{
  "best_metric": 0.9474140798771133,
  "best_model_checkpoint": "xtreme_s_xlsr_minds14_rerun/checkpoint-3600",
  "epoch": 50.0,
  "global_step": 3700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 1.9999999999999996e-07,
      "loss": 2.6424,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.9999999999999993e-07,
      "loss": 2.6429,
      "step": 2
    },
    {
      "epoch": 0.04,
      "learning_rate": 6e-07,
      "loss": 2.6433,
      "step": 3
    },
    {
      "epoch": 0.05,
      "learning_rate": 7.999999999999999e-07,
      "loss": 2.6387,
      "step": 4
    },
    {
      "epoch": 0.07,
      "learning_rate": 1e-06,
      "loss": 2.6385,
      "step": 5
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.2e-06,
      "loss": 2.644,
      "step": 6
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.4e-06,
      "loss": 2.6394,
      "step": 7
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.5999999999999997e-06,
      "loss": 2.6394,
      "step": 8
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.8e-06,
      "loss": 2.6417,
      "step": 9
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-06,
      "loss": 2.6393,
      "step": 10
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.1999999999999997e-06,
      "loss": 2.6364,
      "step": 11
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.4e-06,
      "loss": 2.6414,
      "step": 12
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.5999999999999997e-06,
      "loss": 2.64,
      "step": 13
    },
    {
      "epoch": 0.19,
      "learning_rate": 2.8e-06,
      "loss": 2.636,
      "step": 14
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.9999999999999997e-06,
      "loss": 2.638,
      "step": 15
    },
    {
      "epoch": 0.22,
      "learning_rate": 3.1999999999999994e-06,
      "loss": 2.6363,
      "step": 16
    },
    {
      "epoch": 0.23,
      "learning_rate": 3.4e-06,
      "loss": 2.6375,
      "step": 17
    },
    {
      "epoch": 0.24,
      "learning_rate": 3.6e-06,
      "loss": 2.6434,
      "step": 18
    },
    {
      "epoch": 0.26,
      "learning_rate": 3.7999999999999996e-06,
      "loss": 2.6447,
      "step": 19
    },
    {
      "epoch": 0.27,
      "learning_rate": 4e-06,
      "loss": 2.6387,
      "step": 20
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.2e-06,
      "loss": 2.6414,
      "step": 21
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.399999999999999e-06,
      "loss": 2.645,
      "step": 22
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.599999999999999e-06,
      "loss": 2.6443,
      "step": 23
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.8e-06,
      "loss": 2.6361,
      "step": 24
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 2.6373,
      "step": 25
    },
    {
      "epoch": 0.35,
      "learning_rate": 5.199999999999999e-06,
      "loss": 2.641,
      "step": 26
    },
    {
      "epoch": 0.36,
      "learning_rate": 5.399999999999999e-06,
      "loss": 2.6393,
      "step": 27
    },
    {
      "epoch": 0.38,
      "learning_rate": 5.6e-06,
      "loss": 2.6347,
      "step": 28
    },
    {
      "epoch": 0.39,
      "learning_rate": 5.7999999999999995e-06,
      "loss": 2.6388,
      "step": 29
    },
    {
      "epoch": 0.41,
      "learning_rate": 5.999999999999999e-06,
      "loss": 2.6375,
      "step": 30
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.199999999999999e-06,
      "loss": 2.6396,
      "step": 31
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.399999999999999e-06,
      "loss": 2.6402,
      "step": 32
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.599999999999999e-06,
      "loss": 2.6396,
      "step": 33
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.8e-06,
      "loss": 2.637,
      "step": 34
    },
    {
      "epoch": 0.47,
      "learning_rate": 7e-06,
      "loss": 2.638,
      "step": 35
    },
    {
      "epoch": 0.49,
      "learning_rate": 7.2e-06,
      "loss": 2.6422,
      "step": 36
    },
    {
      "epoch": 0.5,
      "learning_rate": 7.3999999999999995e-06,
      "loss": 2.6398,
      "step": 37
    },
    {
      "epoch": 0.51,
      "learning_rate": 7.599999999999999e-06,
      "loss": 2.6393,
      "step": 38
    },
    {
      "epoch": 0.53,
      "learning_rate": 7.799999999999998e-06,
      "loss": 2.6418,
      "step": 39
    },
    {
      "epoch": 0.54,
      "learning_rate": 8e-06,
      "loss": 2.6419,
      "step": 40
    },
    {
      "epoch": 0.55,
      "learning_rate": 8.2e-06,
      "loss": 2.6427,
      "step": 41
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.4e-06,
      "loss": 2.6361,
      "step": 42
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.599999999999999e-06,
      "loss": 2.6414,
      "step": 43
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.799999999999999e-06,
      "loss": 2.6394,
      "step": 44
    },
    {
      "epoch": 0.61,
      "learning_rate": 8.999999999999999e-06,
      "loss": 2.6396,
      "step": 45
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.199999999999998e-06,
      "loss": 2.6356,
      "step": 46
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.399999999999998e-06,
      "loss": 2.6377,
      "step": 47
    },
    {
      "epoch": 0.65,
      "learning_rate": 9.6e-06,
      "loss": 2.635,
      "step": 48
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.799999999999998e-06,
      "loss": 2.6387,
      "step": 49
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.999999999999999e-06,
      "loss": 2.6381,
      "step": 50
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.02e-05,
      "loss": 2.6401,
      "step": 51
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.0399999999999999e-05,
      "loss": 2.6406,
      "step": 52
    },
    {
      "epoch": 0.72,
      "learning_rate": 1.06e-05,
      "loss": 2.6338,
      "step": 53
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0799999999999998e-05,
      "loss": 2.6373,
      "step": 54
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.1e-05,
      "loss": 2.6427,
      "step": 55
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.12e-05,
      "loss": 2.6423,
      "step": 56
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.14e-05,
      "loss": 2.64,
      "step": 57
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.1599999999999999e-05,
      "loss": 2.6412,
      "step": 58
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.1799999999999999e-05,
      "loss": 2.6427,
      "step": 59
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.1999999999999999e-05,
      "loss": 2.64,
      "step": 60
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.2199999999999998e-05,
      "loss": 2.6379,
      "step": 61
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.2399999999999998e-05,
      "loss": 2.6385,
      "step": 62
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.26e-05,
      "loss": 2.6338,
      "step": 63
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.2799999999999998e-05,
      "loss": 2.6393,
      "step": 64
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.3e-05,
      "loss": 2.635,
      "step": 65
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.3199999999999997e-05,
      "loss": 2.6382,
      "step": 66
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.3399999999999999e-05,
      "loss": 2.6354,
      "step": 67
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.36e-05,
      "loss": 2.6371,
      "step": 68
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.3799999999999998e-05,
      "loss": 2.6358,
      "step": 69
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.4e-05,
      "loss": 2.6315,
      "step": 70
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.4199999999999998e-05,
      "loss": 2.6377,
      "step": 71
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.44e-05,
      "loss": 2.6287,
      "step": 72
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.4599999999999997e-05,
      "loss": 2.6393,
      "step": 73
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.4799999999999999e-05,
      "loss": 2.6279,
      "step": 74
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 2.6467,
      "step": 75
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.5199999999999998e-05,
      "loss": 2.6454,
      "step": 76
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.5399999999999998e-05,
      "loss": 2.6445,
      "step": 77
    },
    {
      "epoch": 1.05,
      "learning_rate": 1.5599999999999996e-05,
      "loss": 2.6421,
      "step": 78
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.5799999999999998e-05,
      "loss": 2.6475,
      "step": 79
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.6e-05,
      "loss": 2.6405,
      "step": 80
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.6199999999999997e-05,
      "loss": 2.6408,
      "step": 81
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.64e-05,
      "loss": 2.6379,
      "step": 82
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.6599999999999997e-05,
      "loss": 2.6297,
      "step": 83
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.68e-05,
      "loss": 2.6317,
