540527344,
      "learning_rate": 1.4590407082533097e-05,
      "loss": 5.0278,
      "step": 8434
    },
    {
      "epoch": 0.8435,
      "grad_norm": 2.8008148670196533,
      "learning_rate": 1.4572256830070497e-05,
      "loss": 4.5095,
      "step": 8435
    },
    {
      "epoch": 0.8436,
      "grad_norm": 7.737863063812256,
      "learning_rate": 1.4554116986719257e-05,
      "loss": 4.8597,
      "step": 8436
    },
    {
      "epoch": 0.8437,
      "grad_norm": 3.0955026149749756,
      "learning_rate": 1.4535987554689712e-05,
      "loss": 4.5859,
      "step": 8437
    },
    {
      "epoch": 0.8438,
      "grad_norm": 1.9457995891571045,
      "learning_rate": 1.4517868536190803e-05,
      "loss": 4.297,
      "step": 8438
    },
    {
      "epoch": 0.8439,
      "grad_norm": 2.0632381439208984,
      "learning_rate": 1.4499759933430346e-05,
      "loss": 4.4786,
      "step": 8439
    },
    {
      "epoch": 0.844,
      "grad_norm": 1.9635549783706665,
      "learning_rate": 1.4481661748614784e-05,
      "loss": 4.3802,
      "step": 8440
    },
    {
      "epoch": 0.8441,
      "grad_norm": 2.859548330307007,
      "learning_rate": 1.4463573983949341e-05,
      "loss": 4.4215,
      "step": 8441
    },
    {
      "epoch": 0.8442,
      "grad_norm": 2.643406867980957,
      "learning_rate": 1.4445496641637967e-05,
      "loss": 4.3371,
      "step": 8442
    },
    {
      "epoch": 0.8443,
      "grad_norm": 4.238451957702637,
      "learning_rate": 1.4427429723883257e-05,
      "loss": 5.1571,
      "step": 8443
    },
    {
      "epoch": 0.8444,
      "grad_norm": 3.1404078006744385,
      "learning_rate": 1.4409373232886702e-05,
      "loss": 4.9146,
      "step": 8444
    },
    {
      "epoch": 0.8445,
      "grad_norm": 3.7619388103485107,
      "learning_rate": 1.439132717084839e-05,
      "loss": 4.8017,
      "step": 8445
    },
    {
      "epoch": 0.8446,
      "grad_norm": 3.2408535480499268,
      "learning_rate": 1.4373291539967182e-05,
      "loss": 4.8374,
      "step": 8446
    },
    {
      "epoch": 0.8447,
      "grad_norm": 3.2047364711761475,
      "learning_rate": 1.4355266342440677e-05,
      "loss": 5.4834,
      "step": 8447
    },
    {
      "epoch": 0.8448,
      "grad_norm": 3.2489891052246094,
      "learning_rate": 1.4337251580465172e-05,
      "loss": 5.0102,
      "step": 8448
    },
    {
      "epoch": 0.8449,
      "grad_norm": 2.897655725479126,
      "learning_rate": 1.4319247256235714e-05,
      "loss": 4.783,
      "step": 8449
    },
    {
      "epoch": 0.845,
      "grad_norm": 2.7230710983276367,
      "learning_rate": 1.4301253371946089e-05,
      "loss": 4.3961,
      "step": 8450
    },
    {
      "epoch": 0.8451,
      "grad_norm": 2.7996606826782227,
      "learning_rate": 1.4283269929788779e-05,
      "loss": 4.8726,
      "step": 8451
    },
    {
      "epoch": 0.8452,
      "grad_norm": 1.87620210647583,
      "learning_rate": 1.426529693195503e-05,
      "loss": 4.3474,
      "step": 8452
    },
    {
      "epoch": 0.8453,
      "grad_norm": 5.8475565910339355,
      "learning_rate": 1.4247334380634792e-05,
      "loss": 4.9823,
      "step": 8453
    },
    {
      "epoch": 0.8454,
      "grad_norm": 2.2325921058654785,
      "learning_rate": 1.4229382278016712e-05,
      "loss": 4.5797,
      "step": 8454
    },
    {
      "epoch": 0.8455,
      "grad_norm": 4.928628444671631,
      "learning_rate": 1.4211440626288286e-05,
      "loss": 5.3097,
      "step": 8455
    },
    {
      "epoch": 0.8456,
      "grad_norm": 2.7446396350860596,
      "learning_rate": 1.4193509427635543e-05,
      "loss": 5.1293,
      "step": 8456
    },
    {
      "epoch": 0.8457,
      "grad_norm": 2.8302688598632812,
      "learning_rate": 1.4175588684243446e-05,
      "loss": 4.7272,
      "step": 8457
    },
    {
      "epoch": 0.8458,
      "grad_norm": 2.257206916809082,
      "learning_rate": 1.4157678398295481e-05,
      "loss": 4.676,
      "step": 8458
    },
    {
      "epoch": 0.8459,
      "grad_norm": 2.3279430866241455,
      "learning_rate": 1.4139778571974049e-05,
      "loss": 4.5821,
      "step": 8459
    },
    {
      "epoch": 0.846,
      "grad_norm": 3.56760835647583,
      "learning_rate": 1.412188920746017e-05,
      "loss": 4.6937,
      "step": 8460
    },
    {
      "epoch": 0.8461,
      "grad_norm": 2.584102153778076,
      "learning_rate": 1.4104010306933557e-05,
      "loss": 4.8198,
      "step": 8461
    },
    {
      "epoch": 0.8462,
      "grad_norm": 3.500945568084717,
      "learning_rate": 1.4086141872572789e-05,
      "loss": 5.5285,
      "step": 8462
    },
    {
      "epoch": 0.8463,
      "grad_norm": 2.287044048309326,
      "learning_rate": 1.406828390655497e-05,
      "loss": 4.6824,
      "step": 8463
    },
    {
      "epoch": 0.8464,
      "grad_norm": 1.7485791444778442,
      "learning_rate": 1.4050436411056123e-05,
      "loss": 4.7904,
      "step": 8464
    },
    {
      "epoch": 0.8465,
      "grad_norm": 3.3748905658721924,
      "learning_rate": 1.40325993882509e-05,
      "loss": 4.6829,
      "step": 8465
    },
    {
      "epoch": 0.8466,
      "grad_norm": 2.3687782287597656,
      "learning_rate": 1.4014772840312663e-05,
      "loss": 4.8949,
      "step": 8466
    },
    {
      "epoch": 0.8467,
      "grad_norm": 2.4991469383239746,
      "learning_rate": 1.3996956769413538e-05,
      "loss": 4.4881,
      "step": 8467
    },
    {
      "epoch": 0.8468,
      "grad_norm": 3.3850317001342773,
      "learning_rate": 1.3979151177724347e-05,
      "loss": 4.9544,
      "step": 8468
    },
    {
      "epoch": 0.8469,
      "grad_norm": 3.972001791000366,
      "learning_rate": 1.3961356067414666e-05,
      "loss": 5.4959,
      "step": 8469
    },
    {
      "epoch": 0.847,
      "grad_norm": 1.9162676334381104,
      "learning_rate": 1.3943571440652769e-05,
      "loss": 4.5638,
      "step": 8470
    },
    {
      "epoch": 0.8471,
      "grad_norm": 1.9648847579956055,
      "learning_rate": 1.3925797299605647e-05,
      "loss": 4.6711,
      "step": 8471
    },
    {
      "epoch": 0.8472,
      "grad_norm": 2.3739423751831055,
      "learning_rate": 1.3908033646439033e-05,
      "loss": 4.7786,
      "step": 8472
    },
    {
      "epoch": 0.8473,
      "grad_norm": 2.076288938522339,
      "learning_rate": 1.3890280483317374e-05,
      "loss": 4.6483,
      "step": 8473
    },
    {
      "epoch": 0.8474,
      "grad_norm": 2.9550273418426514,
      "learning_rate": 1.3872537812403829e-05,
      "loss": 4.6445,
      "step": 8474
    },
    {
      "epoch": 0.8475,
      "grad_norm": 4.1267523765563965,
      "learning_rate": 1.3854805635860336e-05,
      "loss": 4.9901,
      "step": 8475
    },
    {
      "epoch": 0.8476,
      "grad_norm": 7.313724517822266,
      "learning_rate": 1.3837083955847418e-05,
      "loss": 4.8345,
      "step": 8476
    },
    {
      "epoch": 0.8477,
      "grad_norm": 5.185519695281982,
      "learning_rate": 1.3819372774524508e-05,
      "loss": 4.7389,
      "step": 8477
    },
    {
      "epoch": 0.8478,
      "grad_norm": 2.7835001945495605,
      "learning_rate": 1.38016720940496e-05,
      "loss": 5.2362,
      "step": 8478
    },
    {
      "epoch": 0.8479,
      "grad_norm": 3.9391965866088867,
      "learning_rate": 1.3783981916579446e-05,
      "loss": 4.1952,
      "step": 8479
    },
    {
      "epoch": 0.848,
      "grad_norm": 4.8737311363220215,
      "learning_rate": 1.3766302244269624e-05,
      "loss": 4.3331,
      "step": 8480
    },
    {
      "epoch": 0.8481,
      "grad_norm": 2.916710138320923,
      "learning_rate": 1.3748633079274253e-05,
      "loss": 4.8263,
      "step": 8481
    },
    {
      "epoch": 0.8482,
      "grad_norm": 1.7870861291885376,
      "learning_rate": 1.3730974423746335e-05,
      "loss": 4.4934,
      "step": 8482
    },
    {
      "epoch": 0.8483,
      "grad_norm": 2.8641586303710938,
      "learning_rate": 1.3713326279837501e-05,
      "loss": 4.7332,
      "step": 8483
    },
    {
      "epoch": 0.8484,
      "grad_norm": 2.504804849624634,
      "learning_rate": 1.3695688649698124e-05,
      "loss": 4.5666,
      "step": 8484
    },
    {
      "epoch": 0.8485,
      "grad_norm": 2.272033929824829,
      "learning_rate": 1.3678061535477304e-05,
      "loss": 4.6518,
      "step": 8485
    },
    {
      "epoch": 0.8486,
      "grad_norm": 3.24475359916687,
      "learning_rate": 1.3660444939322836e-05,
      "loss": 5.0986,
      "step": 8486
    },
    {
      "epoch": 0.8487,
      "grad_norm": 3.6701200008392334,
      "learning_rate": 1.3642838863381257e-05,
      "loss": 4.8126,
      "step": 8487
    },
    {
      "epoch": 0.8488,
      "grad_norm": 2.3405888080596924,
      "learning_rate": 1.362524330979782e-05,
      "loss": 4.5364,
      "step": 8488
    },
    {
      "epoch": 0.8489,
      "grad_norm": 1.8605806827545166,
      "learning_rate": 1.3607658280716473e-05,
      "loss": 4.3151,
      "step": 8489
    },
    {
      "epoch": 0.849,
      "grad_norm": 3.2683072090148926,
      "learning_rate": 1.3590083778279916e-05,
      "loss": 5.3431,
      "step": 8490
    },
    {
      "epoch": 0.8491,
      "grad_norm": 4.332495212554932,
      "learning_rate": 1.3572519804629536e-05,
      "loss": 5.6552,
      "step": 8491
    },
    {
      "epoch": 0.8492,
      "grad_norm": 7.2649827003479,
      "learning_rate": 1.3554966361905464e-05,
      "loss": 5.913,
      "step": 8492
    },
    {
      "epoch": 0.8493,
      "grad_norm": 5.752014636993408,
      "learning_rate": 1.3537423452246523e-05,
      "loss": 5.0931,
      "step": 8493
    },
    {
      "epoch": 0.8494,
      "grad_norm": 2.6934449672698975,
      "learning_rate": 1.3519891077790236e-05,
      "loss": 4.6956,
      "step": 8494
    },
    {
      "epoch": 0.8495,
      "grad_norm": 2.3991661071777344,
      "learning_rate": 1.3502369240672941e-05,
      "loss": 5.3265,
      "step": 8495
    },
    {
      "epoch": 0.8496,
      "grad_norm": 2.5627100467681885,
      "learning_rate": 1.3484857943029572e-05,
      "loss": 4.9853,
      "step": 8496
    },
    {
      "epoch": 0.8497,
      "grad_norm": 2.291841983795166,
      "learning_rate": 1.34673571869938e-05,
      "loss": 4.9014,
      "step": 8497
    },
    {
      "epoch": 0.8498,
      "grad_norm": 4.921042442321777,
      "learning_rate": 1.3449866974698122e-05,
      "loss": 4.408,
      "step": 8498
    },
    {
      "epoch": 0.8499,
      "grad_norm": 2.5510008335113525,
      "learning_rate": 1.3432387308273575e-05,
      "loss": 4.6561,
      "step": 8499
    },
    {
      "epoch": 0.85,
      "grad_norm": 2.333322525024414,
      "learning_rate": 1.3414918189850089e-05,
      "loss": 4.327,
      "step": 8500
    },
    {
      "epoch": 0.8501,
      "grad_norm": 2.0793213844299316,
      "learning_rate": 1.339745962155613e-05,
      "loss": 4.3647,
      "step": 8501
    },
    {
      "epoch": 0.8502,
      "grad_norm": 2.234926223754883,
      "learning_rate": 1.338001160551906e-05,
      "loss": 4.8129,
      "step": 8502
    },
    {
      "epoch": 0.8503,
      "grad_norm": 1.815430998802185,
      "learning_rate": 1.3362574143864814e-05,
      "loss": 4.5173,
      "step": 8503
    },
    {
      "epoch": 0.8504,
      "grad_norm": 3.5155861377716064,
      "learning_rate": 1.3345147238718126e-05,
      "loss": 4.3156,
      "step": 8504
    },
    {
      "epoch": 0.8505,
      "grad_norm": 2.2945778369903564,
      "learning_rate": 1.3327730892202383e-05,
      "loss": 4.4126,
      "step": 8505
    },
    {
      "epoch": 0.8506,
      "grad_norm": 4.1036272048950195,
      "learning_rate": 1.3310325106439726e-05,
      "loss": 4.5112,
      "step": 8506
    },
    {
      "epoch": 0.8507,
      "grad_norm": 1.9594231843948364,
      "learning_rate": 1.3292929883550998e-05,
      "loss": 4.7003,
      "step": 8507
    },
    {
      "epoch": 0.8508,
      "grad_norm": 2.6589698791503906,
      "learning_rate": 1.327554522565576e-05,
      "loss": 4.4452,
      "step": 8508
    },
    {
      "epoch": 0.8509,
      "grad_norm": 4.780653953552246,
      "learning_rate": 1.3258171134872265e-05,
      "loss": 4.7779,
      "step": 8509
    },
    {
      "epoch": 0.851,
      "grad_norm": 2.708688735961914,
      "learning_rate": 1.3240807613317507e-05,
      "loss": 5.2232,
      "step": 8510
    },
    {
      "epoch": 0.8511,
      "grad_norm": 5.315584182739258,
      "learning_rate": 1.3223454663107172e-05,
      "loss": 4.6174,
      "step": 8511
    },
    {
      "epoch": 0.8512,
      "grad_norm": 4.833944320678711,
      "learning_rate": 1.3206112286355632e-05,
      "loss": 4.1428,
      "step": 8512
    },
    {
      "epoch": 0.8513,
      "grad_norm": 8.856532096862793,
      "learning_rate": 1.3188780485176088e-05,
      "loss": 6.3479,
      "step": 8513
    },
    {
      "epoch": 0.8514,
      "grad_norm": 2.3345205783843994,
      "learning_rate": 1.3171459261680297e-05,
      "loss": 4.5748,
      "step": 8514
    },
    {
      "epoch": 0.8515,
      "grad_norm": 6.388383388519287,
      "learning_rate": 1.3154148617978812e-05,
      "loss": 4.6092,
      "step": 8515
    },
    {
      "epoch": 0.8516,
      "grad_norm": 4.2848334312438965,
      "learning_rate": 1.3136848556180892e-05,
      "loss": 5.0457,
      "step": 8516
    },
    {
      "epoch": 0.8517,
      "grad_norm": 2.4743852615356445,
      "learning_rate": 1.3119559078394461e-05,
      "loss": 4.8899,
      "step": 8517
    },
    {
      "epoch": 0.8518,
      "grad_norm": 5.4313154220581055,
      "learning_rate": 1.3102280186726269e-05,
      "loss": 5.29,
      "step": 8518
    },
    {
      "epoch": 0.8519,
      "grad_norm": 3.023414134979248,
      "learning_rate": 1.3085011883281606e-05,
      "loss": 4.6034,
      "step": 8519
    },
    {
      "epoch": 0.852,
      "grad_norm": 2.0693070888519287,
      "learning_rate": 1.3067754170164614e-05,
      "loss": 4.3616,
      "step": 8520
    },
    {
      "epoch": 0.8521,
      "grad_norm": 6.357508182525635,
      "learning_rate": 1.30505070494781e-05,
      "loss": 6.1411,
      "step": 8521
    },
    {
      "epoch": 0.8522,
      "grad_norm": 2.2661540508270264,
      "learning_rate": 1.3033270523323549e-05,
      "loss": 4.4502,
      "step": 8522
    },
    {
      "epoch": 0.8523,
      "grad_norm": 2.073737859725952,
      "learning_rate": 1.3016044593801202e-05,
      "loss": 4.6395,
      "step": 8523
    },
    {
      "epoch": 0.8524,
      "grad_norm": 2.8492722511291504,
      "learning_rate": 1.2998829263009938e-05,
      "loss": 4.6186,
      "step": 8524
    },
    {
      "epoch": 0.8525,
      "grad_norm": 4.267679691314697,
      "learning_rate": 1.2981624533047432e-05,
      "loss": 4.7541,
      "step": 8525
    },
    {
      "epoch": 0.8526,
      "grad_norm": 2.4987285137176514,
      "learning_rate": 1.296443040601003e-05,
      "loss": 4.7068,
      "step": 8526
    },
    {
      "epoch": 0.8527,
      "grad_norm": 4.056848049163818,
      "learning_rate": 1.294724688399278e-05,
      "loss": 4.6824,
      "step": 8527
    },
    {
      "epoch": 0.8528,
      "grad_norm": 2.686464309692383,
      "learning_rate": 1.293007396908944e-05,
      "loss": 4.5169,
      "step": 8528
    },
    {
      "epoch": 0.8529,
      "grad_norm": 3.7133514881134033,
      "learning_rate": 1.2912911663392469e-05,
      "loss": 4.9086,
      "step": 8529
    },
    {
      "epoch": 0.853,
      "grad_norm": 1.9086830615997314,
      "learning_rate": 1.2895759968993048e-05,
      "loss": 4.5082,
      "step": 8530
    },
    {
      "epoch": 0.8531,
      "grad_norm": 2.331176280975342,
      "learning_rate": 1.2878618887981064e-05,
      "loss": 5.0082,
      "step": 8531
    },
    {
      "epoch": 0.8532,
      "grad_norm": 2.7690670490264893,
      "learning_rate": 1.28614884224451e-05,
      "loss": 4.7595,
      "step": 8532
    },
    {
      "epoch": 0.8533,
      "grad_norm": 3.4200425148010254,
      "learning_rate": 1.2844368574472454e-05,
      "loss": 4.8338,
      "step": 8533
    },
    {
      "epoch": 0.8534,
      "grad_norm": 2.017977714538574,
      "learning_rate": 1.2827259346149122e-05,
      "loss": 4.1347,
      "step": 8534
    },
    {
      "epoch": 0.8535,
      "grad_norm": 11.777688980102539,
      "learning_rate": 1.2810160739559796e-05,
      "loss": 5.1785,
      "step": 8535
    },
    {
      "epoch": 0.8536,
      "grad_norm": 2.207631826400757,
      "learning_rate": 1.279307275678795e-05,
      "loss": 4.7178,
      "step": 8536
    },
    {
      "epoch": 0.8537,
      "grad_norm": 3.542273759841919,
      "learning_rate": 1.2775995399915631e-05,
      "loss": 4.85,
      "step": 8537
    },
    {
      "epoch": 0.8538,
      "grad_norm": 10.820035934448242,
      "learning_rate": 1.2758928671023718e-05,
      "loss": 7.1014,
      "step": 8538
    },
    {
      "epoch": 0.8539,
      "grad_norm": 2.779184579849243,
      "learning_rate": 1.2741872572191682e-05,
      "loss": 4.8245,
      "step": 8539
    },
    {
      "epoch": 0.854,
      "grad_norm": 2.9409196376800537,
      "learning_rate": 1.2724827105497816e-05,
      "loss": 5.1999,
      "step": 8540
    },
    {
      "epoch": 0.8541,
      "grad_norm": 2.74310040473938,
      "learning_rate": 1.2707792273019048e-05,
      "loss": 4.6329,
      "step": 8541
    },
    {
      "epoch": 0.8542,
      "grad_norm": 3.573030710220337,
      "learning_rate": 1.2690768076830972e-05,
      "loss": 4.9917,
      "step": 8542
    },
    {
      "epoch": 0.8543,
      "grad_norm": 3.3721506595611572,
      "learning_rate": 1.2673754519008008e-05,
      "loss": 4.8106,
      "step": 8543
    },
    {
      "epoch": 0.8544,
      "grad_norm": 2.368370532989502,
      "learning_rate": 1.2656751601623118e-05,
      "loss": 4.7299,
      "step": 8544
    },
    {
      "epoch": 0.8545,
      "grad_norm": 4.986358165740967,
      "learning_rate": 1.2639759326748135e-05,
      "loss": 5.7566,
      "step": 8545
    },
    {
      "epoch": 0.8546,
      "grad_norm": 2.925053358078003,
      "learning_rate": 1.262277769645348e-05,
      "loss": 5.017,
      "step": 8546
    },
    {
      "epoch": 0.8547,
      "grad_norm": 3.130060911178589,
      "learning_rate": 1.260580671280832e-05,
      "loss": 5.185,
      "step": 8547
    },
    {
      "epoch": 0.8548,
      "grad_norm": 2.0163793563842773,
      "learning_rate": 1.2588846377880525e-05,
      "loss": 4.9283,
      "step": 8548
    },
    {
      "epoch": 0.8549,
      "grad_norm": 3.7189435958862305,
      "learning_rate": 1.257189669373664e-05,
      "loss": 4.718,
      "step": 8549
    },
    {
      "epoch": 0.855,
      "grad_norm": 2.3411879539489746,
      "learning_rate": 1.2554957662441957e-05,
      "loss": 4.7983,
      "step": 8550
    },
    {
      "epoch": 0.8551,
      "grad_norm": 3.154355525970459,
      "learning_rate": 1.2538029286060426e-05,
      "loss": 5.0299,
      "step": 8551
    },
    {
      "epoch": 0.8552,
      "grad_norm": 5.938164234161377,
      "learning_rate": 1.2521111566654731e-05,
      "loss": 4.6324,
      "step": 8552
    },
    {
      "epoch": 0.8553,
      "grad_norm": 3.5019681453704834,
      "learning_rate": 1.2504204506286243e-05,
      "loss": 4.2435,
      "step": 8553
    },
    {
      "epoch": 0.8554,
      "grad_norm": 2.4241020679473877,
      "learning_rate": 1.2487308107015027e-05,
      "loss": 4.4023,
      "step": 8554
    },
    {
      "epoch": 0.8555,
      "grad_norm": 3.624277114868164,
      "learning_rate": 1.2470422370899838e-05,
      "loss": 4.4476,
      "step": 8555
    },
    {
      "epoch": 0.8556,
      "grad_norm": 4.835483074188232,
      "learning_rate": 1.2453547299998225e-05,
      "loss": 4.8883,
      "step": 8556
    },
    {
      "epoch": 0.8557,
      "grad_norm": 2.470675230026245,
      "learning_rate": 1.243668289636628e-05,
      "loss": 5.3029,
      "step": 8557
    },
    {
      "epoch": 0.8558,
      "grad_norm": 2.5351672172546387,
      "learning_rate": 1.241982916205895e-05,
      "loss": 5.1337,
      "step": 8558
    },
    {
      "epoch": 0.8559,
      "grad_norm": 2.101398468017578,
      "learning_rate": 1.2402986099129765e-05,
      "loss": 4.6825,
      "step": 8559
    },
    {
      "epoch": 0.856,
      "grad_norm": 2.523939609527588,
      "learning_rate": 1.2386153709630988e-05,
      "loss": 5.2254,
      "step": 8560
    },
    {
      "epoch": 0.8561,
      "grad_norm": 2.4997353553771973,
      "learning_rate": 1.2369331995613665e-05,
      "loss": 4.8731,
      "step": 8561
    },
    {
      "epoch": 0.8562,
      "grad_norm": 5.039456844329834,
      "learning_rate": 1.2352520959127379e-05,
      "loss": 5.4307,
      "step": 8562
    },
    {
      "epoch": 0.8563,
      "grad_norm": 3.1753430366516113,
      "learning_rate": 1.2335720602220569e-05,
      "loss": 4.8481,
      "step": 8563
    },
    {
      "epoch": 0.8564,
      "grad_norm": 3.3300609588623047,
      "learning_rate": 1.2318930926940298e-05,
      "loss": 4.8636,
      "step": 8564
    },
    {
      "epoch": 0.8565,
      "grad_norm": 3.0720527172088623,
      "learning_rate": 1.2302151935332329e-05,
      "loss": 4.9421,
      "step": 8565
    },
    {
      "epoch": 0.8566,
      "grad_norm": 2.4076333045959473,
      "learning_rate": 1.228538362944115e-05,
      "loss": 4.8286,
      "step": 8566
    },
    {
      "epoch": 0.8567,
      "grad_norm": 2.7021753787994385,
      "learning_rate": 1.2268626011309858e-05,
      "loss": 4.4647,
      "step": 8567
    },
    {
      "epoch": 0.8568,
      "grad_norm": 2.6486763954162598,
      "learning_rate": 1.22518790829804e-05,
      "loss": 4.5623,
      "step": 8568
    },
    {
      "epoch": 0.8569,
      "grad_norm": 3.447373628616333,
      "learning_rate": 1.2235142846493308e-05,
      "loss": 5.1137,
      "step": 8569
    },
    {
      "epoch": 0.857,
      "grad_norm": 1.7747697830200195,
      "learning_rate": 1.2218417303887842e-05,
      "loss": 4.5631,
      "step": 8570
    },
    {
      "epoch": 0.8571,
      "grad_norm": 3.89992618560791,
      "learning_rate": 1.2201702457201947e-05,
      "loss": 5.2844,
      "step": 8571
    },
    {
      "epoch": 0.8572,
      "grad_norm": 2.6912972927093506,
      "learning_rate": 1.2184998308472295e-05,
      "loss": 5.0885,
      "step": 8572
    },
    {
      "epoch": 0.8573,
      "grad_norm": 2.5314996242523193,
      "learning_rate": 1.2168304859734226e-05,
      "loss": 4.9263,
      "step": 8573
    },
    {
      "epoch": 0.8574,
      "grad_norm": 2.1824238300323486,
      "learning_rate": 1.2151622113021787e-05,
      "loss": 4.6541,
      "step": 8574
    },
    {
      "epoch": 0.8575,
      "grad_norm": 2.2373082637786865,
      "learning_rate": 1.2134950070367723e-05,
      "loss": 4.3035,
      "step": 8575
    },
    {
      "epoch": 0.8576,
      "grad_norm": 2.394096612930298,
      "learning_rate": 1.2118288733803473e-05,
      "loss": 4.7866,
      "step": 8576
    },
    {
      "epoch": 0.8577,
      "grad_norm": 2.244865894317627,
      "learning_rate": 1.2101638105359169e-05,
      "loss": 4.697,
      "step": 8577
    },
    {
      "epoch": 0.8578,
      "grad_norm": 4.5883708000183105,
      "learning_rate": 1.2084998187063613e-05,
      "loss": 4.5578,
      "step": 8578
    },
    {
      "epoch": 0.8579,
      "grad_norm": 4.013721942901611,
      "learning_rate": 1.206836898094439e-05,
      "loss": 4.4906,
      "step": 8579
    },
    {
      "epoch": 0.858,
      "grad_norm": 3.657711982727051,
      "learning_rate": 1.2051750489027647e-05,
      "loss": 4.4345,
      "step": 8580
    },
    {
      "epoch": 0.8581,
      "grad_norm": 3.7616612911224365,
      "learning_rate": 1.2035142713338366e-05,
      "loss": 4.2838,
      "step": 8581
    },
    {
      "epoch": 0.8582,
      "grad_norm": 2.345113515853882,
      "learning_rate": 1.2018545655900081e-05,
      "loss": 5.0299,
      "step": 8582
    },
    {
      "epoch": 0.8583,
      "grad_norm": 5.309663772583008,
      "learning_rate": 1.2001959318735156e-05,
      "loss": 4.6591,
      "step": 8583
    },
    {
      "epoch": 0.8584,
      "grad_norm": 2.5517420768737793,
      "learning_rate": 1.1985383703864583e-05,
      "loss": 4.6142,
      "step": 8584
    },
    {
      "epoch": 0.8585,
      "grad_norm": 2.5022928714752197,
      "learning_rate": 1.196881881330798e-05,
      "loss": 4.4631,
      "step": 8585
    },
    {
      "epoch": 0.8586,
      "grad_norm": 2.1912248134613037,
      "learning_rate": 1.19522646490838e-05,
      "loss": 4.7151,
      "step": 8586
    },
    {
      "epoch": 0.8587,
      "grad_norm": 2.7792141437530518,
      "learning_rate": 1.1935721213209105e-05,
      "loss": 4.6682,
      "step": 8587
    },
    {
      "epoch": 0.8588,
      "grad_norm": 2.6493330001831055,
      "learning_rate": 1.191918850769964e-05,
      "loss": 4.7963,
      "step": 8588
    },
    {
      "epoch": 0.8589,
      "grad_norm": 2.675067186355591,
      "learning_rate": 1.1902666534569883e-05,
      "loss": 4.598,
      "step": 8589
    },
    {
      "epoch": 0.859,
      "grad_norm": 2.5132172107696533,
      "learning_rate": 1.188615529583299e-05,
      "loss": 4.5296,
      "step": 8590
    },
    {
      "epoch": 0.8591,
      "grad_norm": 2.3619301319122314,
      "learning_rate": 1.1869654793500784e-05,
      "loss": 4.5111,
      "step": 8591
    },
    {
      "epoch": 0.8592,
      "grad_norm": 3.230335235595703,
      "learning_rate": 1.1853165029583823e-05,
      "loss": 5.0268,
      "step": 8592
    },
    {
      "epoch": 0.8593,
      "grad_norm": 2.367743968963623,
      "learning_rate": 1.1836686006091313e-05,
      "loss": 4.8022,
      "step": 8593
    },
    {
      "epoch": 0.8594,
      "grad_norm": 2.647876739501953,
      "learning_rate": 1.1820217725031191e-05,
      "loss": 4.4145,
      "step": 8594
    },
    {
      "epoch": 0.8595,
      "grad_norm": 2.2234082221984863,
      "learning_rate": 1.1803760188410073e-05,
      "loss": 4.5483,
      "step": 8595
    },
    {
      "epoch": 0.8596,
      "grad_norm": 2.3570683002471924,
      "learning_rate": 1.1787313398233235e-05,
      "loss": 4.4553,
      "step": 8596
    },
    {
      "epoch": 0.8597,
      "grad_norm": 4.14098596572876,
      "learning_rate": 1.1770877356504683e-05,
      "loss": 5.2595,
      "step": 8597
    },
    {
      "epoch": 0.8598,
      "grad_norm": 5.065312385559082,
      "learning_rate": 1.1754452065227084e-05,
      "loss": 4.6486,
      "step": 8598
    },
    {
      "epoch": 0.8599,
      "grad_norm": 6.448895454406738,
      "learning_rate": 1.1738037526401857e-05,
      "loss": 4.6003,
      "step": 8599
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.894469976425171,
      "learning_rate": 1.1721633742028992e-05,
      "loss": 4.5776,
      "step": 8600
    },
    {
      "epoch": 0.8601,
      "grad_norm": 3.2529428005218506,
      "learning_rate": 1.1705240714107302e-05,
      "loss": 4.7319,
      "step": 8601
    },
    {
      "epoch": 0.8602,
      "grad_norm": 1.950529932975769,
      "learning_rate": 1.168885844463422e-05,
      "loss": 4.4216,
      "step": 8602
    },
    {
      "epoch": 0.8603,
      "grad_norm": 6.09691047668457,
      "learning_rate": 1.1672486935605831e-05,
      "loss": 5.5867,
      "step": 8603
    },
    {
      "epoch": 0.8604,
      "grad_norm": 2.872224807739258,
      "learning_rate": 1.1656126189017014e-05,
      "loss": 3.9085,
      "step": 8604
    },
    {
      "epoch": 0.8605,
      "grad_norm": 3.2087337970733643,
      "learning_rate": 1.1639776206861196e-05,
      "loss": 4.2787,
      "step": 8605
    },
    {
      "epoch": 0.8606,
      "grad_norm": 2.3076794147491455,
      "learning_rate": 1.1623436991130654e-05,
      "loss": 4.2842,
      "step": 8606
    },
    {
      "epoch": 0.8607,
      "grad_norm": 3.1796443462371826,
      "learning_rate": 1.1607108543816248e-05,
      "loss": 4.4728,
      "step": 8607
    },
    {
      "epoch": 0.8608,
      "grad_norm": 2.0173592567443848,
      "learning_rate": 1.159079086690753e-05,
      "loss": 4.4712,
      "step": 8608
    },
    {
      "epoch": 0.8609,
      "grad_norm": 1.8496477603912354,
      "learning_rate": 1.1574483962392767e-05,
      "loss": 4.3998,
      "step": 8609
    },
    {
      "epoch": 0.861,
      "grad_norm": 2.05112886428833,
      "learning_rate": 1.1558187832258926e-05,
      "loss": 4.3878,
      "step": 8610
    },
    {
      "epoch": 0.8611,
      "grad_norm": 6.00363826751709,
      "learning_rate": 1.1541902478491606e-05,
      "loss": 4.8523,
      "step": 8611
    },
    {
      "epoch": 0.8612,
      "grad_norm": 2.0795581340789795,
      "learning_rate": 1.1525627903075165e-05,
      "loss": 4.438,
      "step": 8612
    },
    {
      "epoch": 0.8613,
      "grad_norm": 3.313765048980713,
      "learning_rate": 1.1509364107992583e-05,
      "loss": 5.3135,
      "step": 8613
    },
    {
      "epoch": 0.8614,
      "grad_norm": 3.162602186203003,
      "learning_rate": 1.1493111095225562e-05,
      "loss": 5.7681,
      "step": 8614
    },
    {
      "epoch": 0.8615,
      "grad_norm": 2.600756883621216,
      "learning_rate": 1.1476868866754486e-05,
      "loss": 5.1491,
      "step": 8615
    },
    {
      "epoch": 0.8616,
      "grad_norm": 1.9981344938278198,
      "learning_rate": 1.1460637424558407e-05,
      "loss": 4.4241,
      "step": 8616
    },
    {
      "epoch": 0.8617,
      "grad_norm": 1.8533384799957275,
      "learning_rate": 1.1444416770615118e-05,
      "loss": 4.6877,
      "step": 8617
    },
    {
      "epoch": 0.8618,
      "grad_norm": 3.6249983310699463,
      "learning_rate": 1.1428206906900995e-05,
      "loss": 4.6673,
      "step": 8618
    },
    {
      "epoch": 0.8619,
      "grad_norm": 8.779684066772461,
      "learning_rate": 1.1412007835391236e-05,
      "loss": 4.6827,
      "step": 8619
    },
    {
      "epoch": 0.862,
      "grad_norm": 1.8194797039031982,
      "learning_rate": 1.1395819558059572e-05,
      "loss": 5.0379,
      "step": 8620
    },
    {
      "epoch": 0.8621,
      "grad_norm": 3.2835960388183594,
      "learning_rate": 1.1379642076878527e-05,
      "loss": 4.635,
      "step": 8621
    },
    {
      "epoch": 0.8622,
      "grad_norm": 3.5352842807769775,
      "learning_rate": 1.1363475393819311e-05,
      "loss": 5.7711,
      "step": 8622
    },
    {
      "epoch": 0.8623,
      "grad_norm": 2.809412956237793,
      "learning_rate": 1.1347319510851717e-05,
      "loss": 4.6662,
      "step": 8623
    },
    {
      "epoch": 0.8624,
      "grad_norm": 2.8579893112182617,
      "learning_rate": 1.1331174429944347e-05,
      "loss": 4.6007,
      "step": 8624
    },
    {
      "epoch": 0.8625,
      "grad_norm": 2.66755747795105,
      "learning_rate": 1.1315040153064416e-05,
      "loss": 4.5965,
      "step": 8625
    },
    {
      "epoch": 0.8626,
      "grad_norm": 3.1046228408813477,
      "learning_rate": 1.129891668217783e-05,
      "loss": 4.1187,
      "step": 8626
    },
    {
      "epoch": 0.8627,
      "grad_norm": 1.9175044298171997,
      "learning_rate": 1.1282804019249182e-05,
      "loss": 4.2156,
      "step": 8627
    },
    {
      "epoch": 0.8628,
      "grad_norm": 2.459900379180908,
      "learning_rate": 1.1266702166241772e-05,
      "loss": 4.364,
      "step": 8628
    },
    {
      "epoch": 0.8629,
      "grad_norm": 3.1337573528289795,
      "learning_rate": 1.1250611125117527e-05,
      "loss": 4.711,
      "step": 8629
    },
    {
      "epoch": 0.863,
      "grad_norm": 2.5150258541107178,
      "learning_rate": 1.1234530897837126e-05,
      "loss": 4.1843,
      "step": 8630
    },
    {
      "epoch": 0.8631,
      "grad_norm": 1.9928077459335327,
      "learning_rate": 1.1218461486359877e-05,
      "loss": 4.714,
      "step": 8631
    },
    {
      "epoch": 0.8632,
      "grad_norm": 3.2583253383636475,
      "learning_rate": 1.1202402892643781e-05,
      "loss": 4.565,
      "step": 8632
    },
    {
      "epoch": 0.8633,
      "grad_norm": 2.1132259368896484,
      "learning_rate": 1.1186355118645554e-05,
      "loss": 4.3252,
      "step": 8633
    },
    {
      "epoch": 0.8634,
      "grad_norm": 2.2939703464508057,
      "learning_rate": 1.1170318166320548e-05,
      "loss": 4.4776,
      "step": 8634
    },
    {
      "epoch": 0.8635,
      "grad_norm": 2.1188876628875732,
      "learning_rate": 1.1154292037622838e-05,
      "loss": 4.4861,
      "step": 8635
    },
    {
      "epoch": 0.8636,
      "grad_norm": 2.7523672580718994,
      "learning_rate": 1.1138276734505104e-05,
      "loss": 4.9806,
      "step": 8636
    },
    {
      "epoch": 0.8637,
      "grad_norm": 2.9013826847076416,
      "learning_rate": 1.1122272258918865e-05,
      "loss": 5.4182,
      "step": 8637
    },
    {
      "epoch": 0.8638,
      "grad_norm": 2.3557424545288086,
      "learning_rate": 1.1106278612814125e-05,
      "loss": 4.8302,
      "step": 8638
    },
    {
      "epoch": 0.8639,
      "grad_norm": 2.5498058795928955,
      "learning_rate": 1.1090295798139672e-05,
      "loss": 4.4789,
      "step": 8639
    },
    {
      "epoch": 0.864,
      "grad_norm": 8.350889205932617,
      "learning_rate": 1.1074323816843024e-05,
      "loss": 4.7477,
      "step": 8640
    },
    {
      "epoch": 0.8641,
      "grad_norm": 3.774102210998535,
      "learning_rate": 1.1058362670870249e-05,
      "loss": 4.7661,
      "step": 8641
    },
    {
      "epoch": 0.8642,
      "grad_norm": 4.227261543273926,
      "learning_rate": 1.1042412362166222e-05,
      "loss": 5.3909,
      "step": 8642
    },
    {
      "epoch": 0.8643,
      "grad_norm": 3.0742998123168945,
      "learning_rate": 1.1026472892674378e-05,
      "loss": 4.8029,
      "step": 8643
    },
    {
      "epoch": 0.8644,
      "grad_norm": 2.3233323097229004,
      "learning_rate": 1.1010544264336942e-05,
      "loss": 4.4914,
      "step": 8644
    },
    {
      "epoch": 0.8645,
      "grad_norm": 6.518969535827637,
      "learning_rate": 1.099462647909475e-05,
      "loss": 6.0514,
      "step": 8645
    },
    {
      "epoch": 0.8646,
      "grad_norm": 2.763496160507202,
      "learning_rate": 1.0978719538887349e-05,
      "loss": 4.5261,
      "step": 8646
    },
    {
      "epoch": 0.8647,
      "grad_norm": 3.7983558177948,
      "learning_rate": 1.0962823445652959e-05,
      "loss": 4.5956,
      "step": 8647
    },
    {
      "epoch": 0.8648,
      "grad_norm": 2.3080053329467773,
      "learning_rate": 1.0946938201328416e-05,
      "loss": 4.431,
      "step": 8648
    },
    {
      "epoch": 0.8649,
      "grad_norm": 3.214174509048462,
      "learning_rate": 1.093106380784934e-05,
      "loss": 4.3588,
      "step": 8649
    },
    {
      "epoch": 0.865,
      "grad_norm": 3.758606433868408,
      "learning_rate": 1.0915200267149972e-05,
      "loss": 4.4647,
      "step": 8650
    },
    {
      "epoch": 0.8651,
      "grad_norm": 7.485246658325195,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 5.0657,
      "step": 8651
    },
    {
      "epoch": 0.8652,
      "grad_norm": 2.1062309741973877,
      "learning_rate": 1.08835057518207e-05,
      "loss": 4.3834,
      "step": 8652
    },
    {
      "epoch": 0.8653,
      "grad_norm": 2.7575623989105225,
      "learning_rate": 1.0867674781052684e-05,
      "loss": 4.7477,
      "step": 8653
    },
    {
      "epoch": 0.8654,
      "grad_norm": 2.0400984287261963,
      "learning_rate": 1.0851854670788108e-05,
      "loss": 4.5488,
      "step": 8654
    },
    {
      "epoch": 0.8655,
      "grad_norm": 2.4911065101623535,
      "learning_rate": 1.0836045422954666e-05,
      "loss": 4.4829,
      "step": 8655
    },
    {
      "epoch": 0.8656,
      "grad_norm": 3.3458988666534424,
      "learning_rate": 1.0820247039478604e-05,
      "loss": 4.4337,
      "step": 8656
    },
    {
      "epoch": 0.8657,
      "grad_norm": 2.4806606769561768,
      "learning_rate": 1.0804459522284926e-05,
      "loss": 4.6244,
      "step": 8657
    },
    {
      "epoch": 0.8658,
      "grad_norm": 2.113936185836792,
      "learning_rate": 1.0788682873297307e-05,
      "loss": 4.4769,
      "step": 8658
    },
    {
      "epoch": 0.8659,
      "grad_norm": 5.580718040466309,
      "learning_rate": 1.0772917094438051e-05,
      "loss": 4.609,
      "step": 8659
    },
    {
      "epoch": 0.866,
      "grad_norm": 2.0098400115966797,
      "learning_rate": 1.0757162187628222e-05,
      "loss": 4.4417,
      "step": 8660
    },
    {
      "epoch": 0.8661,
      "grad_norm": 2.9104840755462646,
      "learning_rate": 1.0741418154787442e-05,
      "loss": 5.6193,
      "step": 8661
    },
    {
      "epoch": 0.8662,
      "grad_norm": 2.2695369720458984,
      "learning_rate": 1.0725684997834162e-05,
      "loss": 4.2815,
      "step": 8662
    },
    {
      "epoch": 0.8663,
      "grad_norm": 3.8240807056427,
      "learning_rate": 1.0709962718685318e-05,
      "loss": 5.0672,
      "step": 8663
    },
    {
      "epoch": 0.8664,
      "grad_norm": 4.0284104347229,
      "learning_rate": 1.0694251319256687e-05,
      "loss": 4.9282,
      "step": 8664
    },
    {
      "epoch": 0.8665,
      "grad_norm": 2.5400500297546387,
      "learning_rate": 1.067855080146266e-05,
      "loss": 5.1228,
      "step": 8665
    },
    {
      "epoch": 0.8666,
      "grad_norm": 2.5081071853637695,
      "learning_rate": 1.0662861167216243e-05,
      "loss": 4.5601,
      "step": 8666
    },
    {
      "epoch": 0.8667,
      "grad_norm": 2.73984694480896,
      "learning_rate": 1.0647182418429225e-05,
      "loss": 4.9116,
      "step": 8667
    },
    {
      "epoch": 0.8668,
      "grad_norm": 2.007780075073242,
      "learning_rate": 1.063151455701199e-05,
      "loss": 4.2676,
      "step": 8668
    },
    {
      "epoch": 0.8669,
      "grad_norm": 3.6223928928375244,
      "learning_rate": 1.0615857584873623e-05,
      "loss": 4.2517,
      "step": 8669
    },
    {
      "epoch": 0.867,
      "grad_norm": 2.0585176944732666,
      "learning_rate": 1.0600211503921887e-05,
      "loss": 4.2739,
      "step": 8670
    },
    {
      "epoch": 0.8671,
      "grad_norm": 2.899587631225586,
      "learning_rate": 1.0584576316063188e-05,
      "loss": 4.8656,
      "step": 8671
    },
    {
      "epoch": 0.8672,
      "grad_norm": 2.480729103088379,
      "learning_rate": 1.0568952023202638e-05,
      "loss": 4.7064,
      "step": 8672
    },
    {
      "epoch": 0.8673,
      "grad_norm": 2.7737233638763428,
      "learning_rate": 1.0553338627244025e-05,
      "loss": 4.5228,
      "step": 8673
    },
    {
      "epoch": 0.8674,
      "grad_norm": 3.4459474086761475,
      "learning_rate": 1.053773613008977e-05,
      "loss": 5.349,
      "step": 8674
    },
    {
      "epoch": 0.8675,
      "grad_norm": 2.2043631076812744,
      "learning_rate": 1.0522144533640998e-05,
      "loss": 4.3608,
      "step": 8675
    },
    {
      "epoch": 0.8676,
      "grad_norm": 2.4829611778259277,
      "learning_rate": 1.0506563839797501e-05,
      "loss": 4.5046,
      "step": 8676
    },
    {
      "epoch": 0.8677,
      "grad_norm": 2.21024227142334,
      "learning_rate": 1.0490994050457748e-05,
      "loss": 4.5975,
      "step": 8677
    },
    {
      "epoch": 0.8678,
      "grad_norm": 3.6296205520629883,
      "learning_rate": 1.0475435167518843e-05,
      "loss": 4.7782,
      "step": 8678
    },
    {
      "epoch": 0.8679,
      "grad_norm": 2.724259614944458,
      "learning_rate": 1.0459887192876594e-05,
      "loss": 4.8428,
      "step": 8679
    },
    {
      "epoch": 0.868,
      "grad_norm": 3.078075408935547,
      "learning_rate": 1.0444350128425529e-05,
      "loss": 5.2678,
      "step": 8680
    },
    {
      "epoch": 0.8681,
      "grad_norm": 3.131704568862915,
      "learning_rate": 1.042882397605871e-05,
      "loss": 6.101,
      "step": 8681
    },
    {
      "epoch": 0.8682,
      "grad_norm": 2.2052724361419678,
      "learning_rate": 1.0413308737668005e-05,
      "loss": 4.6728,
      "step": 8682
    },
    {
      "epoch": 0.8683,
      "grad_norm": 2.285208225250244,
      "learning_rate": 1.0397804415143909e-05,
      "loss": 4.3997,
      "step": 8683
    },
    {
      "epoch": 0.8684,
      "grad_norm": 2.538660764694214,
      "learning_rate": 1.0382311010375512e-05,
      "loss": 4.9574,
      "step": 8684
    },
    {
      "epoch": 0.8685,
      "grad_norm": 2.7964277267456055,
      "learning_rate": 1.0366828525250726e-05,
      "loss": 4.3206,
      "step": 8685
    },
    {
      "epoch": 0.8686,
      "grad_norm": 3.102304458618164,
      "learning_rate": 1.0351356961655945e-05,
      "loss": 4.9792,
      "step": 8686
    },
    {
      "epoch": 0.8687,
      "grad_norm": 2.6609561443328857,
      "learning_rate": 1.0335896321476413e-05,
      "loss": 4.4941,
      "step": 8687
    },
    {
      "epoch": 0.8688,
      "grad_norm": 2.4571733474731445,
      "learning_rate": 1.0320446606595934e-05,
      "loss": 4.9159,
      "step": 8688
    },
    {
      "epoch": 0.8689,
      "grad_norm": 5.534883499145508,
      "learning_rate": 1.0305007818897006e-05,
      "loss": 4.8537,
      "step": 8689
    },
    {
      "epoch": 0.869,
      "grad_norm": 2.6681454181671143,
      "learning_rate": 1.028957996026081e-05,
      "loss": 4.2861,
      "step": 8690
    },
    {
      "epoch": 0.8691,
      "grad_norm": 1.9641433954238892,
      "learning_rate": 1.0274163032567163e-05,
      "loss": 4.7946,
      "step": 8691
    },
    {
      "epoch": 0.8692,
      "grad_norm": 2.935879945755005,
      "learning_rate": 1.0258757037694589e-05,
      "loss": 4.8643,
      "step": 8692
    },
    {
      "epoch": 0.8693,
      "grad_norm": 2.7178714275360107,
      "learning_rate": 1.0243361977520249e-05,
      "loss": 4.9444,
      "step": 8693
    },
    {
      "epoch": 0.8694,
      "grad_norm": 2.859490394592285,
      "learning_rate": 1.022797785392e-05,
      "loss": 4.4899,
      "step": 8694
    },
    {
      "epoch": 0.8695,
      "grad_norm": 5.18662166595459,
      "learning_rate": 1.0212604668768343e-05,
      "loss": 4.8032,
      "step": 8695
    },
    {
      "epoch": 0.8696,
      "grad_norm": 2.727688789367676,
      "learning_rate": 1.0197242423938446e-05,
      "loss": 5.5104,
      "step": 8696
    },
    {
      "epoch": 0.8697,
      "grad_norm": 2.2857766151428223,
      "learning_rate": 1.0181891121302145e-05,
      "loss": 4.6634,
      "step": 8697
    },
    {
      "epoch": 0.8698,
      "grad_norm": 2.5160410404205322,
      "learning_rate": 1.0166550762729998e-05,
      "loss": 4.3491,
      "step": 8698
    },
    {
      "epoch": 0.8699,
      "grad_norm": 3.5216283798217773,
      "learning_rate": 1.0151221350091134e-05,
      "loss": 4.9052,
      "step": 8699
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.494596242904663,
      "learning_rate": 1.0135902885253401e-05,
      "loss": 4.7367,
      "step": 8700
    },
    {
      "epoch": 0.8701,
      "grad_norm": 2.649245500564575,
      "learning_rate": 1.0120595370083318e-05,
      "loss": 4.6744,
      "step": 8701
    },
    {
      "epoch": 0.8702,
      "grad_norm": 2.057579517364502,
      "learning_rate": 1.0105298806446028e-05,
      "loss": 4.6272,
      "step": 8702
    },
    {
      "epoch": 0.8703,
      "grad_norm": 2.569733142852783,
      "learning_rate": 1.009001319620545e-05,
      "loss": 4.723,
      "step": 8703
    },
    {
      "epoch": 0.8704,
      "grad_norm": 4.117725372314453,
      "learning_rate": 1.0074738541223993e-05,
      "loss": 4.8231,
      "step": 8704
    },
    {
      "epoch": 0.8705,
      "grad_norm": 4.0609846115112305,
      "learning_rate": 1.0059474843362892e-05,
      "loss": 5.0746,
      "step": 8705
    },
    {
      "epoch": 0.8706,
      "grad_norm": 1.851149082183838,
      "learning_rate": 1.0044222104481971e-05,
      "loss": 4.3597,
      "step": 8706
    },
    {
      "epoch": 0.8707,
      "grad_norm": 2.718358039855957,
      "learning_rate": 1.0028980326439707e-05,
      "loss": 4.8896,
      "step": 8707
    },
    {
      "epoch": 0.8708,
      "grad_norm": 2.083466053009033,
      "learning_rate": 1.0013749511093307e-05,
      "loss": 4.6703,
      "step": 8708
    },
    {
      "epoch": 0.8709,
      "grad_norm": 2.7371761798858643,
      "learning_rate": 9.998529660298539e-06,
      "loss": 5.1291,
      "step": 8709
    },
    {
      "epoch": 0.871,
      "grad_norm": 4.729401111602783,
      "learning_rate": 9.983320775909933e-06,
      "loss": 4.8441,
      "step": 8710
    },
    {
      "epoch": 0.8711,
      "grad_norm": 5.450841903686523,
      "learning_rate": 9.968122859780648e-06,
      "loss": 5.7639,
      "step": 8711
    },
    {
      "epoch": 0.8712,
      "grad_norm": 2.271803855895996,
      "learning_rate": 9.952935913762506e-06,
      "loss": 4.6977,
      "step": 8712
    },
    {
      "epoch": 0.8713,
      "grad_norm": 2.5042195320129395,
      "learning_rate": 9.937759939705971e-06,
      "loss": 4.9083,
      "step": 8713
    },
    {
      "epoch": 0.8714,
      "grad_norm": 2.549596071243286,
      "learning_rate": 9.922594939460194e-06,
      "loss": 4.8199,
      "step": 8714
    },
    {
      "epoch": 0.8715,
      "grad_norm": 2.624634265899658,
      "learning_rate": 9.907440914873e-06,
      "loss": 4.7977,
      "step": 8715
    },
    {
      "epoch": 0.8716,
      "grad_norm": 2.455345869064331,
      "learning_rate": 9.892297867790845e-06,
      "loss": 4.4191,
      "step": 8716
    },
    {
      "epoch": 0.8717,
      "grad_norm": 3.1260228157043457,
      "learning_rate": 9.877165800058874e-06,
      "loss": 4.5715,
      "step": 8717
    },
    {
      "epoch": 0.8718,
      "grad_norm": 2.3313727378845215,
      "learning_rate": 9.86204471352088e-06,
      "loss": 4.7664,
      "step": 8718
    },
    {
      "epoch": 0.8719,
      "grad_norm": 3.2409136295318604,
      "learning_rate": 9.84693461001932e-06,
      "loss": 4.9157,
      "step": 8719
    },
    {
      "epoch": 0.872,
      "grad_norm": 2.99025821685791,
      "learning_rate": 9.831835491395292e-06,
      "loss": 4.5625,
      "step": 8720
    },
    {
      "epoch": 0.8721,
      "grad_norm": 2.885446071624756,
      "learning_rate": 9.816747359488632e-06,
      "loss": 4.6938,
      "step": 8721
    },
    {
      "epoch": 0.8722,
      "grad_norm": 3.209693193435669,
      "learning_rate": 9.801670216137727e-06,
      "loss": 5.4122,
      "step": 8722
    },
    {
      "epoch": 0.8723,
      "grad_norm": 4.007299900054932,
      "learning_rate": 9.786604063179728e-06,
      "loss": 4.9649,
      "step": 8723
    },
    {
      "epoch": 0.8724,
      "grad_norm": 2.2325987815856934,
      "learning_rate": 9.771548902450357e-06,
      "loss": 4.3055,
      "step": 8724
    },
    {
      "epoch": 0.8725,
      "grad_norm": 2.8770947456359863,
      "learning_rate": 9.756504735784067e-06,
      "loss": 4.8277,
      "step": 8725
    },
    {
      "epoch": 0.8726,
      "grad_norm": 3.9321846961975098,
      "learning_rate": 9.74147156501396e-06,
      "loss": 4.8115,
      "step": 8726
    },
    {
      "epoch": 0.8727,
      "grad_norm": 3.394798755645752,
      "learning_rate": 9.726449391971714e-06,
      "loss": 4.5154,
      "step": 8727
    },
    {
      "epoch": 0.8728,
      "grad_norm": 3.2576234340667725,
      "learning_rate": 9.711438218487834e-06,
      "loss": 4.4252,
      "step": 8728
    },
    {
      "epoch": 0.8729,
      "grad_norm": 4.313642501831055,
      "learning_rate": 9.696438046391288e-06,
      "loss": 5.7423,
      "step": 8729
    },
    {
      "epoch": 0.873,
      "grad_norm": 1.7975062131881714,
      "learning_rate": 9.681448877509858e-06,
      "loss": 4.5662,
      "step": 8730
    },
    {
      "epoch": 0.8731,
      "grad_norm": 3.3712804317474365,
      "learning_rate": 9.666470713669918e-06,
      "loss": 4.5369,
      "step": 8731
    },
    {
      "epoch": 0.8732,
      "grad_norm": 2.1813223361968994,
      "learning_rate": 9.651503556696516e-06,
      "loss": 4.884,
      "step": 8732
    },
    {
      "epoch": 0.8733,
      "grad_norm": 5.468778610229492,
      "learning_rate": 9.636547408413355e-06,
      "loss": 4.2959,
      "step": 8733
    },
    {
      "epoch": 0.8734,
      "grad_norm": 5.844184398651123,
      "learning_rate": 9.621602270642781e-06,
      "loss": 5.842,
      "step": 8734
    },
    {
      "epoch": 0.8735,
      "grad_norm": 2.643566131591797,
      "learning_rate": 9.606668145205833e-06,
      "loss": 4.7821,
      "step": 8735
    },
    {
      "epoch": 0.8736,
      "grad_norm": 2.0299408435821533,
      "learning_rate": 9.591745033922173e-06,
      "loss": 4.519,
      "step": 8736
    },
    {
      "epoch": 0.8737,
      "grad_norm": 4.254211902618408,
      "learning_rate": 9.576832938610137e-06,
      "loss": 4.8481,
      "step": 8737
    },
    {
      "epoch": 0.8738,
      "grad_norm": 4.089839458465576,
      "learning_rate": 9.561931861086737e-06,
      "loss": 4.8023,
      "step": 8738
    },
    {
      "epoch": 0.8739,
      "grad_norm": 2.546569585800171,
      "learning_rate": 9.5470418031676e-06,
      "loss": 4.2277,
      "step": 8739
    },
    {
      "epoch": 0.874,
      "grad_norm": 1.8872402906417847,
      "learning_rate": 9.532162766667042e-06,
      "loss": 4.2864,
      "step": 8740
    },
    {
      "epoch": 0.8741,
      "grad_norm": 2.763715982437134,
      "learning_rate": 9.517294753398064e-06,
      "loss": 5.1051,
      "step": 8741
    },
    {
      "epoch": 0.8742,
      "grad_norm": 2.324023723602295,
      "learning_rate": 9.502437765172212e-06,
      "loss": 5.3864,
      "step": 8742
    },
    {
      "epoch": 0.8743,
      "grad_norm": 3.4936015605926514,
      "learning_rate": 9.487591803799856e-06,
      "loss": 4.4165,
      "step": 8743
    },
    {
      "epoch": 0.8744,
      "grad_norm": 2.3895950317382812,
      "learning_rate": 9.47275687108986e-06,
      "loss": 4.6579,
      "step": 8744
    },
    {
      "epoch": 0.8745,
      "grad_norm": 3.1768970489501953,
      "learning_rate": 9.457932968849825e-06,
      "loss": 4.5408,
      "step": 8745
    },
    {
      "epoch": 0.8746,
      "grad_norm": 2.4184420108795166,
      "learning_rate": 9.443120098886061e-06,
      "loss": 4.3241,
      "step": 8746
    },
    {
      "epoch": 0.8747,
      "grad_norm": 5.075557708740234,
      "learning_rate": 9.428318263003378e-06,
      "loss": 5.134,
      "step": 8747
    },
    {
      "epoch": 0.8748,
      "grad_norm": 2.5309062004089355,
      "learning_rate": 9.4135274630054e-06,
      "loss": 4.7426,
      "step": 8748
    },
    {
      "epoch": 0.8749,
      "grad_norm": 3.2121236324310303,
      "learning_rate": 9.398747700694322e-06,
      "loss": 4.8465,
      "step": 8749
    },
    {
      "epoch": 0.875,
      "grad_norm": 2.5411593914031982,
      "learning_rate": 9.383978977871021e-06,
      "loss": 5.2812,
      "step": 8750
    },
    {
      "epoch": 0.8751,
      "grad_norm": 2.3054182529449463,
      "learning_rate": 9.369221296335006e-06,
      "loss": 4.5373,
      "step": 8751
    },
    {
      "epoch": 0.8752,
      "grad_norm": 2.896563768386841,
      "learning_rate": 9.354474657884472e-06,
      "loss": 4.8405,
      "step": 8752
    },
    {
      "epoch": 0.8753,
      "grad_norm": 2.3037235736846924,
      "learning_rate": 9.339739064316233e-06,
      "loss": 4.5514,
      "step": 8753
    },
    {
      "epoch": 0.8754,
      "grad_norm": 2.982818603515625,
      "learning_rate": 9.32501451742579e-06,
      "loss": 4.0718,
      "step": 8754
    },
    {
      "epoch": 0.8755,
      "grad_norm": 2.640371561050415,
      "learning_rate": 9.310301019007285e-06,
      "loss": 4.7887,
      "step": 8755
    },
    {
      "epoch": 0.8756,
      "grad_norm": 10.371726989746094,
      "learning_rate": 9.295598570853514e-06,
      "loss": 5.7836,
      "step": 8756
    },
    {
      "epoch": 0.8757,
      "grad_norm": 2.1169049739837646,
      "learning_rate": 9.280907174755915e-06,
      "loss": 4.203,
      "step": 8757
    },
    {
      "epoch": 0.8758,
      "grad_norm": 2.3381412029266357,
      "learning_rate": 9.266226832504598e-06,
      "loss": 4.4374,
      "step": 8758
    },
    {
      "epoch": 0.8759,
      "grad_norm": 2.9724154472351074,
      "learning_rate": 9.251557545888312e-06,
      "loss": 4.7232,
      "step": 8759
    },
    {
      "epoch": 0.876,
      "grad_norm": 2.162733793258667,
      "learning_rate": 9.236899316694459e-06,
      "loss": 4.4782,
      "step": 8760
    },
    {
      "epoch": 0.8761,
      "grad_norm": 2.7682957649230957,
      "learning_rate": 9.222252146709142e-06,
      "loss": 5.2056,
      "step": 8761
    },
    {
      "epoch": 0.8762,
      "grad_norm": 2.892225742340088,
      "learning_rate": 9.207616037717025e-06,
      "loss": 5.1106,
      "step": 8762
    },
    {
      "epoch": 0.8763,
      "grad_norm": 2.310297966003418,
      "learning_rate": 9.192990991501482e-06,
      "loss": 4.5403,
      "step": 8763
    },
    {
      "epoch": 0.8764,
      "grad_norm": 2.5535264015197754,
      "learning_rate": 9.178377009844563e-06,
      "loss": 5.0011,
      "step": 8764
    },
    {
      "epoch": 0.8765,
      "grad_norm": 2.708742618560791,
      "learning_rate": 9.163774094526889e-06,
      "loss": 4.819,
      "step": 8765
    },
    {
      "epoch": 0.8766,
      "grad_norm": 2.8261663913726807,
      "learning_rate": 9.149182247327837e-06,
      "loss": 4.7367,
      "step": 8766
    },
    {
      "epoch": 0.8767,
      "grad_norm": 2.0129709243774414,
      "learning_rate": 9.134601470025306e-06,
      "loss": 5.1095,
      "step": 8767
    },
    {
      "epoch": 0.8768,
      "grad_norm": 1.8529508113861084,
      "learning_rate": 9.120031764395987e-06,
      "loss": 4.642,
      "step": 8768
    },
    {
      "epoch": 0.8769,
      "grad_norm": 2.3595733642578125,
      "learning_rate": 9.105473132215125e-06,
      "loss": 4.625,
      "step": 8769
    },
    {
      "epoch": 0.877,
      "grad_norm": 3.08668851852417,
      "learning_rate": 9.09092557525666e-06,
      "loss": 4.8503,
      "step": 8770
    },
    {
      "epoch": 0.8771,
      "grad_norm": 2.466050863265991,
      "learning_rate": 9.076389095293148e-06,
      "loss": 4.7133,
      "step": 8771
    },
    {
      "epoch": 0.8772,
      "grad_norm": 3.39971923828125,
      "learning_rate": 9.061863694095828e-06,
      "loss": 4.6332,
      "step": 8772
    },
    {
      "epoch": 0.8773,
      "grad_norm": 3.151628255844116,
      "learning_rate": 9.047349373434566e-06,
      "loss": 5.1136,
      "step": 8773
    },
    {
      "epoch": 0.8774,
      "grad_norm": 2.3871212005615234,
      "learning_rate": 9.0328461350779e-06,
      "loss": 4.6186,
      "step": 8774
    },
    {
      "epoch": 0.8775,
      "grad_norm": 3.517521619796753,
      "learning_rate": 9.018353980792993e-06,
      "loss": 5.0628,
      "step": 8775
    },
    {
      "epoch": 0.8776,
      "grad_norm": 2.2587811946868896,
      "learning_rate": 9.00387291234569e-06,
      "loss": 4.2826,
      "step": 8776
    },
    {
      "epoch": 0.8777,
      "grad_norm": 2.120213508605957,
      "learning_rate": 8.989402931500434e-06,
      "loss": 4.3989,
      "step": 8777
    },
    {
      "epoch": 0.8778,
      "grad_norm": 3.427691698074341,
      "learning_rate": 8.974944040020362e-06,
      "loss": 4.7999,
      "step": 8778
    },
    {
      "epoch": 0.8779,
      "grad_norm": 2.267760753631592,
      "learning_rate": 8.960496239667282e-06,
      "loss": 4.5694,
      "step": 8779
    },
    {
      "epoch": 0.878,
      "grad_norm": 2.228536605834961,
      "learning_rate": 8.946059532201567e-06,
      "loss": 4.2954,
      "step": 8780
    },
    {
      "epoch": 0.8781,
      "grad_norm": 4.906096935272217,
      "learning_rate": 8.931633919382298e-06,
      "loss": 4.5012,
      "step": 8781
    },
    {
      "epoch": 0.8782,
      "grad_norm": 2.55515718460083,
      "learning_rate": 8.917219402967202e-06,
      "loss": 4.3862,
      "step": 8782
    },
    {
      "epoch": 0.8783,
      "grad_norm": 2.9510467052459717,
      "learning_rate": 8.90281598471262e-06,
      "loss": 4.953,
      "step": 8783
    },
    {
      "epoch": 0.8784,
      "grad_norm": 3.9752659797668457,
      "learning_rate": 8.888423666373614e-06,
      "loss": 4.7515,
      "step": 8784
    },
    {
      "epoch": 0.8785,
      "grad_norm": 2.4787232875823975,
      "learning_rate": 8.87404244970378e-06,
      "loss": 4.4391,
      "step": 8785
    },
    {
      "epoch": 0.8786,
      "grad_norm": 6.043068885803223,
      "learning_rate": 8.85967233645547e-06,
      "loss": 5.412,
      "step": 8786
    },
    {
      "epoch": 0.8787,
      "grad_norm": 2.502800226211548,
      "learning_rate": 8.845313328379634e-06,
      "loss": 4.0354,
      "step": 8787
    },
    {
      "epoch": 0.8788,
      "grad_norm": 3.959538698196411,
      "learning_rate": 8.830965427225868e-06,
      "loss": 4.5879,
      "step": 8788
    },
    {
      "epoch": 0.8789,
      "grad_norm": 2.062560558319092,
      "learning_rate": 8.816628634742441e-06,
      "loss": 4.1417,
      "step": 8789
    },
    {
      "epoch": 0.879,
      "grad_norm": 5.228978157043457,
      "learning_rate": 8.80230295267619e-06,
      "loss": 5.1777,
      "step": 8790
    },
    {
      "epoch": 0.8791,
      "grad_norm": 2.512911796569824,
      "learning_rate": 8.787988382772705e-06,
      "loss": 4.8224,
      "step": 8791
    },
    {
      "epoch": 0.8792,
      "grad_norm": 2.596877098083496,
      "learning_rate": 8.77368492677616e-06,
      "loss": 4.151,
      "step": 8792
    },
    {
      "epoch": 0.8793,
      "grad_norm": 2.8246870040893555,
      "learning_rate": 8.759392586429393e-06,
      "loss": 4.593,
      "step": 8793
    },
    {
      "epoch": 0.8794,
      "grad_norm": 2.839712381362915,
      "learning_rate": 8.745111363473868e-06,
      "loss": 4.4628,
      "step": 8794
    },
    {
      "epoch": 0.8795,
      "grad_norm": 2.3297066688537598,
      "learning_rate": 8.730841259649725e-06,
      "loss": 4.4457,
      "step": 8795
    },
    {
      "epoch": 0.8796,
      "grad_norm": 3.7585487365722656,
      "learning_rate": 8.716582276695728e-06,
      "loss": 4.6304,
      "step": 8796
    },
    {
      "epoch": 0.8797,
      "grad_norm": 2.3469479084014893,
      "learning_rate": 8.702334416349278e-06,
      "loss": 5.1458,
      "step": 8797
    },
    {
      "epoch": 0.8798,
      "grad_norm": 2.21311092376709,
      "learning_rate": 8.688097680346453e-06,
      "loss": 4.1124,
      "step": 8798
    },
    {
      "epoch": 0.8799,
      "grad_norm": 2.0984323024749756,
      "learning_rate": 8.67387207042194e-06,
      "loss": 4.4764,
      "step": 8799
    },
    {
      "epoch": 0.88,
      "grad_norm": 3.6739702224731445,
      "learning_rate": 8.6596575883091e-06,
      "loss": 5.6354,
      "step": 8800
    },
    {
      "epoch": 0.8801,
      "grad_norm": 2.693242073059082,
      "learning_rate": 8.645454235739903e-06,
      "loss": 4.718,
      "step": 8801
    },
    {
      "epoch": 0.8802,
      "grad_norm": 6.131505966186523,
      "learning_rate": 8.63126201444503e-06,
      "loss": 4.377,
      "step": 8802
    },
    {
      "epoch": 0.8803,
      "grad_norm": 2.7800419330596924,
      "learning_rate": 8.617080926153698e-06,
      "loss": 5.1356,
      "step": 8803
    },
    {
      "epoch": 0.8804,
      "grad_norm": 3.072803258895874,
      "learning_rate": 8.602910972593892e-06,
      "loss": 4.5791,
      "step": 8804
    },
    {
      "epoch": 0.8805,
      "grad_norm": 4.336158752441406,
      "learning_rate": 8.588752155492119e-06,
      "loss": 4.5522,
      "step": 8805
    },
    {
      "epoch": 0.8806,
      "grad_norm": 2.9067184925079346,
      "learning_rate": 8.574604476573621e-06,
      "loss": 4.2724,
      "step": 8806
    },
    {
      "epoch": 0.8807,
      "grad_norm": 2.4445362091064453,
      "learning_rate": 8.560467937562277e-06,
      "loss": 4.6209,
      "step": 8807
    },
    {
      "epoch": 0.8808,
      "grad_norm": 3.430218458175659,
      "learning_rate": 8.546342540180508e-06,
      "loss": 4.3283,
      "step": 8808
    },
    {
      "epoch": 0.8809,
      "grad_norm": 2.835151433944702,
      "learning_rate": 8.532228286149501e-06,
      "loss": 5.4866,
      "step": 8809
    },
    {
      "epoch": 0.881,
      "grad_norm": 1.9365369081497192,
      "learning_rate": 8.51812517718904e-06,
      "loss": 4.3853,
      "step": 8810
    },
    {
      "epoch": 0.8811,
      "grad_norm": 5.286828994750977,
      "learning_rate": 8.504033215017527e-06,
      "loss": 5.8166,
      "step": 8811
    },
    {
      "epoch": 0.8812,
      "grad_norm": 2.151200294494629,
      "learning_rate": 8.489952401352019e-06,
      "loss": 4.133,
      "step": 8812
    },
    {
      "epoch": 0.8813,
      "grad_norm": 2.3304343223571777,
      "learning_rate": 8.475882737908248e-06,
      "loss": 4.8558,
      "step": 8813
    },
    {
      "epoch": 0.8814,
      "grad_norm": 2.8522188663482666,
      "learning_rate": 8.46182422640054e-06,
      "loss": 4.858,
      "step": 8814
    },
    {
      "epoch": 0.8815,
      "grad_norm": 2.8654606342315674,
      "learning_rate": 8.447776868541879e-06,
      "loss": 4.866,
      "step": 8815
    },
    {
      "epoch": 0.8816,
      "grad_norm": 3.2584755420684814,
      "learning_rate": 8.433740666043898e-06,
      "loss": 4.7322,
      "step": 8816
    },
    {
      "epoch": 0.8817,
      "grad_norm": 2.224492073059082,
      "learning_rate": 8.419715620616874e-06,
      "loss": 4.5702,
      "step": 8817
    },
    {
      "epoch": 0.8818,
      "grad_norm": 3.4625744819641113,
      "learning_rate": 8.405701733969706e-06,
      "loss": 5.2799,
      "step": 8818
    },
    {
      "epoch": 0.8819,
      "grad_norm": 2.2228844165802,
      "learning_rate": 8.39169900780995e-06,
      "loss": 4.5344,
      "step": 8819
    },
    {
      "epoch": 0.882,
      "grad_norm": 2.6673691272735596,
      "learning_rate": 8.377707443843786e-06,
      "loss": 4.574,
      "step": 8820
    },
    {
      "epoch": 0.8821,
      "grad_norm": 10.428030967712402,
      "learning_rate": 8.363727043776038e-06,
      "loss": 4.8223,
      "step": 8821
    },
    {
      "epoch": 0.8822,
      "grad_norm": 2.0214154720306396,
      "learning_rate": 8.34975780931021e-06,
      "loss": 4.4218,
      "step": 8822
    },
    {
      "epoch": 0.8823,
      "grad_norm": 2.2759132385253906,
      "learning_rate": 8.335799742148387e-06,
      "loss": 4.8597,
      "step": 8823
    },
    {
      "epoch": 0.8824,
      "grad_norm": 3.203613519668579,
      "learning_rate": 8.321852843991295e-06,
      "loss": 5.1043,
      "step": 8824
    },
    {
      "epoch": 0.8825,
      "grad_norm": 2.709580183029175,
      "learning_rate": 8.307917116538378e-06,
      "loss": 4.4899,
      "step": 8825