      "step": 84
    },
    {
      "epoch": 1.15,
      "learning_rate": 1.6999999999999996e-05,
      "loss": 2.6317,
      "step": 85
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 2.6345,
      "step": 86
    },
    {
      "epoch": 1.18,
      "learning_rate": 1.74e-05,
      "loss": 2.6321,
      "step": 87
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.7599999999999998e-05,
      "loss": 2.6307,
      "step": 88
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.78e-05,
      "loss": 2.634,
      "step": 89
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.7999999999999997e-05,
      "loss": 2.6302,
      "step": 90
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.82e-05,
      "loss": 2.6256,
      "step": 91
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.8399999999999997e-05,
      "loss": 2.6131,
      "step": 92
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.8599999999999998e-05,
      "loss": 2.6433,
      "step": 93
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.8799999999999996e-05,
      "loss": 2.6461,
      "step": 94
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.9e-05,
      "loss": 2.6454,
      "step": 95
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.92e-05,
      "loss": 2.6505,
      "step": 96
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.9399999999999997e-05,
      "loss": 2.6477,
      "step": 97
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.9599999999999995e-05,
      "loss": 2.6386,
      "step": 98
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.98e-05,
      "loss": 2.6364,
      "step": 99
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 2.6393,
      "step": 100
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.0199999999999996e-05,
      "loss": 2.6283,
      "step": 101
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.04e-05,
      "loss": 2.6292,
      "step": 102
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.06e-05,
      "loss": 2.6377,
      "step": 103
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.0799999999999997e-05,
      "loss": 2.6293,
      "step": 104
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.1e-05,
      "loss": 2.6311,
      "step": 105
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.12e-05,
      "loss": 2.6291,
      "step": 106
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.14e-05,
      "loss": 2.6183,
      "step": 107
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.1599999999999996e-05,
      "loss": 2.6148,
      "step": 108
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.1799999999999998e-05,
      "loss": 2.612,
      "step": 109
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.2e-05,
      "loss": 2.6181,
      "step": 110
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.2199999999999998e-05,
      "loss": 2.6391,
      "step": 111
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.24e-05,
      "loss": 2.6418,
      "step": 112
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.2599999999999997e-05,
      "loss": 2.6408,
      "step": 113
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.28e-05,
      "loss": 2.6507,
      "step": 114
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.2999999999999997e-05,
      "loss": 2.6302,
      "step": 115
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3199999999999998e-05,
      "loss": 2.6418,
      "step": 116
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.34e-05,
      "loss": 2.6365,
      "step": 117
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3599999999999998e-05,
      "loss": 2.6251,
      "step": 118
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.38e-05,
      "loss": 2.6273,
      "step": 119
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.3999999999999997e-05,
      "loss": 2.6346,
      "step": 120
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.42e-05,
      "loss": 2.6205,
      "step": 121
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.4399999999999997e-05,
      "loss": 2.6324,
      "step": 122
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.4599999999999998e-05,
      "loss": 2.6294,
      "step": 123
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.4799999999999996e-05,
      "loss": 2.6115,
      "step": 124
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.4999999999999998e-05,
      "loss": 2.6126,
      "step": 125
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.52e-05,
      "loss": 2.6088,
      "step": 126
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.5399999999999997e-05,
      "loss": 2.6151,
      "step": 127
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.5599999999999995e-05,
      "loss": 2.5967,
      "step": 128
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.5799999999999997e-05,
      "loss": 2.6387,
      "step": 129
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.6e-05,
      "loss": 2.6447,
      "step": 130
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.6199999999999996e-05,
      "loss": 2.6496,
      "step": 131
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.6399999999999995e-05,
      "loss": 2.6378,
      "step": 132
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.66e-05,
      "loss": 2.6484,
      "step": 133
    },
    {
      "epoch": 1.81,
      "learning_rate": 2.6799999999999998e-05,
      "loss": 2.6414,
      "step": 134
    },
    {
      "epoch": 1.82,
      "learning_rate": 2.6999999999999996e-05,
      "loss": 2.639,
      "step": 135
    },
    {
      "epoch": 1.84,
      "learning_rate": 2.72e-05,
      "loss": 2.6478,
      "step": 136
    },
    {
      "epoch": 1.85,
      "learning_rate": 2.74e-05,
      "loss": 2.6325,
      "step": 137
    },
    {
      "epoch": 1.86,
      "learning_rate": 2.7599999999999997e-05,
      "loss": 2.6364,
      "step": 138
    },
    {
      "epoch": 1.88,
      "learning_rate": 2.7799999999999995e-05,
      "loss": 2.6353,
      "step": 139
    },
    {
      "epoch": 1.89,
      "learning_rate": 2.8e-05,
      "loss": 2.6498,
      "step": 140
    },
    {
      "epoch": 1.91,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 2.6522,
      "step": 141
    },
    {
      "epoch": 1.92,
      "learning_rate": 2.8399999999999996e-05,
      "loss": 2.6431,
      "step": 142
    },
    {
      "epoch": 1.93,
      "learning_rate": 2.86e-05,
      "loss": 2.6623,
      "step": 143
    },
    {
      "epoch": 1.95,
      "learning_rate": 2.88e-05,
      "loss": 2.6511,
      "step": 144
    },
    {
      "epoch": 1.96,
      "learning_rate": 2.8999999999999997e-05,
      "loss": 2.656,
      "step": 145
    },
    {
      "epoch": 1.97,
      "learning_rate": 2.9199999999999995e-05,
      "loss": 2.6525,
      "step": 146
    },
    {
      "epoch": 1.99,
      "learning_rate": 2.94e-05,
      "loss": 2.6337,
      "step": 147
    },
    {
      "epoch": 2.0,
      "learning_rate": 2.9599999999999998e-05,
      "loss": 2.6406,
      "step": 148
    },
    {
      "epoch": 2.01,
      "learning_rate": 2.9799999999999996e-05,
      "loss": 2.6266,
      "step": 149
    },
    {
      "epoch": 2.03,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 2.6323,
      "step": 150
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.02e-05,
      "loss": 2.627,
      "step": 151
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.0399999999999997e-05,
      "loss": 2.6319,
      "step": 152
    },
    {
      "epoch": 2.07,
      "learning_rate": 3.06e-05,
      "loss": 2.6337,
      "step": 153
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.0799999999999996e-05,
      "loss": 2.6359,
      "step": 154
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.0999999999999995e-05,
      "loss": 2.6375,
      "step": 155
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.119999999999999e-05,
      "loss": 2.6371,
      "step": 156
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.14e-05,
      "loss": 2.632,
      "step": 157
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.1599999999999996e-05,
      "loss": 2.6438,
      "step": 158
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.1799999999999994e-05,
      "loss": 2.6263,
      "step": 159
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.2e-05,
      "loss": 2.6342,
      "step": 160
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.22e-05,
      "loss": 2.6353,
      "step": 161
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.2399999999999995e-05,
      "loss": 2.634,
      "step": 162
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.259999999999999e-05,
      "loss": 2.6285,
      "step": 163
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.28e-05,
      "loss": 2.6129,
      "step": 164
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 2.585,
      "step": 165
    },
    {
      "epoch": 2.24,
      "learning_rate": 3.3199999999999994e-05,
      "loss": 2.5956,
      "step": 166
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.34e-05,
      "loss": 2.6361,
      "step": 167
    },
    {
      "epoch": 2.27,
      "learning_rate": 3.36e-05,
      "loss": 2.6416,
      "step": 168
    },
    {
      "epoch": 2.28,