    },
    {
      "epoch": 0.8826,
      "grad_norm": 4.097067832946777,
      "learning_rate": 8.293992561487596e-06,
      "loss": 5.0854,
      "step": 8826
    },
    {
      "epoch": 0.8827,
      "grad_norm": 2.685413360595703,
      "learning_rate": 8.280079180535672e-06,
      "loss": 4.3906,
      "step": 8827
    },
    {
      "epoch": 0.8828,
      "grad_norm": 2.937159299850464,
      "learning_rate": 8.26617697537786e-06,
      "loss": 4.4163,
      "step": 8828
    },
    {
      "epoch": 0.8829,
      "grad_norm": 3.3287947177886963,
      "learning_rate": 8.252285947708139e-06,
      "loss": 4.8213,
      "step": 8829
    },
    {
      "epoch": 0.883,
      "grad_norm": 2.311931610107422,
      "learning_rate": 8.238406099219077e-06,
      "loss": 5.3765,
      "step": 8830
    },
    {
      "epoch": 0.8831,
      "grad_norm": 2.3820695877075195,
      "learning_rate": 8.224537431601886e-06,
      "loss": 4.5504,
      "step": 8831
    },
    {
      "epoch": 0.8832,
      "grad_norm": 2.2117152214050293,
      "learning_rate": 8.21067994654644e-06,
      "loss": 4.7873,
      "step": 8832
    },
    {
      "epoch": 0.8833,
      "grad_norm": 2.1899890899658203,
      "learning_rate": 8.196833645741186e-06,
      "loss": 4.8569,
      "step": 8833
    },
    {
      "epoch": 0.8834,
      "grad_norm": 2.6373932361602783,
      "learning_rate": 8.182998530873298e-06,
      "loss": 4.9869,
      "step": 8834
    },
    {
      "epoch": 0.8835,
      "grad_norm": 3.1250391006469727,
      "learning_rate": 8.169174603628538e-06,
      "loss": 5.0161,
      "step": 8835
    },
    {
      "epoch": 0.8836,
      "grad_norm": 2.3784611225128174,
      "learning_rate": 8.15536186569129e-06,
      "loss": 4.7068,
      "step": 8836
    },
    {
      "epoch": 0.8837,
      "grad_norm": 5.070887565612793,
      "learning_rate": 8.1415603187446e-06,
      "loss": 4.4996,
      "step": 8837
    },
    {
      "epoch": 0.8838,
      "grad_norm": 3.0801916122436523,
      "learning_rate": 8.127769964470156e-06,
      "loss": 4.8837,
      "step": 8838
    },
    {
      "epoch": 0.8839,
      "grad_norm": 2.702174186706543,
      "learning_rate": 8.113990804548244e-06,
      "loss": 4.1756,
      "step": 8839
    },
    {
      "epoch": 0.884,
      "grad_norm": 12.548107147216797,
      "learning_rate": 8.100222840657878e-06,
      "loss": 5.8423,
      "step": 8840
    },
    {
      "epoch": 0.8841,
      "grad_norm": 2.308267116546631,
      "learning_rate": 8.086466074476563e-06,
      "loss": 4.379,
      "step": 8841
    },
    {
      "epoch": 0.8842,
      "grad_norm": 4.790067672729492,
      "learning_rate": 8.072720507680565e-06,
      "loss": 4.8778,
      "step": 8842
    },
    {
      "epoch": 0.8843,
      "grad_norm": 3.0347652435302734,
      "learning_rate": 8.058986141944724e-06,
      "loss": 4.4973,
      "step": 8843
    },
    {
      "epoch": 0.8844,
      "grad_norm": 2.178286552429199,
      "learning_rate": 8.045262978942513e-06,
      "loss": 4.7445,
      "step": 8844
    },
    {
      "epoch": 0.8845,
      "grad_norm": 3.6863443851470947,
      "learning_rate": 8.031551020346128e-06,
      "loss": 4.4694,
      "step": 8845
    },
    {
      "epoch": 0.8846,
      "grad_norm": 2.088005542755127,
      "learning_rate": 8.017850267826232e-06,
      "loss": 5.1005,
      "step": 8846
    },
    {
      "epoch": 0.8847,
      "grad_norm": 4.256787300109863,
      "learning_rate": 8.004160723052312e-06,
      "loss": 5.4287,
      "step": 8847
    },
    {
      "epoch": 0.8848,
      "grad_norm": 2.6739354133605957,
      "learning_rate": 7.990482387692311e-06,
      "loss": 4.8302,
      "step": 8848
    },
    {
      "epoch": 0.8849,
      "grad_norm": 2.894029140472412,
      "learning_rate": 7.976815263412963e-06,
      "loss": 4.7312,
      "step": 8849
    },
    {
      "epoch": 0.885,
      "grad_norm": 2.58744740486145,
      "learning_rate": 7.963159351879556e-06,
      "loss": 4.7161,
      "step": 8850
    },
    {
      "epoch": 0.8851,
      "grad_norm": 3.004188060760498,
      "learning_rate": 7.949514654755962e-06,
      "loss": 4.8242,
      "step": 8851
    },
    {
      "epoch": 0.8852,
      "grad_norm": 2.783155918121338,
      "learning_rate": 7.935881173704819e-06,
      "loss": 4.5729,
      "step": 8852
    },
    {
      "epoch": 0.8853,
      "grad_norm": 5.1716766357421875,
      "learning_rate": 7.922258910387282e-06,
      "loss": 5.3958,
      "step": 8853
    },
    {
      "epoch": 0.8854,
      "grad_norm": 2.577392101287842,
      "learning_rate": 7.908647866463203e-06,
      "loss": 4.6871,
      "step": 8854
    },
    {
      "epoch": 0.8855,
      "grad_norm": 2.170377016067505,
      "learning_rate": 7.895048043591036e-06,
      "loss": 4.57,
      "step": 8855
    },
    {
      "epoch": 0.8856,
      "grad_norm": 1.7770757675170898,
      "learning_rate": 7.881459443427886e-06,
      "loss": 4.2012,
      "step": 8856
    },
    {
      "epoch": 0.8857,
      "grad_norm": 3.4575588703155518,
      "learning_rate": 7.867882067629472e-06,
      "loss": 4.373,
      "step": 8857
    },
    {
      "epoch": 0.8858,
      "grad_norm": 2.1096558570861816,
      "learning_rate": 7.854315917850163e-06,
      "loss": 4.1839,
      "step": 8858
    },
    {
      "epoch": 0.8859,
      "grad_norm": 2.7760400772094727,
      "learning_rate": 7.840760995742946e-06,
      "loss": 4.7035,
      "step": 8859
    },
    {
      "epoch": 0.886,
      "grad_norm": 2.51037335395813,
      "learning_rate": 7.827217302959467e-06,
      "loss": 4.3947,
      "step": 8860
    },
    {
      "epoch": 0.8861,
      "grad_norm": 2.530010223388672,
      "learning_rate": 7.81368484114996e-06,
      "loss": 4.3801,
      "step": 8861
    },
    {
      "epoch": 0.8862,
      "grad_norm": 4.767881870269775,
      "learning_rate": 7.800163611963318e-06,
      "loss": 4.6897,
      "step": 8862
    },
    {
      "epoch": 0.8863,
      "grad_norm": 4.390204429626465,
      "learning_rate": 7.786653617047079e-06,
      "loss": 5.0188,
      "step": 8863
    },
    {
      "epoch": 0.8864,
      "grad_norm": 2.9131581783294678,
      "learning_rate": 7.77315485804736e-06,
      "loss": 4.6434,
      "step": 8864
    },
    {
      "epoch": 0.8865,
      "grad_norm": 2.2848124504089355,
      "learning_rate": 7.75966733660901e-06,
      "loss": 5.2347,
      "step": 8865
    },
    {
      "epoch": 0.8866,
      "grad_norm": 2.4950950145721436,
      "learning_rate": 7.746191054375362e-06,
      "loss": 4.3732,
      "step": 8866
    },
    {
      "epoch": 0.8867,
      "grad_norm": 2.745161533355713,
      "learning_rate": 7.73272601298851e-06,
      "loss": 5.2298,
      "step": 8867
    },
    {
      "epoch": 0.8868,
      "grad_norm": 3.0328330993652344,
      "learning_rate": 7.719272214089145e-06,
      "loss": 4.3103,
      "step": 8868
    },
    {
      "epoch": 0.8869,
      "grad_norm": 2.1104753017425537,
      "learning_rate": 7.7058296593165e-06,
      "loss": 4.5008,
      "step": 8869
    },
    {
      "epoch": 0.887,
      "grad_norm": 8.631593704223633,
      "learning_rate": 7.692398350308594e-06,
      "loss": 4.3234,
      "step": 8870
    },
    {
      "epoch": 0.8871,
      "grad_norm": 2.1757302284240723,
      "learning_rate": 7.67897828870191e-06,
      "loss": 4.6937,
      "step": 8871
    },
    {
      "epoch": 0.8872,
      "grad_norm": 2.2005538940429688,
      "learning_rate": 7.665569476131706e-06,
      "loss": 4.3408,
      "step": 8872
    },
    {
      "epoch": 0.8873,
      "grad_norm": 3.5463802814483643,
      "learning_rate": 7.652171914231776e-06,
      "loss": 4.1053,
      "step": 8873
    },
    {
      "epoch": 0.8874,
      "grad_norm": 4.783703327178955,
      "learning_rate": 7.638785604634579e-06,
      "loss": 4.3819,
      "step": 8874
    },
    {
      "epoch": 0.8875,
      "grad_norm": 2.7134149074554443,
      "learning_rate": 7.625410548971191e-06,
      "loss": 4.3682,
      "step": 8875
    },
    {
      "epoch": 0.8876,
      "grad_norm": 1.9780843257904053,
      "learning_rate": 7.612046748871327e-06,
      "loss": 4.6811,
      "step": 8876
    },
    {
      "epoch": 0.8877,
      "grad_norm": 1.8874635696411133,
      "learning_rate": 7.59869420596333e-06,
      "loss": 4.2981,
      "step": 8877
    },
    {
      "epoch": 0.8878,
      "grad_norm": 2.267974853515625,
      "learning_rate": 7.585352921874156e-06,
      "loss": 4.2512,
      "step": 8878
    },
    {
      "epoch": 0.8879,
      "grad_norm": 2.255434989929199,
      "learning_rate": 7.572022898229403e-06,
      "loss": 4.8496,
      "step": 8879
    },
    {
      "epoch": 0.888,
      "grad_norm": 3.6688151359558105,
      "learning_rate": 7.558704136653305e-06,
      "loss": 4.9578,
      "step": 8880
    },
    {
      "epoch": 0.8881,
      "grad_norm": 3.4965503215789795,
      "learning_rate": 7.545396638768698e-06,
      "loss": 5.4071,
      "step": 8881
    },
    {
      "epoch": 0.8882,
      "grad_norm": 2.7984414100646973,
      "learning_rate": 7.5321004061970405e-06,
      "loss": 4.5589,
      "step": 8882
    },
    {
      "epoch": 0.8883,
      "grad_norm": 4.428894996643066,
      "learning_rate": 7.518815440558513e-06,
      "loss": 5.4352,
      "step": 8883
    },
    {
      "epoch": 0.8884,
      "grad_norm": 3.4463980197906494,
      "learning_rate": 7.505541743471756e-06,
      "loss": 4.5624,
      "step": 8884
    },
    {
      "epoch": 0.8885,
      "grad_norm": 3.8881261348724365,
      "learning_rate": 7.492279316554207e-06,
      "loss": 4.1961,
      "step": 8885
    },
    {
      "epoch": 0.8886,
      "grad_norm": 4.636345863342285,
      "learning_rate": 7.479028161421797e-06,
      "loss": 5.1421,
      "step": 8886
    },
    {
      "epoch": 0.8887,
      "grad_norm": 3.346356153488159,
      "learning_rate": 7.465788279689156e-06,
      "loss": 4.6843,
      "step": 8887
    },
    {
      "epoch": 0.8888,
      "grad_norm": 2.3888537883758545,
      "learning_rate": 7.45255967296955e-06,
      "loss": 4.8294,
      "step": 8888
    },
    {
      "epoch": 0.8889,
      "grad_norm": 2.3268802165985107,
      "learning_rate": 7.439342342874789e-06,
      "loss": 4.5319,
      "step": 8889
    },
    {
      "epoch": 0.889,
      "grad_norm": 4.291236877441406,
      "learning_rate": 7.426136291015417e-06,
      "loss": 4.7186,
      "step": 8890
    },
    {
      "epoch": 0.8891,
      "grad_norm": 3.4129977226257324,
      "learning_rate": 7.412941519000527e-06,
      "loss": 4.6681,
      "step": 8891
    },
    {
      "epoch": 0.8892,
      "grad_norm": 2.6915132999420166,
      "learning_rate": 7.399758028437864e-06,
      "loss": 4.7773,
      "step": 8892
    },
    {
      "epoch": 0.8893,
      "grad_norm": 2.4664089679718018,
      "learning_rate": 7.386585820933811e-06,
      "loss": 4.915,
      "step": 8893
    },
    {
      "epoch": 0.8894,
      "grad_norm": 2.5627193450927734,
      "learning_rate": 7.3734248980933395e-06,
      "loss": 5.1998,
      "step": 8894
    },
    {
      "epoch": 0.8895,
      "grad_norm": 2.4949874877929688,
      "learning_rate": 7.360275261520078e-06,
      "loss": 4.7773,
      "step": 8895
    },
    {
      "epoch": 0.8896,
      "grad_norm": 2.7768681049346924,
      "learning_rate": 7.347136912816277e-06,
      "loss": 4.6733,
      "step": 8896
    },
    {
      "epoch": 0.8897,
      "grad_norm": 2.3214306831359863,
      "learning_rate": 7.3340098535827905e-06,
      "loss": 4.6154,
      "step": 8897
    },
    {
      "epoch": 0.8898,
      "grad_norm": 2.1289539337158203,
      "learning_rate": 7.320894085419116e-06,
      "loss": 4.6629,
      "step": 8898
    },
    {
      "epoch": 0.8899,
      "grad_norm": 2.7651705741882324,
      "learning_rate": 7.3077896099233765e-06,
      "loss": 4.8732,
      "step": 8899
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.928316354751587,
      "learning_rate": 7.2946964286923046e-06,
      "loss": 4.4806,
      "step": 8900
    },
    {
      "epoch": 0.8901,
      "grad_norm": 2.57163405418396,
      "learning_rate": 7.281614543321269e-06,
      "loss": 4.9074,
      "step": 8901
    },
    {
      "epoch": 0.8902,
      "grad_norm": 1.7538559436798096,
      "learning_rate": 7.268543955404239e-06,
      "loss": 4.3581,
      "step": 8902
    },
    {
      "epoch": 0.8903,
      "grad_norm": 4.891363620758057,
      "learning_rate": 7.255484666533874e-06,
      "loss": 4.9015,
      "step": 8903
    },
    {
      "epoch": 0.8904,
      "grad_norm": 6.567727565765381,
      "learning_rate": 7.242436678301367e-06,
      "loss": 5.6405,
      "step": 8904
    },
    {
      "epoch": 0.8905,
      "grad_norm": 2.3211724758148193,
      "learning_rate": 7.2293999922965705e-06,
      "loss": 4.2358,
      "step": 8905
    },
    {
      "epoch": 0.8906,
      "grad_norm": 4.595614910125732,
      "learning_rate": 7.216374610108012e-06,
      "loss": 4.8458,
      "step": 8906
    },
    {
      "epoch": 0.8907,
      "grad_norm": 2.6091065406799316,
      "learning_rate": 7.203360533322734e-06,
      "loss": 4.1106,
      "step": 8907
    },
    {
      "epoch": 0.8908,
      "grad_norm": 3.2151567935943604,
      "learning_rate": 7.190357763526523e-06,
      "loss": 4.6962,
      "step": 8908
    },
    {
      "epoch": 0.8909,
      "grad_norm": 2.7808351516723633,
      "learning_rate": 7.177366302303667e-06,
      "loss": 5.0021,
      "step": 8909
    },
    {
      "epoch": 0.891,
      "grad_norm": 2.699664831161499,
      "learning_rate": 7.164386151237179e-06,
      "loss": 4.521,
      "step": 8910
    },
    {
      "epoch": 0.8911,
      "grad_norm": 2.8216097354888916,
      "learning_rate": 7.151417311908648e-06,
      "loss": 5.1347,
      "step": 8911
    },
    {
      "epoch": 0.8912,
      "grad_norm": 3.2264506816864014,
      "learning_rate": 7.138459785898266e-06,
      "loss": 4.6347,
      "step": 8912
    },
    {
      "epoch": 0.8913,
      "grad_norm": 3.938873529434204,
      "learning_rate": 7.125513574784903e-06,
      "loss": 5.4142,
      "step": 8913
    },
    {
      "epoch": 0.8914,
      "grad_norm": 2.8794922828674316,
      "learning_rate": 7.112578680145954e-06,
      "loss": 4.8163,
      "step": 8914
    },
    {
      "epoch": 0.8915,
      "grad_norm": 2.54301118850708,
      "learning_rate": 7.099655103557556e-06,
      "loss": 4.6502,
      "step": 8915
    },
    {
      "epoch": 0.8916,
      "grad_norm": 3.7699501514434814,
      "learning_rate": 7.086742846594385e-06,
      "loss": 4.885,
      "step": 8916
    },
    {
      "epoch": 0.8917,
      "grad_norm": 2.1137380599975586,
      "learning_rate": 7.07384191082977e-06,
      "loss": 4.6508,
      "step": 8917
    },
    {
      "epoch": 0.8918,
      "grad_norm": 2.4012317657470703,
      "learning_rate": 7.060952297835633e-06,
      "loss": 4.7864,
      "step": 8918
    },
    {
      "epoch": 0.8919,
      "grad_norm": 2.2190561294555664,
      "learning_rate": 7.048074009182548e-06,
      "loss": 4.578,
      "step": 8919
    },
    {
      "epoch": 0.892,
      "grad_norm": 2.7912960052490234,
      "learning_rate": 7.035207046439673e-06,
      "loss": 4.5946,
      "step": 8920
    },
    {
      "epoch": 0.8921,
      "grad_norm": 2.401081085205078,
      "learning_rate": 7.022351411174866e-06,
      "loss": 4.357,
      "step": 8921
    },
    {
      "epoch": 0.8922,
      "grad_norm": 3.3186697959899902,
      "learning_rate": 7.009507104954493e-06,
      "loss": 4.2477,
      "step": 8922
    },
    {
      "epoch": 0.8923,
      "grad_norm": 2.135366439819336,
      "learning_rate": 6.996674129343606e-06,
      "loss": 4.9409,
      "step": 8923
    },
    {
      "epoch": 0.8924,
      "grad_norm": 3.5601491928100586,
      "learning_rate": 6.9838524859058616e-06,
      "loss": 5.0861,
      "step": 8924
    },
    {
      "epoch": 0.8925,
      "grad_norm": 3.685581922531128,
      "learning_rate": 6.971042176203535e-06,
      "loss": 4.432,
      "step": 8925
    },
    {
      "epoch": 0.8926,
      "grad_norm": 3.92299747467041,
      "learning_rate": 6.958243201797554e-06,
      "loss": 5.0101,
      "step": 8926
    },
    {
      "epoch": 0.8927,
      "grad_norm": 2.9631314277648926,
      "learning_rate": 6.945455564247394e-06,
      "loss": 4.3608,
      "step": 8927
    },
    {
      "epoch": 0.8928,
      "grad_norm": 6.923827648162842,
      "learning_rate": 6.932679265111231e-06,
      "loss": 4.9047,
      "step": 8928
    },
    {
      "epoch": 0.8929,
      "grad_norm": 2.2660651206970215,
      "learning_rate": 6.919914305945774e-06,
      "loss": 4.5786,
      "step": 8929
    },
    {
      "epoch": 0.893,
      "grad_norm": 2.480043411254883,
      "learning_rate": 6.907160688306425e-06,
      "loss": 4.769,
      "step": 8930
    },
    {
      "epoch": 0.8931,
      "grad_norm": 1.8473604917526245,
      "learning_rate": 6.894418413747183e-06,
      "loss": 4.6485,
      "step": 8931
    },
    {
      "epoch": 0.8932,
      "grad_norm": 7.965359687805176,
      "learning_rate": 6.881687483820609e-06,
      "loss": 7.1398,
      "step": 8932
    },
    {
      "epoch": 0.8933,
      "grad_norm": 2.9369471073150635,
      "learning_rate": 6.868967900077972e-06,
      "loss": 5.3093,
      "step": 8933
    },
    {
      "epoch": 0.8934,
      "grad_norm": 1.974391222000122,
      "learning_rate": 6.856259664069098e-06,
      "loss": 4.1849,
      "step": 8934
    },
    {
      "epoch": 0.8935,
      "grad_norm": 3.1651744842529297,
      "learning_rate": 6.8435627773424495e-06,
      "loss": 5.3875,
      "step": 8935
    },
    {
      "epoch": 0.8936,
      "grad_norm": 2.4528918266296387,
      "learning_rate": 6.830877241445111e-06,
      "loss": 4.6323,
      "step": 8936
    },
    {
      "epoch": 0.8937,
      "grad_norm": 2.3573384284973145,
      "learning_rate": 6.818203057922757e-06,
      "loss": 4.3065,
      "step": 8937
    },
    {
      "epoch": 0.8938,
      "grad_norm": 3.8156325817108154,
      "learning_rate": 6.805540228319718e-06,
      "loss": 5.4196,
      "step": 8938
    },
    {
      "epoch": 0.8939,
      "grad_norm": 3.370450735092163,
      "learning_rate": 6.7928887541789055e-06,
      "loss": 4.5451,
      "step": 8939
    },
    {
      "epoch": 0.894,
      "grad_norm": 2.690833330154419,
      "learning_rate": 6.780248637041875e-06,
      "loss": 4.9654,
      "step": 8940
    },
    {
      "epoch": 0.8941,
      "grad_norm": 3.504821300506592,
      "learning_rate": 6.767619878448783e-06,
      "loss": 4.5494,
      "step": 8941
    },
    {
      "epoch": 0.8942,
      "grad_norm": 6.583511829376221,
      "learning_rate": 6.755002479938411e-06,
      "loss": 4.9028,
      "step": 8942
    },
    {
      "epoch": 0.8943,
      "grad_norm": 2.5813591480255127,
      "learning_rate": 6.742396443048138e-06,
      "loss": 4.7263,
      "step": 8943
    },
    {
      "epoch": 0.8944,
      "grad_norm": 2.7122225761413574,
      "learning_rate": 6.729801769313981e-06,
      "loss": 4.882,
      "step": 8944
    },
    {
      "epoch": 0.8945,
      "grad_norm": 5.964580535888672,
      "learning_rate": 6.717218460270536e-06,
      "loss": 5.3094,
      "step": 8945
    },
    {
      "epoch": 0.8946,
      "grad_norm": 2.2921864986419678,
      "learning_rate": 6.704646517451107e-06,
      "loss": 4.7982,
      "step": 8946
    },
    {
      "epoch": 0.8947,
      "grad_norm": 3.775726556777954,
      "learning_rate": 6.692085942387483e-06,
      "loss": 4.7926,
      "step": 8947
    },
    {
      "epoch": 0.8948,
      "grad_norm": 2.172813653945923,
      "learning_rate": 6.679536736610137e-06,
      "loss": 4.5813,
      "step": 8948
    },
    {
      "epoch": 0.8949,
      "grad_norm": 2.895630359649658,
      "learning_rate": 6.666998901648203e-06,
      "loss": 4.7405,
      "step": 8949
    },
    {
      "epoch": 0.895,
      "grad_norm": 2.4028589725494385,
      "learning_rate": 6.654472439029313e-06,
      "loss": 4.7573,
      "step": 8950
    },
    {
      "epoch": 0.8951,
      "grad_norm": 2.510836601257324,
      "learning_rate": 6.6419573502798374e-06,
      "loss": 4.8448,
      "step": 8951
    },
    {
      "epoch": 0.8952,
      "grad_norm": 2.987708330154419,
      "learning_rate": 6.629453636924643e-06,
      "loss": 4.954,
      "step": 8952
    },
    {
      "epoch": 0.8953,
      "grad_norm": 2.885619640350342,
      "learning_rate": 6.616961300487324e-06,
      "loss": 4.6267,
      "step": 8953
    },
    {
      "epoch": 0.8954,
      "grad_norm": 4.7632856369018555,
      "learning_rate": 6.604480342490004e-06,
      "loss": 5.4725,
      "step": 8954
    },
    {
      "epoch": 0.8955,
      "grad_norm": 2.5840213298797607,
      "learning_rate": 6.592010764453449e-06,
      "loss": 4.6496,
      "step": 8955
    },
    {
      "epoch": 0.8956,
      "grad_norm": 2.3137705326080322,
      "learning_rate": 6.579552567897051e-06,
      "loss": 4.6765,
      "step": 8956
    },
    {
      "epoch": 0.8957,
      "grad_norm": 2.5076022148132324,
      "learning_rate": 6.5671057543387985e-06,
      "loss": 4.8883,
      "step": 8957
    },
    {
      "epoch": 0.8958,
      "grad_norm": 2.356159210205078,
      "learning_rate": 6.554670325295298e-06,
      "loss": 4.8734,
      "step": 8958
    },
    {
      "epoch": 0.8959,
      "grad_norm": 4.514456272125244,
      "learning_rate": 6.542246282281772e-06,
      "loss": 4.7585,
      "step": 8959
    },
    {
      "epoch": 0.896,
      "grad_norm": 3.893876314163208,
      "learning_rate": 6.529833626812043e-06,
      "loss": 4.9597,
      "step": 8960
    },
    {
      "epoch": 0.8961,
      "grad_norm": 4.233180522918701,
      "learning_rate": 6.517432360398556e-06,
      "loss": 4.7282,
      "step": 8961
    },
    {
      "epoch": 0.8962,
      "grad_norm": 2.410092353820801,
      "learning_rate": 6.5050424845523815e-06,
      "loss": 4.6684,
      "step": 8962
    },
    {
      "epoch": 0.8963,
      "grad_norm": 5.6860737800598145,
      "learning_rate": 6.492664000783166e-06,
      "loss": 4.4853,
      "step": 8963
    },
    {
      "epoch": 0.8964,
      "grad_norm": 4.552877426147461,
      "learning_rate": 6.480296910599237e-06,
      "loss": 4.9761,
      "step": 8964
    },
    {
      "epoch": 0.8965,
      "grad_norm": 2.143646001815796,
      "learning_rate": 6.467941215507433e-06,
      "loss": 4.7992,
      "step": 8965
    },
    {
      "epoch": 0.8966,
      "grad_norm": 3.0564205646514893,
      "learning_rate": 6.455596917013273e-06,
      "loss": 4.7205,
      "step": 8966
    },
    {
      "epoch": 0.8967,
      "grad_norm": 4.102299690246582,
      "learning_rate": 6.443264016620887e-06,
      "loss": 4.9219,
      "step": 8967
    },
    {
      "epoch": 0.8968,
      "grad_norm": 2.6727349758148193,
      "learning_rate": 6.430942515832983e-06,
      "loss": 4.4265,
      "step": 8968
    },
    {
      "epoch": 0.8969,
      "grad_norm": 2.039792776107788,
      "learning_rate": 6.418632416150927e-06,
      "loss": 4.9033,
      "step": 8969
    },
    {
      "epoch": 0.897,
      "grad_norm": 2.7340493202209473,
      "learning_rate": 6.406333719074619e-06,
      "loss": 4.4755,
      "step": 8970
    },
    {
      "epoch": 0.8971,
      "grad_norm": 2.2124836444854736,
      "learning_rate": 6.394046426102674e-06,
      "loss": 4.7025,
      "step": 8971
    },
    {
      "epoch": 0.8972,
      "grad_norm": 3.26025652885437,
      "learning_rate": 6.381770538732224e-06,
      "loss": 4.9631,
      "step": 8972
    },
    {
      "epoch": 0.8973,
      "grad_norm": 3.2716753482818604,
      "learning_rate": 6.3695060584590625e-06,
      "loss": 4.8155,
      "step": 8973
    },
    {
      "epoch": 0.8974,
      "grad_norm": 2.039505958557129,
      "learning_rate": 6.357252986777595e-06,
      "loss": 4.6014,
      "step": 8974
    },
    {
      "epoch": 0.8975,
      "grad_norm": 6.806495666503906,
      "learning_rate": 6.345011325180772e-06,
      "loss": 4.9514,
      "step": 8975
    },
    {
      "epoch": 0.8976,
      "grad_norm": 2.032172203063965,
      "learning_rate": 6.332781075160243e-06,
      "loss": 4.4953,
      "step": 8976
    },
    {
      "epoch": 0.8977,
      "grad_norm": 2.069721221923828,
      "learning_rate": 6.320562238206218e-06,
      "loss": 4.8177,
      "step": 8977
    },
    {
      "epoch": 0.8978,
      "grad_norm": 7.433823108673096,
      "learning_rate": 6.308354815807527e-06,
      "loss": 5.3885,
      "step": 8978
    },
    {
      "epoch": 0.8979,
      "grad_norm": 2.4086196422576904,
      "learning_rate": 6.296158809451602e-06,
      "loss": 4.6847,
      "step": 8979
    },
    {
      "epoch": 0.898,
      "grad_norm": 3.105590343475342,
      "learning_rate": 6.283974220624489e-06,
      "loss": 5.0809,
      "step": 8980
    },
    {
      "epoch": 0.8981,
      "grad_norm": 2.0045838356018066,
      "learning_rate": 6.2718010508108545e-06,
      "loss": 4.5695,
      "step": 8981
    },
    {
      "epoch": 0.8982,
      "grad_norm": 2.0486457347869873,
      "learning_rate": 6.259639301493947e-06,
      "loss": 4.3533,
      "step": 8982
    },
    {
      "epoch": 0.8983,
      "grad_norm": 3.2638332843780518,
      "learning_rate": 6.2474889741556575e-06,
      "loss": 4.814,
      "step": 8983
    },
    {
      "epoch": 0.8984,
      "grad_norm": 9.431929588317871,
      "learning_rate": 6.235350070276447e-06,
      "loss": 4.4612,
      "step": 8984
    },
    {
      "epoch": 0.8985,
      "grad_norm": 5.351754665374756,
      "learning_rate": 6.223222591335409e-06,
      "loss": 5.2369,
      "step": 8985
    },
    {
      "epoch": 0.8986,
      "grad_norm": 2.495537757873535,
      "learning_rate": 6.21110653881023e-06,
      "loss": 4.6657,
      "step": 8986
    },
    {
      "epoch": 0.8987,
      "grad_norm": 2.4522995948791504,
      "learning_rate": 6.1990019141772605e-06,
      "loss": 4.9826,
      "step": 8987
    },
    {
      "epoch": 0.8988,
      "grad_norm": 6.8166184425354,
      "learning_rate": 6.186908718911344e-06,
      "loss": 4.9533,
      "step": 8988
    },
    {
      "epoch": 0.8989,
      "grad_norm": 2.687861442565918,
      "learning_rate": 6.174826954486068e-06,
      "loss": 4.5193,
      "step": 8989
    },
    {
      "epoch": 0.899,
      "grad_norm": 1.862525463104248,
      "learning_rate": 6.1627566223735e-06,
      "loss": 4.5434,
      "step": 8990
    },
    {
      "epoch": 0.8991,
      "grad_norm": 2.2695696353912354,
      "learning_rate": 6.1506977240444074e-06,
      "loss": 4.7943,
      "step": 8991
    },
    {
      "epoch": 0.8992,
      "grad_norm": 3.5589981079101562,
      "learning_rate": 6.138650260968137e-06,
      "loss": 4.9337,
      "step": 8992
    },
    {
      "epoch": 0.8993,
      "grad_norm": 3.216830253601074,
      "learning_rate": 6.126614234612593e-06,
      "loss": 4.9656,
      "step": 8993
    },
    {
      "epoch": 0.8994,
      "grad_norm": 3.40544056892395,
      "learning_rate": 6.1145896464443685e-06,
      "loss": 4.6567,
      "step": 8994
    },
    {
      "epoch": 0.8995,
      "grad_norm": 2.092055082321167,
      "learning_rate": 6.102576497928614e-06,
      "loss": 4.5361,
      "step": 8995
    },
    {
      "epoch": 0.8996,
      "grad_norm": 2.227360963821411,
      "learning_rate": 6.090574790529091e-06,
      "loss": 4.745,
      "step": 8996
    },
    {
      "epoch": 0.8997,
      "grad_norm": 2.7571942806243896,
      "learning_rate": 6.078584525708176e-06,
      "loss": 4.4547,
      "step": 8997
    },
    {
      "epoch": 0.8998,
      "grad_norm": 3.526732921600342,
      "learning_rate": 6.066605704926831e-06,
      "loss": 5.1164,
      "step": 8998
    },
    {
      "epoch": 0.8999,
      "grad_norm": 2.743854284286499,
      "learning_rate": 6.054638329644657e-06,
      "loss": 4.7299,
      "step": 8999
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.4050204753875732,
      "learning_rate": 6.042682401319844e-06,
      "loss": 4.9346,
      "step": 9000
    },
    {
      "epoch": 0.9001,
      "grad_norm": 2.497135877609253,
      "learning_rate": 6.030737921409169e-06,
      "loss": 4.4046,
      "step": 9001
    },
    {
      "epoch": 0.9002,
      "grad_norm": 3.193295955657959,
      "learning_rate": 6.018804891368035e-06,
      "loss": 5.4832,
      "step": 9002
    },
    {
      "epoch": 0.9003,
      "grad_norm": 2.2235870361328125,
      "learning_rate": 6.006883312650457e-06,
      "loss": 4.5252,
      "step": 9003
    },
    {
      "epoch": 0.9004,
      "grad_norm": 2.4079129695892334,
      "learning_rate": 5.994973186709041e-06,
      "loss": 4.5816,
      "step": 9004
    },
    {
      "epoch": 0.9005,
      "grad_norm": 2.6099767684936523,
      "learning_rate": 5.98307451499498e-06,
      "loss": 4.7906,
      "step": 9005
    },
    {
      "epoch": 0.9006,
      "grad_norm": 11.038399696350098,
      "learning_rate": 5.971187298958103e-06,
      "loss": 4.6758,
      "step": 9006
    },
    {
      "epoch": 0.9007,
      "grad_norm": 3.646012306213379,
      "learning_rate": 5.9593115400468636e-06,
      "loss": 5.3202,
      "step": 9007
    },
    {
      "epoch": 0.9008,
      "grad_norm": 2.198439598083496,
      "learning_rate": 5.947447239708215e-06,
      "loss": 4.3398,
      "step": 9008
    },
    {
      "epoch": 0.9009,
      "grad_norm": 4.354560852050781,
      "learning_rate": 5.935594399387856e-06,
      "loss": 5.0069,
      "step": 9009
    },
    {
      "epoch": 0.901,
      "grad_norm": 2.466897487640381,
      "learning_rate": 5.923753020529999e-06,
      "loss": 4.6235,
      "step": 9010
    },
    {
      "epoch": 0.9011,
      "grad_norm": 3.630133867263794,
      "learning_rate": 5.911923104577455e-06,
      "loss": 4.582,
      "step": 9011
    },
    {
      "epoch": 0.9012,
      "grad_norm": 2.6662099361419678,
      "learning_rate": 5.900104652971694e-06,
      "loss": 4.226,
      "step": 9012
    },
    {
      "epoch": 0.9013,
      "grad_norm": 4.389400482177734,
      "learning_rate": 5.888297667152731e-06,
      "loss": 5.1589,
      "step": 9013
    },
    {
      "epoch": 0.9014,
      "grad_norm": 1.829481840133667,
      "learning_rate": 5.8765021485592376e-06,
      "loss": 4.6615,
      "step": 9014
    },
    {
      "epoch": 0.9015,
      "grad_norm": 3.1666388511657715,
      "learning_rate": 5.864718098628441e-06,
      "loss": 4.5842,
      "step": 9015
    },
    {
      "epoch": 0.9016,
      "grad_norm": 2.3674471378326416,
      "learning_rate": 5.852945518796205e-06,
      "loss": 4.4636,
      "step": 9016
    },
    {
      "epoch": 0.9017,
      "grad_norm": 3.65934681892395,
      "learning_rate": 5.8411844104969916e-06,
      "loss": 4.6697,
      "step": 9017
    },
    {
      "epoch": 0.9018,
      "grad_norm": 2.439911127090454,
      "learning_rate": 5.829434775163833e-06,
      "loss": 4.657,
      "step": 9018
    },
    {
      "epoch": 0.9019,
      "grad_norm": 2.1941773891448975,
      "learning_rate": 5.8176966142283965e-06,
      "loss": 4.3775,
      "step": 9019
    },
    {
      "epoch": 0.902,
      "grad_norm": 4.13905668258667,
      "learning_rate": 5.805969929120947e-06,
      "loss": 4.8469,
      "step": 9020
    },
    {
      "epoch": 0.9021,
      "grad_norm": 3.381314754486084,
      "learning_rate": 5.7942547212703315e-06,
      "loss": 5.0062,
      "step": 9021
    },
    {
      "epoch": 0.9022,
      "grad_norm": 4.505680084228516,
      "learning_rate": 5.78255099210403e-06,
      "loss": 6.4142,
      "step": 9022
    },
    {
      "epoch": 0.9023,
      "grad_norm": 1.792452096939087,
      "learning_rate": 5.770858743048091e-06,
      "loss": 4.7555,
      "step": 9023
    },
    {
      "epoch": 0.9024,
      "grad_norm": 3.0838587284088135,
      "learning_rate": 5.759177975527186e-06,
      "loss": 4.2252,
      "step": 9024
    },
    {
      "epoch": 0.9025,
      "grad_norm": 3.1067423820495605,
      "learning_rate": 5.747508690964598e-06,
      "loss": 4.3809,
      "step": 9025
    },
    {
      "epoch": 0.9026,
      "grad_norm": 2.3819100856781006,
      "learning_rate": 5.735850890782157e-06,
      "loss": 4.2085,
      "step": 9026
    },
    {
      "epoch": 0.9027,
      "grad_norm": 2.6879513263702393,
      "learning_rate": 5.724204576400371e-06,
      "loss": 4.7583,
      "step": 9027
    },
    {
      "epoch": 0.9028,
      "grad_norm": 2.3721179962158203,
      "learning_rate": 5.7125697492382835e-06,
      "loss": 4.6049,
      "step": 9028
    },
    {
      "epoch": 0.9029,
      "grad_norm": 3.734668493270874,
      "learning_rate": 5.700946410713548e-06,
      "loss": 4.4525,
      "step": 9029
    },
    {
      "epoch": 0.903,
      "grad_norm": 3.0057320594787598,
      "learning_rate": 5.6893345622424874e-06,
      "loss": 4.6629,
      "step": 9030
    },
    {
      "epoch": 0.9031,
      "grad_norm": 4.317963600158691,
      "learning_rate": 5.6777342052399045e-06,
      "loss": 4.9794,
      "step": 9031
    },
    {
      "epoch": 0.9032,
      "grad_norm": 3.1108334064483643,
      "learning_rate": 5.666145341119322e-06,
      "loss": 5.2411,
      "step": 9032
    },
    {
      "epoch": 0.9033,
      "grad_norm": 4.031155109405518,
      "learning_rate": 5.654567971292757e-06,
      "loss": 5.0312,
      "step": 9033
    },
    {
      "epoch": 0.9034,
      "grad_norm": 4.333821773529053,
      "learning_rate": 5.643002097170924e-06,
      "loss": 5.0851,
      "step": 9034
    },
    {
      "epoch": 0.9035,
      "grad_norm": 3.9448437690734863,
      "learning_rate": 5.6314477201630745e-06,
      "loss": 4.8299,
      "step": 9035
    },
    {
      "epoch": 0.9036,
      "grad_norm": 4.547377109527588,
      "learning_rate": 5.619904841677059e-06,
      "loss": 4.7925,
      "step": 9036
    },
    {
      "epoch": 0.9037,
      "grad_norm": 3.252718210220337,
      "learning_rate": 5.608373463119354e-06,
      "loss": 4.9472,
      "step": 9037
    },
    {
      "epoch": 0.9038,
      "grad_norm": 2.3564529418945312,
      "learning_rate": 5.5968535858950345e-06,
      "loss": 4.499,
      "step": 9038
    },
    {
      "epoch": 0.9039,
      "grad_norm": 2.105703830718994,
      "learning_rate": 5.585345211407733e-06,
      "loss": 4.4868,
      "step": 9039
    },
    {
      "epoch": 0.904,
      "grad_norm": 2.0261294841766357,
      "learning_rate": 5.573848341059739e-06,
      "loss": 4.6303,
      "step": 9040
    },
    {
      "epoch": 0.9041,
      "grad_norm": 2.189669609069824,
      "learning_rate": 5.562362976251901e-06,
      "loss": 4.6734,
      "step": 9041
    },
    {
      "epoch": 0.9042,
      "grad_norm": 5.216747283935547,
      "learning_rate": 5.550889118383673e-06,
      "loss": 4.5749,
      "step": 9042
    },
    {
      "epoch": 0.9043,
      "grad_norm": 1.8307996988296509,
      "learning_rate": 5.5394267688531066e-06,
      "loss": 4.4316,
      "step": 9043
    },
    {
      "epoch": 0.9044,
      "grad_norm": 2.3457224369049072,
      "learning_rate": 5.52797592905685e-06,
      "loss": 4.5826,
      "step": 9044
    },
    {
      "epoch": 0.9045,
      "grad_norm": 2.4600014686584473,
      "learning_rate": 5.516536600390188e-06,
      "loss": 4.4705,
      "step": 9045
    },
    {
      "epoch": 0.9046,
      "grad_norm": 3.620086669921875,
      "learning_rate": 5.505108784246926e-06,
      "loss": 4.3431,
      "step": 9046
    },
    {
      "epoch": 0.9047,
      "grad_norm": 4.507942199707031,
      "learning_rate": 5.49369248201953e-06,
      "loss": 4.8736,
      "step": 9047
    },
    {
      "epoch": 0.9048,
      "grad_norm": 2.8256115913391113,
      "learning_rate": 5.482287695099031e-06,
      "loss": 5.2288,
      "step": 9048
    },
    {
      "epoch": 0.9049,
      "grad_norm": 3.3946759700775146,
      "learning_rate": 5.470894424875062e-06,
      "loss": 4.5601,
      "step": 9049
    },
    {
      "epoch": 0.905,
      "grad_norm": 2.5011260509490967,
      "learning_rate": 5.4595126727359e-06,
      "loss": 4.7496,
      "step": 9050
    },
    {
      "epoch": 0.9051,
      "grad_norm": 3.4786741733551025,
      "learning_rate": 5.448142440068316e-06,
      "loss": 4.968,
      "step": 9051
    },
    {
      "epoch": 0.9052,
      "grad_norm": 3.68015193939209,
      "learning_rate": 5.436783728257788e-06,
      "loss": 4.4851,
      "step": 9052
    },
    {
      "epoch": 0.9053,
      "grad_norm": 1.9913949966430664,
      "learning_rate": 5.425436538688322e-06,
      "loss": 4.7212,
      "step": 9053
    },
    {
      "epoch": 0.9054,
      "grad_norm": 2.562159538269043,
      "learning_rate": 5.414100872742534e-06,
      "loss": 5.2286,
      "step": 9054
    },
    {
      "epoch": 0.9055,
      "grad_norm": 2.949126958847046,
      "learning_rate": 5.402776731801662e-06,
      "loss": 4.7647,
      "step": 9055
    },
    {
      "epoch": 0.9056,
      "grad_norm": 3.5785090923309326,
      "learning_rate": 5.39146411724547e-06,
      "loss": 5.0857,
      "step": 9056
    },
    {
      "epoch": 0.9057,
      "grad_norm": 4.6134724617004395,
      "learning_rate": 5.380163030452412e-06,
      "loss": 4.621,
      "step": 9057
    },
    {
      "epoch": 0.9058,
      "grad_norm": 2.143657684326172,
      "learning_rate": 5.368873472799474e-06,
      "loss": 4.6355,
      "step": 9058
    },
    {
      "epoch": 0.9059,
      "grad_norm": 3.5191171169281006,
      "learning_rate": 5.357595445662267e-06,
      "loss": 4.8912,
      "step": 9059
    },
    {
      "epoch": 0.906,
      "grad_norm": 2.8404064178466797,
      "learning_rate": 5.346328950414969e-06,
      "loss": 5.0338,
      "step": 9060
    },
    {
      "epoch": 0.9061,
      "grad_norm": 2.0741360187530518,
      "learning_rate": 5.335073988430372e-06,
      "loss": 4.694,
      "step": 9061
    },
    {
      "epoch": 0.9062,
      "grad_norm": 2.2198848724365234,
      "learning_rate": 5.3238305610798565e-06,
      "loss": 4.3256,
      "step": 9062
    },
    {
      "epoch": 0.9063,
      "grad_norm": 2.536454677581787,
      "learning_rate": 5.312598669733404e-06,
      "loss": 4.2989,
      "step": 9063
    },
    {
      "epoch": 0.9064,
      "grad_norm": 2.4382641315460205,
      "learning_rate": 5.301378315759598e-06,
      "loss": 5.0219,
      "step": 9064
    },
    {
      "epoch": 0.9065,
      "grad_norm": 2.481003999710083,
      "learning_rate": 5.290169500525577e-06,
      "loss": 4.422,
      "step": 9065
    },
    {
      "epoch": 0.9066,
      "grad_norm": 3.0810980796813965,
      "learning_rate": 5.278972225397127e-06,
      "loss": 4.6938,
      "step": 9066
    },
    {
      "epoch": 0.9067,
      "grad_norm": 2.306452512741089,
      "learning_rate": 5.267786491738569e-06,
      "loss": 4.651,
      "step": 9067
    },
    {
      "epoch": 0.9068,
      "grad_norm": 2.4790782928466797,
      "learning_rate": 5.256612300912911e-06,
      "loss": 4.7956,
      "step": 9068
    },
    {
      "epoch": 0.9069,
      "grad_norm": 2.4088592529296875,
      "learning_rate": 5.245449654281631e-06,
      "loss": 4.5785,
      "step": 9069
    },
    {
      "epoch": 0.907,
      "grad_norm": 2.4826362133026123,
      "learning_rate": 5.2342985532049084e-06,
      "loss": 4.2726,
      "step": 9070
    },
    {
      "epoch": 0.9071,
      "grad_norm": 2.112189769744873,
      "learning_rate": 5.223158999041444e-06,
      "loss": 4.6473,
      "step": 9071
    },
    {
      "epoch": 0.9072,
      "grad_norm": 4.209081649780273,
      "learning_rate": 5.212030993148553e-06,
      "loss": 4.5463,
      "step": 9072
    },
    {
      "epoch": 0.9073,
      "grad_norm": 3.8482577800750732,
      "learning_rate": 5.200914536882185e-06,
      "loss": 4.8954,
      "step": 9073
    },
    {
      "epoch": 0.9074,
      "grad_norm": 4.320698261260986,
      "learning_rate": 5.189809631596798e-06,
      "loss": 4.6663,
      "step": 9074
    },
    {
      "epoch": 0.9075,
      "grad_norm": 2.357785224914551,
      "learning_rate": 5.178716278645535e-06,
      "loss": 4.6487,
      "step": 9075
    },
    {
      "epoch": 0.9076,
      "grad_norm": 3.0770223140716553,
      "learning_rate": 5.167634479380068e-06,
      "loss": 5.131,
      "step": 9076
    },
    {
      "epoch": 0.9077,
      "grad_norm": 2.7916502952575684,
      "learning_rate": 5.1565642351506845e-06,
      "loss": 4.8723,
      "step": 9077
    },
    {
      "epoch": 0.9078,
      "grad_norm": 6.042062282562256,
      "learning_rate": 5.145505547306251e-06,
      "loss": 4.6283,
      "step": 9078
    },
    {
      "epoch": 0.9079,
      "grad_norm": 7.802548408508301,
      "learning_rate": 5.134458417194254e-06,
      "loss": 4.6693,
      "step": 9079
    },
    {
      "epoch": 0.908,
      "grad_norm": 2.1963398456573486,
      "learning_rate": 5.1234228461607304e-06,
      "loss": 4.7469,
      "step": 9080
    },
    {
      "epoch": 0.9081,
      "grad_norm": 4.549583435058594,
      "learning_rate": 5.1123988355503475e-06,
      "loss": 4.403,
      "step": 9081
    },
    {
      "epoch": 0.9082,
      "grad_norm": 1.9896841049194336,
      "learning_rate": 5.101386386706342e-06,
      "loss": 4.5346,
      "step": 9082
    },
    {
      "epoch": 0.9083,
      "grad_norm": 4.106112480163574,
      "learning_rate": 5.0903855009705514e-06,
      "loss": 5.1207,
      "step": 9083
    },
    {
      "epoch": 0.9084,
      "grad_norm": 2.14424729347229,
      "learning_rate": 5.079396179683383e-06,
      "loss": 4.7198,
      "step": 9084
    },
    {
      "epoch": 0.9085,
      "grad_norm": 3.4593822956085205,
      "learning_rate": 5.068418424183874e-06,
      "loss": 4.9291,
      "step": 9085
    },
    {
      "epoch": 0.9086,
      "grad_norm": 2.9166955947875977,
      "learning_rate": 5.057452235809624e-06,
      "loss": 5.0108,
      "step": 9086
    },
    {
      "epoch": 0.9087,
      "grad_norm": 2.7720582485198975,
      "learning_rate": 5.046497615896806e-06,
      "loss": 4.9364,
      "step": 9087
    },
    {
      "epoch": 0.9088,
      "grad_norm": 3.5702896118164062,
      "learning_rate": 5.035554565780265e-06,
      "loss": 4.0689,
      "step": 9088
    },
    {
      "epoch": 0.9089,
      "grad_norm": 5.883121967315674,
      "learning_rate": 5.024623086793323e-06,
      "loss": 5.1912,
      "step": 9089
    },
    {
      "epoch": 0.909,
      "grad_norm": 2.9915964603424072,
      "learning_rate": 5.013703180267959e-06,
      "loss": 5.1889,
      "step": 9090
    },
    {
      "epoch": 0.9091,
      "grad_norm": 12.275453567504883,
      "learning_rate": 5.002794847534764e-06,
      "loss": 4.9991,
      "step": 9091
    },
    {
      "epoch": 0.9092,
      "grad_norm": 6.334139347076416,
      "learning_rate": 4.991898089922819e-06,
      "loss": 4.6316,
      "step": 9092
    },
    {
      "epoch": 0.9093,
      "grad_norm": 3.099318027496338,
      "learning_rate": 4.981012908759941e-06,
      "loss": 4.1733,
      "step": 9093
    },
    {
      "epoch": 0.9094,
      "grad_norm": 2.2805707454681396,
      "learning_rate": 4.97013930537239e-06,
      "loss": 4.556,
      "step": 9094
    },
    {
      "epoch": 0.9095,
      "grad_norm": 1.9960435628890991,
      "learning_rate": 4.959277281085129e-06,
      "loss": 4.6225,
      "step": 9095
    },
    {
      "epoch": 0.9096,
      "grad_norm": 2.7273688316345215,
      "learning_rate": 4.948426837221631e-06,
      "loss": 4.3687,
      "step": 9096
    },
    {
      "epoch": 0.9097,
      "grad_norm": 2.005902051925659,
      "learning_rate": 4.937587975103996e-06,
      "loss": 4.6847,
      "step": 9097
    },
    {
      "epoch": 0.9098,
      "grad_norm": 2.375631332397461,
      "learning_rate": 4.926760696052934e-06,
      "loss": 4.3162,
      "step": 9098
    },
    {
      "epoch": 0.9099,
      "grad_norm": 3.2998411655426025,
      "learning_rate": 4.915945001387667e-06,
      "loss": 4.789,
      "step": 9099
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.3485021591186523,
      "learning_rate": 4.905140892426097e-06,
      "loss": 4.5618,
      "step": 9100
    },
    {
      "epoch": 0.9101,
      "grad_norm": 2.2597708702087402,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 4.5259,
      "step": 9101
    },
    {
      "epoch": 0.9102,
      "grad_norm": 2.650508165359497,
      "learning_rate": 4.8835674368783666e-06,
      "loss": 4.7891,
      "step": 9102
    },
    {
      "epoch": 0.9103,
      "grad_norm": 2.677785873413086,
      "learning_rate": 4.872798092920872e-06,
      "loss": 4.739,
      "step": 9103
    },
    {
      "epoch": 0.9104,
      "grad_norm": 2.5112764835357666,
      "learning_rate": 4.862040339924378e-06,
      "loss": 4.6577,
      "step": 9104
    },
    {
      "epoch": 0.9105,
      "grad_norm": 2.1044223308563232,
      "learning_rate": 4.8512941791996726e-06,
      "loss": 4.697,
      "step": 9105
    },
    {
      "epoch": 0.9106,
      "grad_norm": 2.3266870975494385,
      "learning_rate": 4.840559612056183e-06,
      "loss": 4.3713,
      "step": 9106
    },
    {
      "epoch": 0.9107,
      "grad_norm": 4.086468696594238,
      "learning_rate": 4.829836639801843e-06,
      "loss": 4.6904,
      "step": 9107
    },
    {
      "epoch": 0.9108,
      "grad_norm": 2.561049222946167,
      "learning_rate": 4.819125263743229e-06,
      "loss": 5.1508,
      "step": 9108
    },
    {
      "epoch": 0.9109,
      "grad_norm": 2.9999258518218994,
      "learning_rate": 4.808425485185486e-06,
      "loss": 5.286,
      "step": 9109
    },
    {
      "epoch": 0.911,
      "grad_norm": 2.4396305084228516,
      "learning_rate": 4.797737305432337e-06,
      "loss": 4.6825,
      "step": 9110
    },
    {
      "epoch": 0.9111,
      "grad_norm": 3.437838554382324,
      "learning_rate": 4.7870607257861415e-06,
      "loss": 5.0634,
      "step": 9111
    },
    {
      "epoch": 0.9112,
      "grad_norm": 2.444394588470459,
      "learning_rate": 4.776395747547757e-06,
      "loss": 4.7877,
      "step": 9112
    },
    {
      "epoch": 0.9113,
      "grad_norm": 2.6457762718200684,
      "learning_rate": 4.765742372016735e-06,
      "loss": 4.7994,
      "step": 9113
    },
    {
      "epoch": 0.9114,
      "grad_norm": 2.2688164710998535,
      "learning_rate": 4.755100600491102e-06,
      "loss": 4.968,
      "step": 9114
    },
    {
      "epoch": 0.9115,
      "grad_norm": 2.0134451389312744,
      "learning_rate": 4.744470434267567e-06,
      "loss": 4.5503,
      "step": 9115
    },
    {
      "epoch": 0.9116,
      "grad_norm": 2.268183708190918,
      "learning_rate": 4.733851874641382e-06,
      "loss": 4.0059,
      "step": 9116
    },
    {
      "epoch": 0.9117,
      "grad_norm": 2.566993474960327,
      "learning_rate": 4.723244922906356e-06,
      "loss": 4.6486,
      "step": 9117
    },
    {
      "epoch": 0.9118,
      "grad_norm": 2.506260395050049,
      "learning_rate": 4.712649580354933e-06,
      "loss": 4.6988,
      "step": 9118
    },
    {
      "epoch": 0.9119,
      "grad_norm": 2.842834711074829,
      "learning_rate": 4.702065848278126e-06,
      "loss": 4.8598,
      "step": 9119
    },
    {
      "epoch": 0.912,
      "grad_norm": 2.1821181774139404,
      "learning_rate": 4.691493727965512e-06,
      "loss": 4.3606,
      "step": 9120
    },
    {
      "epoch": 0.9121,
      "grad_norm": 3.061953544616699,
      "learning_rate": 4.680933220705308e-06,
      "loss": 4.6764,
      "step": 9121
    },
    {
      "epoch": 0.9122,
      "grad_norm": 1.8601408004760742,
      "learning_rate": 4.670384327784239e-06,
      "loss": 4.5475,
      "step": 9122
    },
    {
      "epoch": 0.9123,
      "grad_norm": 2.468593120574951,
      "learning_rate": 4.659847050487687e-06,
      "loss": 5.2841,
      "step": 9123
    },
    {
      "epoch": 0.9124,
      "grad_norm": 3.080652952194214,
      "learning_rate": 4.64932139009957e-06,
      "loss": 4.9396,
      "step": 9124
    },
    {
      "epoch": 0.9125,
      "grad_norm": 2.8477859497070312,
      "learning_rate": 4.638807347902408e-06,
      "loss": 5.1647,
      "step": 9125
    },
    {
      "epoch": 0.9126,
      "grad_norm": 2.053297758102417,
      "learning_rate": 4.628304925177318e-06,
      "loss": 4.8695,
      "step": 9126
    },
    {
      "epoch": 0.9127,
      "grad_norm": 3.358342170715332,
      "learning_rate": 4.617814123203967e-06,
      "loss": 4.8265,
      "step": 9127
    },
    {
      "epoch": 0.9128,
      "grad_norm": 2.619318962097168,
      "learning_rate": 4.607334943260655e-06,
      "loss": 4.6325,
      "step": 9128
    },
    {
      "epoch": 0.9129,
      "grad_norm": 2.510906934738159,
      "learning_rate": 4.596867386624215e-06,
      "loss": 4.6701,
      "step": 9129
    },
    {
      "epoch": 0.913,
      "grad_norm": 2.3702080249786377,
      "learning_rate": 4.586411454570083e-06,
      "loss": 4.6844,
      "step": 9130
    },
    {
      "epoch": 0.9131,
      "grad_norm": 3.6825809478759766,
      "learning_rate": 4.575967148372317e-06,
      "loss": 5.0943,
      "step": 9131
    },
    {
      "epoch": 0.9132,
      "grad_norm": 2.6989400386810303,
      "learning_rate": 4.5655344693034895e-06,
      "loss": 4.5046,
      "step": 9132
    },
    {
      "epoch": 0.9133,
      "grad_norm": 2.56664776802063,
      "learning_rate": 4.555113418634805e-06,
      "loss": 4.0729,
      "step": 9133
    },
    {
      "epoch": 0.9134,
      "grad_norm": 2.230647563934326,
      "learning_rate": 4.544703997636046e-06,
      "loss": 4.5469,
      "step": 9134
    },
    {
      "epoch": 0.9135,
      "grad_norm": 4.356062412261963,
      "learning_rate": 4.534306207575545e-06,
      "loss": 4.5431,
      "step": 9135
    },
    {
      "epoch": 0.9136,
      "grad_norm": 3.4258625507354736,
      "learning_rate": 4.523920049720265e-06,
      "loss": 4.6079,
      "step": 9136
    },
    {
      "epoch": 0.9137,
      "grad_norm": 2.4256718158721924,
      "learning_rate": 4.513545525335705e-06,
      "loss": 4.8606,
      "step": 9137
    },
    {
      "epoch": 0.9138,
      "grad_norm": 2.822880983352661,
      "learning_rate": 4.5031826356859876e-06,
      "loss": 5.5544,
      "step": 9138
    },
    {
      "epoch": 0.9139,
      "grad_norm": 2.5311453342437744,
      "learning_rate": 4.492831382033791e-06,
      "loss": 4.5988,
      "step": 9139
    },
    {
      "epoch": 0.914,
      "grad_norm": 2.5664784908294678,
      "learning_rate": 4.482491765640395e-06,
      "loss": 4.5659,
      "step": 9140
    },
    {
      "epoch": 0.9141,
      "grad_norm": 2.6475608348846436,
      "learning_rate": 4.4721637877656375e-06,
      "loss": 4.6124,
      "step": 9141
    },
    {
      "epoch": 0.9142,
      "grad_norm": 2.323791742324829,
      "learning_rate": 4.461847449667955e-06,
      "loss": 4.7648,
      "step": 9142
    },
    {
      "epoch": 0.9143,
      "grad_norm": 2.086947202682495,
      "learning_rate": 4.451542752604365e-06,
      "loss": 4.5287,
      "step": 9143
    },
    {
      "epoch": 0.9144,
      "grad_norm": 3.0087625980377197,
      "learning_rate": 4.4412496978304515e-06,
      "loss": 5.5191,
      "step": 9144
    },
    {
      "epoch": 0.9145,
      "grad_norm": 2.704641819000244,
      "learning_rate": 4.4309682866004125e-06,
      "loss": 5.321,
      "step": 9145
    },
    {
      "epoch": 0.9146,
      "grad_norm": 1.8723156452178955,
      "learning_rate": 4.420698520166988e-06,
      "loss": 4.6483,
      "step": 9146
    },
    {
      "epoch": 0.9147,
      "grad_norm": 2.6842761039733887,
      "learning_rate": 4.410440399781534e-06,
      "loss": 4.9016,
      "step": 9147
    },
    {
      "epoch": 0.9148,
      "grad_norm": 2.669562339782715,
      "learning_rate": 4.400193926693952e-06,
      "loss": 4.735,
      "step": 9148
    },
    {
      "epoch": 0.9149,
      "grad_norm": 2.1441805362701416,
      "learning_rate": 4.389959102152774e-06,
      "loss": 4.6086,
      "step": 9149
    },
    {
      "epoch": 0.915,
      "grad_norm": 2.962510824203491,
      "learning_rate": 4.379735927405038e-06,
      "loss": 4.6983,
      "step": 9150
    },
    {
      "epoch": 0.9151,
      "grad_norm": 2.164241075515747,
      "learning_rate": 4.369524403696457e-06,
      "loss": 4.8489,
      "step": 9151
    },
    {
      "epoch": 0.9152,
      "grad_norm": 3.3566935062408447,
      "learning_rate": 4.3593245322712474e-06,
      "loss": 4.2586,
      "step": 9152
    },
    {
      "epoch": 0.9153,
      "grad_norm": 4.216036796569824,
      "learning_rate": 4.349136314372204e-06,
      "loss": 4.5262,
      "step": 9153
    },
    {
      "epoch": 0.9154,
      "grad_norm": 3.327827215194702,
      "learning_rate": 4.338959751240801e-06,
      "loss": 5.0816,
      "step": 9154
    },
    {
      "epoch": 0.9155,
      "grad_norm": 4.159848690032959,
      "learning_rate": 4.328794844116946e-06,
      "loss": 4.7579,
      "step": 9155
    },
    {
      "epoch": 0.9156,
      "grad_norm": 2.603423595428467,
      "learning_rate": 4.318641594239259e-06,
      "loss": 5.1009,
      "step": 9156
    },
    {
      "epoch": 0.9157,
      "grad_norm": 3.5895111560821533,
      "learning_rate": 4.308500002844862e-06,
      "loss": 5.0528,
      "step": 9157
    },
    {
      "epoch": 0.9158,
      "grad_norm": 1.9938623905181885,
      "learning_rate": 4.2983700711694665e-06,
      "loss": 4.4729,
      "step": 9158
    },
    {
      "epoch": 0.9159,
      "grad_norm": 2.9584238529205322,
      "learning_rate": 4.288251800447385e-06,
      "loss": 4.2596,
      "step": 9159
    },
    {
      "epoch": 0.916,
      "grad_norm": 2.037435531616211,
      "learning_rate": 4.278145191911509e-06,
      "loss": 5.0697,
      "step": 9160
    },
    {
      "epoch": 0.9161,
      "grad_norm": 2.1519784927368164,
      "learning_rate": 4.268050246793276e-06,
      "loss": 4.7158,
      "step": 9161
    },
    {
      "epoch": 0.9162,
      "grad_norm": 3.5219976902008057,
      "learning_rate": 4.257966966322735e-06,
      "loss": 5.1305,
      "step": 9162
    },
    {
      "epoch": 0.9163,
      "grad_norm": 4.727957725524902,
      "learning_rate": 4.247895351728504e-06,
      "loss": 4.9555,
      "step": 9163
    },
    {
      "epoch": 0.9164,
      "grad_norm": 2.7784841060638428,
      "learning_rate": 4.237835404237778e-06,
      "loss": 4.6549,
      "step": 9164
    },
    {
      "epoch": 0.9165,
      "grad_norm": 2.670410394668579,
      "learning_rate": 4.227787125076332e-06,
      "loss": 4.7025,
      "step": 9165
    },
    {
      "epoch": 0.9166,
      "grad_norm": 3.2106099128723145,
      "learning_rate": 4.217750515468522e-06,
      "loss": 4.8097,
      "step": 9166
    },
    {
      "epoch": 0.9167,
      "grad_norm": 2.2988593578338623,
      "learning_rate": 4.207725576637256e-06,
      "loss": 4.8422,
      "step": 9167
    },
    {
      "epoch": 0.9168,
      "grad_norm": 5.1974287033081055,
      "learning_rate": 4.197712309804058e-06,
      "loss": 5.118,
      "step": 9168
    },
    {
      "epoch": 0.9169,
      "grad_norm": 2.489136219024658,
      "learning_rate": 4.187710716189042e-06,
      "loss": 4.8494,
      "step": 9169
    },
    {
      "epoch": 0.917,
      "grad_norm": 2.118682861328125,
      "learning_rate": 4.177720797010831e-06,
      "loss": 4.1838,
      "step": 9170
    },
    {
      "epoch": 0.9171,
      "grad_norm": 2.127063274383545,
      "learning_rate": 4.167742553486675e-06,
      "loss": 4.6761,
      "step": 9171
    },
    {
      "epoch": 0.9172,
      "grad_norm": 7.5147929191589355,
      "learning_rate": 4.1577759868324125e-06,
      "loss": 4.8618,
      "step": 9172
    },
    {
      "epoch": 0.9173,
      "grad_norm": 1.8622329235076904,
      "learning_rate": 4.147821098262405e-06,
      "loss": 4.6328,
      "step": 9173
    },
    {
      "epoch": 0.9174,
      "grad_norm": 2.222730875015259,
      "learning_rate": 4.137877888989672e-06,
      "loss": 4.347,
      "step": 9174
    },
    {
      "epoch": 0.9175,
      "grad_norm": 2.1870129108428955,
      "learning_rate": 4.127946360225721e-06,
      "loss": 4.348,
      "step": 9175
    },
    {
      "epoch": 0.9176,
      "grad_norm": 3.1893186569213867,
      "learning_rate": 4.118026513180695e-06,
      "loss": 4.7354,
      "step": 9176
    },
    {
      "epoch": 0.9177,
      "grad_norm": 2.2987842559814453,
      "learning_rate": 4.108118349063306e-06,
      "loss": 4.581,
      "step": 9177
    },
    {
      "epoch": 0.9178,
      "grad_norm": 4.04417085647583,
      "learning_rate": 4.09822186908082e-06,
      "loss": 4.6359,
      "step": 9178
    },
    {
      "epoch": 0.9179,
      "grad_norm": 2.4476358890533447,
      "learning_rate": 4.088337074439097e-06,
      "loss": 4.6035,
      "step": 9179
    },
    {
      "epoch": 0.918,
      "grad_norm": 1.986764669418335,
      "learning_rate": 4.078463966342571e-06,
      "loss": 4.7388,
      "step": 9180
    },
    {
      "epoch": 0.9181,
      "grad_norm": 2.598003387451172,
      "learning_rate": 4.068602545994249e-06,
      "loss": 4.5884,
      "step": 9181
    },
    {
      "epoch": 0.9182,
      "grad_norm": 3.833692789077759,
      "learning_rate": 4.0587528145957235e-06,
      "loss": 4.6807,
      "step": 9182
    },
    {
      "epoch": 0.9183,
      "grad_norm": 1.9615905284881592,
      "learning_rate": 4.048914773347134e-06,
      "loss": 5.0641,
      "step": 9183
    },
    {
      "epoch": 0.9184,
      "grad_norm": 2.2766940593719482,
      "learning_rate": 4.039088423447235e-06,
      "loss": 4.772,
      "step": 9184
    },
    {
      "epoch": 0.9185,
      "grad_norm": 3.1344587802886963,
      "learning_rate": 4.029273766093333e-06,
      "loss": 4.5276,
      "step": 9185
    },
    {
      "epoch": 0.9186,
      "grad_norm": 4.628110885620117,
      "learning_rate": 4.019470802481307e-06,
      "loss": 4.6302,
      "step": 9186
    },
    {
      "epoch": 0.9187,
      "grad_norm": 3.4311788082122803,
      "learning_rate": 4.009679533805633e-06,
      "loss": 4.4292,
      "step": 9187
    },
    {
      "epoch": 0.9188,
      "grad_norm": 2.377767562866211,
      "learning_rate": 3.999899961259335e-06,
      "loss": 5.0942,
      "step": 9188
    },
    {
      "epoch": 0.9189,
      "grad_norm": 4.419395446777344,
      "learning_rate": 3.990132086034026e-06,
      "loss": 5.0901,
      "step": 9189
    },
    {
      "epoch": 0.919,
      "grad_norm": 2.450157880783081,
      "learning_rate": 3.9803759093199e-06,
      "loss": 4.6012,
      "step": 9190
    },
    {
      "epoch": 0.9191,
      "grad_norm": 2.244874954223633,
      "learning_rate": 3.970631432305694e-06,
      "loss": 4.8386,
      "step": 9191
    },
    {
      "epoch": 0.9192,
      "grad_norm": 3.0571413040161133,
      "learning_rate": 3.96089865617878e-06,
      "loss": 5.022,
      "step": 9192
    },
    {
      "epoch": 0.9193,
      "grad_norm": 3.267869710922241,
      "learning_rate": 3.951177582125021e-06,
      "loss": 4.4259,
      "step": 9193
    },
    {
      "epoch": 0.9194,
      "grad_norm": 5.390868663787842,
      "learning_rate": 3.941468211328947e-06,
      "loss": 5.4492,
      "step": 9194
    },
    {
      "epoch": 0.9195,
      "grad_norm": 2.7123026847839355,
      "learning_rate": 3.931770544973601e-06,
      "loss": 4.8025,
      "step": 9195
    },
    {
      "epoch": 0.9196,
      "grad_norm": 2.4685356616973877,
      "learning_rate": 3.922084584240582e-06,
      "loss": 4.6471,
      "step": 9196
    },
    {
      "epoch": 0.9197,
      "grad_norm": 2.1837048530578613,
      "learning_rate": 3.912410330310156e-06,
      "loss": 4.5731,
      "step": 9197
    },
    {
      "epoch": 0.9198,
      "grad_norm": 2.945399045944214,
      "learning_rate": 3.902747784361038e-06,
      "loss": 4.4126,
      "step": 9198
    },
    {
      "epoch": 0.9199,
      "grad_norm": 2.668567657470703,
      "learning_rate": 3.893096947570618e-06,
      "loss": 4.5292,
      "step": 9199
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.2640717029571533,
      "learning_rate": 3.883457821114811e-06,
      "loss": 4.3485,
      "step": 9200
    },
    {
      "epoch": 0.9201,
      "grad_norm": 3.354422092437744,
      "learning_rate": 3.873830406168111e-06,
      "loss": 4.6838,
      "step": 9201
    },
    {
      "epoch": 0.9202,
      "grad_norm": 1.9158464670181274,
      "learning_rate": 3.864214703903601e-06,
      "loss": 4.3497,
      "step": 9202
    },
    {
      "epoch": 0.9203,
      "grad_norm": 2.880230665206909,
      "learning_rate": 3.8546107154929235e-06,
      "loss": 4.6596,
      "step": 9203
    },
    {
      "epoch": 0.9204,
      "grad_norm": 2.244296073913574,
      "learning_rate": 3.845018442106285e-06,
      "loss": 4.9461,
      "step": 9204
    },
    {
      "epoch": 0.9205,
      "grad_norm": 3.0276880264282227,
      "learning_rate": 3.835437884912474e-06,
      "loss": 5.8668,
      "step": 9205
    },
    {
      "epoch": 0.9206,
      "grad_norm": 2.574514150619507,
      "learning_rate": 3.825869045078867e-06,
      "loss": 4.6341,
      "step": 9206
    },
    {
      "epoch": 0.9207,
      "grad_norm": 2.7609071731567383,
      "learning_rate": 3.816311923771387e-06,
      "loss": 4.5853,
      "step": 9207
    },
    {
      "epoch": 0.9208,
      "grad_norm": 5.75250768661499,
      "learning_rate": 3.806766522154548e-06,
      "loss": 4.8733,
      "step": 9208
    },
    {
      "epoch": 0.9209,
      "grad_norm": 7.796431541442871,
      "learning_rate": 3.797232841391407e-06,
      "loss": 6.3484,
      "step": 9209
    },
    {
      "epoch": 0.921,
      "grad_norm": 5.758908271789551,
      "learning_rate": 3.787710882643658e-06,
      "loss": 5.9336,
      "step": 9210
    },
    {
      "epoch": 0.9211,
      "grad_norm": 2.4271738529205322,
      "learning_rate": 3.7782006470714616e-06,
      "loss": 4.6736,
      "step": 9211
    },
    {
      "epoch": 0.9212,
      "grad_norm": 4.371510982513428,
      "learning_rate": 3.7687021358336683e-06,
      "loss": 4.6986,
      "step": 9212
    },
    {
      "epoch": 0.9213,
      "grad_norm": 3.814988374710083,
      "learning_rate": 3.759215350087619e-06,
      "loss": 4.9998,
      "step": 9213
    },
    {
      "epoch": 0.9214,
      "grad_norm": 3.702383279800415,
      "learning_rate": 3.7497402909892342e-06,
      "loss": 4.9874,
      "step": 9214
    },
    {
      "epoch": 0.9215,
      "grad_norm": 2.228795289993286,
      "learning_rate": 3.7402769596930563e-06,
      "loss": 4.2525,
      "step": 9215
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.8918994665145874,
      "learning_rate": 3.730825357352119e-06,
      "loss": 5.099,
      "step": 9216
    },
    {
      "epoch": 0.9217,
      "grad_norm": 2.7498319149017334,
      "learning_rate": 3.721385485118123e-06,
      "loss": 5.2561,
      "step": 9217
    },
    {
      "epoch": 0.9218,
      "grad_norm": 2.964390277862549,