      "learning_rate": 3.3799999999999995e-05,
      "loss": 2.6389,
      "step": 169
    },
    {
      "epoch": 2.3,
      "learning_rate": 3.399999999999999e-05,
      "loss": 2.624,
      "step": 170
    },
    {
      "epoch": 2.31,
      "learning_rate": 3.42e-05,
      "loss": 2.6308,
      "step": 171
    },
    {
      "epoch": 2.32,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 2.6331,
      "step": 172
    },
    {
      "epoch": 2.34,
      "learning_rate": 3.4599999999999994e-05,
      "loss": 2.6285,
      "step": 173
    },
    {
      "epoch": 2.35,
      "learning_rate": 3.48e-05,
      "loss": 2.6344,
      "step": 174
    },
    {
      "epoch": 2.36,
      "learning_rate": 3.5e-05,
      "loss": 2.6295,
      "step": 175
    },
    {
      "epoch": 2.38,
      "learning_rate": 3.5199999999999995e-05,
      "loss": 2.6262,
      "step": 176
    },
    {
      "epoch": 2.39,
      "learning_rate": 3.539999999999999e-05,
      "loss": 2.6003,
      "step": 177
    },
    {
      "epoch": 2.41,
      "learning_rate": 3.56e-05,
      "loss": 2.6134,
      "step": 178
    },
    {
      "epoch": 2.42,
      "learning_rate": 3.5799999999999996e-05,
      "loss": 2.6182,
      "step": 179
    },
    {
      "epoch": 2.43,
      "learning_rate": 3.5999999999999994e-05,
      "loss": 2.6151,
      "step": 180
    },
    {
      "epoch": 2.45,
      "learning_rate": 3.62e-05,
      "loss": 2.5875,
      "step": 181
    },
    {
      "epoch": 2.46,
      "learning_rate": 3.64e-05,
      "loss": 2.5949,
      "step": 182
    },
    {
      "epoch": 2.47,
      "learning_rate": 3.6599999999999995e-05,
      "loss": 2.5829,
      "step": 183
    },
    {
      "epoch": 2.49,
      "learning_rate": 3.679999999999999e-05,
      "loss": 2.549,
      "step": 184
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.7e-05,
      "loss": 2.6056,
      "step": 185
    },
    {
      "epoch": 2.51,
      "learning_rate": 3.7199999999999996e-05,
      "loss": 2.6415,
      "step": 186
    },
    {
      "epoch": 2.53,
      "learning_rate": 3.7399999999999994e-05,
      "loss": 2.6073,
      "step": 187
    },
    {
      "epoch": 2.54,
      "learning_rate": 3.759999999999999e-05,
      "loss": 2.608,
      "step": 188
    },
    {
      "epoch": 2.55,
      "learning_rate": 3.78e-05,
      "loss": 2.5839,
      "step": 189
    },
    {
      "epoch": 2.57,
      "learning_rate": 3.8e-05,
      "loss": 2.6358,
      "step": 190
    },
    {
      "epoch": 2.58,
      "learning_rate": 3.8199999999999993e-05,
      "loss": 2.6465,
      "step": 191
    },
    {
      "epoch": 2.59,
      "learning_rate": 3.84e-05,
      "loss": 2.5963,
      "step": 192
    },
    {
      "epoch": 2.61,
      "learning_rate": 3.86e-05,
      "loss": 2.6281,
      "step": 193
    },
    {
      "epoch": 2.62,
      "learning_rate": 3.8799999999999994e-05,
      "loss": 2.6172,
      "step": 194
    },
    {
      "epoch": 2.64,
      "learning_rate": 3.9e-05,
      "loss": 2.5859,
      "step": 195
    },
    {
      "epoch": 2.65,
      "learning_rate": 3.919999999999999e-05,
      "loss": 2.618,
      "step": 196
    },
    {
      "epoch": 2.66,
      "learning_rate": 3.9399999999999995e-05,
      "loss": 2.6335,
      "step": 197
    },
    {
      "epoch": 2.68,
      "learning_rate": 3.96e-05,
      "loss": 2.6102,
      "step": 198
    },
    {
      "epoch": 2.69,
      "learning_rate": 3.979999999999999e-05,
      "loss": 2.5717,
      "step": 199
    },
    {
      "epoch": 2.7,
      "learning_rate": 3.9999999999999996e-05,
      "loss": 2.551,
      "step": 200
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.12012195121951219,
      "eval_f1": 0.040745442824033036,
      "eval_loss": 2.585480213165283,
      "eval_runtime": 125.4403,
      "eval_samples_per_second": 13.074,
      "eval_steps_per_second": 0.821,
      "step": 200
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.02e-05,
      "loss": 2.5389,
      "step": 201
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.039999999999999e-05,
      "loss": 2.5376,
      "step": 202
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.06e-05,
      "loss": 2.6092,
      "step": 203
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.08e-05,
      "loss": 2.6448,
      "step": 204
    },
    {
      "epoch": 2.77,
      "learning_rate": 4.0999999999999994e-05,
      "loss": 2.5844,
      "step": 205
    },
    {
      "epoch": 2.78,
      "learning_rate": 4.12e-05,
      "loss": 2.5996,
      "step": 206
    },
    {
      "epoch": 2.8,
      "learning_rate": 4.14e-05,
      "loss": 2.5904,
      "step": 207
    },
    {
      "epoch": 2.81,
      "learning_rate": 4.1599999999999995e-05,
      "loss": 2.6111,
      "step": 208
    },
    {
      "epoch": 2.82,
      "learning_rate": 4.18e-05,
      "loss": 2.6169,
      "step": 209
    },
    {
      "epoch": 2.84,
      "learning_rate": 4.2e-05,
      "loss": 2.5966,
      "step": 210
    },
    {
      "epoch": 2.85,
      "learning_rate": 4.2199999999999996e-05,
      "loss": 2.601,
      "step": 211
    },
    {
      "epoch": 2.86,
      "learning_rate": 4.24e-05,
      "loss": 2.6214,
      "step": 212
    },
    {
      "epoch": 2.88,
      "learning_rate": 4.259999999999999e-05,
      "loss": 2.5995,
      "step": 213
    },
    {
      "epoch": 2.89,
      "learning_rate": 4.28e-05,
      "loss": 2.5751,
      "step": 214
    },
    {
      "epoch": 2.91,
      "learning_rate": 4.3e-05,
      "loss": 2.5673,
      "step": 215
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.319999999999999e-05,
      "loss": 2.5665,
      "step": 216
    },
    {
      "epoch": 2.93,
      "learning_rate": 4.34e-05,
      "loss": 2.6174,
      "step": 217
    },
    {
      "epoch": 2.95,
      "learning_rate": 4.3599999999999996e-05,
      "loss": 2.5674,
      "step": 218
    },
    {
      "epoch": 2.96,
      "learning_rate": 4.3799999999999994e-05,
      "loss": 2.5886,
      "step": 219
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.4e-05,
      "loss": 2.5122,
      "step": 220
    },
    {
      "epoch": 2.99,
      "learning_rate": 4.42e-05,
      "loss": 2.5953,
      "step": 221
    },
    {
      "epoch": 3.0,
      "learning_rate": 4.4399999999999995e-05,
      "loss": 2.5498,
      "step": 222
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.46e-05,
      "loss": 2.5118,
      "step": 223
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.48e-05,
      "loss": 2.5294,
      "step": 224
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 2.5166,
      "step": 225
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.5199999999999994e-05,
      "loss": 2.5378,
      "step": 226
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.539999999999999e-05,
      "loss": 2.5504,
      "step": 227
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.56e-05,
      "loss": 2.5433,
      "step": 228
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.5799999999999995e-05,
      "loss": 2.6383,
      "step": 229
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.599999999999999e-05,
      "loss": 2.5849,
      "step": 230
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.62e-05,
      "loss": 2.6266,
      "step": 231
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.6399999999999996e-05,
      "loss": 2.5249,
      "step": 232
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.6599999999999994e-05,
      "loss": 2.5782,
      "step": 233
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.68e-05,
      "loss": 2.611,
      "step": 234
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.7e-05,
      "loss": 2.6289,
      "step": 235
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.7199999999999995e-05,
      "loss": 2.5187,
      "step": 236
    },
    {
      "epoch": 3.2,
      "learning_rate": 4.7399999999999993e-05,
      "loss": 2.5624,
      "step": 237
    },
    {
      "epoch": 3.22,
      "learning_rate": 4.76e-05,
      "loss": 2.5188,
      "step": 238
    },
    {
      "epoch": 3.23,
      "learning_rate": 4.7799999999999996e-05,
      "loss": 2.5248,
      "step": 239
    },
    {
      "epoch": 3.24,
      "learning_rate": 4.7999999999999994e-05,
      "loss": 2.4677,
      "step": 240
    },
    {
      "epoch": 3.26,
      "learning_rate": 4.82e-05,
      "loss": 2.5576,
      "step": 241
    },
    {
      "epoch": 3.27,
      "learning_rate": 4.84e-05,
      "loss": 2.6681,
      "step": 242
    },
    {
      "epoch": 3.28,
      "learning_rate": 4.8599999999999995e-05,
      "loss": 2.5615,
      "step": 243
    },
    {
      "epoch": 3.3,
      "learning_rate": 4.8799999999999994e-05,
      "loss": 2.6146,
      "step": 244
    },
    {
      "epoch": 3.31,
      "learning_rate": 4.899999999999999e-05,
      "loss": 2.5792,
      "step": 245
    },
    {
      "epoch": 3.32,
      "learning_rate": 4.9199999999999997e-05,
      "loss": 2.5896,
      "step": 246
    },
    {
      "epoch": 3.34,
      "learning_rate": 4.9399999999999995e-05,
      "loss": 2.5957,
      "step": 247
    },
    {
      "epoch": 3.35,
      "learning_rate": 4.959999999999999e-05,
      "loss": 2.5715,
      "step": 248
    },
    {
      "epoch": 3.36,
      "learning_rate": 4.98e-05,
      "loss": 2.5662,
      "step": 249
    },
    {
      "epoch": 3.38,
      "learning_rate": 4.9999999999999996e-05,
      "loss": 2.5826,
      "step": 250
    },
    {
      "epoch": 3.39,
      "learning_rate": 5.0199999999999994e-05,
      "loss": 2.5727,
      "step": 251
    },
    {
      "epoch": 3.41,
      "learning_rate": 5.04e-05,
      "loss": 2.5583,
      "step": 252
    },
    {
      "epoch": 3.42,
      "learning_rate": 5.06e-05,
      "loss": 2.5106,
      "step": 253
    },
    {
      "epoch": 3.43,
      "learning_rate": 5.0799999999999995e-05,
      "loss": 2.574,
      "step": 254
    },
    {
      "epoch": 3.45,
      "learning_rate": 5.1e-05,
      "loss": 2.4493,
      "step": 255
    },
    {
      "epoch": 3.46,
      "learning_rate": 5.119999999999999e-05,
      "loss": 2.5145,
      "step": 256
    },
    {
      "epoch": 3.47,
      "learning_rate": 5.1399999999999996e-05,
      "loss": 2.4472,
      "step": 257
    },
    {
      "epoch": 3.49,
      "learning_rate": 5.1599999999999994e-05,
      "loss": 2.4566,
      "step": 258
    },
    {
      "epoch": 3.5,
      "learning_rate": 5.179999999999999e-05,
      "loss": 2.5023,
      "step": 259
    },
    {
      "epoch": 3.51,
      "learning_rate": 5.2e-05,
      "loss": 2.5159,
      "step": 260
    },
    {
      "epoch": 3.53,
      "learning_rate": 5.2199999999999995e-05,
      "loss": 2.4929,
      "step": 261
    },
    {
      "epoch": 3.54,
      "learning_rate": 5.239999999999999e-05,
      "loss": 2.5133,
      "step": 262
    },
    {
      "epoch": 3.55,
      "learning_rate": 5.26e-05,
      "loss": 2.4245,
      "step": 263
    },
    {
      "epoch": 3.57,
      "learning_rate": 5.279999999999999e-05,
      "loss": 2.4907,
      "step": 264
    },
    {
      "epoch": 3.58,
      "learning_rate": 5.2999999999999994e-05,
      "loss": 2.5539,
      "step": 265
    },
    {
      "epoch": 3.59,
      "learning_rate": 5.32e-05,
      "loss": 2.4136,
      "step": 266
    },
    {
      "epoch": 3.61,
      "learning_rate": 5.339999999999999e-05,
      "loss": 2.5215,
      "step": 267
    },
    {
      "epoch": 3.62,
      "learning_rate": 5.3599999999999995e-05,
      "loss": 2.512,
      "step": 268
    },
    {
      "epoch": 3.64,
      "learning_rate": 5.38e-05,
      "loss": 2.5115,
      "step": 269
    },
    {
      "epoch": 3.65,
      "learning_rate": 5.399999999999999e-05,
      "loss": 2.4778,
      "step": 270
    },
    {
      "epoch": 3.66,
      "learning_rate": 5.4199999999999996e-05,
      "loss": 2.4272,
      "step": 271
    },
    {
      "epoch": 3.68,
      "learning_rate": 5.44e-05,
      "loss": 2.4864,
      "step": 272
    },
    {
      "epoch": 3.69,
      "learning_rate": 5.459999999999999e-05,
      "loss": 2.4845,
      "step": 273
    },
    {
      "epoch": 3.7,
      "learning_rate": 5.48e-05,
      "loss": 2.4285,
      "step": 274
    },
    {
      "epoch": 3.72,
      "learning_rate": 5.499999999999999e-05,
      "loss": 2.4581,
      "step": 275
    },
    {
      "epoch": 3.73,
      "learning_rate": 5.519999999999999e-05,
      "loss": 2.3295,
      "step": 276
    },
    {
      "epoch": 3.74,
      "learning_rate": 5.54e-05,
      "loss": 2.5608,
      "step": 277
    },
    {
      "epoch": 3.76,
      "learning_rate": 5.559999999999999e-05,
      "loss": 2.5338,
      "step": 278
    },
    {
      "epoch": 3.77,
      "learning_rate": 5.5799999999999994e-05,
      "loss": 2.4393,
      "step": 279
    },
    {
      "epoch": 3.78,
      "learning_rate": 5.6e-05,
      "loss": 2.4816,
      "step": 280
    },
    {
      "epoch": 3.8,
      "learning_rate": 5.619999999999999e-05,
      "loss": 2.3688,
      "step": 281
    },
    {
      "epoch": 3.81,
      "learning_rate": 5.6399999999999995e-05,
      "loss": 2.398,
      "step": 282
    },
    {
      "epoch": 3.82,
      "learning_rate": 5.66e-05,
      "loss": 2.4456,
      "step": 283
    },
    {
      "epoch": 3.84,
      "learning_rate": 5.679999999999999e-05,
      "loss": 2.5291,
      "step": 284
    },
    {
      "epoch": 3.85,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 2.4221,
      "step": 285
    },
    {
      "epoch": 3.86,
      "learning_rate": 5.72e-05,
      "loss": 2.4566,
      "step": 286
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.739999999999999e-05,
      "loss": 2.3871,
      "step": 287
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.76e-05,
      "loss": 2.3566,
      "step": 288
    },
    {
      "epoch": 3.91,
      "learning_rate": 5.78e-05,
      "loss": 2.3178,
      "step": 289
    },
    {
      "epoch": 3.92,
      "learning_rate": 5.7999999999999994e-05,
      "loss": 2.3941,
      "step": 290
    },
    {
      "epoch": 3.93,
      "learning_rate": 5.82e-05,
      "loss": 2.3332,
      "step": 291
    },
    {
      "epoch": 3.95,
      "learning_rate": 5.839999999999999e-05,
      "loss": 2.4014,
      "step": 292
    },
    {
      "epoch": 3.96,
      "learning_rate": 5.8599999999999995e-05,
      "loss": 2.4147,
      "step": 293
    },
    {
      "epoch": 3.97,
      "learning_rate": 5.88e-05,
      "loss": 2.4667,
      "step": 294
    },
    {
      "epoch": 3.99,
      "learning_rate": 5.899999999999999e-05,
      "loss": 2.338,
      "step": 295
    },
    {
      "epoch": 4.0,
      "learning_rate": 5.9199999999999996e-05,
      "loss": 2.2476,
      "step": 296
    },
    {
      "epoch": 4.01,
      "learning_rate": 5.94e-05,
      "loss": 2.4051,
      "step": 297
    },
    {
      "epoch": 4.03,
      "learning_rate": 5.959999999999999e-05,
      "loss": 2.4817,
      "step": 298
    },
    {
      "epoch": 4.04,
      "learning_rate": 5.98e-05,
      "loss": 2.4715,
      "step": 299
    },
    {
      "epoch": 4.05,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 2.2758,
      "step": 300
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.019999999999999e-05,
      "loss": 2.2442,
      "step": 301
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.04e-05,
      "loss": 2.294,
      "step": 302
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.0599999999999996e-05,
      "loss": 2.2415,
      "step": 303
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.0799999999999994e-05,
      "loss": 2.256,
      "step": 304
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.1e-05,
      "loss": 2.3082,
      "step": 305
    },
    {
      "epoch": 4.14,
      "learning_rate": 6.12e-05,
      "loss": 2.2657,
      "step": 306
    },
    {
      "epoch": 4.15,
      "learning_rate": 6.139999999999999e-05,
      "loss": 2.2412,
      "step": 307
    },
    {
      "epoch": 4.16,
      "learning_rate": 6.159999999999999e-05,
      "loss": 2.1984,
      "step": 308
    },
    {
      "epoch": 4.18,
      "learning_rate": 6.18e-05,
      "loss": 2.2006,
      "step": 309
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.199999999999999e-05,
      "loss": 2.2552,
      "step": 310
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.22e-05,
      "loss": 2.2136,
      "step": 311
    },
    {
      "epoch": 4.22,
      "learning_rate": 6.239999999999999e-05,
      "loss": 2.2209,
      "step": 312
    },
    {
      "epoch": 4.23,
      "learning_rate": 6.259999999999999e-05,
      "loss": 2.1844,
      "step": 313
    },
    {
      "epoch": 4.24,
      "learning_rate": 6.28e-05,
      "loss": 2.1825,
      "step": 314
    },
    {
      "epoch": 4.26,
      "learning_rate": 6.299999999999999e-05,
      "loss": 2.3587,
      "step": 315
    },
    {
      "epoch": 4.27,
      "learning_rate": 6.319999999999999e-05,
      "loss": 2.2979,
      "step": 316
    },
    {
      "epoch": 4.28,
      "learning_rate": 6.34e-05,
      "loss": 2.2733,
      "step": 317
    },
    {
      "epoch": 4.3,
      "learning_rate": 6.359999999999999e-05,
      "loss": 2.2113,
      "step": 318
    },
    {
      "epoch": 4.31,
      "learning_rate": 6.379999999999999e-05,
      "loss": 2.1144,
      "step": 319
    },
    {
      "epoch": 4.32,
      "learning_rate": 6.4e-05,
      "loss": 2.1475,
      "step": 320
    },
    {
      "epoch": 4.34,
      "learning_rate": 6.419999999999999e-05,
      "loss": 2.0578,
      "step": 321
    },
    {
      "epoch": 4.35,
      "learning_rate": 6.44e-05,
      "loss": 2.1867,
      "step": 322
    },
    {
      "epoch": 4.36,
      "learning_rate": 6.459999999999998e-05,
      "loss": 2.0722,
      "step": 323
    },
    {
      "epoch": 4.38,
      "learning_rate": 6.479999999999999e-05,
      "loss": 2.2211,
      "step": 324
    },
    {
      "epoch": 4.39,
      "learning_rate": 6.5e-05,
      "loss": 2.1083,
      "step": 325
    },
    {
      "epoch": 4.41,
      "learning_rate": 6.519999999999999e-05,
      "loss": 2.2044,
      "step": 326
    },
    {
      "epoch": 4.42,
      "learning_rate": 6.539999999999999e-05,
      "loss": 2.1204,
      "step": 327
    },
    {
      "epoch": 4.43,
      "learning_rate": 6.56e-05,
      "loss": 2.2203,
      "step": 328
    },
    {
      "epoch": 4.45,
      "learning_rate": 6.579999999999999e-05,
      "loss": 2.0703,
      "step": 329
    },
    {
      "epoch": 4.46,
      "learning_rate": 6.599999999999999e-05,