      "learning_rate": 3.711957344141237e-06,
      "loss": 4.6836,
      "step": 9218
    },
    {
      "epoch": 0.9219,
      "grad_norm": 4.907686233520508,
      "learning_rate": 3.7025409355702976e-06,
      "loss": 4.6891,
      "step": 9219
    },
    {
      "epoch": 0.922,
      "grad_norm": 2.4909181594848633,
      "learning_rate": 3.693136260552632e-06,
      "loss": 4.2523,
      "step": 9220
    },
    {
      "epoch": 0.9221,
      "grad_norm": 3.6965808868408203,
      "learning_rate": 3.68374332023419e-06,
      "loss": 4.9436,
      "step": 9221
    },
    {
      "epoch": 0.9222,
      "grad_norm": 2.7386386394500732,
      "learning_rate": 3.6743621157594555e-06,
      "loss": 5.3661,
      "step": 9222
    },
    {
      "epoch": 0.9223,
      "grad_norm": 2.6203908920288086,
      "learning_rate": 3.664992648271526e-06,
      "loss": 4.5546,
      "step": 9223
    },
    {
      "epoch": 0.9224,
      "grad_norm": 2.0776917934417725,
      "learning_rate": 3.6556349189120097e-06,
      "loss": 4.8146,
      "step": 9224
    },
    {
      "epoch": 0.9225,
      "grad_norm": 4.042715549468994,
      "learning_rate": 3.6462889288211512e-06,
      "loss": 4.9091,
      "step": 9225
    },
    {
      "epoch": 0.9226,
      "grad_norm": 2.939584255218506,
      "learning_rate": 3.6369546791377052e-06,
      "loss": 4.1754,
      "step": 9226
    },
    {
      "epoch": 0.9227,
      "grad_norm": 4.613167762756348,
      "learning_rate": 3.627632170999029e-06,
      "loss": 4.6556,
      "step": 9227
    },
    {
      "epoch": 0.9228,
      "grad_norm": 3.3909854888916016,
      "learning_rate": 3.6183214055410586e-06,
      "loss": 4.5824,
      "step": 9228
    },
    {
      "epoch": 0.9229,
      "grad_norm": 4.002730846405029,
      "learning_rate": 3.6090223838982417e-06,
      "loss": 4.6153,
      "step": 9229
    },
    {
      "epoch": 0.923,
      "grad_norm": 4.971938133239746,
      "learning_rate": 3.5997351072036945e-06,
      "loss": 5.2175,
      "step": 9230
    },
    {
      "epoch": 0.9231,
      "grad_norm": 2.4571313858032227,
      "learning_rate": 3.5904595765890005e-06,
      "loss": 4.8858,
      "step": 9231
    },
    {
      "epoch": 0.9232,
      "grad_norm": 5.602231502532959,
      "learning_rate": 3.5811957931843554e-06,
      "loss": 5.752,
      "step": 9232
    },
    {
      "epoch": 0.9233,
      "grad_norm": 2.607621669769287,
      "learning_rate": 3.5719437581185454e-06,
      "loss": 4.5092,
      "step": 9233
    },
    {
      "epoch": 0.9234,
      "grad_norm": 3.6863081455230713,
      "learning_rate": 3.562703472518869e-06,
      "loss": 4.5987,
      "step": 9234
    },
    {
      "epoch": 0.9235,
      "grad_norm": 3.4029366970062256,
      "learning_rate": 3.553474937511281e-06,
      "loss": 4.3056,
      "step": 9235
    },
    {
      "epoch": 0.9236,
      "grad_norm": 4.8242292404174805,
      "learning_rate": 3.5442581542201923e-06,
      "loss": 5.6538,
      "step": 9236
    },
    {
      "epoch": 0.9237,
      "grad_norm": 2.3755390644073486,
      "learning_rate": 3.5350531237686724e-06,
      "loss": 4.7639,
      "step": 9237
    },
    {
      "epoch": 0.9238,
      "grad_norm": 2.5690205097198486,
      "learning_rate": 3.5258598472783233e-06,
      "loss": 4.9446,
      "step": 9238
    },
    {
      "epoch": 0.9239,
      "grad_norm": 4.5282673835754395,
      "learning_rate": 3.516678325869316e-06,
      "loss": 4.7929,
      "step": 9239
    },
    {
      "epoch": 0.924,
      "grad_norm": 2.046513557434082,
      "learning_rate": 3.5075085606604e-06,
      "loss": 4.6175,
      "step": 9240
    },
    {
      "epoch": 0.9241,
      "grad_norm": 2.726705312728882,
      "learning_rate": 3.4983505527688586e-06,
      "loss": 5.3431,
      "step": 9241
    },
    {
      "epoch": 0.9242,
      "grad_norm": 7.129289150238037,
      "learning_rate": 3.489204303310578e-06,
      "loss": 4.5192,
      "step": 9242
    },
    {
      "epoch": 0.9243,
      "grad_norm": 3.205247640609741,
      "learning_rate": 3.480069813400022e-06,
      "loss": 4.742,
      "step": 9243
    },
    {
      "epoch": 0.9244,
      "grad_norm": 1.8748856782913208,
      "learning_rate": 3.470947084150167e-06,
      "loss": 4.3777,
      "step": 9244
    },
    {
      "epoch": 0.9245,
      "grad_norm": 3.500152587890625,
      "learning_rate": 3.461836116672612e-06,
      "loss": 4.9603,
      "step": 9245
    },
    {
      "epoch": 0.9246,
      "grad_norm": 3.1695122718811035,
      "learning_rate": 3.452736912077503e-06,
      "loss": 4.7181,
      "step": 9246
    },
    {
      "epoch": 0.9247,
      "grad_norm": 3.7021267414093018,
      "learning_rate": 3.443649471473531e-06,
      "loss": 4.979,
      "step": 9247
    },
    {
      "epoch": 0.9248,
      "grad_norm": 9.373987197875977,
      "learning_rate": 3.434573795967988e-06,
      "loss": 5.995,
      "step": 9248
    },
    {
      "epoch": 0.9249,
      "grad_norm": 2.5462281703948975,
      "learning_rate": 3.425509886666711e-06,
      "loss": 4.9935,
      "step": 9249
    },
    {
      "epoch": 0.925,
      "grad_norm": 6.158668518066406,
      "learning_rate": 3.4164577446741174e-06,
      "loss": 4.6637,
      "step": 9250
    },
    {
      "epoch": 0.9251,
      "grad_norm": 4.005648612976074,
      "learning_rate": 3.40741737109318e-06,
      "loss": 5.109,
      "step": 9251
    },
    {
      "epoch": 0.9252,
      "grad_norm": 5.715511322021484,
      "learning_rate": 3.3983887670254177e-06,
      "loss": 4.7982,
      "step": 9252
    },
    {
      "epoch": 0.9253,
      "grad_norm": 3.077838659286499,
      "learning_rate": 3.389371933570995e-06,
      "loss": 4.4437,
      "step": 9253
    },
    {
      "epoch": 0.9254,
      "grad_norm": 3.6403582096099854,
      "learning_rate": 3.380366871828522e-06,
      "loss": 4.997,
      "step": 9254
    },
    {
      "epoch": 0.9255,
      "grad_norm": 4.286044597625732,
      "learning_rate": 3.3713735828952987e-06,
      "loss": 5.7006,
      "step": 9255
    },
    {
      "epoch": 0.9256,
      "grad_norm": 2.566572427749634,
      "learning_rate": 3.3623920678670597e-06,
      "loss": 4.339,
      "step": 9256
    },
    {
      "epoch": 0.9257,
      "grad_norm": 2.1513936519622803,
      "learning_rate": 3.3534223278382405e-06,
      "loss": 4.9726,
      "step": 9257
    },
    {
      "epoch": 0.9258,
      "grad_norm": 2.50138521194458,
      "learning_rate": 3.344464363901756e-06,
      "loss": 4.2006,
      "step": 9258
    },
    {
      "epoch": 0.9259,
      "grad_norm": 1.7635844945907593,
      "learning_rate": 3.3355181771490772e-06,
      "loss": 4.6123,
      "step": 9259
    },
    {
      "epoch": 0.926,
      "grad_norm": 2.663350820541382,
      "learning_rate": 3.3265837686703106e-06,
      "loss": 4.5184,
      "step": 9260
    },
    {
      "epoch": 0.9261,
      "grad_norm": 2.9554014205932617,
      "learning_rate": 3.3176611395540626e-06,
      "loss": 5.1237,
      "step": 9261
    },
    {
      "epoch": 0.9262,
      "grad_norm": 2.039865016937256,
      "learning_rate": 3.3087502908875413e-06,
      "loss": 4.4014,
      "step": 9262
    },
    {
      "epoch": 0.9263,
      "grad_norm": 3.0619871616363525,
      "learning_rate": 3.2998512237565005e-06,
      "loss": 4.9829,
      "step": 9263
    },
    {
      "epoch": 0.9264,
      "grad_norm": 4.374544620513916,
      "learning_rate": 3.290963939245262e-06,
      "loss": 4.3646,
      "step": 9264
    },
    {
      "epoch": 0.9265,
      "grad_norm": 2.0360350608825684,
      "learning_rate": 3.2820884384367146e-06,
      "loss": 4.6108,
      "step": 9265
    },
    {
      "epoch": 0.9266,
      "grad_norm": 2.8853256702423096,
      "learning_rate": 3.273224722412327e-06,
      "loss": 4.4213,
      "step": 9266
    },
    {
      "epoch": 0.9267,
      "grad_norm": 2.3181614875793457,
      "learning_rate": 3.2643727922520906e-06,
      "loss": 4.4622,
      "step": 9267
    },
    {
      "epoch": 0.9268,
      "grad_norm": 3.3000802993774414,
      "learning_rate": 3.2555326490346095e-06,
      "loss": 4.9724,
      "step": 9268
    },
    {
      "epoch": 0.9269,
      "grad_norm": 2.275583267211914,
      "learning_rate": 3.246704293837011e-06,
      "loss": 4.1368,
      "step": 9269
    },
    {
      "epoch": 0.927,
      "grad_norm": 2.011828899383545,
      "learning_rate": 3.2378877277350116e-06,
      "loss": 4.4617,
      "step": 9270
    },
    {
      "epoch": 0.9271,
      "grad_norm": 2.1078450679779053,
      "learning_rate": 3.2290829518028862e-06,
      "loss": 4.3581,
      "step": 9271
    },
    {
      "epoch": 0.9272,
      "grad_norm": 5.879441738128662,
      "learning_rate": 3.2202899671134546e-06,
      "loss": 5.951,
      "step": 9272
    },
    {
      "epoch": 0.9273,
      "grad_norm": 3.584883689880371,
      "learning_rate": 3.2115087747381366e-06,
      "loss": 4.8458,
      "step": 9273
    },
    {
      "epoch": 0.9274,
      "grad_norm": 2.5250298976898193,
      "learning_rate": 3.2027393757468773e-06,
      "loss": 4.6163,
      "step": 9274
    },
    {
      "epoch": 0.9275,
      "grad_norm": 5.112006664276123,
      "learning_rate": 3.19398177120821e-06,
      "loss": 4.372,
      "step": 9275
    },
    {
      "epoch": 0.9276,
      "grad_norm": 3.4228105545043945,
      "learning_rate": 3.1852359621892367e-06,
      "loss": 5.3701,
      "step": 9276
    },
    {
      "epoch": 0.9277,
      "grad_norm": 2.435572862625122,
      "learning_rate": 3.1765019497555616e-06,
      "loss": 4.5844,
      "step": 9277
    },
    {
      "epoch": 0.9278,
      "grad_norm": 4.596418857574463,
      "learning_rate": 3.1677797349714544e-06,
      "loss": 4.702,
      "step": 9278
    },
    {
      "epoch": 0.9279,
      "grad_norm": 2.542449712753296,
      "learning_rate": 3.1590693188996323e-06,
      "loss": 4.4366,
      "step": 9279
    },
    {
      "epoch": 0.928,
      "grad_norm": 2.382068157196045,
      "learning_rate": 3.1503707026014906e-06,
      "loss": 4.9937,
      "step": 9280
    },
    {
      "epoch": 0.9281,
      "grad_norm": 2.5492029190063477,
      "learning_rate": 3.1416838871368924e-06,
      "loss": 4.7507,
      "step": 9281
    },
    {
      "epoch": 0.9282,
      "grad_norm": 3.1058993339538574,
      "learning_rate": 3.1330088735643025e-06,
      "loss": 5.8566,
      "step": 9282
    },
    {
      "epoch": 0.9283,
      "grad_norm": 2.3727829456329346,
      "learning_rate": 3.124345662940764e-06,
      "loss": 4.5268,
      "step": 9283
    },
    {
      "epoch": 0.9284,
      "grad_norm": 3.5301403999328613,
      "learning_rate": 3.115694256321855e-06,
      "loss": 4.6634,
      "step": 9284
    },
    {
      "epoch": 0.9285,
      "grad_norm": 7.406538009643555,
      "learning_rate": 3.1070546547617095e-06,
      "loss": 5.4706,
      "step": 9285
    },
    {
      "epoch": 0.9286,
      "grad_norm": 2.613391399383545,
      "learning_rate": 3.0984268593130528e-06,
      "loss": 4.5712,
      "step": 9286
    },
    {
      "epoch": 0.9287,
      "grad_norm": 3.8792965412139893,
      "learning_rate": 3.0898108710271434e-06,
      "loss": 5.0935,
      "step": 9287
    },
    {
      "epoch": 0.9288,
      "grad_norm": 2.6975319385528564,
      "learning_rate": 3.081206690953831e-06,
      "loss": 4.7049,
      "step": 9288
    },
    {
      "epoch": 0.9289,
      "grad_norm": 2.7136754989624023,
      "learning_rate": 3.072614320141487e-06,
      "loss": 4.5565,
      "step": 9289
    },
    {
      "epoch": 0.929,
      "grad_norm": 4.149767875671387,
      "learning_rate": 3.064033759637064e-06,
      "loss": 5.0208,
      "step": 9290
    },
    {
      "epoch": 0.9291,
      "grad_norm": 2.51459002494812,
      "learning_rate": 3.0554650104861136e-06,
      "loss": 4.3599,
      "step": 9291
    },
    {
      "epoch": 0.9292,
      "grad_norm": 1.9615026712417603,
      "learning_rate": 3.046908073732668e-06,
      "loss": 4.4255,
      "step": 9292
    },
    {
      "epoch": 0.9293,
      "grad_norm": 1.8522286415100098,
      "learning_rate": 3.0383629504194046e-06,
      "loss": 4.3466,
      "step": 9293
    },
    {
      "epoch": 0.9294,
      "grad_norm": 2.015774726867676,
      "learning_rate": 3.0298296415874894e-06,
      "loss": 4.4212,
      "step": 9294
    },
    {
      "epoch": 0.9295,
      "grad_norm": 3.5141727924346924,
      "learning_rate": 3.0213081482766805e-06,
      "loss": 4.5217,
      "step": 9295
    },
    {
      "epoch": 0.9296,
      "grad_norm": 3.399407148361206,
      "learning_rate": 3.012798471525324e-06,
      "loss": 4.4867,
      "step": 9296
    },
    {
      "epoch": 0.9297,
      "grad_norm": 3.0361483097076416,
      "learning_rate": 3.0043006123702697e-06,
      "loss": 4.4199,
      "step": 9297
    },
    {
      "epoch": 0.9298,
      "grad_norm": 2.3871870040893555,
      "learning_rate": 2.9958145718469777e-06,
      "loss": 4.463,
      "step": 9298
    },
    {
      "epoch": 0.9299,
      "grad_norm": 2.4290106296539307,
      "learning_rate": 2.9873403509894203e-06,
      "loss": 4.3738,
      "step": 9299
    },
    {
      "epoch": 0.93,
      "grad_norm": 3.303361415863037,
      "learning_rate": 2.978877950830172e-06,
      "loss": 4.2829,
      "step": 9300
    },
    {
      "epoch": 0.9301,
      "grad_norm": 2.775813341140747,
      "learning_rate": 2.970427372400353e-06,
      "loss": 4.3093,
      "step": 9301
    },
    {
      "epoch": 0.9302,
      "grad_norm": 3.08012056350708,
      "learning_rate": 2.9619886167296384e-06,
      "loss": 4.6251,
      "step": 9302
    },
    {
      "epoch": 0.9303,
      "grad_norm": 1.9737370014190674,
      "learning_rate": 2.953561684846262e-06,
      "loss": 4.5838,
      "step": 9303
    },
    {
      "epoch": 0.9304,
      "grad_norm": 5.4719648361206055,
      "learning_rate": 2.9451465777770247e-06,
      "loss": 4.3769,
      "step": 9304
    },
    {
      "epoch": 0.9305,
      "grad_norm": 3.083738088607788,
      "learning_rate": 2.936743296547273e-06,
      "loss": 4.5713,
      "step": 9305
    },
    {
      "epoch": 0.9306,
      "grad_norm": 4.390488624572754,
      "learning_rate": 2.928351842180921e-06,
      "loss": 4.8368,
      "step": 9306
    },
    {
      "epoch": 0.9307,
      "grad_norm": 1.9234371185302734,
      "learning_rate": 2.9199722157004616e-06,
      "loss": 4.4601,
      "step": 9307
    },
    {
      "epoch": 0.9308,
      "grad_norm": 4.372253894805908,
      "learning_rate": 2.9116044181269007e-06,
      "loss": 4.9663,
      "step": 9308
    },
    {
      "epoch": 0.9309,
      "grad_norm": 5.670618534088135,
      "learning_rate": 2.9032484504798452e-06,
      "loss": 5.1081,
      "step": 9309
    },
    {
      "epoch": 0.931,
      "grad_norm": 2.7642855644226074,
      "learning_rate": 2.8949043137774355e-06,
      "loss": 4.5926,
      "step": 9310
    },
    {
      "epoch": 0.9311,
      "grad_norm": 3.024535655975342,
      "learning_rate": 2.8865720090364034e-06,
      "loss": 4.5905,
      "step": 9311
    },
    {
      "epoch": 0.9312,
      "grad_norm": 2.8902363777160645,
      "learning_rate": 2.878251537271981e-06,
      "loss": 4.6634,
      "step": 9312
    },
    {
      "epoch": 0.9313,
      "grad_norm": 3.6779298782348633,
      "learning_rate": 2.8699428994980017e-06,
      "loss": 4.7538,
      "step": 9313
    },
    {
      "epoch": 0.9314,
      "grad_norm": 2.710714101791382,
      "learning_rate": 2.8616460967268667e-06,
      "loss": 4.5269,
      "step": 9314
    },
    {
      "epoch": 0.9315,
      "grad_norm": 3.2925331592559814,
      "learning_rate": 2.8533611299694783e-06,
      "loss": 4.8485,
      "step": 9315
    },
    {
      "epoch": 0.9316,
      "grad_norm": 3.348539352416992,
      "learning_rate": 2.845088000235396e-06,
      "loss": 4.6649,
      "step": 9316
    },
    {
      "epoch": 0.9317,
      "grad_norm": 2.3825795650482178,
      "learning_rate": 2.836826708532603e-06,
      "loss": 4.908,
      "step": 9317
    },
    {
      "epoch": 0.9318,
      "grad_norm": 2.481471061706543,
      "learning_rate": 2.8285772558677705e-06,
      "loss": 4.6363,
      "step": 9318
    },
    {
      "epoch": 0.9319,
      "grad_norm": 1.9198254346847534,
      "learning_rate": 2.8203396432460506e-06,
      "loss": 4.5706,
      "step": 9319
    },
    {
      "epoch": 0.932,
      "grad_norm": 2.027737855911255,
      "learning_rate": 2.8121138716711404e-06,
      "loss": 4.5746,
      "step": 9320
    },
    {
      "epoch": 0.9321,
      "grad_norm": 2.878157615661621,
      "learning_rate": 2.8038999421453826e-06,
      "loss": 4.9974,
      "step": 9321
    },
    {
      "epoch": 0.9322,
      "grad_norm": 3.3677985668182373,
      "learning_rate": 2.7956978556695767e-06,
      "loss": 5.4789,
      "step": 9322
    },
    {
      "epoch": 0.9323,
      "grad_norm": 2.2448723316192627,
      "learning_rate": 2.7875076132431344e-06,
      "loss": 4.813,
      "step": 9323
    },
    {
      "epoch": 0.9324,
      "grad_norm": 2.4188406467437744,
      "learning_rate": 2.7793292158640126e-06,
      "loss": 4.8326,
      "step": 9324
    },
    {
      "epoch": 0.9325,
      "grad_norm": 2.340517044067383,
      "learning_rate": 2.771162664528726e-06,
      "loss": 4.5244,
      "step": 9325
    },
    {
      "epoch": 0.9326,
      "grad_norm": 3.7820563316345215,
      "learning_rate": 2.7630079602323442e-06,
      "loss": 4.9096,
      "step": 9326
    },
    {
      "epoch": 0.9327,
      "grad_norm": 2.5736684799194336,
      "learning_rate": 2.7548651039684846e-06,
      "loss": 4.519,
      "step": 9327
    },
    {
      "epoch": 0.9328,
      "grad_norm": 3.7934603691101074,
      "learning_rate": 2.746734096729342e-06,
      "loss": 4.562,
      "step": 9328
    },
    {
      "epoch": 0.9329,
      "grad_norm": 4.47434663772583,
      "learning_rate": 2.738614939505646e-06,
      "loss": 4.5002,
      "step": 9329
    },
    {
      "epoch": 0.933,
      "grad_norm": 3.1846325397491455,
      "learning_rate": 2.7305076332867054e-06,
      "loss": 4.5868,
      "step": 9330
    },
    {
      "epoch": 0.9331,
      "grad_norm": 2.8509202003479004,
      "learning_rate": 2.7224121790603517e-06,
      "loss": 5.2118,
      "step": 9331
    },
    {
      "epoch": 0.9332,
      "grad_norm": 3.2100718021392822,
      "learning_rate": 2.714328577812997e-06,
      "loss": 4.3424,
      "step": 9332
    },
    {
      "epoch": 0.9333,
      "grad_norm": 3.5030643939971924,
      "learning_rate": 2.7062568305295965e-06,
      "loss": 4.6585,
      "step": 9333
    },
    {
      "epoch": 0.9334,
      "grad_norm": 3.790889263153076,
      "learning_rate": 2.6981969381936978e-06,
      "loss": 5.2772,
      "step": 9334
    },
    {
      "epoch": 0.9335,
      "grad_norm": 3.737830400466919,
      "learning_rate": 2.690148901787337e-06,
      "loss": 4.7174,
      "step": 9335
    },
    {
      "epoch": 0.9336,
      "grad_norm": 1.8036746978759766,
      "learning_rate": 2.6821127222911857e-06,
      "loss": 4.1646,
      "step": 9336
    },
    {
      "epoch": 0.9337,
      "grad_norm": 2.153590679168701,
      "learning_rate": 2.6740884006843825e-06,
      "loss": 4.552,
      "step": 9337
    },
    {
      "epoch": 0.9338,
      "grad_norm": 5.9091081619262695,
      "learning_rate": 2.66607593794469e-06,
      "loss": 5.9409,
      "step": 9338
    },
    {
      "epoch": 0.9339,
      "grad_norm": 4.545868396759033,
      "learning_rate": 2.6580753350484046e-06,
      "loss": 4.5371,
      "step": 9339
    },
    {
      "epoch": 0.934,
      "grad_norm": 1.9378727674484253,
      "learning_rate": 2.650086592970358e-06,
      "loss": 4.4666,
      "step": 9340
    },
    {
      "epoch": 0.9341,
      "grad_norm": 1.808810830116272,
      "learning_rate": 2.6421097126839712e-06,
      "loss": 4.4231,
      "step": 9341
    },
    {
      "epoch": 0.9342,
      "grad_norm": 5.608677387237549,
      "learning_rate": 2.6341446951612005e-06,
      "loss": 5.8552,
      "step": 9342
    },
    {
      "epoch": 0.9343,
      "grad_norm": 2.313518762588501,
      "learning_rate": 2.6261915413725578e-06,
      "loss": 4.7723,
      "step": 9343
    },
    {
      "epoch": 0.9344,
      "grad_norm": 3.85229229927063,
      "learning_rate": 2.618250252287113e-06,
      "loss": 4.7215,
      "step": 9344
    },
    {
      "epoch": 0.9345,
      "grad_norm": 3.3754544258117676,
      "learning_rate": 2.610320828872481e-06,
      "loss": 4.5676,
      "step": 9345
    },
    {
      "epoch": 0.9346,
      "grad_norm": 2.613022565841675,
      "learning_rate": 2.6024032720948443e-06,
      "loss": 4.8271,
      "step": 9346
    },
    {
      "epoch": 0.9347,
      "grad_norm": 2.072347640991211,
      "learning_rate": 2.59449758291892e-06,
      "loss": 4.6764,
      "step": 9347
    },
    {
      "epoch": 0.9348,
      "grad_norm": 1.8827027082443237,
      "learning_rate": 2.5866037623080153e-06,
      "loss": 4.3346,
      "step": 9348
    },
    {
      "epoch": 0.9349,
      "grad_norm": 2.2022411823272705,
      "learning_rate": 2.5787218112239496e-06,
      "loss": 4.9336,
      "step": 9349
    },
    {
      "epoch": 0.935,
      "grad_norm": 3.507707357406616,
      "learning_rate": 2.570851730627122e-06,
      "loss": 4.5849,
      "step": 9350
    },
    {
      "epoch": 0.9351,
      "grad_norm": 14.51577091217041,
      "learning_rate": 2.5629935214764865e-06,
      "loss": 5.5587,
      "step": 9351
    },
    {
      "epoch": 0.9352,
      "grad_norm": 2.468204975128174,
      "learning_rate": 2.5551471847295228e-06,
      "loss": 4.6299,
      "step": 9352
    },
    {
      "epoch": 0.9353,
      "grad_norm": 2.479893207550049,
      "learning_rate": 2.5473127213422763e-06,
      "loss": 4.7981,
      "step": 9353
    },
    {
      "epoch": 0.9354,
      "grad_norm": 1.6207401752471924,
      "learning_rate": 2.5394901322694067e-06,
      "loss": 4.5881,
      "step": 9354
    },
    {
      "epoch": 0.9355,
      "grad_norm": 1.9914628267288208,
      "learning_rate": 2.531679418464006e-06,
      "loss": 4.4945,
      "step": 9355
    },
    {
      "epoch": 0.9356,
      "grad_norm": 2.457672119140625,
      "learning_rate": 2.5238805808778242e-06,
      "loss": 4.2241,
      "step": 9356
    },
    {
      "epoch": 0.9357,
      "grad_norm": 6.0731072425842285,
      "learning_rate": 2.516093620461124e-06,
      "loss": 6.2362,
      "step": 9357
    },
    {
      "epoch": 0.9358,
      "grad_norm": 2.4246933460235596,
      "learning_rate": 2.508318538162702e-06,
      "loss": 4.3492,
      "step": 9358
    },
    {
      "epoch": 0.9359,
      "grad_norm": 4.11655330657959,
      "learning_rate": 2.5005553349299547e-06,
      "loss": 4.8221,
      "step": 9359
    },
    {
      "epoch": 0.936,
      "grad_norm": 3.127011299133301,
      "learning_rate": 2.4928040117087827e-06,
      "loss": 4.6231,
      "step": 9360
    },
    {
      "epoch": 0.9361,
      "grad_norm": 3.19864821434021,
      "learning_rate": 2.4850645694436736e-06,
      "loss": 4.6199,
      "step": 9361
    },
    {
      "epoch": 0.9362,
      "grad_norm": 2.94156551361084,
      "learning_rate": 2.4773370090776626e-06,
      "loss": 4.5866,
      "step": 9362
    },
    {
      "epoch": 0.9363,
      "grad_norm": 3.57938551902771,
      "learning_rate": 2.4696213315523074e-06,
      "loss": 4.9242,
      "step": 9363
    },
    {
      "epoch": 0.9364,
      "grad_norm": 1.875213623046875,
      "learning_rate": 2.4619175378077565e-06,
      "loss": 4.4553,
      "step": 9364
    },
    {
      "epoch": 0.9365,
      "grad_norm": 2.0333666801452637,
      "learning_rate": 2.4542256287826914e-06,
      "loss": 4.1171,
      "step": 9365
    },
    {
      "epoch": 0.9366,
      "grad_norm": 2.1773273944854736,
      "learning_rate": 2.446545605414341e-06,
      "loss": 4.7505,
      "step": 9366
    },
    {
      "epoch": 0.9367,
      "grad_norm": 3.4172916412353516,
      "learning_rate": 2.4388774686385007e-06,
      "loss": 4.4234,
      "step": 9367
    },
    {
      "epoch": 0.9368,
      "grad_norm": 2.1550278663635254,
      "learning_rate": 2.4312212193895125e-06,
      "loss": 4.4427,
      "step": 9368
    },
    {
      "epoch": 0.9369,
      "grad_norm": 3.1958119869232178,
      "learning_rate": 2.4235768586002517e-06,
      "loss": 4.5103,
      "step": 9369
    },
    {
      "epoch": 0.937,
      "grad_norm": 4.205552577972412,
      "learning_rate": 2.415944387202174e-06,
      "loss": 4.9308,
      "step": 9370
    },
    {
      "epoch": 0.9371,
      "grad_norm": 2.9482011795043945,
      "learning_rate": 2.4083238061252567e-06,
      "loss": 4.3684,
      "step": 9371
    },
    {
      "epoch": 0.9372,
      "grad_norm": 2.1869781017303467,
      "learning_rate": 2.40071511629808e-06,
      "loss": 4.3952,
      "step": 9372
    },
    {
      "epoch": 0.9373,
      "grad_norm": 3.786635398864746,
      "learning_rate": 2.3931183186477026e-06,
      "loss": 4.5351,
      "step": 9373
    },
    {
      "epoch": 0.9374,
      "grad_norm": 3.4171741008758545,
      "learning_rate": 2.385533414099783e-06,
      "loss": 5.53,
      "step": 9374
    },
    {
      "epoch": 0.9375,
      "grad_norm": 2.6627140045166016,
      "learning_rate": 2.3779604035785273e-06,
      "loss": 4.271,
      "step": 9375
    },
    {
      "epoch": 0.9376,
      "grad_norm": 3.1944336891174316,
      "learning_rate": 2.3703992880066638e-06,
      "loss": 4.5494,
      "step": 9376
    },
    {
      "epoch": 0.9377,
      "grad_norm": 2.0422165393829346,
      "learning_rate": 2.3628500683055222e-06,
      "loss": 4.5636,
      "step": 9377
    },
    {
      "epoch": 0.9378,
      "grad_norm": 4.693737983703613,
      "learning_rate": 2.355312745394922e-06,
      "loss": 4.9877,
      "step": 9378
    },
    {
      "epoch": 0.9379,
      "grad_norm": 2.442382335662842,
      "learning_rate": 2.3477873201932734e-06,
      "loss": 4.4837,
      "step": 9379
    },
    {
      "epoch": 0.938,
      "grad_norm": 3.2792530059814453,
      "learning_rate": 2.3402737936175425e-06,
      "loss": 4.743,
      "step": 9380
    },
    {
      "epoch": 0.9381,
      "grad_norm": 2.7105815410614014,
      "learning_rate": 2.332772166583208e-06,
      "loss": 4.6394,
      "step": 9381
    },
    {
      "epoch": 0.9382,
      "grad_norm": 5.650467395782471,
      "learning_rate": 2.325282440004339e-06,
      "loss": 4.795,
      "step": 9382
    },
    {
      "epoch": 0.9383,
      "grad_norm": 2.0999221801757812,
      "learning_rate": 2.3178046147935175e-06,
      "loss": 4.5902,
      "step": 9383
    },
    {
      "epoch": 0.9384,
      "grad_norm": 2.0304763317108154,
      "learning_rate": 2.3103386918619018e-06,
      "loss": 4.667,
      "step": 9384
    },
    {
      "epoch": 0.9385,
      "grad_norm": 2.529110908508301,
      "learning_rate": 2.3028846721191876e-06,
      "loss": 5.0854,
      "step": 9385
    },
    {
      "epoch": 0.9386,
      "grad_norm": 2.6403019428253174,
      "learning_rate": 2.295442556473637e-06,
      "loss": 4.3404,
      "step": 9386
    },
    {
      "epoch": 0.9387,
      "grad_norm": 2.3628714084625244,
      "learning_rate": 2.288012345832047e-06,
      "loss": 4.3087,
      "step": 9387
    },
    {
      "epoch": 0.9388,
      "grad_norm": 2.9622507095336914,
      "learning_rate": 2.2805940410997484e-06,
      "loss": 4.3002,
      "step": 9388
    },
    {
      "epoch": 0.9389,
      "grad_norm": 2.955436944961548,
      "learning_rate": 2.273187643180652e-06,
      "loss": 4.1116,
      "step": 9389
    },
    {
      "epoch": 0.939,
      "grad_norm": 2.1757419109344482,
      "learning_rate": 2.2657931529772136e-06,
      "loss": 4.9996,
      "step": 9390
    },
    {
      "epoch": 0.9391,
      "grad_norm": 2.9594545364379883,
      "learning_rate": 2.2584105713904125e-06,
      "loss": 4.8327,
      "step": 9391
    },
    {
      "epoch": 0.9392,
      "grad_norm": 2.3914287090301514,
      "learning_rate": 2.2510398993198067e-06,
      "loss": 4.4732,
      "step": 9392
    },
    {
      "epoch": 0.9393,
      "grad_norm": 2.3030831813812256,
      "learning_rate": 2.2436811376634893e-06,
      "loss": 4.5008,
      "step": 9393
    },
    {
      "epoch": 0.9394,
      "grad_norm": 2.1857550144195557,
      "learning_rate": 2.2363342873180757e-06,
      "loss": 4.2336,
      "step": 9394
    },
    {
      "epoch": 0.9395,
      "grad_norm": 5.359841346740723,
      "learning_rate": 2.2289993491788064e-06,
      "loss": 5.2703,
      "step": 9395
    },
    {
      "epoch": 0.9396,
      "grad_norm": 1.7778347730636597,
      "learning_rate": 2.2216763241393767e-06,
      "loss": 4.321,
      "step": 9396
    },
    {
      "epoch": 0.9397,
      "grad_norm": 4.960620403289795,
      "learning_rate": 2.2143652130921176e-06,
      "loss": 5.1073,
      "step": 9397
    },
    {
      "epoch": 0.9398,
      "grad_norm": 2.4724602699279785,
      "learning_rate": 2.2070660169278166e-06,
      "loss": 4.5852,
      "step": 9398
    },
    {
      "epoch": 0.9399,
      "grad_norm": 2.279224395751953,
      "learning_rate": 2.1997787365358958e-06,
      "loss": 4.8171,
      "step": 9399
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.97253155708313,
      "learning_rate": 2.192503372804278e-06,
      "loss": 4.9154,
      "step": 9400
    },
    {
      "epoch": 0.9401,
      "grad_norm": 2.9456558227539062,
      "learning_rate": 2.1852399266194314e-06,
      "loss": 4.7664,
      "step": 9401
    },
    {
      "epoch": 0.9402,
      "grad_norm": 3.0257251262664795,
      "learning_rate": 2.177988398866415e-06,
      "loss": 4.4607,
      "step": 9402
    },
    {
      "epoch": 0.9403,
      "grad_norm": 4.50240421295166,
      "learning_rate": 2.1707487904287672e-06,
      "loss": 5.4227,
      "step": 9403
    },
    {
      "epoch": 0.9404,
      "grad_norm": 2.4963648319244385,
      "learning_rate": 2.163521102188648e-06,
      "loss": 4.5804,
      "step": 9404
    },
    {
      "epoch": 0.9405,
      "grad_norm": 3.0942533016204834,
      "learning_rate": 2.156305335026698e-06,
      "loss": 4.4012,
      "step": 9405
    },
    {
      "epoch": 0.9406,
      "grad_norm": 2.438601016998291,
      "learning_rate": 2.1491014898221582e-06,
      "loss": 4.7518,
      "step": 9406
    },
    {
      "epoch": 0.9407,
      "grad_norm": 2.7706239223480225,
      "learning_rate": 2.141909567452793e-06,
      "loss": 4.9047,
      "step": 9407
    },
    {
      "epoch": 0.9408,
      "grad_norm": 5.768016815185547,
      "learning_rate": 2.134729568794902e-06,
      "loss": 5.0452,
      "step": 9408
    },
    {
      "epoch": 0.9409,
      "grad_norm": 2.9968643188476562,
      "learning_rate": 2.1275614947233624e-06,
      "loss": 5.135,
      "step": 9409
    },
    {
      "epoch": 0.941,
      "grad_norm": 2.73126220703125,
      "learning_rate": 2.120405346111576e-06,
      "loss": 4.7672,
      "step": 9410
    },
    {
      "epoch": 0.9411,
      "grad_norm": 2.182042121887207,
      "learning_rate": 2.1132611238315003e-06,
      "loss": 4.3674,
      "step": 9411
    },
    {
      "epoch": 0.9412,
      "grad_norm": 7.575430393218994,
      "learning_rate": 2.1061288287536285e-06,
      "loss": 4.6379,
      "step": 9412
    },
    {
      "epoch": 0.9413,
      "grad_norm": 2.3881120681762695,
      "learning_rate": 2.0990084617470206e-06,
      "loss": 4.6286,
      "step": 9413
    },
    {
      "epoch": 0.9414,