      "loss": 2.1777,
      "step": 330
    },
    {
      "epoch": 4.47,
      "learning_rate": 6.62e-05,
      "loss": 2.0376,
      "step": 331
    },
    {
      "epoch": 4.49,
      "learning_rate": 6.639999999999999e-05,
      "loss": 2.0562,
      "step": 332
    },
    {
      "epoch": 4.5,
      "learning_rate": 6.659999999999999e-05,
      "loss": 2.2317,
      "step": 333
    },
    {
      "epoch": 4.51,
      "learning_rate": 6.68e-05,
      "loss": 1.9297,
      "step": 334
    },
    {
      "epoch": 4.53,
      "learning_rate": 6.68e-05,
      "loss": 1.9537,
      "step": 335
    },
    {
      "epoch": 4.54,
      "learning_rate": 6.699999999999999e-05,
      "loss": 2.0425,
      "step": 336
    },
    {
      "epoch": 4.55,
      "learning_rate": 6.72e-05,
      "loss": 1.8857,
      "step": 337
    },
    {
      "epoch": 4.57,
      "learning_rate": 6.739999999999998e-05,
      "loss": 2.0898,
      "step": 338
    },
    {
      "epoch": 4.58,
      "learning_rate": 6.759999999999999e-05,
      "loss": 1.9779,
      "step": 339
    },
    {
      "epoch": 4.59,
      "learning_rate": 6.78e-05,
      "loss": 2.2227,
      "step": 340
    },
    {
      "epoch": 4.61,
      "learning_rate": 6.799999999999999e-05,
      "loss": 1.9317,
      "step": 341
    },
    {
      "epoch": 4.62,
      "learning_rate": 6.819999999999999e-05,
      "loss": 2.2124,
      "step": 342
    },
    {
      "epoch": 4.64,
      "learning_rate": 6.84e-05,
      "loss": 1.9054,
      "step": 343
    },
    {
      "epoch": 4.65,
      "learning_rate": 6.859999999999999e-05,
      "loss": 1.9361,
      "step": 344
    },
    {
      "epoch": 4.66,
      "learning_rate": 6.879999999999999e-05,
      "loss": 2.1215,
      "step": 345
    },
    {
      "epoch": 4.68,
      "learning_rate": 6.9e-05,
      "loss": 2.0735,
      "step": 346
    },
    {
      "epoch": 4.69,
      "learning_rate": 6.919999999999999e-05,
      "loss": 1.9301,
      "step": 347
    },
    {
      "epoch": 4.7,
      "learning_rate": 6.939999999999999e-05,
      "loss": 1.975,
      "step": 348
    },
    {
      "epoch": 4.72,
      "learning_rate": 6.96e-05,
      "loss": 1.9047,
      "step": 349
    },
    {
      "epoch": 4.73,
      "learning_rate": 6.979999999999999e-05,
      "loss": 2.0832,
      "step": 350
    },
    {
      "epoch": 4.74,
      "learning_rate": 7e-05,
      "loss": 1.8544,
      "step": 351
    },
    {
      "epoch": 4.76,
      "learning_rate": 7.02e-05,
      "loss": 1.9242,
      "step": 352
    },
    {
      "epoch": 4.77,
      "learning_rate": 7.039999999999999e-05,
      "loss": 1.8664,
      "step": 353
    },
    {
      "epoch": 4.78,
      "learning_rate": 7.06e-05,
      "loss": 1.8887,
      "step": 354
    },
    {
      "epoch": 4.8,
      "learning_rate": 7.079999999999999e-05,
      "loss": 1.889,
      "step": 355
    },
    {
      "epoch": 4.81,
      "learning_rate": 7.099999999999999e-05,
      "loss": 1.875,
      "step": 356
    },
    {
      "epoch": 4.82,
      "learning_rate": 7.12e-05,
      "loss": 1.7739,
      "step": 357
    },
    {
      "epoch": 4.84,
      "learning_rate": 7.139999999999999e-05,
      "loss": 1.7445,
      "step": 358
    },
    {
      "epoch": 4.85,
      "learning_rate": 7.159999999999999e-05,
      "loss": 1.8671,
      "step": 359
    },
    {
      "epoch": 4.86,
      "learning_rate": 7.18e-05,
      "loss": 1.7867,
      "step": 360
    },
    {
      "epoch": 4.88,
      "learning_rate": 7.199999999999999e-05,
      "loss": 1.7686,
      "step": 361
    },
    {
      "epoch": 4.89,
      "learning_rate": 7.219999999999999e-05,
      "loss": 1.6284,
      "step": 362
    },
    {
      "epoch": 4.91,
      "learning_rate": 7.24e-05,
      "loss": 1.791,
      "step": 363
    },
    {
      "epoch": 4.92,
      "learning_rate": 7.259999999999999e-05,
      "loss": 1.8432,
      "step": 364
    },
    {
      "epoch": 4.93,
      "learning_rate": 7.28e-05,
      "loss": 1.9891,
      "step": 365
    },
    {
      "epoch": 4.95,
      "learning_rate": 7.3e-05,
      "loss": 1.9106,
      "step": 366
    },
    {
      "epoch": 4.96,
      "learning_rate": 7.319999999999999e-05,
      "loss": 1.727,
      "step": 367
    },
    {
      "epoch": 4.97,
      "learning_rate": 7.34e-05,
      "loss": 1.7622,
      "step": 368
    },
    {
      "epoch": 4.99,
      "learning_rate": 7.359999999999999e-05,
      "loss": 1.6442,
      "step": 369
    },
    {
      "epoch": 5.0,
      "learning_rate": 7.379999999999999e-05,
      "loss": 1.2379,
      "step": 370
    },
    {
      "epoch": 5.01,
      "learning_rate": 7.4e-05,
      "loss": 1.9131,
      "step": 371
    },
    {
      "epoch": 5.03,
      "learning_rate": 7.419999999999999e-05,
      "loss": 1.7918,
      "step": 372
    },
    {
      "epoch": 5.04,
      "learning_rate": 7.439999999999999e-05,
      "loss": 1.6588,
      "step": 373
    },
    {
      "epoch": 5.05,
      "learning_rate": 7.46e-05,
      "loss": 1.6115,
      "step": 374
    },
    {
      "epoch": 5.07,
      "learning_rate": 7.479999999999999e-05,
      "loss": 1.4868,
      "step": 375
    },
    {
      "epoch": 5.08,
      "learning_rate": 7.5e-05,
      "loss": 1.6101,
      "step": 376
    },
    {
      "epoch": 5.09,
      "learning_rate": 7.519999999999998e-05,
      "loss": 1.6173,
      "step": 377
    },
    {
      "epoch": 5.11,
      "learning_rate": 7.54e-05,
      "loss": 1.6393,
      "step": 378
    },
    {
      "epoch": 5.12,
      "learning_rate": 7.56e-05,
      "loss": 1.896,
      "step": 379
    },
    {
      "epoch": 5.14,
      "learning_rate": 7.579999999999999e-05,
      "loss": 1.4267,
      "step": 380
    },
    {
      "epoch": 5.15,
      "learning_rate": 7.6e-05,
      "loss": 1.5306,
      "step": 381
    },
    {
      "epoch": 5.16,
      "learning_rate": 7.62e-05,
      "loss": 1.6724,
      "step": 382
    },
    {
      "epoch": 5.18,
      "learning_rate": 7.639999999999999e-05,
      "loss": 1.7283,
      "step": 383
    },
    {
      "epoch": 5.19,
      "learning_rate": 7.66e-05,
      "loss": 1.8373,
      "step": 384
    },
    {
      "epoch": 5.2,
      "learning_rate": 7.68e-05,
      "loss": 1.7769,
      "step": 385
    },
    {
      "epoch": 5.22,
      "learning_rate": 7.699999999999999e-05,
      "loss": 1.5922,
      "step": 386
    },
    {
      "epoch": 5.23,
      "learning_rate": 7.72e-05,
      "loss": 1.5844,
      "step": 387
    },
    {
      "epoch": 5.24,
      "learning_rate": 7.74e-05,
      "loss": 1.6893,
      "step": 388
    },
    {
      "epoch": 5.26,
      "learning_rate": 7.759999999999999e-05,
      "loss": 1.7527,
      "step": 389
    },
    {
      "epoch": 5.27,
      "learning_rate": 7.780000000000001e-05,
      "loss": 1.4456,
      "step": 390
    },
    {
      "epoch": 5.28,
      "learning_rate": 7.8e-05,
      "loss": 1.591,
      "step": 391
    },
    {
      "epoch": 5.3,
      "learning_rate": 7.819999999999999e-05,
      "loss": 1.512,
      "step": 392
    },
    {
      "epoch": 5.31,
      "learning_rate": 7.839999999999998e-05,
      "loss": 1.4948,
      "step": 393
    },
    {
      "epoch": 5.32,
      "learning_rate": 7.86e-05,
      "loss": 1.2908,
      "step": 394
    },
    {
      "epoch": 5.34,
      "learning_rate": 7.879999999999999e-05,
      "loss": 1.4818,
      "step": 395
    },
    {
      "epoch": 5.35,
      "learning_rate": 7.899999999999998e-05,
      "loss": 1.6264,
      "step": 396
    },
    {
      "epoch": 5.36,
      "learning_rate": 7.92e-05,
      "loss": 1.6992,
      "step": 397
    },
    {
      "epoch": 5.38,
      "learning_rate": 7.939999999999999e-05,
      "loss": 1.5083,
      "step": 398
    },
    {
      "epoch": 5.39,
      "learning_rate": 7.959999999999998e-05,
      "loss": 1.6165,
      "step": 399
    },
    {
      "epoch": 5.41,
      "learning_rate": 7.98e-05,
      "loss": 1.6934,
      "step": 400
    },
    {
      "epoch": 5.41,
      "eval_accuracy": 0.6085365853658536,
      "eval_f1": 0.5861868311714201,
      "eval_loss": 1.5071511268615723,
      "eval_runtime": 125.2449,
      "eval_samples_per_second": 13.094,
      "eval_steps_per_second": 0.822,
      "step": 400
    },
    {
      "epoch": 5.42,
      "learning_rate": 7.999999999999999e-05,
      "loss": 1.4664,
      "step": 401
    },
    {
      "epoch": 5.43,
      "learning_rate": 8.019999999999998e-05,
      "loss": 1.5484,
      "step": 402
    },
    {
      "epoch": 5.45,
      "learning_rate": 8.04e-05,
      "loss": 1.4594,
      "step": 403
    },
    {
      "epoch": 5.46,
      "learning_rate": 8.06e-05,
      "loss": 1.3655,
      "step": 404
    },
    {
      "epoch": 5.47,
      "learning_rate": 8.079999999999999e-05,
      "loss": 1.3421,
      "step": 405
    },
    {
      "epoch": 5.49,
      "learning_rate": 8.1e-05,
      "loss": 1.4723,
      "step": 406
    },
    {
      "epoch": 5.5,
      "learning_rate": 8.12e-05,
      "loss": 1.5759,
      "step": 407
    },
    {
      "epoch": 5.51,
      "learning_rate": 8.139999999999999e-05,
      "loss": 1.6416,
      "step": 408
    },
    {
      "epoch": 5.53,
      "learning_rate": 8.16e-05,
      "loss": 1.4298,
      "step": 409
    },
    {
      "epoch": 5.54,
      "learning_rate": 8.18e-05,
      "loss": 1.3123,
      "step": 410
    },
    {
      "epoch": 5.55,
      "learning_rate": 8.199999999999999e-05,
      "loss": 1.3655,
      "step": 411
    },
    {
      "epoch": 5.57,
      "learning_rate": 8.22e-05,
      "loss": 1.3117,
      "step": 412
    },
    {
      "epoch": 5.58,
      "learning_rate": 8.24e-05,
      "loss": 1.1209,
      "step": 413
    },
    {
      "epoch": 5.59,
      "learning_rate": 8.259999999999999e-05,
      "loss": 1.418,
      "step": 414
    },
    {
      "epoch": 5.61,
      "learning_rate": 8.28e-05,
      "loss": 1.3417,
      "step": 415
    },
    {
      "epoch": 5.62,
      "learning_rate": 8.3e-05,
      "loss": 1.5942,
      "step": 416
    },
    {
      "epoch": 5.64,
      "learning_rate": 8.319999999999999e-05,
      "loss": 1.3329,
      "step": 417
    },
    {
      "epoch": 5.65,
      "learning_rate": 8.34e-05,
      "loss": 1.6013,
      "step": 418
    },
    {
      "epoch": 5.66,
      "learning_rate": 8.36e-05,
      "loss": 1.4991,
      "step": 419
    },
    {
      "epoch": 5.68,
      "learning_rate": 8.379999999999999e-05,
      "loss": 1.5366,
      "step": 420
    },
    {
      "epoch": 5.69,
      "learning_rate": 8.4e-05,
      "loss": 1.4418,
      "step": 421
    },
    {
      "epoch": 5.7,
      "learning_rate": 8.42e-05,
      "loss": 1.4736,
      "step": 422
    },
    {
      "epoch": 5.72,
      "learning_rate": 8.439999999999999e-05,
      "loss": 1.4342,
      "step": 423
    },
    {
      "epoch": 5.73,
      "learning_rate": 8.459999999999998e-05,
      "loss": 1.3454,
      "step": 424
    },
    {
      "epoch": 5.74,
      "learning_rate": 8.48e-05,
      "loss": 1.6085,
      "step": 425
    },
    {
      "epoch": 5.76,
      "learning_rate": 8.499999999999999e-05,
      "loss": 1.5582,
      "step": 426
    },
    {
      "epoch": 5.77,
      "learning_rate": 8.519999999999998e-05,
      "loss": 1.3238,
      "step": 427
    },
    {
      "epoch": 5.78,
      "learning_rate": 8.54e-05,
      "loss": 1.2761,
      "step": 428
    },
    {
      "epoch": 5.8,
      "learning_rate": 8.56e-05,
      "loss": 1.3632,
      "step": 429
    },
    {
      "epoch": 5.81,
      "learning_rate": 8.579999999999998e-05,
      "loss": 1.1634,
      "step": 430
    },
    {
      "epoch": 5.82,
      "learning_rate": 8.6e-05,
      "loss": 1.1411,
      "step": 431
    },
    {
      "epoch": 5.84,
      "learning_rate": 8.62e-05,
      "loss": 1.3476,
      "step": 432
    },
    {
      "epoch": 5.85,
      "learning_rate": 8.639999999999999e-05,
      "loss": 1.1318,
      "step": 433
    },
    {
      "epoch": 5.86,
      "learning_rate": 8.659999999999999e-05,
      "loss": 1.3659,
      "step": 434
    },
    {
      "epoch": 5.88,
      "learning_rate": 8.68e-05,
      "loss": 1.3049,
      "step": 435
    },
    {
      "epoch": 5.89,
      "learning_rate": 8.699999999999999e-05,
      "loss": 1.2728,
      "step": 436
    },
    {
      "epoch": 5.91,
      "learning_rate": 8.719999999999999e-05,
      "loss": 1.3106,
      "step": 437
    },
    {
      "epoch": 5.92,
      "learning_rate": 8.74e-05,
      "loss": 1.1085,
      "step": 438
    },
    {
      "epoch": 5.93,
      "learning_rate": 8.759999999999999e-05,
      "loss": 1.4459,
      "step": 439
    },
    {
      "epoch": 5.95,
      "learning_rate": 8.779999999999999e-05,
      "loss": 1.3631,
      "step": 440
    },
    {
      "epoch": 5.96,
      "learning_rate": 8.8e-05,
      "loss": 1.5037,
      "step": 441
    },
    {
      "epoch": 5.97,
      "learning_rate": 8.819999999999999e-05,
      "loss": 1.3511,
      "step": 442
    },
    {
      "epoch": 5.99,
      "learning_rate": 8.84e-05,
      "loss": 1.2469,
      "step": 443
    },
    {
      "epoch": 6.0,
      "learning_rate": 8.86e-05,
      "loss": 1.2176,
      "step": 444
    },
    {
      "epoch": 6.01,
      "learning_rate": 8.879999999999999e-05,
      "loss": 1.1247,
      "step": 445
    },
    {
      "epoch": 6.03,
      "learning_rate": 8.9e-05,
      "loss": 1.0558,
      "step": 446
    },
    {
      "epoch": 6.04,
      "learning_rate": 8.92e-05,
      "loss": 1.2339,
      "step": 447
    },
    {
      "epoch": 6.05,
      "learning_rate": 8.939999999999999e-05,
      "loss": 1.1128,
      "step": 448
    },
    {
      "epoch": 6.07,
      "learning_rate": 8.96e-05,
      "loss": 1.0906,
      "step": 449
    },
    {
      "epoch": 6.08,
      "learning_rate": 8.98e-05,
      "loss": 1.0071,
      "step": 450
    },
    {
      "epoch": 6.09,
      "learning_rate": 8.999999999999999e-05,
      "loss": 1.2068,
      "step": 451
    },
    {
      "epoch": 6.11,
      "learning_rate": 9.02e-05,
      "loss": 1.0462,
      "step": 452
    },
    {
      "epoch": 6.12,
      "learning_rate": 9.039999999999999e-05,
      "loss": 1.1117,
      "step": 453
    },
    {
      "epoch": 6.14,
      "learning_rate": 9.059999999999999e-05,
      "loss": 1.0696,
      "step": 454
    },
    {
      "epoch": 6.15,
      "learning_rate": 9.079999999999998e-05,
      "loss": 0.9539,
      "step": 455
    },
    {
      "epoch": 6.16,
      "learning_rate": 9.099999999999999e-05,
      "loss": 1.2348,
      "step": 456
    },
    {
      "epoch": 6.18,
      "learning_rate": 9.12e-05,
      "loss": 1.0503,
      "step": 457
    },
    {
      "epoch": 6.19,
      "learning_rate": 9.139999999999999e-05,
      "loss": 1.2834,
      "step": 458
    },
    {
      "epoch": 6.2,
      "learning_rate": 9.159999999999999e-05,
      "loss": 1.1434,
      "step": 459
    },
    {
      "epoch": 6.22,
      "learning_rate": 9.18e-05,
      "loss": 1.1374,
      "step": 460
    },
    {
      "epoch": 6.23,
      "learning_rate": 9.199999999999999e-05,
      "loss": 1.1996,
      "step": 461
    },
    {
      "epoch": 6.24,
      "learning_rate": 9.219999999999999e-05,
      "loss": 1.2158,
      "step": 462
    },
    {
      "epoch": 6.26,
      "learning_rate": 9.24e-05,
      "loss": 1.2133,
      "step": 463
    },
    {
      "epoch": 6.27,
      "learning_rate": 9.259999999999999e-05,
      "loss": 1.3433,
      "step": 464
    },
    {
      "epoch": 6.28,
      "learning_rate": 9.279999999999999e-05,
      "loss": 0.9525,
      "step": 465
    },
    {
      "epoch": 6.3,
      "learning_rate": 9.3e-05,
      "loss": 0.8775,
      "step": 466
    },
    {
      "epoch": 6.31,
      "learning_rate": 9.319999999999999e-05,
      "loss": 1.1374,
      "step": 467
    },
    {
      "epoch": 6.32,
      "learning_rate": 9.34e-05,
      "loss": 0.8626,
      "step": 468
    },
    {
      "epoch": 6.34,
      "learning_rate": 9.36e-05,
      "loss": 1.1012,
      "step": 469
    },
    {
      "epoch": 6.35,
      "learning_rate": 9.379999999999999e-05,
      "loss": 1.096,
      "step": 470
    },
    {
      "epoch": 6.36,
      "learning_rate": 9.4e-05,
      "loss": 1.0571,
      "step": 471
    },
    {
      "epoch": 6.38,
      "learning_rate": 9.419999999999999e-05,
      "loss": 1.0057,
      "step": 472
    },
    {
      "epoch": 6.39,
      "learning_rate": 9.439999999999999e-05,
      "loss": 1.0507,
      "step": 473
    },
    {
      "epoch": 6.41,
      "learning_rate": 9.46e-05,
      "loss": 1.0086,
      "step": 474
    },
    {
      "epoch": 6.42,
      "learning_rate": 9.479999999999999e-05,
      "loss": 0.9422,
      "step": 475
    },
    {
      "epoch": 6.43,
      "learning_rate": 9.499999999999999e-05,
      "loss": 1.2987,
      "step": 476
    },
    {
      "epoch": 6.45,
      "learning_rate": 9.52e-05,
      "loss": 1.147,
      "step": 477
    },
    {
      "epoch": 6.46,
      "learning_rate": 9.539999999999999e-05,
      "loss": 0.7816,
      "step": 478
    },
    {
      "epoch": 6.47,
      "learning_rate": 9.559999999999999e-05,
      "loss": 0.8796,
      "step": 479
    },
    {
      "epoch": 6.49,
      "learning_rate": 9.58e-05,
      "loss": 0.8271,
      "step": 480
    },
    {
      "epoch": 6.5,
      "learning_rate": 9.599999999999999e-05,
      "loss": 1.1841,
      "step": 481
    },
    {
      "epoch": 6.51,
      "learning_rate": 9.62e-05,
      "loss": 0.8443,
      "step": 482
    },
    {
      "epoch": 6.53,
      "learning_rate": 9.64e-05,
      "loss": 1.0368,
      "step": 483
    },
    {
      "epoch": 6.54,
      "learning_rate": 9.659999999999999e-05,
      "loss": 0.7301,
      "step": 484
    },
    {
      "epoch": 6.55,
      "learning_rate": 9.68e-05,
      "loss": 0.867,
      "step": 485
    },
    {
      "epoch": 6.57,
      "learning_rate": 9.699999999999999e-05,
      "loss": 0.8702,
      "step": 486
    },
    {
      "epoch": 6.58,
      "learning_rate": 9.719999999999999e-05,
      "loss": 0.7351,
      "step": 487
    },
    {
      "epoch": 6.59,
      "learning_rate": 9.74e-05,
      "loss": 0.9688,
      "step": 488
    },
    {
      "epoch": 6.61,
      "learning_rate": 9.759999999999999e-05,
      "loss": 0.9619,
      "step": 489
    },
    {
      "epoch": 6.62,
      "learning_rate": 9.779999999999999e-05,
      "loss": 0.9119,
      "step": 490
    },
    {
      "epoch": 6.64,