      "grad_norm": 5.124207496643066,
      "learning_rate": 2.0919000236792607e-06,
      "loss": 5.4183,
      "step": 9414
    },
    {
      "epoch": 0.9415,
      "grad_norm": 2.857678174972534,
      "learning_rate": 2.084803515416511e-06,
      "loss": 4.6532,
      "step": 9415
    },
    {
      "epoch": 0.9416,
      "grad_norm": 3.090207576751709,
      "learning_rate": 2.0777189378234143e-06,
      "loss": 4.6376,
      "step": 9416
    },
    {
      "epoch": 0.9417,
      "grad_norm": 2.273547887802124,
      "learning_rate": 2.0706462917632673e-06,
      "loss": 4.5685,
      "step": 9417
    },
    {
      "epoch": 0.9418,
      "grad_norm": 2.4750823974609375,
      "learning_rate": 2.0635855780978044e-06,
      "loss": 4.8136,
      "step": 9418
    },
    {
      "epoch": 0.9419,
      "grad_norm": 2.669255256652832,
      "learning_rate": 2.0565367976873584e-06,
      "loss": 4.6695,
      "step": 9419
    },
    {
      "epoch": 0.942,
      "grad_norm": 3.110020637512207,
      "learning_rate": 2.049499951390832e-06,
      "loss": 4.6963,
      "step": 9420
    },
    {
      "epoch": 0.9421,
      "grad_norm": 4.546471118927002,
      "learning_rate": 2.0424750400655947e-06,
      "loss": 4.821,
      "step": 9421
    },
    {
      "epoch": 0.9422,
      "grad_norm": 3.5888283252716064,
      "learning_rate": 2.0354620645676504e-06,
      "loss": 4.5393,
      "step": 9422
    },
    {
      "epoch": 0.9423,
      "grad_norm": 2.3555026054382324,
      "learning_rate": 2.0284610257514937e-06,
      "loss": 4.4977,
      "step": 9423
    },
    {
      "epoch": 0.9424,
      "grad_norm": 2.466609239578247,
      "learning_rate": 2.021471924470175e-06,
      "loss": 4.7029,
      "step": 9424
    },
    {
      "epoch": 0.9425,
      "grad_norm": 2.173981189727783,
      "learning_rate": 2.014494761575314e-06,
      "loss": 4.6494,
      "step": 9425
    },
    {
      "epoch": 0.9426,
      "grad_norm": 1.8746991157531738,
      "learning_rate": 2.0075295379170412e-06,
      "loss": 4.5696,
      "step": 9426
    },
    {
      "epoch": 0.9427,
      "grad_norm": 1.9238684177398682,
      "learning_rate": 2.0005762543440445e-06,
      "loss": 4.8083,
      "step": 9427
    },
    {
      "epoch": 0.9428,
      "grad_norm": 2.839219808578491,
      "learning_rate": 1.993634911703579e-06,
      "loss": 4.5451,
      "step": 9428
    },
    {
      "epoch": 0.9429,
      "grad_norm": 2.3258368968963623,
      "learning_rate": 1.986705510841402e-06,
      "loss": 4.5995,
      "step": 9429
    },
    {
      "epoch": 0.943,
      "grad_norm": 2.3660807609558105,
      "learning_rate": 1.979788052601861e-06,
      "loss": 4.9181,
      "step": 9430
    },
    {
      "epoch": 0.9431,
      "grad_norm": 5.484026908874512,
      "learning_rate": 1.9728825378278246e-06,
      "loss": 5.7043,
      "step": 9431
    },
    {
      "epoch": 0.9432,
      "grad_norm": 1.7968721389770508,
      "learning_rate": 1.965988967360688e-06,
      "loss": 4.5912,
      "step": 9432
    },
    {
      "epoch": 0.9433,
      "grad_norm": 2.480408191680908,
      "learning_rate": 1.9591073420404337e-06,
      "loss": 4.6489,
      "step": 9433
    },
    {
      "epoch": 0.9434,
      "grad_norm": 2.3764307498931885,
      "learning_rate": 1.9522376627055583e-06,
      "loss": 4.3949,
      "step": 9434
    },
    {
      "epoch": 0.9435,
      "grad_norm": 3.5544254779815674,
      "learning_rate": 1.945379930193125e-06,
      "loss": 5.1436,
      "step": 9435
    },
    {
      "epoch": 0.9436,
      "grad_norm": 2.1838855743408203,
      "learning_rate": 1.9385341453386995e-06,
      "loss": 4.7755,
      "step": 9436
    },
    {
      "epoch": 0.9437,
      "grad_norm": 2.0911455154418945,
      "learning_rate": 1.931700308976436e-06,
      "loss": 4.4478,
      "step": 9437
    },
    {
      "epoch": 0.9438,
      "grad_norm": 2.9659035205841064,
      "learning_rate": 1.924878421939036e-06,
      "loss": 5.3578,
      "step": 9438
    },
    {
      "epoch": 0.9439,
      "grad_norm": 2.2431843280792236,
      "learning_rate": 1.918068485057689e-06,
      "loss": 4.6091,
      "step": 9439
    },
    {
      "epoch": 0.944,
      "grad_norm": 3.283208131790161,
      "learning_rate": 1.911270499162199e-06,
      "loss": 5.1318,
      "step": 9440
    },
    {
      "epoch": 0.9441,
      "grad_norm": 1.9328850507736206,
      "learning_rate": 1.904484465080847e-06,
      "loss": 4.8957,
      "step": 9441
    },
    {
      "epoch": 0.9442,
      "grad_norm": 2.3331964015960693,
      "learning_rate": 1.8977103836405053e-06,
      "loss": 4.2549,
      "step": 9442
    },
    {
      "epoch": 0.9443,
      "grad_norm": 3.6937992572784424,
      "learning_rate": 1.8909482556666024e-06,
      "loss": 5.2684,
      "step": 9443
    },
    {
      "epoch": 0.9444,
      "grad_norm": 2.949183940887451,
      "learning_rate": 1.8841980819830351e-06,
      "loss": 4.5759,
      "step": 9444
    },
    {
      "epoch": 0.9445,
      "grad_norm": 3.744528293609619,
      "learning_rate": 1.8774598634123232e-06,
      "loss": 4.8246,
      "step": 9445
    },
    {
      "epoch": 0.9446,
      "grad_norm": 2.4998929500579834,
      "learning_rate": 1.8707336007754873e-06,
      "loss": 4.4726,
      "step": 9446
    },
    {
      "epoch": 0.9447,
      "grad_norm": 3.086763858795166,
      "learning_rate": 1.8640192948921054e-06,
      "loss": 4.8931,
      "step": 9447
    },
    {
      "epoch": 0.9448,
      "grad_norm": 2.928745985031128,
      "learning_rate": 1.8573169465802898e-06,
      "loss": 4.3399,
      "step": 9448
    },
    {
      "epoch": 0.9449,
      "grad_norm": 3.372084140777588,
      "learning_rate": 1.8506265566567094e-06,
      "loss": 4.8864,
      "step": 9449
    },
    {
      "epoch": 0.945,
      "grad_norm": 1.9645206928253174,
      "learning_rate": 1.8439481259365675e-06,
      "loss": 4.5575,
      "step": 9450
    },
    {
      "epoch": 0.9451,
      "grad_norm": 5.065884590148926,
      "learning_rate": 1.8372816552336026e-06,
      "loss": 5.1216,
      "step": 9451
    },
    {
      "epoch": 0.9452,
      "grad_norm": 2.2697036266326904,
      "learning_rate": 1.8306271453601199e-06,
      "loss": 4.3394,
      "step": 9452
    },
    {
      "epoch": 0.9453,
      "grad_norm": 2.393897771835327,
      "learning_rate": 1.8239845971269266e-06,
      "loss": 5.0644,
      "step": 9453
    },
    {
      "epoch": 0.9454,
      "grad_norm": 2.77726674079895,
      "learning_rate": 1.8173540113434194e-06,
      "loss": 4.2422,
      "step": 9454
    },
    {
      "epoch": 0.9455,
      "grad_norm": 1.9107123613357544,
      "learning_rate": 1.8107353888175083e-06,
      "loss": 4.414,
      "step": 9455
    },
    {
      "epoch": 0.9456,
      "grad_norm": 3.4532485008239746,
      "learning_rate": 1.8041287303556364e-06,
      "loss": 5.0041,
      "step": 9456
    },
    {
      "epoch": 0.9457,
      "grad_norm": 3.2456955909729004,
      "learning_rate": 1.7975340367628268e-06,
      "loss": 5.3208,
      "step": 9457
    },
    {
      "epoch": 0.9458,
      "grad_norm": 2.7346293926239014,
      "learning_rate": 1.7909513088426255e-06,
      "loss": 5.196,
      "step": 9458
    },
    {
      "epoch": 0.9459,
      "grad_norm": 5.035188674926758,
      "learning_rate": 1.7843805473970798e-06,
      "loss": 4.8361,
      "step": 9459
    },
    {
      "epoch": 0.946,
      "grad_norm": 1.9533096551895142,
      "learning_rate": 1.7778217532268714e-06,
      "loss": 4.6665,
      "step": 9460
    },
    {
      "epoch": 0.9461,
      "grad_norm": 3.7833163738250732,
      "learning_rate": 1.771274927131139e-06,
      "loss": 4.6095,
      "step": 9461
    },
    {
      "epoch": 0.9462,
      "grad_norm": 4.780084609985352,
      "learning_rate": 1.7647400699075888e-06,
      "loss": 4.5772,
      "step": 9462
    },
    {
      "epoch": 0.9463,
      "grad_norm": 3.5357582569122314,
      "learning_rate": 1.7582171823524951e-06,
      "loss": 4.3934,
      "step": 9463
    },
    {
      "epoch": 0.9464,
      "grad_norm": 3.5470008850097656,
      "learning_rate": 1.751706265260611e-06,
      "loss": 4.621,
      "step": 9464
    },
    {
      "epoch": 0.9465,
      "grad_norm": 2.3228259086608887,
      "learning_rate": 1.7452073194253238e-06,
      "loss": 4.8184,
      "step": 9465
    },
    {
      "epoch": 0.9466,
      "grad_norm": 2.185107707977295,
      "learning_rate": 1.7387203456384782e-06,
      "loss": 4.7479,
      "step": 9466
    },
    {
      "epoch": 0.9467,
      "grad_norm": 4.759929180145264,
      "learning_rate": 1.7322453446905085e-06,
      "loss": 4.9237,
      "step": 9467
    },
    {
      "epoch": 0.9468,
      "grad_norm": 1.9068293571472168,
      "learning_rate": 1.7257823173703503e-06,
      "loss": 4.5998,
      "step": 9468
    },
    {
      "epoch": 0.9469,
      "grad_norm": 2.279529571533203,
      "learning_rate": 1.719331264465529e-06,
      "loss": 4.6152,
      "step": 9469
    },
    {
      "epoch": 0.947,
      "grad_norm": 2.017210006713867,
      "learning_rate": 1.712892186762083e-06,
      "loss": 4.589,
      "step": 9470
    },
    {
      "epoch": 0.9471,
      "grad_norm": 2.538273572921753,
      "learning_rate": 1.706465085044584e-06,
      "loss": 4.7545,
      "step": 9471
    },
    {
      "epoch": 0.9472,
      "grad_norm": 3.371514081954956,
      "learning_rate": 1.7000499600961505e-06,
      "loss": 4.7501,
      "step": 9472
    },
    {
      "epoch": 0.9473,
      "grad_norm": 4.2482218742370605,
      "learning_rate": 1.6936468126984572e-06,
      "loss": 4.9807,
      "step": 9473
    },
    {
      "epoch": 0.9474,
      "grad_norm": 2.1732914447784424,
      "learning_rate": 1.6872556436317022e-06,
      "loss": 4.3914,
      "step": 9474
    },
    {
      "epoch": 0.9475,
      "grad_norm": 4.450615882873535,
      "learning_rate": 1.680876453674629e-06,
      "loss": 4.9762,
      "step": 9475
    },
    {
      "epoch": 0.9476,
      "grad_norm": 2.7535548210144043,
      "learning_rate": 1.6745092436045494e-06,
      "loss": 5.0027,
      "step": 9476
    },
    {
      "epoch": 0.9477,
      "grad_norm": 4.975569725036621,
      "learning_rate": 1.6681540141972429e-06,
      "loss": 5.0689,
      "step": 9477
    },
    {
      "epoch": 0.9478,
      "grad_norm": 3.678255796432495,
      "learning_rate": 1.661810766227112e-06,
      "loss": 4.8783,
      "step": 9478
    },
    {
      "epoch": 0.9479,
      "grad_norm": 2.545604944229126,
      "learning_rate": 1.6554795004670388e-06,
      "loss": 5.4584,
      "step": 9479
    },
    {
      "epoch": 0.948,
      "grad_norm": 9.26956844329834,
      "learning_rate": 1.6491602176884724e-06,
      "loss": 6.4725,
      "step": 9480
    },
    {
      "epoch": 0.9481,
      "grad_norm": 1.9873579740524292,
      "learning_rate": 1.6428529186614195e-06,
      "loss": 5.1216,
      "step": 9481
    },
    {
      "epoch": 0.9482,
      "grad_norm": 3.0501084327697754,
      "learning_rate": 1.636557604154365e-06,
      "loss": 4.7904,
      "step": 9482
    },
    {
      "epoch": 0.9483,
      "grad_norm": 2.333813190460205,
      "learning_rate": 1.6302742749344291e-06,
      "loss": 4.1458,
      "step": 9483
    },
    {
      "epoch": 0.9484,
      "grad_norm": 3.362046003341675,
      "learning_rate": 1.6240029317671658e-06,
      "loss": 4.9192,
      "step": 9484
    },
    {
      "epoch": 0.9485,
      "grad_norm": 2.650442600250244,
      "learning_rate": 1.6177435754167415e-06,
      "loss": 4.6292,
      "step": 9485
    },
    {
      "epoch": 0.9486,
      "grad_norm": 2.2744109630584717,
      "learning_rate": 1.611496206645835e-06,
      "loss": 4.86,
      "step": 9486
    },
    {
      "epoch": 0.9487,
      "grad_norm": 2.298954963684082,
      "learning_rate": 1.605260826215682e-06,
      "loss": 4.6472,
      "step": 9487
    },
    {
      "epoch": 0.9488,
      "grad_norm": 3.399641752243042,
      "learning_rate": 1.5990374348860305e-06,
      "loss": 5.0441,
      "step": 9488
    },
    {
      "epoch": 0.9489,
      "grad_norm": 3.3995461463928223,
      "learning_rate": 1.5928260334151845e-06,
      "loss": 5.1081,
      "step": 9489
    },
    {
      "epoch": 0.949,
      "grad_norm": 4.223649978637695,
      "learning_rate": 1.5866266225599834e-06,
      "loss": 4.3395,
      "step": 9490
    },
    {
      "epoch": 0.9491,
      "grad_norm": 3.1783013343811035,
      "learning_rate": 1.580439203075812e-06,
      "loss": 5.0665,
      "step": 9491
    },
    {
      "epoch": 0.9492,
      "grad_norm": 5.30206298828125,
      "learning_rate": 1.574263775716578e-06,
      "loss": 4.455,
      "step": 9492
    },
    {
      "epoch": 0.9493,
      "grad_norm": 2.3763182163238525,
      "learning_rate": 1.5681003412347572e-06,
      "loss": 4.2885,
      "step": 9493
    },
    {
      "epoch": 0.9494,
      "grad_norm": 2.6354944705963135,
      "learning_rate": 1.561948900381327e-06,
      "loss": 4.7192,
      "step": 9494
    },
    {
      "epoch": 0.9495,
      "grad_norm": 3.025266408920288,
      "learning_rate": 1.555809453905821e-06,
      "loss": 4.8138,
      "step": 9495
    },
    {
      "epoch": 0.9496,
      "grad_norm": 7.00272274017334,
      "learning_rate": 1.5496820025563409e-06,
      "loss": 6.2327,
      "step": 9496
    },
    {
      "epoch": 0.9497,
      "grad_norm": 2.556164264678955,
      "learning_rate": 1.543566547079467e-06,
      "loss": 4.444,
      "step": 9497
    },
    {
      "epoch": 0.9498,
      "grad_norm": 2.2946693897247314,
      "learning_rate": 1.5374630882203588e-06,
      "loss": 4.6669,
      "step": 9498
    },
    {
      "epoch": 0.9499,
      "grad_norm": 2.306303024291992,
      "learning_rate": 1.5313716267226996e-06,
      "loss": 4.7882,
      "step": 9499
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.7785639762878418,
      "learning_rate": 1.5252921633287177e-06,
      "loss": 4.5197,
      "step": 9500
    },
    {
      "epoch": 0.9501,
      "grad_norm": 3.5591158866882324,
      "learning_rate": 1.5192246987791981e-06,
      "loss": 4.9234,
      "step": 9501
    },
    {
      "epoch": 0.9502,
      "grad_norm": 2.2953104972839355,
      "learning_rate": 1.5131692338134052e-06,
      "loss": 5.1982,
      "step": 9502
    },
    {
      "epoch": 0.9503,
      "grad_norm": 2.5673248767852783,
      "learning_rate": 1.5071257691692153e-06,
      "loss": 4.3251,
      "step": 9503
    },
    {
      "epoch": 0.9504,
      "grad_norm": 4.173130035400391,
      "learning_rate": 1.501094305582984e-06,
      "loss": 5.4445,
      "step": 9504
    },
    {
      "epoch": 0.9505,
      "grad_norm": 2.7288734912872314,
      "learning_rate": 1.4950748437896234e-06,
      "loss": 4.7272,
      "step": 9505
    },
    {
      "epoch": 0.9506,
      "grad_norm": 4.904592037200928,
      "learning_rate": 1.4890673845226133e-06,
      "loss": 4.6203,
      "step": 9506
    },
    {
      "epoch": 0.9507,
      "grad_norm": 2.430365800857544,
      "learning_rate": 1.4830719285139127e-06,
      "loss": 4.2653,
      "step": 9507
    },
    {
      "epoch": 0.9508,
      "grad_norm": 2.6594860553741455,
      "learning_rate": 1.4770884764940706e-06,
      "loss": 4.4409,
      "step": 9508
    },
    {
      "epoch": 0.9509,
      "grad_norm": 4.34389066696167,
      "learning_rate": 1.4711170291921484e-06,
      "loss": 4.554,
      "step": 9509
    },
    {
      "epoch": 0.951,
      "grad_norm": 2.653215169906616,
      "learning_rate": 1.4651575873357416e-06,
      "loss": 5.2454,
      "step": 9510
    },
    {
      "epoch": 0.9511,
      "grad_norm": 3.1571524143218994,
      "learning_rate": 1.4592101516509914e-06,
      "loss": 4.4751,
      "step": 9511
    },
    {
      "epoch": 0.9512,
      "grad_norm": 2.1949100494384766,
      "learning_rate": 1.4532747228625854e-06,
      "loss": 4.7413,
      "step": 9512
    },
    {
      "epoch": 0.9513,
      "grad_norm": 2.808135747909546,
      "learning_rate": 1.4473513016937223e-06,
      "loss": 4.8831,
      "step": 9513
    },
    {
      "epoch": 0.9514,
      "grad_norm": 2.7646570205688477,
      "learning_rate": 1.4414398888661695e-06,
      "loss": 4.9323,
      "step": 9514
    },
    {
      "epoch": 0.9515,
      "grad_norm": 2.0542361736297607,
      "learning_rate": 1.4355404851001952e-06,
      "loss": 4.7971,
      "step": 9515
    },
    {
      "epoch": 0.9516,
      "grad_norm": 2.342839479446411,
      "learning_rate": 1.4296530911146466e-06,
      "loss": 4.3347,
      "step": 9516
    },
    {
      "epoch": 0.9517,
      "grad_norm": 2.1744391918182373,
      "learning_rate": 1.423777707626872e-06,
      "loss": 4.6423,
      "step": 9517
    },
    {
      "epoch": 0.9518,
      "grad_norm": 3.1098899841308594,
      "learning_rate": 1.4179143353527547e-06,
      "loss": 4.9266,
      "step": 9518
    },
    {
      "epoch": 0.9519,
      "grad_norm": 2.034024477005005,
      "learning_rate": 1.412062975006767e-06,
      "loss": 4.6332,
      "step": 9519
    },
    {
      "epoch": 0.952,
      "grad_norm": 2.624039888381958,
      "learning_rate": 1.4062236273018392e-06,
      "loss": 4.0721,
      "step": 9520
    },
    {
      "epoch": 0.9521,
      "grad_norm": 2.709433078765869,
      "learning_rate": 1.400396292949513e-06,
      "loss": 4.6368,
      "step": 9521
    },
    {
      "epoch": 0.9522,
      "grad_norm": 3.871262550354004,
      "learning_rate": 1.394580972659798e-06,
      "loss": 4.7984,
      "step": 9522
    },
    {
      "epoch": 0.9523,
      "grad_norm": 4.900648593902588,
      "learning_rate": 1.3887776671412943e-06,
      "loss": 4.9611,
      "step": 9523
    },
    {
      "epoch": 0.9524,
      "grad_norm": 3.9606266021728516,
      "learning_rate": 1.3829863771011253e-06,
      "loss": 4.7285,
      "step": 9524
    },
    {
      "epoch": 0.9525,
      "grad_norm": 2.3209073543548584,
      "learning_rate": 1.377207103244904e-06,
      "loss": 4.3653,
      "step": 9525
    },
    {
      "epoch": 0.9526,
      "grad_norm": 2.6698365211486816,
      "learning_rate": 1.3714398462768563e-06,
      "loss": 4.984,
      "step": 9526
    },
    {
      "epoch": 0.9527,
      "grad_norm": 2.8274075984954834,
      "learning_rate": 1.3656846068996976e-06,
      "loss": 5.0797,
      "step": 9527
    },
    {
      "epoch": 0.9528,
      "grad_norm": 3.459195613861084,
      "learning_rate": 1.359941385814667e-06,
      "loss": 4.4135,
      "step": 9528
    },
    {
      "epoch": 0.9529,
      "grad_norm": 4.349889755249023,
      "learning_rate": 1.3542101837215826e-06,
      "loss": 4.6985,
      "step": 9529
    },
    {
      "epoch": 0.953,
      "grad_norm": 2.147223472595215,
      "learning_rate": 1.3484910013187524e-06,
      "loss": 4.5058,
      "step": 9530
    },
    {
      "epoch": 0.9531,
      "grad_norm": 2.4205567836761475,
      "learning_rate": 1.3427838393030633e-06,
      "loss": 4.8833,
      "step": 9531
    },
    {
      "epoch": 0.9532,
      "grad_norm": 6.780369758605957,
      "learning_rate": 1.3370886983698927e-06,
      "loss": 5.828,
      "step": 9532
    },
    {
      "epoch": 0.9533,
      "grad_norm": 1.8646618127822876,
      "learning_rate": 1.3314055792131964e-06,
      "loss": 4.2158,
      "step": 9533
    },
    {
      "epoch": 0.9534,
      "grad_norm": 2.6848061084747314,
      "learning_rate": 1.3257344825254315e-06,
      "loss": 4.3399,
      "step": 9534
    },
    {
      "epoch": 0.9535,
      "grad_norm": 4.176382064819336,
      "learning_rate": 1.320075408997612e-06,
      "loss": 5.0692,
      "step": 9535
    },
    {
      "epoch": 0.9536,
      "grad_norm": 3.987025022506714,
      "learning_rate": 1.3144283593192752e-06,
      "loss": 4.0143,
      "step": 9536
    },
    {
      "epoch": 0.9537,
      "grad_norm": 5.813756465911865,
      "learning_rate": 1.308793334178493e-06,
      "loss": 5.3231,
      "step": 9537
    },
    {
      "epoch": 0.9538,
      "grad_norm": 2.7163357734680176,
      "learning_rate": 1.303170334261883e-06,
      "loss": 4.8549,
      "step": 9538
    },
    {
      "epoch": 0.9539,
      "grad_norm": 5.203600883483887,
      "learning_rate": 1.2975593602545965e-06,
      "loss": 4.5204,
      "step": 9539
    },
    {
      "epoch": 0.954,
      "grad_norm": 2.1740546226501465,
      "learning_rate": 1.2919604128402874e-06,
      "loss": 4.8628,
      "step": 9540
    },
    {
      "epoch": 0.9541,
      "grad_norm": 6.84915018081665,
      "learning_rate": 1.2863734927012095e-06,
      "loss": 5.0018,
      "step": 9541
    },
    {
      "epoch": 0.9542,
      "grad_norm": 2.4268946647644043,
      "learning_rate": 1.280798600518085e-06,
      "loss": 4.2612,
      "step": 9542
    },
    {
      "epoch": 0.9543,
      "grad_norm": 2.543292760848999,
      "learning_rate": 1.275235736970193e-06,
      "loss": 4.4006,
      "step": 9543
    },
    {
      "epoch": 0.9544,
      "grad_norm": 3.27103590965271,
      "learning_rate": 1.2696849027353796e-06,
      "loss": 4.6634,
      "step": 9544
    },
    {
      "epoch": 0.9545,
      "grad_norm": 3.5738797187805176,
      "learning_rate": 1.26414609848996e-06,
      "loss": 4.6719,
      "step": 9545
    },
    {
      "epoch": 0.9546,
      "grad_norm": 2.364168882369995,
      "learning_rate": 1.2586193249088608e-06,
      "loss": 4.9545,
      "step": 9546
    },
    {
      "epoch": 0.9547,
      "grad_norm": 3.848349094390869,
      "learning_rate": 1.2531045826654653e-06,
      "loss": 4.7464,
      "step": 9547
    },
    {
      "epoch": 0.9548,
      "grad_norm": 2.1608731746673584,
      "learning_rate": 1.2476018724317585e-06,
      "loss": 4.9157,
      "step": 9548
    },
    {
      "epoch": 0.9549,
      "grad_norm": 2.159154176712036,
      "learning_rate": 1.2421111948782149e-06,
      "loss": 4.8656,
      "step": 9549
    },
    {
      "epoch": 0.955,
      "grad_norm": 2.8660593032836914,
      "learning_rate": 1.236632550673844e-06,
      "loss": 4.3729,
      "step": 9550
    },
    {
      "epoch": 0.9551,
      "grad_norm": 2.5124638080596924,
      "learning_rate": 1.231165940486234e-06,
      "loss": 4.7877,
      "step": 9551
    },
    {
      "epoch": 0.9552,
      "grad_norm": 2.923078775405884,
      "learning_rate": 1.225711364981441e-06,
      "loss": 5.6085,
      "step": 9552
    },
    {
      "epoch": 0.9553,
      "grad_norm": 3.119931697845459,
      "learning_rate": 1.2202688248241112e-06,
      "loss": 4.9756,
      "step": 9553
    },
    {
      "epoch": 0.9554,
      "grad_norm": 3.402493715286255,
      "learning_rate": 1.2148383206773916e-06,
      "loss": 4.8037,
      "step": 9554
    },
    {
      "epoch": 0.9555,
      "grad_norm": 8.912856101989746,
      "learning_rate": 1.2094198532029755e-06,
      "loss": 4.8044,
      "step": 9555
    },
    {
      "epoch": 0.9556,
      "grad_norm": 2.378034830093384,
      "learning_rate": 1.20401342306109e-06,
      "loss": 4.9802,
      "step": 9556
    },
    {
      "epoch": 0.9557,
      "grad_norm": 2.387657880783081,
      "learning_rate": 1.1986190309104861e-06,
      "loss": 4.7176,
      "step": 9557
    },
    {
      "epoch": 0.9558,
      "grad_norm": 3.687439441680908,
      "learning_rate": 1.193236677408449e-06,
      "loss": 5.5767,
      "step": 9558
    },
    {
      "epoch": 0.9559,
      "grad_norm": 1.9080379009246826,
      "learning_rate": 1.1878663632108322e-06,
      "loss": 4.6076,
      "step": 9559
    },
    {
      "epoch": 0.956,
      "grad_norm": 3.6504673957824707,
      "learning_rate": 1.1825080889719563e-06,
      "loss": 4.6313,
      "step": 9560
    },
    {
      "epoch": 0.9561,
      "grad_norm": 2.135080575942993,
      "learning_rate": 1.1771618553447216e-06,
      "loss": 4.483,
      "step": 9561
    },
    {
      "epoch": 0.9562,
      "grad_norm": 6.340640544891357,
      "learning_rate": 1.1718276629805625e-06,
      "loss": 5.0054,
      "step": 9562
    },
    {
      "epoch": 0.9563,
      "grad_norm": 3.299903631210327,
      "learning_rate": 1.1665055125294033e-06,
      "loss": 4.9427,
      "step": 9563
    },
    {
      "epoch": 0.9564,
      "grad_norm": 3.8345725536346436,
      "learning_rate": 1.16119540463977e-06,
      "loss": 4.5304,
      "step": 9564
    },
    {
      "epoch": 0.9565,
      "grad_norm": 2.2439749240875244,
      "learning_rate": 1.155897339958667e-06,
      "loss": 4.5798,
      "step": 9565
    },
    {
      "epoch": 0.9566,
      "grad_norm": 2.272599458694458,
      "learning_rate": 1.1506113191316447e-06,
      "loss": 4.2515,
      "step": 9566
    },
    {
      "epoch": 0.9567,
      "grad_norm": 2.0300192832946777,
      "learning_rate": 1.1453373428027992e-06,
      "loss": 4.7114,
      "step": 9567
    },
    {
      "epoch": 0.9568,
      "grad_norm": 3.2012999057769775,
      "learning_rate": 1.1400754116147271e-06,
      "loss": 5.0032,
      "step": 9568
    },
    {
      "epoch": 0.9569,
      "grad_norm": 2.172713041305542,
      "learning_rate": 1.134825526208605e-06,
      "loss": 4.7033,
      "step": 9569
    },
    {
      "epoch": 0.957,
      "grad_norm": 3.6553449630737305,
      "learning_rate": 1.1295876872240873e-06,
      "loss": 4.8334,
      "step": 9570
    },
    {
      "epoch": 0.9571,
      "grad_norm": 2.2038259506225586,
      "learning_rate": 1.1243618952994195e-06,
      "loss": 4.7943,
      "step": 9571
    },
    {
      "epoch": 0.9572,
      "grad_norm": 2.0079944133758545,
      "learning_rate": 1.1191481510713253e-06,
      "loss": 4.5793,
      "step": 9572
    },
    {
      "epoch": 0.9573,
      "grad_norm": 2.082160711288452,
      "learning_rate": 1.1139464551750856e-06,
      "loss": 4.6617,
      "step": 9573
    },
    {
      "epoch": 0.9574,
      "grad_norm": 2.8472373485565186,
      "learning_rate": 1.1087568082445264e-06,
      "loss": 4.4485,
      "step": 9574
    },
    {
      "epoch": 0.9575,
      "grad_norm": 4.859907150268555,
      "learning_rate": 1.103579210911976e-06,
      "loss": 4.7378,
      "step": 9575
    },
    {
      "epoch": 0.9576,
      "grad_norm": 3.568706512451172,
      "learning_rate": 1.0984136638083177e-06,
      "loss": 4.8774,
      "step": 9576
    },
    {
      "epoch": 0.9577,
      "grad_norm": 3.544193744659424,
      "learning_rate": 1.0932601675629595e-06,
      "loss": 4.528,
      "step": 9577
    },
    {
      "epoch": 0.9578,
      "grad_norm": 2.202871084213257,
      "learning_rate": 1.0881187228038215e-06,
      "loss": 4.1112,
      "step": 9578
    },
    {
      "epoch": 0.9579,
      "grad_norm": 2.5909364223480225,
      "learning_rate": 1.0829893301573913e-06,
      "loss": 4.9597,
      "step": 9579
    },
    {
      "epoch": 0.958,
      "grad_norm": 2.840143918991089,
      "learning_rate": 1.0778719902486689e-06,
      "loss": 5.6856,
      "step": 9580
    },
    {
      "epoch": 0.9581,
      "grad_norm": 3.860750436782837,
      "learning_rate": 1.0727667037011668e-06,
      "loss": 5.7544,
      "step": 9581
    },
    {
      "epoch": 0.9582,
      "grad_norm": 3.455402135848999,
      "learning_rate": 1.0676734711369762e-06,
      "loss": 5.0038,
      "step": 9582
    },
    {
      "epoch": 0.9583,
      "grad_norm": 4.306360244750977,
      "learning_rate": 1.0625922931766785e-06,
      "loss": 4.6092,
      "step": 9583
    },
    {
      "epoch": 0.9584,
      "grad_norm": 3.1552987098693848,
      "learning_rate": 1.0575231704393895e-06,
      "loss": 4.9655,
      "step": 9584
    },
    {
      "epoch": 0.9585,
      "grad_norm": 3.924555778503418,
      "learning_rate": 1.052466103542793e-06,
      "loss": 4.6646,
      "step": 9585
    },
    {
      "epoch": 0.9586,
      "grad_norm": 4.16031551361084,
      "learning_rate": 1.0474210931030514e-06,
      "loss": 4.9289,
      "step": 9586
    },
    {
      "epoch": 0.9587,
      "grad_norm": 3.099954843521118,
      "learning_rate": 1.0423881397349068e-06,
      "loss": 4.4387,
      "step": 9587
    },
    {
      "epoch": 0.9588,
      "grad_norm": 7.285055637359619,
      "learning_rate": 1.0373672440515902e-06,
      "loss": 5.3596,
      "step": 9588
    },
    {
      "epoch": 0.9589,
      "grad_norm": 2.664480209350586,
      "learning_rate": 1.0323584066648795e-06,
      "loss": 4.8184,
      "step": 9589
    },
    {
      "epoch": 0.959,
      "grad_norm": 3.8709323406219482,
      "learning_rate": 1.0273616281851083e-06,
      "loss": 4.7364,
      "step": 9590
    },
    {
      "epoch": 0.9591,
      "grad_norm": 1.9227275848388672,
      "learning_rate": 1.0223769092211012e-06,
      "loss": 4.4235,
      "step": 9591
    },
    {
      "epoch": 0.9592,
      "grad_norm": 4.432983875274658,
      "learning_rate": 1.0174042503802493e-06,
      "loss": 4.2359,
      "step": 9592
    },
    {
      "epoch": 0.9593,
      "grad_norm": 6.510991096496582,
      "learning_rate": 1.0124436522684243e-06,
      "loss": 4.9306,
      "step": 9593
    },
    {
      "epoch": 0.9594,
      "grad_norm": 3.2869598865509033,
      "learning_rate": 1.007495115490087e-06,
      "loss": 4.0961,
      "step": 9594
    },
    {
      "epoch": 0.9595,
      "grad_norm": 2.9503278732299805,
      "learning_rate": 1.002558640648199e-06,
      "loss": 4.3084,
      "step": 9595
    },
    {
      "epoch": 0.9596,
      "grad_norm": 2.3795762062072754,
      "learning_rate": 9.976342283442463e-07,
      "loss": 4.8516,
      "step": 9596
    },
    {
      "epoch": 0.9597,
      "grad_norm": 2.558971405029297,
      "learning_rate": 9.927218791782599e-07,
      "loss": 5.3894,
      "step": 9597
    },
    {
      "epoch": 0.9598,
      "grad_norm": 3.4094033241271973,
      "learning_rate": 9.878215937487834e-07,
      "loss": 4.8727,
      "step": 9598
    },
    {
      "epoch": 0.9599,
      "grad_norm": 3.554710865020752,
      "learning_rate": 9.829333726529056e-07,
      "loss": 4.4879,
      "step": 9599
    },
    {
      "epoch": 0.96,
      "grad_norm": 4.691343307495117,
      "learning_rate": 9.78057216486261e-07,
      "loss": 4.2668,
      "step": 9600
    },
    {
      "epoch": 0.9601,
      "grad_norm": 3.677272319793701,
      "learning_rate": 9.731931258429638e-07,
      "loss": 4.7293,
      "step": 9601
    },
    {
      "epoch": 0.9602,
      "grad_norm": 2.8406364917755127,
      "learning_rate": 9.683411013157174e-07,
      "loss": 5.1527,
      "step": 9602
    },
    {
      "epoch": 0.9603,
      "grad_norm": 1.9724221229553223,
      "learning_rate": 9.635011434957152e-07,
      "loss": 4.4049,
      "step": 9603
    },
    {
      "epoch": 0.9604,
      "grad_norm": 2.78570294380188,
      "learning_rate": 9.58673252972675e-07,
      "loss": 4.0559,
      "step": 9604
    },
    {
      "epoch": 0.9605,
      "grad_norm": 1.93950355052948,
      "learning_rate": 9.538574303348813e-07,
      "loss": 4.8292,
      "step": 9605
    },
    {
      "epoch": 0.9606,
      "grad_norm": 2.54172420501709,
      "learning_rate": 9.490536761691204e-07,
      "loss": 4.3735,
      "step": 9606
    },
    {
      "epoch": 0.9607,
      "grad_norm": 2.6683552265167236,
      "learning_rate": 9.442619910607131e-07,
      "loss": 4.3207,
      "step": 9607
    },
    {
      "epoch": 0.9608,
      "grad_norm": 2.930227041244507,
      "learning_rate": 9.394823755935145e-07,
      "loss": 4.5946,
      "step": 9608
    },
    {
      "epoch": 0.9609,
      "grad_norm": 3.238466501235962,
      "learning_rate": 9.347148303499142e-07,
      "loss": 5.2105,