      "learning_rate": 9.799999999999998e-05,
      "loss": 0.9434,
      "step": 491
    },
    {
      "epoch": 6.65,
      "learning_rate": 9.819999999999999e-05,
      "loss": 1.018,
      "step": 492
    },
    {
      "epoch": 6.66,
      "learning_rate": 9.839999999999999e-05,
      "loss": 1.0254,
      "step": 493
    },
    {
      "epoch": 6.68,
      "learning_rate": 9.859999999999998e-05,
      "loss": 1.0968,
      "step": 494
    },
    {
      "epoch": 6.69,
      "learning_rate": 9.879999999999999e-05,
      "loss": 0.9505,
      "step": 495
    },
    {
      "epoch": 6.7,
      "learning_rate": 9.9e-05,
      "loss": 0.8652,
      "step": 496
    },
    {
      "epoch": 6.72,
      "learning_rate": 9.919999999999999e-05,
      "loss": 0.9814,
      "step": 497
    },
    {
      "epoch": 6.73,
      "learning_rate": 9.939999999999999e-05,
      "loss": 0.9181,
      "step": 498
    },
    {
      "epoch": 6.74,
      "learning_rate": 9.96e-05,
      "loss": 1.01,
      "step": 499
    },
    {
      "epoch": 6.76,
      "learning_rate": 9.979999999999999e-05,
      "loss": 0.7266,
      "step": 500
    },
    {
      "epoch": 6.77,
      "learning_rate": 9.999999999999999e-05,
      "loss": 0.7895,
      "step": 501
    },
    {
      "epoch": 6.78,
      "learning_rate": 0.0001002,
      "loss": 0.611,
      "step": 502
    },
    {
      "epoch": 6.8,
      "learning_rate": 0.00010039999999999999,
      "loss": 1.1868,
      "step": 503
    },
    {
      "epoch": 6.81,
      "learning_rate": 0.00010059999999999999,
      "loss": 0.8831,
      "step": 504
    },
    {
      "epoch": 6.82,
      "learning_rate": 0.0001008,
      "loss": 0.8871,
      "step": 505
    },
    {
      "epoch": 6.84,
      "learning_rate": 0.00010099999999999999,
      "loss": 0.9081,
      "step": 506
    },
    {
      "epoch": 6.85,
      "learning_rate": 0.0001012,
      "loss": 0.9286,
      "step": 507
    },
    {
      "epoch": 6.86,
      "learning_rate": 0.0001014,
      "loss": 0.7412,
      "step": 508
    },
    {
      "epoch": 6.88,
      "learning_rate": 0.00010159999999999999,
      "loss": 1.0034,
      "step": 509
    },
    {
      "epoch": 6.89,
      "learning_rate": 0.00010179999999999998,
      "loss": 1.0609,
      "step": 510
    },
    {
      "epoch": 6.91,
      "learning_rate": 0.000102,
      "loss": 0.747,
      "step": 511
    },
    {
      "epoch": 6.92,
      "learning_rate": 0.00010219999999999999,
      "loss": 1.0779,
      "step": 512
    },
    {
      "epoch": 6.93,
      "learning_rate": 0.00010239999999999998,
      "loss": 0.9145,
      "step": 513
    },
    {
      "epoch": 6.95,
      "learning_rate": 0.0001026,
      "loss": 1.1139,
      "step": 514
    },
    {
      "epoch": 6.96,
      "learning_rate": 0.00010279999999999999,
      "loss": 1.0289,
      "step": 515
    },
    {
      "epoch": 6.97,
      "learning_rate": 0.00010299999999999998,
      "loss": 0.9738,
      "step": 516
    },
    {
      "epoch": 6.99,
      "learning_rate": 0.00010319999999999999,
      "loss": 0.8002,
      "step": 517
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00010339999999999999,
      "loss": 1.2751,
      "step": 518
    },
    {
      "epoch": 7.01,
      "learning_rate": 0.00010359999999999998,
      "loss": 0.7601,
      "step": 519
    },
    {
      "epoch": 7.03,
      "learning_rate": 0.00010379999999999999,
      "loss": 0.6615,
      "step": 520
    },
    {
      "epoch": 7.04,
      "learning_rate": 0.000104,
      "loss": 0.7271,
      "step": 521
    },
    {
      "epoch": 7.05,
      "learning_rate": 0.00010419999999999998,
      "loss": 0.5911,
      "step": 522
    },
    {
      "epoch": 7.07,
      "learning_rate": 0.00010439999999999999,
      "loss": 0.6788,
      "step": 523
    },
    {
      "epoch": 7.08,
      "learning_rate": 0.0001046,
      "loss": 0.7115,
      "step": 524
    },
    {
      "epoch": 7.09,
      "learning_rate": 0.00010479999999999999,
      "loss": 0.8169,
      "step": 525
    },
    {
      "epoch": 7.11,
      "learning_rate": 0.00010499999999999999,
      "loss": 0.8788,
      "step": 526
    },
    {
      "epoch": 7.12,
      "learning_rate": 0.0001052,
      "loss": 0.6788,
      "step": 527
    },
    {
      "epoch": 7.14,
      "learning_rate": 0.00010539999999999999,
      "loss": 0.9598,
      "step": 528
    },
    {
      "epoch": 7.15,
      "learning_rate": 0.00010559999999999998,
      "loss": 0.7591,
      "step": 529
    },
    {
      "epoch": 7.16,
      "learning_rate": 0.0001058,
      "loss": 0.9786,
      "step": 530
    },
    {
      "epoch": 7.18,
      "learning_rate": 0.00010599999999999999,
      "loss": 0.8107,
      "step": 531
    },
    {
      "epoch": 7.19,
      "learning_rate": 0.00010619999999999998,
      "loss": 1.1233,
      "step": 532
    },
    {
      "epoch": 7.2,
      "learning_rate": 0.0001064,
      "loss": 0.7778,
      "step": 533
    },
    {
      "epoch": 7.22,
      "learning_rate": 0.00010659999999999999,
      "loss": 0.7913,
      "step": 534
    },
    {
      "epoch": 7.23,
      "learning_rate": 0.00010679999999999998,
      "loss": 0.8252,
      "step": 535
    },
    {
      "epoch": 7.24,
      "learning_rate": 0.000107,
      "loss": 0.7701,
      "step": 536
    },
    {
      "epoch": 7.26,
      "learning_rate": 0.00010719999999999999,
      "loss": 0.9638,
      "step": 537
    },
    {
      "epoch": 7.27,
      "learning_rate": 0.00010739999999999998,
      "loss": 0.8369,
      "step": 538
    },
    {
      "epoch": 7.28,
      "learning_rate": 0.0001076,
      "loss": 0.5674,
      "step": 539
    },
    {
      "epoch": 7.3,
      "learning_rate": 0.00010779999999999999,
      "loss": 0.9723,
      "step": 540
    },
    {
      "epoch": 7.31,
      "learning_rate": 0.00010799999999999998,
      "loss": 0.8142,
      "step": 541
    },
    {
      "epoch": 7.32,
      "learning_rate": 0.0001082,
      "loss": 0.6756,
      "step": 542
    },
    {
      "epoch": 7.34,
      "learning_rate": 0.00010839999999999999,
      "loss": 0.6929,
      "step": 543
    },
    {
      "epoch": 7.35,
      "learning_rate": 0.00010859999999999998,
      "loss": 0.6101,
      "step": 544
    },
    {
      "epoch": 7.36,
      "learning_rate": 0.0001088,
      "loss": 0.8679,
      "step": 545
    },
    {
      "epoch": 7.38,
      "learning_rate": 0.00010899999999999999,
      "loss": 0.8262,
      "step": 546
    },
    {
      "epoch": 7.39,
      "learning_rate": 0.00010919999999999998,
      "loss": 0.6156,
      "step": 547
    },
    {
      "epoch": 7.41,
      "learning_rate": 0.00010939999999999998,
      "loss": 0.5569,
      "step": 548
    },
    {
      "epoch": 7.42,
      "learning_rate": 0.0001096,
      "loss": 0.8671,
      "step": 549
    },
    {
      "epoch": 7.43,
      "learning_rate": 0.00010979999999999999,
      "loss": 0.8231,
      "step": 550
    },
    {
      "epoch": 7.45,
      "learning_rate": 0.00010999999999999998,
      "loss": 0.638,
      "step": 551
    },
    {
      "epoch": 7.46,
      "learning_rate": 0.0001102,
      "loss": 0.8226,
      "step": 552
    },
    {
      "epoch": 7.47,
      "learning_rate": 0.00011039999999999999,
      "loss": 0.838,
      "step": 553
    },
    {
      "epoch": 7.49,
      "learning_rate": 0.00011059999999999998,
      "loss": 0.759,
      "step": 554
    },
    {
      "epoch": 7.5,
      "learning_rate": 0.0001108,
      "loss": 0.7866,
      "step": 555
    },
    {
      "epoch": 7.51,
      "learning_rate": 0.00011099999999999999,
      "loss": 0.6658,
      "step": 556
    },
    {
      "epoch": 7.53,
      "learning_rate": 0.00011119999999999998,
      "loss": 0.4325,
      "step": 557
    },
    {
      "epoch": 7.54,
      "learning_rate": 0.0001114,
      "loss": 0.4884,
      "step": 558
    },
    {
      "epoch": 7.55,
      "learning_rate": 0.00011159999999999999,
      "loss": 0.5291,
      "step": 559
    },
    {
      "epoch": 7.57,
      "learning_rate": 0.00011179999999999998,
      "loss": 0.6159,
      "step": 560
    },
    {
      "epoch": 7.58,
      "learning_rate": 0.000112,
      "loss": 0.9663,
      "step": 561
    },
    {
      "epoch": 7.59,
      "learning_rate": 0.00011219999999999999,
      "loss": 1.1896,
      "step": 562
    },
    {
      "epoch": 7.61,
      "learning_rate": 0.00011239999999999998,
      "loss": 0.7645,
      "step": 563
    },
    {
      "epoch": 7.62,
      "learning_rate": 0.0001126,
      "loss": 0.8655,
      "step": 564
    },
    {
      "epoch": 7.64,
      "learning_rate": 0.00011279999999999999,
      "loss": 1.0667,
      "step": 565
    },
    {
      "epoch": 7.65,
      "learning_rate": 0.00011299999999999998,
      "loss": 0.6522,
      "step": 566
    },
    {
      "epoch": 7.66,
      "learning_rate": 0.0001132,
      "loss": 0.8428,
      "step": 567
    },
    {
      "epoch": 7.68,
      "learning_rate": 0.00011339999999999999,
      "loss": 0.9477,
      "step": 568
    },
    {
      "epoch": 7.69,
      "learning_rate": 0.00011359999999999998,
      "loss": 0.8015,
      "step": 569
    },
    {
      "epoch": 7.7,
      "learning_rate": 0.0001138,
      "loss": 0.8128,
      "step": 570
    },