      "step": 9609
    },
    {
      "epoch": 0.961,
      "grad_norm": 2.229964256286621,
      "learning_rate": 9.299593559108033e-07,
      "loss": 4.8862,
      "step": 9610
    },
    {
      "epoch": 0.9611,
      "grad_norm": 3.328033924102783,
      "learning_rate": 9.252159528556403e-07,
      "loss": 4.7587,
      "step": 9611
    },
    {
      "epoch": 0.9612,
      "grad_norm": 5.062195777893066,
      "learning_rate": 9.204846217623853e-07,
      "loss": 4.3176,
      "step": 9612
    },
    {
      "epoch": 0.9613,
      "grad_norm": 4.638591766357422,
      "learning_rate": 9.157653632075436e-07,
      "loss": 4.7355,
      "step": 9613
    },
    {
      "epoch": 0.9614,
      "grad_norm": 3.205320119857788,
      "learning_rate": 9.110581777661331e-07,
      "loss": 4.6381,
      "step": 9614
    },
    {
      "epoch": 0.9615,
      "grad_norm": 3.7490015029907227,
      "learning_rate": 9.063630660117173e-07,
      "loss": 4.4421,
      "step": 9615
    },
    {
      "epoch": 0.9616,
      "grad_norm": 3.8164825439453125,
      "learning_rate": 9.016800285163718e-07,
      "loss": 4.6132,
      "step": 9616
    },
    {
      "epoch": 0.9617,
      "grad_norm": 3.6080875396728516,
      "learning_rate": 8.970090658507291e-07,
      "loss": 5.1702,
      "step": 9617
    },
    {
      "epoch": 0.9618,
      "grad_norm": 2.537747621536255,
      "learning_rate": 8.923501785839117e-07,
      "loss": 4.1275,
      "step": 9618
    },
    {
      "epoch": 0.9619,
      "grad_norm": 2.598057985305786,
      "learning_rate": 8.877033672835988e-07,
      "loss": 4.7906,
      "step": 9619
    },
    {
      "epoch": 0.962,
      "grad_norm": 4.463070869445801,
      "learning_rate": 8.830686325160042e-07,
      "loss": 4.5813,
      "step": 9620
    },
    {
      "epoch": 0.9621,
      "grad_norm": 2.533937692642212,
      "learning_rate": 8.784459748458318e-07,
      "loss": 4.7616,
      "step": 9621
    },
    {
      "epoch": 0.9622,
      "grad_norm": 2.4644477367401123,
      "learning_rate": 8.73835394836342e-07,
      "loss": 4.8877,
      "step": 9622
    },
    {
      "epoch": 0.9623,
      "grad_norm": 2.675982713699341,
      "learning_rate": 8.692368930493521e-07,
      "loss": 4.8332,
      "step": 9623
    },
    {
      "epoch": 0.9624,
      "grad_norm": 2.625013828277588,
      "learning_rate": 8.646504700451252e-07,
      "loss": 4.6375,
      "step": 9624
    },
    {
      "epoch": 0.9625,
      "grad_norm": 3.1000261306762695,
      "learning_rate": 8.600761263825474e-07,
      "loss": 4.4149,
      "step": 9625
    },
    {
      "epoch": 0.9626,
      "grad_norm": 1.9144244194030762,
      "learning_rate": 8.555138626189618e-07,
      "loss": 4.4239,
      "step": 9626
    },
    {
      "epoch": 0.9627,
      "grad_norm": 3.697803497314453,
      "learning_rate": 8.509636793102682e-07,
      "loss": 4.7672,
      "step": 9627
    },
    {
      "epoch": 0.9628,
      "grad_norm": 2.9986116886138916,
      "learning_rate": 8.46425577010912e-07,
      "loss": 4.6655,
      "step": 9628
    },
    {
      "epoch": 0.9629,
      "grad_norm": 2.4681148529052734,
      "learning_rate": 8.418995562738285e-07,
      "loss": 4.4195,
      "step": 9629
    },
    {
      "epoch": 0.963,
      "grad_norm": 1.8155205249786377,
      "learning_rate": 8.373856176505101e-07,
      "loss": 4.4995,
      "step": 9630
    },
    {
      "epoch": 0.9631,
      "grad_norm": 3.7881417274475098,
      "learning_rate": 8.328837616909613e-07,
      "loss": 4.9351,
      "step": 9631
    },
    {
      "epoch": 0.9632,
      "grad_norm": 3.2108476161956787,
      "learning_rate": 8.283939889437209e-07,
      "loss": 4.7508,
      "step": 9632
    },
    {
      "epoch": 0.9633,
      "grad_norm": 2.020840883255005,
      "learning_rate": 8.239162999558403e-07,
      "loss": 4.5073,
      "step": 9633
    },
    {
      "epoch": 0.9634,
      "grad_norm": 7.536710262298584,
      "learning_rate": 8.194506952729386e-07,
      "loss": 5.021,
      "step": 9634
    },
    {
      "epoch": 0.9635,
      "grad_norm": 3.9658026695251465,
      "learning_rate": 8.14997175439125e-07,
      "loss": 4.8223,
      "step": 9635
    },
    {
      "epoch": 0.9636,
      "grad_norm": 3.515228271484375,
      "learning_rate": 8.105557409970432e-07,
      "loss": 5.169,
      "step": 9636
    },
    {
      "epoch": 0.9637,
      "grad_norm": 1.9565832614898682,
      "learning_rate": 8.061263924878604e-07,
      "loss": 4.431,
      "step": 9637
    },
    {
      "epoch": 0.9638,
      "grad_norm": 3.414125919342041,
      "learning_rate": 8.017091304513003e-07,
      "loss": 4.9959,
      "step": 9638
    },
    {
      "epoch": 0.9639,
      "grad_norm": 2.326160430908203,
      "learning_rate": 7.973039554255768e-07,
      "loss": 4.4392,
      "step": 9639
    },
    {
      "epoch": 0.964,
      "grad_norm": 3.2497613430023193,
      "learning_rate": 7.929108679474607e-07,
      "loss": 4.6863,
      "step": 9640
    },
    {
      "epoch": 0.9641,
      "grad_norm": 3.5559351444244385,
      "learning_rate": 7.885298685522235e-07,
      "loss": 5.0091,
      "step": 9641
    },
    {
      "epoch": 0.9642,
      "grad_norm": 2.424622058868408,
      "learning_rate": 7.841609577736719e-07,
      "loss": 4.5939,
      "step": 9642
    },
    {
      "epoch": 0.9643,
      "grad_norm": 7.706921577453613,
      "learning_rate": 7.798041361441688e-07,
      "loss": 4.819,
      "step": 9643
    },
    {
      "epoch": 0.9644,
      "grad_norm": 2.4842426776885986,
      "learning_rate": 7.754594041945562e-07,
      "loss": 4.7909,
      "step": 9644
    },
    {
      "epoch": 0.9645,
      "grad_norm": 5.065613746643066,
      "learning_rate": 7.711267624542329e-07,
      "loss": 4.5289,
      "step": 9645
    },
    {
      "epoch": 0.9646,
      "grad_norm": 2.5570602416992188,
      "learning_rate": 7.668062114511321e-07,
      "loss": 4.9198,
      "step": 9646
    },
    {
      "epoch": 0.9647,
      "grad_norm": 3.249743700027466,
      "learning_rate": 7.624977517116772e-07,
      "loss": 4.8842,
      "step": 9647
    },
    {
      "epoch": 0.9648,
      "grad_norm": 2.5774166584014893,
      "learning_rate": 7.582013837608592e-07,
      "loss": 4.941,
      "step": 9648
    },
    {
      "epoch": 0.9649,
      "grad_norm": 2.615800380706787,
      "learning_rate": 7.539171081221596e-07,
      "loss": 4.3392,
      "step": 9649
    },
    {
      "epoch": 0.965,
      "grad_norm": 3.2272307872772217,
      "learning_rate": 7.496449253176274e-07,
      "loss": 5.3396,
      "step": 9650
    },
    {
      "epoch": 0.9651,
      "grad_norm": 3.200477123260498,
      "learning_rate": 7.453848358678017e-07,
      "loss": 5.0928,
      "step": 9651
    },
    {
      "epoch": 0.9652,
      "grad_norm": 3.949146032333374,
      "learning_rate": 7.411368402917563e-07,
      "loss": 4.5697,
      "step": 9652
    },
    {
      "epoch": 0.9653,
      "grad_norm": 2.4901528358459473,
      "learning_rate": 7.369009391070992e-07,
      "loss": 4.93,
      "step": 9653
    },
    {
      "epoch": 0.9654,
      "grad_norm": 1.8086869716644287,
      "learning_rate": 7.326771328299731e-07,
      "loss": 4.2383,
      "step": 9654
    },
    {
      "epoch": 0.9655,
      "grad_norm": 3.639287233352661,
      "learning_rate": 7.284654219750331e-07,
      "loss": 4.9896,
      "step": 9655
    },
    {
      "epoch": 0.9656,
      "grad_norm": 3.6130197048187256,
      "learning_rate": 7.242658070554464e-07,
      "loss": 4.8309,
      "step": 9656
    },
    {
      "epoch": 0.9657,
      "grad_norm": 4.962070941925049,
      "learning_rate": 7.200782885829482e-07,
      "loss": 4.9368,
      "step": 9657
    },
    {
      "epoch": 0.9658,
      "grad_norm": 3.153433322906494,
      "learning_rate": 7.159028670677526e-07,
      "loss": 4.6807,
      "step": 9658
    },
    {
      "epoch": 0.9659,
      "grad_norm": 3.8484551906585693,
      "learning_rate": 7.117395430186414e-07,
      "loss": 4.6669,
      "step": 9659
    },
    {
      "epoch": 0.966,
      "grad_norm": 2.2524986267089844,
      "learning_rate": 7.075883169428754e-07,
      "loss": 4.6759,
      "step": 9660
    },
    {
      "epoch": 0.9661,
      "grad_norm": 5.648280620574951,
      "learning_rate": 7.034491893463058e-07,
      "loss": 5.0471,
      "step": 9661
    },
    {
      "epoch": 0.9662,
      "grad_norm": 4.546449184417725,
      "learning_rate": 6.9932216073324e-07,
      "loss": 5.6882,
      "step": 9662
    },
    {
      "epoch": 0.9663,
      "grad_norm": 4.753981113433838,
      "learning_rate": 6.952072316065761e-07,
      "loss": 4.6681,
      "step": 9663
    },
    {
      "epoch": 0.9664,
      "grad_norm": 2.1184849739074707,
      "learning_rate": 6.911044024676683e-07,
      "loss": 4.316,
      "step": 9664
    },
    {
      "epoch": 0.9665,
      "grad_norm": 2.6384665966033936,
      "learning_rate": 6.870136738164612e-07,
      "loss": 4.9821,
      "step": 9665
    },
    {
      "epoch": 0.9666,
      "grad_norm": 3.1980552673339844,
      "learning_rate": 6.829350461514006e-07,
      "loss": 4.7407,
      "step": 9666
    },
    {
      "epoch": 0.9667,
      "grad_norm": 4.7751617431640625,
      "learning_rate": 6.788685199694222e-07,
      "loss": 5.1305,
      "step": 9667
    },
    {
      "epoch": 0.9668,
      "grad_norm": 2.026799440383911,
      "learning_rate": 6.748140957660631e-07,
      "loss": 4.682,
      "step": 9668
    },
    {
      "epoch": 0.9669,
      "grad_norm": 4.734689235687256,
      "learning_rate": 6.707717740353059e-07,
      "loss": 5.0551,
      "step": 9669
    },
    {
      "epoch": 0.967,
      "grad_norm": 3.1623756885528564,
      "learning_rate": 6.66741555269712e-07,
      "loss": 4.4754,
      "step": 9670
    },
    {
      "epoch": 0.9671,
      "grad_norm": 2.8266830444335938,
      "learning_rate": 6.627234399603555e-07,
      "loss": 4.8183,
      "step": 9671
    },
    {
      "epoch": 0.9672,
      "grad_norm": 2.1909987926483154,
      "learning_rate": 6.587174285968223e-07,
      "loss": 4.4464,
      "step": 9672
    },
    {
      "epoch": 0.9673,
      "grad_norm": 2.493046283721924,
      "learning_rate": 6.547235216672443e-07,
      "loss": 5.2471,
      "step": 9673
    },
    {
      "epoch": 0.9674,
      "grad_norm": 2.182823896408081,
      "learning_rate": 6.507417196582544e-07,
      "loss": 4.7658,
      "step": 9674
    },
    {
      "epoch": 0.9675,
      "grad_norm": 3.359356164932251,
      "learning_rate": 6.4677202305502e-07,
      "loss": 4.6634,
      "step": 9675
    },
    {
      "epoch": 0.9676,
      "grad_norm": 3.980367422103882,
      "learning_rate": 6.428144323412544e-07,
      "loss": 4.7576,
      "step": 9676
    },
    {
      "epoch": 0.9677,
      "grad_norm": 2.379879951477051,
      "learning_rate": 6.388689479991605e-07,
      "loss": 4.6077,
      "step": 9677
    },
    {
      "epoch": 0.9678,
      "grad_norm": 2.3508687019348145,
      "learning_rate": 6.349355705094984e-07,
      "loss": 4.2611,
      "step": 9678
    },
    {
      "epoch": 0.9679,
      "grad_norm": 3.2963504791259766,
      "learning_rate": 6.310143003515179e-07,
      "loss": 4.6007,
      "step": 9679
    },
    {
      "epoch": 0.968,
      "grad_norm": 4.080836772918701,
      "learning_rate": 6.271051380030368e-07,
      "loss": 4.853,
      "step": 9680
    },
    {
      "epoch": 0.9681,
      "grad_norm": 5.622200965881348,
      "learning_rate": 6.232080839403631e-07,
      "loss": 6.0587,
      "step": 9681
    },
    {
      "epoch": 0.9682,
      "grad_norm": 2.686493158340454,
      "learning_rate": 6.193231386383391e-07,
      "loss": 5.3021,
      "step": 9682
    },
    {
      "epoch": 0.9683,
      "grad_norm": 3.695403814315796,
      "learning_rate": 6.154503025703417e-07,
      "loss": 5.4067,
      "step": 9683
    },
    {
      "epoch": 0.9684,
      "grad_norm": 3.896665573120117,
      "learning_rate": 6.115895762082602e-07,
      "loss": 5.5751,
      "step": 9684
    },
    {
      "epoch": 0.9685,
      "grad_norm": 2.098836898803711,
      "learning_rate": 6.07740960022507e-07,
      "loss": 4.5725,
      "step": 9685
    },
    {
      "epoch": 0.9686,
      "grad_norm": 3.0687875747680664,
      "learning_rate": 6.039044544820404e-07,
      "loss": 4.8339,
      "step": 9686
    },
    {
      "epoch": 0.9687,
      "grad_norm": 2.2516472339630127,
      "learning_rate": 6.000800600542977e-07,
      "loss": 4.8844,
      "step": 9687
    },
    {
      "epoch": 0.9688,
      "grad_norm": 2.714294195175171,
      "learning_rate": 5.96267777205295e-07,
      "loss": 5.7633,
      "step": 9688
    },
    {
      "epoch": 0.9689,
      "grad_norm": 3.246706008911133,
      "learning_rate": 5.924676063995382e-07,
      "loss": 5.5266,
      "step": 9689
    },
    {
      "epoch": 0.969,
      "grad_norm": 2.3354403972625732,
      "learning_rate": 5.886795481000795e-07,
      "loss": 4.6457,
      "step": 9690
    },
    {
      "epoch": 0.9691,
      "grad_norm": 2.2547500133514404,
      "learning_rate": 5.849036027684606e-07,
      "loss": 4.7023,
      "step": 9691
    },
    {
      "epoch": 0.9692,
      "grad_norm": 2.512458562850952,
      "learning_rate": 5.811397708647803e-07,
      "loss": 4.7722,
      "step": 9692
    },
    {
      "epoch": 0.9693,
      "grad_norm": 4.5449371337890625,
      "learning_rate": 5.773880528476494e-07,
      "loss": 4.5332,
      "step": 9693
    },
    {
      "epoch": 0.9694,
      "grad_norm": 2.3287694454193115,
      "learning_rate": 5.736484491742134e-07,
      "loss": 4.5032,
      "step": 9694
    },
    {
      "epoch": 0.9695,
      "grad_norm": 3.7720487117767334,
      "learning_rate": 5.699209603001076e-07,
      "loss": 5.2151,
      "step": 9695
    },
    {
      "epoch": 0.9696,
      "grad_norm": 3.4182004928588867,
      "learning_rate": 5.662055866795357e-07,
      "loss": 4.3348,
      "step": 9696
    },
    {
      "epoch": 0.9697,
      "grad_norm": 3.6004865169525146,
      "learning_rate": 5.62502328765202e-07,
      "loss": 5.0933,
      "step": 9697
    },
    {
      "epoch": 0.9698,
      "grad_norm": 3.0032708644866943,
      "learning_rate": 5.588111870083346e-07,
      "loss": 4.9201,
      "step": 9698
    },
    {
      "epoch": 0.9699,
      "grad_norm": 2.958181858062744,
      "learning_rate": 5.551321618586736e-07,
      "loss": 4.736,
      "step": 9699
    },
    {
      "epoch": 0.97,
      "grad_norm": 2.7372219562530518,
      "learning_rate": 5.514652537645271e-07,
      "loss": 4.8486,
      "step": 9700
    },
    {
      "epoch": 0.9701,
      "grad_norm": 5.410106182098389,
      "learning_rate": 5.478104631726711e-07,
      "loss": 4.8949,
      "step": 9701
    },
    {
      "epoch": 0.9702,
      "grad_norm": 3.0754952430725098,
      "learning_rate": 5.441677905284381e-07,
      "loss": 5.1289,
      "step": 9702
    },
    {
      "epoch": 0.9703,
      "grad_norm": 2.191885471343994,
      "learning_rate": 5.405372362756734e-07,
      "loss": 4.9034,
      "step": 9703
    },
    {
      "epoch": 0.9704,
      "grad_norm": 2.0280890464782715,
      "learning_rate": 5.369188008567672e-07,
      "loss": 5.0073,
      "step": 9704
    },
    {
      "epoch": 0.9705,
      "grad_norm": 2.413590669631958,
      "learning_rate": 5.333124847125892e-07,
      "loss": 4.8733,
      "step": 9705
    },
    {
      "epoch": 0.9706,
      "grad_norm": 3.6264238357543945,
      "learning_rate": 5.297182882825879e-07,
      "loss": 5.2592,
      "step": 9706
    },
    {
      "epoch": 0.9707,
      "grad_norm": 3.132889986038208,
      "learning_rate": 5.261362120046686e-07,
      "loss": 5.3714,
      "step": 9707
    },
    {
      "epoch": 0.9708,
      "grad_norm": 2.912715196609497,
      "learning_rate": 5.225662563153266e-07,
      "loss": 5.2383,
      "step": 9708
    },
    {
      "epoch": 0.9709,
      "grad_norm": 2.187523126602173,
      "learning_rate": 5.190084216495361e-07,
      "loss": 4.6357,
      "step": 9709
    },
    {
      "epoch": 0.971,
      "grad_norm": 2.420184373855591,
      "learning_rate": 5.154627084408059e-07,
      "loss": 4.836,
      "step": 9710
    },
    {
      "epoch": 0.9711,
      "grad_norm": 2.5369925498962402,
      "learning_rate": 5.119291171211793e-07,
      "loss": 4.9853,
      "step": 9711
    },
    {
      "epoch": 0.9712,
      "grad_norm": 7.364004135131836,
      "learning_rate": 5.084076481212119e-07,
      "loss": 4.7148,
      "step": 9712
    },
    {
      "epoch": 0.9713,
      "grad_norm": 2.4158525466918945,
      "learning_rate": 5.048983018699827e-07,
      "loss": 5.0212,
      "step": 9713
    },
    {
      "epoch": 0.9714,
      "grad_norm": 6.142767906188965,
      "learning_rate": 5.01401078795094e-07,
      "loss": 5.683,
      "step": 9714
    },
    {
      "epoch": 0.9715,
      "grad_norm": 2.718400478363037,
      "learning_rate": 4.979159793226718e-07,
      "loss": 4.6852,
      "step": 9715
    },
    {
      "epoch": 0.9716,
      "grad_norm": 2.0877411365509033,
      "learning_rate": 4.944430038773762e-07,
      "loss": 4.9114,
      "step": 9716
    },
    {
      "epoch": 0.9717,
      "grad_norm": 2.250077247619629,
      "learning_rate": 4.909821528823577e-07,
      "loss": 4.039,
      "step": 9717
    },
    {
      "epoch": 0.9718,
      "grad_norm": 3.5980639457702637,
      "learning_rate": 4.875334267593234e-07,
      "loss": 4.9338,
      "step": 9718
    },
    {
      "epoch": 0.9719,
      "grad_norm": 2.8433101177215576,
      "learning_rate": 4.840968259284817e-07,
      "loss": 5.1216,
      "step": 9719
    },
    {
      "epoch": 0.972,
      "grad_norm": 2.510732650756836,
      "learning_rate": 4.806723508085864e-07,
      "loss": 4.9094,
      "step": 9720
    },
    {
      "epoch": 0.9721,
      "grad_norm": 3.8394088745117188,
      "learning_rate": 4.772600018168816e-07,
      "loss": 4.465,
      "step": 9721
    },
    {
      "epoch": 0.9722,
      "grad_norm": 1.978750228881836,
      "learning_rate": 4.738597793691679e-07,
      "loss": 4.5018,
      "step": 9722
    },
    {
      "epoch": 0.9723,
      "grad_norm": 4.755558967590332,
      "learning_rate": 4.704716838797363e-07,
      "loss": 5.1167,
      "step": 9723
    },
    {
      "epoch": 0.9724,
      "grad_norm": 2.5247271060943604,
      "learning_rate": 4.670957157614453e-07,
      "loss": 5.172,
      "step": 9724
    },
    {
      "epoch": 0.9725,
      "grad_norm": 2.662130117416382,
      "learning_rate": 4.6373187542561035e-07,
      "loss": 5.1252,
      "step": 9725
    },
    {
      "epoch": 0.9726,
      "grad_norm": 2.7873380184173584,
      "learning_rate": 4.6038016328211476e-07,
      "loss": 4.4904,
      "step": 9726
    },
    {
      "epoch": 0.9727,
      "grad_norm": 6.8220438957214355,
      "learning_rate": 4.570405797393762e-07,
      "loss": 4.934,
      "step": 9727
    },
    {
      "epoch": 0.9728,
      "grad_norm": 2.461125135421753,
      "learning_rate": 4.5371312520429144e-07,
      "loss": 4.6628,
      "step": 9728
    },
    {
      "epoch": 0.9729,
      "grad_norm": 2.9312119483947754,
      "learning_rate": 4.503978000823028e-07,
      "loss": 5.0216,
      "step": 9729
    },
    {
      "epoch": 0.973,
      "grad_norm": 1.9926015138626099,
      "learning_rate": 4.4709460477737607e-07,
      "loss": 4.4138,
      "step": 9730
    },
    {
      "epoch": 0.9731,
      "grad_norm": 3.3452398777008057,
      "learning_rate": 4.438035396920004e-07,
      "loss": 4.6155,
      "step": 9731
    },
    {
      "epoch": 0.9732,
      "grad_norm": 1.9572876691818237,
      "learning_rate": 4.405246052271772e-07,
      "loss": 4.7439,
      "step": 9732
    },
    {
      "epoch": 0.9733,
      "grad_norm": 2.717686414718628,
      "learning_rate": 4.372578017824314e-07,
      "loss": 4.7346,
      "step": 9733
    },
    {
      "epoch": 0.9734,
      "grad_norm": 2.832899808883667,
      "learning_rate": 4.3400312975581117e-07,
      "loss": 4.5086,
      "step": 9734
    },
    {
      "epoch": 0.9735,
      "grad_norm": 4.906620025634766,
      "learning_rate": 4.307605895439104e-07,
      "loss": 4.2274,
      "step": 9735
    },
    {
      "epoch": 0.9736,
      "grad_norm": 2.3595972061157227,
      "learning_rate": 4.275301815417909e-07,
      "loss": 4.276,
      "step": 9736
    },
    {
      "epoch": 0.9737,
      "grad_norm": 3.9504284858703613,
      "learning_rate": 4.2431190614309335e-07,
      "loss": 4.5222,
      "step": 9737
    },
    {
      "epoch": 0.9738,
      "grad_norm": 2.791163206100464,
      "learning_rate": 4.2110576373993736e-07,
      "loss": 5.1622,
      "step": 9738
    },
    {
      "epoch": 0.9739,
      "grad_norm": 2.784057378768921,
      "learning_rate": 4.179117547229883e-07,
      "loss": 4.398,
      "step": 9739
    },
    {
      "epoch": 0.974,
      "grad_norm": 2.49216365814209,
      "learning_rate": 4.1472987948143473e-07,
      "loss": 4.3282,
      "step": 9740
    },
    {
      "epoch": 0.9741,
      "grad_norm": 4.2563958168029785,
      "learning_rate": 4.115601384029666e-07,
      "loss": 4.3781,
      "step": 9741
    },
    {
      "epoch": 0.9742,
      "grad_norm": 4.72770881652832,
      "learning_rate": 4.084025318738083e-07,
      "loss": 4.854,
      "step": 9742
    },
    {
      "epoch": 0.9743,
      "grad_norm": 2.898397207260132,
      "learning_rate": 4.0525706027870756e-07,
      "loss": 4.5996,
      "step": 9743
    },
    {
      "epoch": 0.9744,
      "grad_norm": 3.4211301803588867,
      "learning_rate": 4.021237240009468e-07,
      "loss": 5.8483,
      "step": 9744
    },
    {
      "epoch": 0.9745,
      "grad_norm": 2.2762949466705322,
      "learning_rate": 3.9900252342228717e-07,
      "loss": 4.1455,
      "step": 9745
    },
    {
      "epoch": 0.9746,
      "grad_norm": 2.8175644874572754,
      "learning_rate": 3.958934589230467e-07,
      "loss": 4.6254,
      "step": 9746
    },
    {
      "epoch": 0.9747,
      "grad_norm": 2.6470305919647217,
      "learning_rate": 3.9279653088205584e-07,
      "loss": 4.4096,
      "step": 9747
    },
    {
      "epoch": 0.9748,
      "grad_norm": 2.438822031021118,
      "learning_rate": 3.8971173967666807e-07,
      "loss": 5.796,
      "step": 9748
    },
    {
      "epoch": 0.9749,
      "grad_norm": 2.433861017227173,
      "learning_rate": 3.866390856827495e-07,
      "loss": 4.1721,
      "step": 9749
    },
    {
      "epoch": 0.975,
      "grad_norm": 2.0878403186798096,
      "learning_rate": 3.835785692747118e-07,
      "loss": 4.5077,
      "step": 9750
    },
    {
      "epoch": 0.9751,
      "grad_norm": 2.6638567447662354,
      "learning_rate": 3.805301908254455e-07,
      "loss": 4.6289,
      "step": 9751
    },
    {
      "epoch": 0.9752,
      "grad_norm": 2.4110045433044434,
      "learning_rate": 3.774939507063979e-07,
      "loss": 4.4045,
      "step": 9752
    },
    {
      "epoch": 0.9753,
      "grad_norm": 3.064218759536743,
      "learning_rate": 3.744698492875398e-07,
      "loss": 4.6328,
      "step": 9753
    },
    {
      "epoch": 0.9754,
      "grad_norm": 2.5982799530029297,
      "learning_rate": 3.7145788693732086e-07,
      "loss": 4.6624,
      "step": 9754
    },
    {
      "epoch": 0.9755,
      "grad_norm": 2.114973783493042,
      "learning_rate": 3.6845806402275863e-07,
      "loss": 4.4816,
      "step": 9755
    },
    {
      "epoch": 0.9756,
      "grad_norm": 2.477809190750122,
      "learning_rate": 3.654703809093607e-07,
      "loss": 4.6041,
      "step": 9756
    },
    {
      "epoch": 0.9757,
      "grad_norm": 2.814594268798828,
      "learning_rate": 3.6249483796116924e-07,
      "loss": 4.702,
      "step": 9757
    },
    {
      "epoch": 0.9758,
      "grad_norm": 3.882528781890869,
      "learning_rate": 3.595314355407609e-07,
      "loss": 4.9051,
      "step": 9758
    },
    {
      "epoch": 0.9759,
      "grad_norm": 3.532662868499756,
      "learning_rate": 3.565801740092023e-07,
      "loss": 4.2702,
      "step": 9759
    },
    {
      "epoch": 0.976,
      "grad_norm": 3.3904504776000977,
      "learning_rate": 3.536410537260948e-07,
      "loss": 5.3706,
      "step": 9760
    },
    {
      "epoch": 0.9761,
      "grad_norm": 2.8507134914398193,
      "learning_rate": 3.50714075049563e-07,
      "loss": 5.2042,
      "step": 9761
    },
    {
      "epoch": 0.9762,
      "grad_norm": 4.718573093414307,
      "learning_rate": 3.4779923833626606e-07,
      "loss": 5.3049,
      "step": 9762
    },
    {
      "epoch": 0.9763,
      "grad_norm": 3.0755441188812256,
      "learning_rate": 3.4489654394134205e-07,
      "loss": 4.5771,
      "step": 9763
    },
    {
      "epoch": 0.9764,
      "grad_norm": 3.4173433780670166,
      "learning_rate": 3.4200599221848595e-07,
      "loss": 5.4714,
      "step": 9764
    },
    {
      "epoch": 0.9765,
      "grad_norm": 4.018089771270752,
      "learning_rate": 3.3912758351991593e-07,
      "loss": 5.4804,
      "step": 9765
    },
    {
      "epoch": 0.9766,
      "grad_norm": 3.4911372661590576,
      "learning_rate": 3.362613181963403e-07,
      "loss": 4.7053,
      "step": 9766
    },
    {
      "epoch": 0.9767,
      "grad_norm": 2.650181770324707,
      "learning_rate": 3.3340719659701313e-07,
      "loss": 4.8522,
      "step": 9767
    },
    {
      "epoch": 0.9768,
      "grad_norm": 4.8136773109436035,
      "learning_rate": 3.305652190696895e-07,
      "loss": 5.5643,
      "step": 9768
    },
    {
      "epoch": 0.9769,
      "grad_norm": 2.9781501293182373,
      "learning_rate": 3.277353859606813e-07,
      "loss": 4.262,
      "step": 9769
    },
    {
      "epoch": 0.977,
      "grad_norm": 2.3484249114990234,
      "learning_rate": 3.249176976147683e-07,
      "loss": 4.723,
      "step": 9770
    },
    {
      "epoch": 0.9771,
      "grad_norm": 3.0045523643493652,
      "learning_rate": 3.2211215437528694e-07,
      "loss": 4.6349,
      "step": 9771
    },
    {
      "epoch": 0.9772,
      "grad_norm": 2.3881494998931885,
      "learning_rate": 3.1931875658408604e-07,
      "loss": 4.6795,
      "step": 9772
    },
    {
      "epoch": 0.9773,
      "grad_norm": 4.544253349304199,
      "learning_rate": 3.1653750458152666e-07,
      "loss": 5.7955,
      "step": 9773
    },
    {
      "epoch": 0.9774,
      "grad_norm": 3.323106527328491,
      "learning_rate": 3.137683987065043e-07,
      "loss": 4.5087,
      "step": 9774
    },
    {
      "epoch": 0.9775,
      "grad_norm": 2.905632972717285,
      "learning_rate": 3.1101143929641585e-07,
      "loss": 4.7002,
      "step": 9775
    },
    {
      "epoch": 0.9776,
      "grad_norm": 3.3028743267059326,
      "learning_rate": 3.0826662668720364e-07,
      "loss": 4.6955,
      "step": 9776
    },
    {
      "epoch": 0.9777,
      "grad_norm": 3.8477025032043457,
      "learning_rate": 3.0553396121330013e-07,
      "loss": 4.5167,
      "step": 9777
    },
    {
      "epoch": 0.9778,
      "grad_norm": 3.8664932250976562,
      "learning_rate": 3.0281344320768347e-07,
      "loss": 5.4605,
      "step": 9778
    },
    {
      "epoch": 0.9779,
      "grad_norm": 2.741093397140503,
      "learning_rate": 3.001050730018218e-07,
      "loss": 4.9055,
      "step": 9779
    },
    {
      "epoch": 0.978,
      "grad_norm": 6.887925148010254,
      "learning_rate": 2.974088509257511e-07,
      "loss": 5.8611,
      "step": 9780
    },
    {
      "epoch": 0.9781,
      "grad_norm": 2.712599039077759,
      "learning_rate": 2.947247773079753e-07,
      "loss": 4.6335,
      "step": 9781
    },
    {
      "epoch": 0.9782,
      "grad_norm": 3.2429018020629883,
      "learning_rate": 2.9205285247555505e-07,
      "loss": 4.4756,
      "step": 9782
    },
    {
      "epoch": 0.9783,
      "grad_norm": 4.141704082489014,
      "learning_rate": 2.893930767540298e-07,
      "loss": 4.3997,
      "step": 9783
    },
    {
      "epoch": 0.9784,
      "grad_norm": 3.270480155944824,
      "learning_rate": 2.867454504675182e-07,
      "loss": 4.6368,
      "step": 9784
    },
    {
      "epoch": 0.9785,
      "grad_norm": 2.561439037322998,
      "learning_rate": 2.841099739386066e-07,
      "loss": 4.3316,
      "step": 9785
    },
    {
      "epoch": 0.9786,
      "grad_norm": 3.890403985977173,
      "learning_rate": 2.81486647488427e-07,
      "loss": 4.8029,
      "step": 9786
    },
    {
      "epoch": 0.9787,
      "grad_norm": 1.5956625938415527,
      "learning_rate": 2.7887547143662373e-07,
      "loss": 4.5256,
      "step": 9787
    },
    {
      "epoch": 0.9788,
      "grad_norm": 2.302388906478882,
      "learning_rate": 2.762764461013423e-07,
      "loss": 4.7399,
      "step": 9788
    },
    {
      "epoch": 0.9789,
      "grad_norm": 2.6624538898468018,
      "learning_rate": 2.73689571799296e-07,
      "loss": 4.5523,
      "step": 9789
    },
    {
      "epoch": 0.979,
      "grad_norm": 2.1988682746887207,
      "learning_rate": 2.7111484884567717e-07,
      "loss": 4.5955,
      "step": 9790
    },
    {
      "epoch": 0.9791,
      "grad_norm": 4.148911952972412,
      "learning_rate": 2.685522775541904e-07,
      "loss": 4.7277,
      "step": 9791
    },
    {
      "epoch": 0.9792,
      "grad_norm": 3.547826051712036,
      "learning_rate": 2.660018582370971e-07,
      "loss": 4.7018,
      "step": 9792
    },
    {
      "epoch": 0.9793,
      "grad_norm": 2.214524745941162,
      "learning_rate": 2.6346359120514863e-07,
      "loss": 4.3069,
      "step": 9793
    },
    {
      "epoch": 0.9794,
      "grad_norm": 3.223715305328369,
      "learning_rate": 2.609374767676309e-07,
      "loss": 4.9228,
      "step": 9794
    },
    {
      "epoch": 0.9795,
      "grad_norm": 2.088124990463257,
      "learning_rate": 2.584235152323422e-07,
      "loss": 5.0107,
      "step": 9795
    },
    {
      "epoch": 0.9796,
      "grad_norm": 1.8132582902908325,
      "learning_rate": 2.5592170690560414e-07,
      "loss": 4.4104,
      "step": 9796
    },
    {
      "epoch": 0.9797,
      "grad_norm": 2.1241111755371094,
      "learning_rate": 2.534320520922506e-07,
      "loss": 4.6723,
      "step": 9797
    },
    {
      "epoch": 0.9798,
      "grad_norm": 2.9235599040985107,
      "learning_rate": 2.5095455109562795e-07,
      "loss": 5.2406,
      "step": 9798
    },
    {
      "epoch": 0.9799,
      "grad_norm": 2.3081910610198975,
      "learning_rate": 2.484892042176279e-07,
      "loss": 4.5038,
      "step": 9799
    },
    {
      "epoch": 0.98,
      "grad_norm": 3.0082898139953613,
      "learning_rate": 2.4603601175864356e-07,
      "loss": 4.9348,
      "step": 9800
    },
    {
      "epoch": 0.9801,
      "grad_norm": 2.340907573699951,
      "learning_rate": 2.4359497401758024e-07,
      "loss": 4.556,
      "step": 9801
    },
    {
      "epoch": 0.9802,
      "grad_norm": 2.9353604316711426,
      "learning_rate": 2.4116609129187786e-07,
      "loss": 5.2837,
      "step": 9802
    },
    {
      "epoch": 0.9803,
      "grad_norm": 2.736825704574585,
      "learning_rate": 2.387493638774774e-07,
      "loss": 5.3594,
      "step": 9803
    },
    {
      "epoch": 0.9804,
      "grad_norm": 4.442468166351318,
      "learning_rate": 2.363447920688655e-07,
      "loss": 5.3209,
      "step": 9804
    },
    {
      "epoch": 0.9805,
      "grad_norm": 2.213515281677246,
      "learning_rate": 2.339523761590301e-07,
      "loss": 4.5463,
      "step": 9805
    },
    {
      "epoch": 0.9806,
      "grad_norm": 2.1810455322265625,
      "learning_rate": 2.315721164394713e-07,
      "loss": 4.5119,
      "step": 9806
    },
    {
      "epoch": 0.9807,
      "grad_norm": 4.475069999694824,
      "learning_rate": 2.2920401320022378e-07,
      "loss": 4.5683,
      "step": 9807
    },
    {
      "epoch": 0.9808,
      "grad_norm": 2.1861443519592285,
      "learning_rate": 2.2684806672982338e-07,
      "loss": 4.8634,
      "step": 9808
    },
    {
      "epoch": 0.9809,
      "grad_norm": 3.8096046447753906,
      "learning_rate": 2.2450427731534053e-07,
      "loss": 4.4348,
      "step": 9809
    },
    {
      "epoch": 0.981,
      "grad_norm": 2.35546612739563,
      "learning_rate": 2.2217264524236892e-07,
      "loss": 4.8773,
      "step": 9810
    },
    {
      "epoch": 0.9811,
      "grad_norm": 2.485518217086792,
      "learning_rate": 2.1985317079500356e-07,
      "loss": 4.6821,
      "step": 9811
    },
    {
      "epoch": 0.9812,
      "grad_norm": 2.0807547569274902,
      "learning_rate": 2.175458542558517e-07,
      "loss": 4.9048,
      "step": 9812
    },
    {
      "epoch": 0.9813,
      "grad_norm": 2.265542507171631,
      "learning_rate": 2.1525069590607737e-07,
      "loss": 4.401,
      "step": 9813
    },
    {
      "epoch": 0.9814,
      "grad_norm": 3.4469051361083984,
      "learning_rate": 2.1296769602532352e-07,
      "loss": 4.7759,
      "step": 9814
    },
    {
      "epoch": 0.9815,
      "grad_norm": 2.395524740219116,
      "learning_rate": 2.106968548917676e-07,
      "loss": 4.7591,
      "step": 9815
    },
    {
      "epoch": 0.9816,
      "grad_norm": 2.2297983169555664,
      "learning_rate": 2.0843817278209942e-07,
      "loss": 4.4514,
      "step": 9816
    },
    {
      "epoch": 0.9817,
      "grad_norm": 2.8747174739837646,
      "learning_rate": 2.0619164997155438e-07,
      "loss": 4.8248,
      "step": 9817
    },
    {
      "epoch": 0.9818,
      "grad_norm": 5.435572624206543,
      "learning_rate": 2.0395728673383574e-07,
      "loss": 4.5053,
      "step": 9818
    },
    {
      "epoch": 0.9819,
      "grad_norm": 3.7711896896362305,
      "learning_rate": 2.017350833412146e-07,
      "loss": 4.5411,
      "step": 9819
    },
    {
      "epoch": 0.982,
      "grad_norm": 1.8618392944335938,
      "learning_rate": 1.9952504006446325e-07,
      "loss": 4.7427,
      "step": 9820
    },
    {
      "epoch": 0.9821,
      "grad_norm": 3.10318660736084,
      "learning_rate": 1.973271571728441e-07,
      "loss": 4.5485,
      "step": 9821
    },
    {
      "epoch": 0.9822,
      "grad_norm": 4.121397018432617,
      "learning_rate": 1.9514143493417625e-07,
      "loss": 6.0067,
      "step": 9822
    },
    {
      "epoch": 0.9823,
      "grad_norm": 3.0983707904815674,
      "learning_rate": 1.9296787361480216e-07,
      "loss": 4.4965,
      "step": 9823
    },
    {
      "epoch": 0.9824,
      "grad_norm": 6.16081428527832,
      "learning_rate": 1.908064734795323e-07,
      "loss": 4.6482,
      "step": 9824
    },
    {
      "epoch": 0.9825,
      "grad_norm": 1.7999615669250488,
      "learning_rate": 1.8865723479173368e-07,
      "loss": 4.8866,
      "step": 9825
    },
    {
      "epoch": 0.9826,
      "grad_norm": 3.586911916732788,
      "learning_rate": 1.86520157813308e-07,
      "loss": 5.2344,
      "step": 9826
    },
    {
      "epoch": 0.9827,
      "grad_norm": 1.886051893234253,
      "learning_rate": 1.8439524280462472e-07,
      "loss": 4.7813,
      "step": 9827
    },
    {
      "epoch": 0.9828,
      "grad_norm": 4.234555721282959,
      "learning_rate": 1.8228249002461006e-07,
      "loss": 5.5216,
      "step": 9828
    },
    {
      "epoch": 0.9829,
      "grad_norm": 2.2953739166259766,
      "learning_rate": 1.8018189973069143e-07,
      "loss": 4.4324,
      "step": 9829
    },
    {
      "epoch": 0.983,
      "grad_norm": 3.444871187210083,
      "learning_rate": 1.7809347217881966e-07,
      "loss": 4.7339,
      "step": 9830
    },
    {
      "epoch": 0.9831,
      "grad_norm": 3.499309539794922,
      "learning_rate": 1.7601720762346897e-07,
      "loss": 4.5676,
      "step": 9831
    },
    {
      "epoch": 0.9832,
      "grad_norm": 1.8492666482925415,
      "learning_rate": 1.7395310631762585e-07,
      "loss": 4.7847,
      "step": 9832
    },
    {
      "epoch": 0.9833,
      "grad_norm": 3.971142292022705,
      "learning_rate": 1.7190116851280026e-07,
      "loss": 4.6555,
      "step": 9833
    },
    {
      "epoch": 0.9834,
      "grad_norm": 3.391052007675171,
      "learning_rate": 1.698613944589922e-07,
      "loss": 4.412,
      "step": 9834
    },
    {
      "epoch": 0.9835,
      "grad_norm": 3.2661936283111572,
      "learning_rate": 1.678337844047695e-07,
      "loss": 4.2515,
      "step": 9835
    },
    {
      "epoch": 0.9836,
      "grad_norm": 2.2302567958831787,
      "learning_rate": 1.6581833859716788e-07,
      "loss": 5.0355,
      "step": 9836
    },
    {
      "epoch": 0.9837,
      "grad_norm": 3.093904733657837,
      "learning_rate": 1.6381505728176872e-07,
      "loss": 5.2035,
      "step": 9837
    },
    {
      "epoch": 0.9838,
      "grad_norm": 2.5757362842559814,
      "learning_rate": 1.618239407026767e-07,
      "loss": 5.2657,
      "step": 9838
    },
    {
      "epoch": 0.9839,
      "grad_norm": 2.0016655921936035,
      "learning_rate": 1.598449891024978e-07,
      "loss": 4.5221,
      "step": 9839
    },
    {
      "epoch": 0.984,
      "grad_norm": 3.311858892440796,
      "learning_rate": 1.578782027223502e-07,
      "loss": 4.3485,
      "step": 9840
    },
    {
      "epoch": 0.9841,
      "grad_norm": 2.5445520877838135,
      "learning_rate": 1.5592358180189782e-07,
      "loss": 4.8502,
      "step": 9841
    },
    {
      "epoch": 0.9842,
      "grad_norm": 2.857891321182251,
      "learning_rate": 1.5398112657929453e-07,
      "loss": 4.8917,
      "step": 9842
    },
    {
      "epoch": 0.9843,
      "grad_norm": 3.7521004676818848,
      "learning_rate": 1.520508372912288e-07,
      "loss": 4.278,
      "step": 9843
    },
    {
      "epoch": 0.9844,
      "grad_norm": 3.396583080291748,
      "learning_rate": 1.5013271417290143e-07,
      "loss": 4.7875,
      "step": 9844
    },
    {
      "epoch": 0.9845,
      "grad_norm": 1.893633246421814,
      "learning_rate": 1.4822675745801429e-07,
      "loss": 4.0561,
      "step": 9845
    },
    {
      "epoch": 0.9846,
      "grad_norm": 2.356534719467163,
      "learning_rate": 1.4633296737882607e-07,
      "loss": 4.5965,
      "step": 9846
    },
    {
      "epoch": 0.9847,
      "grad_norm": 5.545508861541748,
      "learning_rate": 1.4445134416607442e-07,
      "loss": 4.8509,
      "step": 9847
    },
    {
      "epoch": 0.9848,
      "grad_norm": 3.6404128074645996,
      "learning_rate": 1.425818880490315e-07,
      "loss": 5.0533,
      "step": 9848
    },
    {
      "epoch": 0.9849,
      "grad_norm": 2.247498035430908,
      "learning_rate": 1.4072459925548177e-07,
      "loss": 4.5687,
      "step": 9849
    },
    {
      "epoch": 0.985,
      "grad_norm": 2.679359197616577,
      "learning_rate": 1.3887947801173307e-07,
      "loss": 5.0313,
      "step": 9850
    },
    {
      "epoch": 0.9851,
      "grad_norm": 4.380936622619629,
      "learning_rate": 1.3704652454261668e-07,
      "loss": 4.9512,
      "step": 9851
    },
    {
      "epoch": 0.9852,
      "grad_norm": 2.2424380779266357,
      "learning_rate": 1.3522573907145398e-07,
      "loss": 4.2603,
      "step": 9852
    },
    {
      "epoch": 0.9853,
      "grad_norm": 4.923818111419678,
      "learning_rate": 1.33417121820123e-07,
      "loss": 5.1244,
      "step": 9853
    },
    {
      "epoch": 0.9854,
      "grad_norm": 2.3601889610290527,
      "learning_rate": 1.3162067300898084e-07,
      "loss": 4.8932,
      "step": 9854
    },
    {
      "epoch": 0.9855,
      "grad_norm": 2.5459423065185547,
      "learning_rate": 1.2983639285693017e-07,
      "loss": 5.3879,
      "step": 9855
    },
    {
      "epoch": 0.9856,
      "grad_norm": 2.29267954826355,
      "learning_rate": 1.2806428158138596e-07,
      "loss": 4.9543,
      "step": 9856
    },
    {
      "epoch": 0.9857,
      "grad_norm": 3.1315526962280273,
      "learning_rate": 1.2630433939825327e-07,
      "loss": 4.8888,
      "step": 9857
    },
    {
      "epoch": 0.9858,
      "grad_norm": 2.8238024711608887,
      "learning_rate": 1.2455656652198277e-07,
      "loss": 4.6702,
      "step": 9858
    },
    {
      "epoch": 0.9859,
      "grad_norm": 2.10599422454834,
      "learning_rate": 1.2282096316554858e-07,
      "loss": 4.5912,
      "step": 9859
    },
    {
      "epoch": 0.986,
      "grad_norm": 2.394620895385742,
      "learning_rate": 1.2109752954042597e-07,
      "loss": 4.795,
      "step": 9860
    },
    {
      "epoch": 0.9861,
      "grad_norm": 3.119081974029541,
      "learning_rate": 1.193862658566025e-07,
      "loss": 4.5565,
      "step": 9861
    },
    {
      "epoch": 0.9862,
      "grad_norm": 3.6596274375915527,
      "learning_rate": 1.1768717232257809e-07,
      "loss": 4.7392,
      "step": 9862
    },
    {
      "epoch": 0.9863,
      "grad_norm": 5.357091426849365,
      "learning_rate": 1.160002491454093e-07,
      "loss": 4.8162,
      "step": 9863
    },
    {
      "epoch": 0.9864,
      "grad_norm": 2.374302864074707,
      "learning_rate": 1.1432549653063174e-07,
      "loss": 4.4578,
      "step": 9864
    },
    {
      "epoch": 0.9865,
      "grad_norm": 2.744828224182129,
      "learning_rate": 1.1266291468229328e-07,
      "loss": 4.5268,
      "step": 9865
    },
    {
      "epoch": 0.9866,
      "grad_norm": 3.1052768230438232,
      "learning_rate": 1.1101250380300965e-07,
      "loss": 4.6349,
      "step": 9866
    },
    {
      "epoch": 0.9867,
      "grad_norm": 3.479790210723877,
      "learning_rate": 1.0937426409384221e-07,
      "loss": 5.4145,
      "step": 9867
    },
    {
      "epoch": 0.9868,
      "grad_norm": 2.327057123184204,
      "learning_rate": 1.0774819575442019e-07,
      "loss": 5.0031,
      "step": 9868
    },
    {
      "epoch": 0.9869,
      "grad_norm": 2.3993160724639893,
      "learning_rate": 1.0613429898287398e-07,
      "loss": 4.1378,
      "step": 9869
    },
    {
      "epoch": 0.987,
      "grad_norm": 2.333427906036377,
      "learning_rate": 1.0453257397585737e-07,
      "loss": 4.4389,
      "step": 9870
    },
    {
      "epoch": 0.9871,
      "grad_norm": 3.122091770172119,
      "learning_rate": 1.0294302092853647e-07,
      "loss": 4.9547,
      "step": 9871
    },
    {
      "epoch": 0.9872,
      "grad_norm": 2.612783432006836,
      "learning_rate": 1.013656400345786e-07,
      "loss": 4.9103,
      "step": 9872
    },
    {
      "epoch": 0.9873,
      "grad_norm": 2.0622427463531494,
      "learning_rate": 9.980043148619667e-08,
      "loss": 4.6106,
      "step": 9873
    },
    {
      "epoch": 0.9874,
      "grad_norm": 2.3087518215179443,
      "learning_rate": 9.824739547410477e-08,
      "loss": 4.6634,
      "step": 9874
    },
    {
      "epoch": 0.9875,
      "grad_norm": 3.3235652446746826,
      "learning_rate": 9.670653218752934e-08,
      "loss": 5.2669,
      "step": 9875
    },
    {
      "epoch": 0.9876,
      "grad_norm": 3.1525871753692627,
      "learning_rate": 9.517784181422019e-08,
      "loss": 4.6416,
      "step": 9876
    },
    {
      "epoch": 0.9877,
      "grad_norm": 2.2521140575408936,
      "learning_rate": 9.366132454046162e-08,
      "loss": 4.5412,
      "step": 9877
    },
    {
      "epoch": 0.9878,
      "grad_norm": 2.2093920707702637,
      "learning_rate": 9.215698055100585e-08,
      "loss": 4.7757,
      "step": 9878
    },
    {
      "epoch": 0.9879,
      "grad_norm": 2.1817262172698975,
      "learning_rate": 9.066481002918403e-08,
      "loss": 4.6241,
      "step": 9879
    },
    {
      "epoch": 0.988,
      "grad_norm": 2.395268201828003,
      "learning_rate": 8.918481315678406e-08,
      "loss": 4.8899,
      "step": 9880
    },
    {
      "epoch": 0.9881,
      "grad_norm": 2.5404646396636963,
      "learning_rate": 8.771699011416168e-08,
      "loss": 4.6619,
      "step": 9881
    },
    {
      "epoch": 0.9882,
      "grad_norm": 4.645028591156006,
      "learning_rate": 8.626134108016271e-08,
      "loss": 4.6305,
      "step": 9882
    },
    {
      "epoch": 0.9883,
      "grad_norm": 2.3250772953033447,
      "learning_rate": 8.481786623214528e-08,
      "loss": 4.7203,
      "step": 9883
    },
    {
      "epoch": 0.9884,
      "grad_norm": 4.716635704040527,
      "learning_rate": 8.33865657459909e-08,
      "loss": 5.4553,
      "step": 9884
    },
    {
      "epoch": 0.9885,
      "grad_norm": 3.150696039199829,
      "learning_rate": 8.196743979610455e-08,
      "loss": 5.1652,
      "step": 9885
    },
    {
      "epoch": 0.9886,
      "grad_norm": 2.3665196895599365,
      "learning_rate": 8.056048855540343e-08,
      "loss": 4.185,
      "step": 9886
    },
    {
      "epoch": 0.9887,
      "grad_norm": 2.137617588043213,
      "learning_rate": 7.91657121953171e-08,
      "loss": 4.3105,
      "step": 9887
    },
    {
      "epoch": 0.9888,
      "grad_norm": 2.4076502323150635,
      "learning_rate": 7.778311088579849e-08,
      "loss": 4.5869,
      "step": 9888
    },
    {
      "epoch": 0.9889,
      "grad_norm": 3.6087758541107178,
      "learning_rate": 7.641268479531283e-08,
      "loss": 4.7959,
      "step": 9889
    },
    {
      "epoch": 0.989,
      "grad_norm": 2.3434479236602783,
      "learning_rate": 7.505443409083768e-08,
      "loss": 4.4177,
      "step": 9890
    },
    {
      "epoch": 0.9891,
      "grad_norm": 2.414767026901245,
      "learning_rate": 7.370835893788508e-08,
      "loss": 4.6715,
      "step": 9891
    },
    {
      "epoch": 0.9892,
      "grad_norm": 3.9745774269104004,
      "learning_rate": 7.237445950044608e-08,
      "loss": 4.9604,
      "step": 9892
    },
    {
      "epoch": 0.9893,
      "grad_norm": 1.8253811597824097,
      "learning_rate": 7.105273594107953e-08,
      "loss": 4.496,
      "step": 9893
    },
    {
      "epoch": 0.9894,
      "grad_norm": 2.7239441871643066,
      "learning_rate": 6.974318842081217e-08,
      "loss": 5.0743,
      "step": 9894
    },
    {
      "epoch": 0.9895,
      "grad_norm": 2.8999578952789307,
      "learning_rate": 6.844581709921638e-08,
      "loss": 5.2247,
      "step": 9895
    },
    {
      "epoch": 0.9896,
      "grad_norm": 3.105931282043457,
      "learning_rate": 6.71606221343768e-08,
      "loss": 4.7535,
      "step": 9896
    },
    {
      "epoch": 0.9897,
      "grad_norm": 2.7016327381134033,
      "learning_rate": 6.588760368287928e-08,
      "loss": 4.7933,
      "step": 9897
    },
    {
      "epoch": 0.9898,
      "grad_norm": 2.4604697227478027,
      "learning_rate": 6.462676189985528e-08,
      "loss": 4.5532,
      "step": 9898
    },
    {
      "epoch": 0.9899,
      "grad_norm": 3.445120334625244,
      "learning_rate": 6.337809693891527e-08,
      "loss": 4.4749,
      "step": 9899
    },
    {
      "epoch": 0.99,
      "grad_norm": 3.5657262802124023,
      "learning_rate": 6.214160895222643e-08,
      "loss": 4.4061,
      "step": 9900
    },
    {
      "epoch": 0.9901,
      "grad_norm": 3.6904711723327637,
      "learning_rate": 6.09172980904238e-08,
      "loss": 4.8502,
      "step": 9901
    },
    {
      "epoch": 0.9902,
      "grad_norm": 3.251526117324829,
      "learning_rate": 5.970516450271025e-08,
      "loss": 5.2371,
      "step": 9902
    },
    {
      "epoch": 0.9903,
      "grad_norm": 3.4338841438293457,
      "learning_rate": 5.850520833676765e-08,
      "loss": 4.771,
      "step": 9903
    },
    {
      "epoch": 0.9904,
      "grad_norm": 4.578663349151611,
      "learning_rate": 5.7317429738812376e-08,
      "loss": 4.5829,
      "step": 9904
    },
    {
      "epoch": 0.9905,
      "grad_norm": 3.8174855709075928,
      "learning_rate": 5.6141828853573106e-08,
      "loss": 4.5685,
      "step": 9905
    },
    {
      "epoch": 0.9906,
      "grad_norm": 4.939512252807617,
      "learning_rate": 5.497840582429081e-08,
      "loss": 4.8352,
      "step": 9906
    },
    {
      "epoch": 0.9907,
      "grad_norm": 2.447418689727783,
      "learning_rate": 5.382716079271877e-08,
      "loss": 5.0728,
      "step": 9907
    },
    {
      "epoch": 0.9908,
      "grad_norm": 3.116201639175415,
      "learning_rate": 5.268809389913365e-08,
      "loss": 4.9111,
      "step": 9908
    },
    {
      "epoch": 0.9909,
      "grad_norm": 3.005060911178589,
      "learning_rate": 5.1561205282335547e-08,
      "loss": 5.0062,
      "step": 9909
    },
    {
      "epoch": 0.991,
      "grad_norm": 3.075676441192627,
      "learning_rate": 5.0446495079636836e-08,
      "loss": 4.2806,
      "step": 9910
    },
    {
      "epoch": 0.9911,
      "grad_norm": 2.853245973587036,
      "learning_rate": 4.934396342684e-08,
      "loss": 4.6982,
      "step": 9911
    },
    {
      "epoch": 0.9912,
      "grad_norm": 3.1791656017303467,
      "learning_rate": 4.825361045831534e-08,
      "loss": 4.1147,
      "step": 9912
    },
    {
      "epoch": 0.9913,
      "grad_norm": 2.992802143096924,
      "learning_rate": 4.717543630688992e-08,
      "loss": 5.1181,
      "step": 9913
    },
    {
      "epoch": 0.9914,
      "grad_norm": 3.7850968837738037,
      "learning_rate": 4.610944110394755e-08,
      "loss": 5.2897,
      "step": 9914
    },
    {
      "epoch": 0.9915,
      "grad_norm": 2.802032947540283,
      "learning_rate": 4.5055624979384316e-08,
      "loss": 4.7073,
      "step": 9915
    },
    {
      "epoch": 0.9916,
      "grad_norm": 3.4285244941711426,
      "learning_rate": 4.401398806159751e-08,
      "loss": 4.7315,
      "step": 9916
    },
    {
      "epoch": 0.9917,
      "grad_norm": 7.42349910736084,
      "learning_rate": 4.298453047749673e-08,
      "loss": 5.0827,
      "step": 9917
    },
    {
      "epoch": 0.9918,
      "grad_norm": 2.284435272216797,
      "learning_rate": 4.196725235253718e-08,
      "loss": 4.7733,
      "step": 9918
    },
    {
      "epoch": 0.9919,
      "grad_norm": 1.9034217596054077,
      "learning_rate": 4.096215381066415e-08,
      "loss": 4.6281,
      "step": 9919
    },
    {
      "epoch": 0.992,
      "grad_norm": 2.6102898120880127,
      "learning_rate": 3.996923497434635e-08,
      "loss": 4.7243,
      "step": 9920
    },
    {
      "epoch": 0.9921,
      "grad_norm": 4.962637901306152,
      "learning_rate": 3.898849596456478e-08,
      "loss": 5.127,
      "step": 9921
    },
    {
      "epoch": 0.9922,
      "grad_norm": 1.9710845947265625,
      "learning_rate": 3.8019936900812735e-08,
      "loss": 4.4821,
      "step": 9922
    },
    {
      "epoch": 0.9923,
      "grad_norm": 2.273564338684082,
      "learning_rate": 3.7063557901129144e-08,
      "loss": 4.437,
      "step": 9923
    },
    {
      "epoch": 0.9924,
      "grad_norm": 3.8513505458831787,
      "learning_rate": 3.61193590820208e-08,
      "loss": 5.3146,
      "step": 9924
    },
    {
      "epoch": 0.9925,
      "grad_norm": 3.774867534637451,
      "learning_rate": 3.518734055855122e-08,
      "loss": 4.4811,
      "step": 9925
    },
    {
      "epoch": 0.9926,
      "grad_norm": 2.3654048442840576,
      "learning_rate": 3.4267502444274015e-08,
      "loss": 4.8422,
      "step": 9926
    },
    {
      "epoch": 0.9927,
      "grad_norm": 3.915952682495117,
      "learning_rate": 3.33598448512773e-08,
      "loss": 4.7015,
      "step": 9927
    },
    {
      "epoch": 0.9928,
      "grad_norm": 2.4404032230377197,
      "learning_rate": 3.246436789015039e-08,
      "loss": 4.8264,
      "step": 9928
    },
    {
      "epoch": 0.9929,
      "grad_norm": 10.227304458618164,
      "learning_rate": 3.1581071670006015e-08,
      "loss": 6.0935,
      "step": 9929
    },
    {
      "epoch": 0.993,
      "grad_norm": 3.9312493801116943,
      "learning_rate": 3.070995629846918e-08,
      "loss": 5.7813,
      "step": 9930
    },
    {
      "epoch": 0.9931,
      "grad_norm": 2.3485300540924072,
      "learning_rate": 2.985102188168831e-08,
      "loss": 4.3363,
      "step": 9931
    },
    {
      "epoch": 0.9932,
      "grad_norm": 2.1240134239196777,
      "learning_rate": 2.900426852431304e-08,
      "loss": 4.4854,
      "step": 9932
    },
    {
      "epoch": 0.9933,
      "grad_norm": 2.197031021118164,
      "learning_rate": 2.8169696329527483e-08,
      "loss": 4.4416,
      "step": 9933
    },
    {
      "epoch": 0.9934,
      "grad_norm": 2.934856414794922,
      "learning_rate": 2.7347305399016977e-08,
      "loss": 4.6333,
      "step": 9934
    },
    {
      "epoch": 0.9935,
      "grad_norm": 2.3473281860351562,
      "learning_rate": 2.6537095832990245e-08,
      "loss": 4.3089,
      "step": 9935
    },
    {
      "epoch": 0.9936,
      "grad_norm": 1.8296701908111572,
      "learning_rate": 2.573906773016832e-08,
      "loss": 4.5068,
      "step": 9936
    },
    {
      "epoch": 0.9937,
      "grad_norm": 2.336520195007324,
      "learning_rate": 2.4953221187784537e-08,
      "loss": 4.8232,
      "step": 9937
    },
    {
      "epoch": 0.9938,
      "grad_norm": 2.461291551589966,
      "learning_rate": 2.417955630159563e-08,
      "loss": 4.417,
      "step": 9938
    },
    {
      "epoch": 0.9939,
      "grad_norm": 2.9947621822357178,
      "learning_rate": 2.3418073165870636e-08,
      "loss": 4.8735,
      "step": 9939
    },
    {
      "epoch": 0.994,
      "grad_norm": 3.345306634902954,
      "learning_rate": 2.266877187339089e-08,
      "loss": 4.8007,
      "step": 9940
    },
    {
      "epoch": 0.9941,
      "grad_norm": 1.891263484954834,
      "learning_rate": 2.193165251545004e-08,
      "loss": 4.4547,
      "step": 9941
    },
    {
      "epoch": 0.9942,
      "grad_norm": 1.8324517011642456,
      "learning_rate": 2.1206715181876226e-08,
      "loss": 4.4872,
      "step": 9942
    },
    {
      "epoch": 0.9943,
      "grad_norm": 3.038400411605835,
      "learning_rate": 2.0493959960998787e-08,
      "loss": 5.2132,
      "step": 9943
    },
    {
      "epoch": 0.9944,
      "grad_norm": 2.19549298286438,
      "learning_rate": 1.9793386939659376e-08,
      "loss": 4.9037,
      "step": 9944
    },
    {
      "epoch": 0.9945,
      "grad_norm": 2.541832208633423,
      "learning_rate": 1.9104996203223037e-08,
      "loss": 4.7842,
      "step": 9945
    },
    {
      "epoch": 0.9946,
      "grad_norm": 2.2710750102996826,
      "learning_rate": 1.842878783557822e-08,
      "loss": 4.5225,
      "step": 9946
    },
    {
      "epoch": 0.9947,
      "grad_norm": 4.198962688446045,
      "learning_rate": 1.7764761919103477e-08,
      "loss": 4.7327,
      "step": 9947
    },
    {
      "epoch": 0.9948,
      "grad_norm": 3.9533348083496094,
      "learning_rate": 1.7112918534711865e-08,
      "loss": 4.8168,
      "step": 9948
    },
    {
      "epoch": 0.9949,
      "grad_norm": 2.297637939453125,
      "learning_rate": 1.6473257761828732e-08,
      "loss": 4.3765,
      "step": 9949
    },
    {
      "epoch": 0.995,
      "grad_norm": 4.791310787200928,
      "learning_rate": 1.584577967840284e-08,
      "loss": 4.8057,
      "step": 9950
    },
    {
      "epoch": 0.9951,
      "grad_norm": 2.5754003524780273,
      "learning_rate": 1.5230484360873044e-08,
      "loss": 4.8104,
      "step": 9951
    },
    {
      "epoch": 0.9952,
      "grad_norm": 3.153583288192749,
      "learning_rate": 1.4627371884234909e-08,
      "loss": 4.5522,
      "step": 9952
    },
    {
      "epoch": 0.9953,
      "grad_norm": 6.566007137298584,
      "learning_rate": 1.4036442321962995e-08,
      "loss": 4.6673,
      "step": 9953
    },
    {
      "epoch": 0.9954,
      "grad_norm": 7.9379448890686035,
      "learning_rate": 1.3457695746055265e-08,
      "loss": 4.7743,
      "step": 9954
    },
    {
      "epoch": 0.9955,
      "grad_norm": 3.7844440937042236,
      "learning_rate": 1.2891132227033087e-08,
      "loss": 5.411,
      "step": 9955
    },
    {
      "epoch": 0.9956,
      "grad_norm": 1.9911317825317383,
      "learning_rate": 1.2336751833941229e-08,
      "loss": 4.5826,
      "step": 9956
    },
    {
      "epoch": 0.9957,
      "grad_norm": 3.101064920425415,
      "learning_rate": 1.1794554634314558e-08,
      "loss": 4.4659,
      "step": 9957
    },
    {
      "epoch": 0.9958,
      "grad_norm": 3.5076169967651367,
      "learning_rate": 1.126454069423355e-08,
      "loss": 4.4649,
      "step": 9958
    },
    {
      "epoch": 0.9959,
      "grad_norm": 2.5324745178222656,
      "learning_rate": 1.074671007825767e-08,
      "loss": 4.758,
      "step": 9959
    },
    {
      "epoch": 0.996,
      "grad_norm": 5.700284004211426,
      "learning_rate": 1.0241062849503103e-08,
      "loss": 4.8784,
      "step": 9960
    },
    {
      "epoch": 0.9961,
      "grad_norm": 5.601644515991211,
      "learning_rate": 9.747599069576119e-09,
      "loss": 4.6213,
      "step": 9961
    },
    {
      "epoch": 0.9962,
      "grad_norm": 2.597376585006714,
      "learning_rate": 9.2663187986064e-09,
      "loss": 4.4225,
      "step": 9962
    },
    {
      "epoch": 0.9963,
      "grad_norm": 3.2968835830688477,
      "learning_rate": 8.797222095224823e-09,
      "loss": 4.2811,
      "step": 9963
    },
    {
      "epoch": 0.9964,
      "grad_norm": 2.970116138458252,
      "learning_rate": 8.340309016585668e-09,
      "loss": 4.3051,
      "step": 9964
    },
    {
      "epoch": 0.9965,
      "grad_norm": 5.342408180236816,
      "learning_rate": 7.895579618388827e-09,
      "loss": 4.5239,
      "step": 9965
    },
    {
      "epoch": 0.9966,
      "grad_norm": 2.063608169555664,
      "learning_rate": 7.463033954802078e-09,
      "loss": 4.2868,
      "step": 9966
    },
    {
      "epoch": 0.9967,
      "grad_norm": 3.4945361614227295,
      "learning_rate": 7.042672078527713e-09,
      "loss": 4.8681,
      "step": 9967
    },
    {
      "epoch": 0.9968,
      "grad_norm": 1.7331583499908447,
      "learning_rate": 6.634494040802519e-09,
      "loss": 4.5535,
      "step": 9968
    },
    {
      "epoch": 0.9969,
      "grad_norm": 2.8115317821502686,
      "learning_rate": 6.238499891353389e-09,
      "loss": 4.8663,
      "step": 9969
    },
    {
      "epoch": 0.997,
      "grad_norm": 2.4858334064483643,
      "learning_rate": 5.854689678419512e-09,
      "loss": 5.0763,
      "step": 9970
    },
    {
      "epoch": 0.9971,
      "grad_norm": 9.86035442352295,
      "learning_rate": 5.483063448785686e-09,
      "loss": 4.8235,
      "step": 9971
    },
    {
      "epoch": 0.9972,
      "grad_norm": 2.8495895862579346,
      "learning_rate": 5.123621247726806e-09,
      "loss": 4.5948,
      "step": 9972
    },
    {
      "epoch": 0.9973,
      "grad_norm": 2.6102986335754395,
      "learning_rate": 4.776363119030069e-09,
      "loss": 4.3045,
      "step": 9973
    },
    {
      "epoch": 0.9974,
      "grad_norm": 2.2726497650146484,
      "learning_rate": 4.4412891050171765e-09,
      "loss": 4.385,
      "step": 9974
    },
    {
      "epoch": 0.9975,
      "grad_norm": 4.023268699645996,
      "learning_rate": 4.1183992465221315e-09,
      "loss": 4.9677,
      "step": 9975
    },
    {
      "epoch": 0.9976,
      "grad_norm": 2.3202779293060303,
      "learning_rate": 3.807693582869032e-09,
      "loss": 4.3788,
      "step": 9976
    },
    {
      "epoch": 0.9977,
      "grad_norm": 2.5922510623931885,
      "learning_rate": 3.509172151938689e-09,
      "loss": 4.7692,
      "step": 9977
    },
    {
      "epoch": 0.9978,
      "grad_norm": 3.3551433086395264,
      "learning_rate": 3.222834990090906e-09,
      "loss": 4.6336,
      "step": 9978
    },
    {
      "epoch": 0.9979,
      "grad_norm": 2.978111505508423,
      "learning_rate": 2.948682132208891e-09,
      "loss": 4.7165,
      "step": 9979
    },
    {
      "epoch": 0.998,
      "grad_norm": 3.078913450241089,
      "learning_rate": 2.6867136117214587e-09,
      "loss": 4.8982,
      "step": 9980
    },
    {
      "epoch": 0.9981,
      "grad_norm": 2.505378007888794,
      "learning_rate": 2.4369294605253166e-09,
      "loss": 4.4701,
      "step": 9981
    },
    {
      "epoch": 0.9982,
      "grad_norm": 4.858293056488037,
      "learning_rate": 2.1993297090627806e-09,
      "loss": 4.3035,
      "step": 9982
    },
    {
      "epoch": 0.9983,
      "grad_norm": 4.943415641784668,
      "learning_rate": 1.973914386288467e-09,
      "loss": 4.6759,
      "step": 9983
    },
    {
      "epoch": 0.9984,
      "grad_norm": 2.7249300479888916,
      "learning_rate": 1.7606835196692927e-09,
      "loss": 5.0134,
      "step": 9984
    },
    {
      "epoch": 0.9985,
      "grad_norm": 1.877061367034912,
      "learning_rate": 1.5596371351733752e-09,
      "loss": 4.812,
      "step": 9985
    },
    {
      "epoch": 0.9986,
      "grad_norm": 2.2341458797454834,
      "learning_rate": 1.3707752573255405e-09,
      "loss": 4.3556,
      "step": 9986
    },
    {
      "epoch": 0.9987,
      "grad_norm": 2.330948829650879,
      "learning_rate": 1.1940979091074056e-09,
      "loss": 4.4176,
      "step": 9987
    },
    {
      "epoch": 0.9988,
      "grad_norm": 2.3270182609558105,
      "learning_rate": 1.029605112068399e-09,
      "loss": 4.4286,
      "step": 9988
    },
    {
      "epoch": 0.9989,
      "grad_norm": 4.0920939445495605,
      "learning_rate": 8.772968862369446e-10,
      "loss": 4.6479,
      "step": 9989
    },
    {
      "epoch": 0.999,
      "grad_norm": 2.8853251934051514,
      "learning_rate": 7.371732501759709e-10,
      "loss": 4.6289,
      "step": 9990
    },
    {
      "epoch": 0.9991,
      "grad_norm": 4.703173637390137,
      "learning_rate": 6.092342209607083e-10,
      "loss": 4.8764,
      "step": 9991
    },
    {
      "epoch": 0.9992,
      "grad_norm": 2.196643829345703,
      "learning_rate": 4.934798141786879e-10,
      "loss": 5.0498,
      "step": 9992
    },
    {
      "epoch": 0.9993,
      "grad_norm": 2.886936664581299,
      "learning_rate": 3.899100439408443e-10,
      "loss": 4.4846,
      "step": 9993
    },
    {
      "epoch": 0.9994,
      "grad_norm": 2.0517802238464355,
      "learning_rate": 2.985249228593112e-10,
      "loss": 4.3164,
      "step": 9994
    },
    {
      "epoch": 0.9995,
      "grad_norm": 2.1550631523132324,
      "learning_rate": 2.1932446206962553e-10,
      "loss": 4.6878,
      "step": 9995
    },
    {
      "epoch": 0.9996,
      "grad_norm": 2.8274049758911133,
      "learning_rate": 1.5230867123072756e-10,
      "loss": 4.7392,
      "step": 9996
    },
    {
      "epoch": 0.9997,
      "grad_norm": 2.403284788131714,
      "learning_rate": 9.74775584916543e-11,
      "loss": 4.5104,
      "step": 9997
    },
    {
      "epoch": 0.9998,
      "grad_norm": 2.870192050933838,
      "learning_rate": 5.483113054705058e-11,
      "loss": 4.9291,
      "step": 9998
    },
    {
      "epoch": 0.9999,
      "grad_norm": 2.9844560623168945,
      "learning_rate": 2.4369392592760164e-11,
      "loss": 4.4728,
      "step": 9999
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.904883623123169,
      "learning_rate": 6.092348336927955e-12,
      "loss": 5.4201,
      "step": 10000
    }
  ],
  "logging_steps": 1,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.412376018944e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}