    {
      "epoch": 7.72,
      "learning_rate": 0.00011399999999999999,
      "loss": 0.8675,
      "step": 571
    },
    {
      "epoch": 7.73,
      "learning_rate": 0.00011419999999999998,
      "loss": 0.7635,
      "step": 572
    },
    {
      "epoch": 7.74,
      "learning_rate": 0.0001144,
      "loss": 0.8018,
      "step": 573
    },
    {
      "epoch": 7.76,
      "learning_rate": 0.0001146,
      "loss": 0.7982,
      "step": 574
    },
    {
      "epoch": 7.77,
      "learning_rate": 0.00011479999999999999,
      "loss": 0.55,
      "step": 575
    },
    {
      "epoch": 7.78,
      "learning_rate": 0.000115,
      "loss": 0.4027,
      "step": 576
    },
    {
      "epoch": 7.8,
      "learning_rate": 0.0001152,
      "loss": 0.7638,
      "step": 577
    },
    {
      "epoch": 7.81,
      "learning_rate": 0.00011539999999999999,
      "loss": 0.6963,
      "step": 578
    },
    {
      "epoch": 7.82,
      "learning_rate": 0.0001156,
      "loss": 0.7307,
      "step": 579
    },
    {
      "epoch": 7.84,
      "learning_rate": 0.0001158,
      "loss": 0.8597,
      "step": 580
    },
    {
      "epoch": 7.85,
      "learning_rate": 0.00011599999999999999,
      "loss": 0.6919,
      "step": 581
    },
    {
      "epoch": 7.86,
      "learning_rate": 0.00011619999999999998,
      "loss": 0.7231,
      "step": 582
    },
    {
      "epoch": 7.88,
      "learning_rate": 0.0001164,
      "loss": 0.6268,
      "step": 583
    },
    {
      "epoch": 7.89,
      "learning_rate": 0.00011659999999999999,
      "loss": 0.7772,
      "step": 584
    },
    {
      "epoch": 7.91,
      "learning_rate": 0.00011679999999999998,
      "loss": 0.6456,
      "step": 585
    },
    {
      "epoch": 7.92,
      "learning_rate": 0.000117,
      "loss": 0.8508,
      "step": 586
    },
    {
      "epoch": 7.93,
      "learning_rate": 0.00011719999999999999,
      "loss": 0.6442,
      "step": 587
    },
    {
      "epoch": 7.95,
      "learning_rate": 0.00011739999999999998,
      "loss": 0.7843,
      "step": 588
    },
    {
      "epoch": 7.96,
      "learning_rate": 0.0001176,
      "loss": 0.6512,
      "step": 589
    },
    {
      "epoch": 7.97,
      "learning_rate": 0.00011779999999999999,
      "loss": 0.6872,
      "step": 590
    },
    {
      "epoch": 7.99,
      "learning_rate": 0.00011799999999999998,
      "loss": 0.6607,
      "step": 591
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.0001182,
      "loss": 1.1916,
      "step": 592
    },
    {
      "epoch": 8.01,
      "learning_rate": 0.00011839999999999999,
      "loss": 0.6402,
      "step": 593
    },
    {
      "epoch": 8.03,
      "learning_rate": 0.00011859999999999998,
      "loss": 0.644,
      "step": 594
    },
    {
      "epoch": 8.04,
      "learning_rate": 0.0001188,
      "loss": 0.52,
      "step": 595
    },
    {
      "epoch": 8.05,
      "learning_rate": 0.00011899999999999999,
      "loss": 0.4968,
      "step": 596
    },
    {
      "epoch": 8.07,
      "learning_rate": 0.00011919999999999998,
      "loss": 0.5015,
      "step": 597
    },
    {
      "epoch": 8.08,
      "learning_rate": 0.0001194,
      "loss": 0.4596,
      "step": 598
    },
    {
      "epoch": 8.09,
      "learning_rate": 0.0001196,
      "loss": 0.4895,
      "step": 599
    },
    {
      "epoch": 8.11,
      "learning_rate": 0.00011979999999999998,
      "loss": 0.5914,
      "step": 600
    },
    {
      "epoch": 8.11,
      "eval_accuracy": 0.823170731707317,
      "eval_f1": 0.8269670145694816,
      "eval_loss": 0.727419912815094,
      "eval_runtime": 125.2484,
      "eval_samples_per_second": 13.094,
      "eval_steps_per_second": 0.822,
      "step": 600
    },
    {
      "epoch": 8.12,
      "learning_rate": 0.00011999999999999999,
      "loss": 0.5681,
      "step": 601
    },
    {
      "epoch": 8.14,
      "learning_rate": 0.0001202,
      "loss": 0.5451,
      "step": 602
    },
    {
      "epoch": 8.15,
      "learning_rate": 0.00012039999999999999,
      "loss": 0.7452,
      "step": 603
    },
    {
      "epoch": 8.16,
      "learning_rate": 0.00012059999999999999,
      "loss": 0.7043,
      "step": 604
    },
    {
      "epoch": 8.18,
      "learning_rate": 0.0001208,
      "loss": 0.7543,
      "step": 605
    },
    {
      "epoch": 8.19,
      "learning_rate": 0.00012099999999999999,
      "loss": 0.6062,
      "step": 606
    },
    {
      "epoch": 8.2,
      "learning_rate": 0.00012119999999999999,
      "loss": 0.6695,
      "step": 607
    },
    {
      "epoch": 8.22,
      "learning_rate": 0.0001214,
      "loss": 0.8872,
      "step": 608
    },
    {
      "epoch": 8.23,
      "learning_rate": 0.00012159999999999999,
      "loss": 0.7177,
      "step": 609
    },
    {
      "epoch": 8.24,
      "learning_rate": 0.00012179999999999999,
      "loss": 0.6349,
      "step": 610
    },
    {
      "epoch": 8.26,
      "learning_rate": 0.000122,
      "loss": 0.8606,
      "step": 611
    },
    {
      "epoch": 8.27,
      "learning_rate": 0.0001222,
      "loss": 0.6162,
      "step": 612
    },
    {
      "epoch": 8.28,
      "learning_rate": 0.0001224,
      "loss": 0.4797,
      "step": 613
    },
    {
      "epoch": 8.3,
      "learning_rate": 0.0001226,
      "loss": 0.7336,
      "step": 614
    },
    {
      "epoch": 8.31,
      "learning_rate": 0.00012279999999999998,
      "loss": 0.4168,
      "step": 615
    },
    {
      "epoch": 8.32,
      "learning_rate": 0.00012299999999999998,
      "loss": 0.3948,
      "step": 616
    },
    {
      "epoch": 8.34,
      "learning_rate": 0.00012319999999999999,
      "loss": 0.6227,
      "step": 617
    },
    {
      "epoch": 8.35,
      "learning_rate": 0.0001234,
      "loss": 0.601,
      "step": 618
    },
    {
      "epoch": 8.36,
      "learning_rate": 0.0001236,
      "loss": 0.6976,
      "step": 619
    },
    {
      "epoch": 8.38,
      "learning_rate": 0.0001238,
      "loss": 0.6604,
      "step": 620
    },
    {
      "epoch": 8.39,
      "learning_rate": 0.00012399999999999998,
      "loss": 0.5822,
      "step": 621
    },
    {
      "epoch": 8.41,
      "learning_rate": 0.00012419999999999998,
      "loss": 0.5903,
      "step": 622
    },
    {
      "epoch": 8.42,
      "learning_rate": 0.0001244,
      "loss": 0.541,
      "step": 623
    },
    {
      "epoch": 8.43,
      "learning_rate": 0.0001246,
      "loss": 0.8225,
      "step": 624
    },
    {
      "epoch": 8.45,
      "learning_rate": 0.00012479999999999997,
      "loss": 0.4016,
      "step": 625
    },
    {
      "epoch": 8.46,
      "learning_rate": 0.000125,
      "loss": 0.5248,
      "step": 626
    },
    {
      "epoch": 8.47,
      "learning_rate": 0.00012519999999999998,
      "loss": 0.5524,
      "step": 627
    },
    {
      "epoch": 8.49,
      "learning_rate": 0.00012539999999999999,
      "loss": 0.8061,
      "step": 628
    },
    {
      "epoch": 8.5,
      "learning_rate": 0.0001256,
      "loss": 0.799,
      "step": 629
    },
    {
      "epoch": 8.51,
      "learning_rate": 0.0001258,
      "loss": 0.6971,
      "step": 630
    },
    {
      "epoch": 8.53,
      "learning_rate": 0.00012599999999999997,
      "loss": 0.4261,
      "step": 631
    },
    {
      "epoch": 8.54,
      "learning_rate": 0.0001262,
      "loss": 0.4811,
      "step": 632
    },
    {
      "epoch": 8.55,
      "learning_rate": 0.00012639999999999998,
      "loss": 0.5132,
      "step": 633
    },
    {
      "epoch": 8.57,
      "learning_rate": 0.0001266,
      "loss": 0.7005,
      "step": 634
    },
    {
      "epoch": 8.58,
      "learning_rate": 0.0001268,
      "loss": 0.4786,
      "step": 635
    },
    {
      "epoch": 8.59,
      "learning_rate": 0.000127,
      "loss": 0.5653,
      "step": 636
    },
    {
      "epoch": 8.61,
      "learning_rate": 0.00012719999999999997,
      "loss": 0.5302,
      "step": 637
    },
    {
      "epoch": 8.62,
      "learning_rate": 0.0001274,
      "loss": 0.3899,
      "step": 638
    },
    {
      "epoch": 8.64,
      "learning_rate": 0.00012759999999999998,
      "loss": 0.5071,
      "step": 639
    },
    {
      "epoch": 8.65,
      "learning_rate": 0.0001278,
      "loss": 0.7406,
      "step": 640
    },
    {
      "epoch": 8.66,
      "learning_rate": 0.000128,
      "loss": 0.4734,
      "step": 641
    },
    {
      "epoch": 8.68,
      "learning_rate": 0.0001282,
      "loss": 0.5941,
      "step": 642
    },
    {
      "epoch": 8.69,
      "learning_rate": 0.00012839999999999998,
      "loss": 0.4753,
      "step": 643
    },
    {
      "epoch": 8.7,
      "learning_rate": 0.00012859999999999998,
      "loss": 0.5457,
      "step": 644
    },
    {
      "epoch": 8.72,
      "learning_rate": 0.0001288,
      "loss": 0.3933,
      "step": 645
    },
    {
      "epoch": 8.73,
      "learning_rate": 0.000129,
      "loss": 0.8251,
      "step": 646
    },
    {
      "epoch": 8.74,
      "learning_rate": 0.00012919999999999997,
      "loss": 0.3093,
      "step": 647
    },
    {
      "epoch": 8.76,
      "learning_rate": 0.0001294,
      "loss": 0.509,
      "step": 648
    },
    {
      "epoch": 8.77,
      "learning_rate": 0.00012959999999999998,
      "loss": 0.6583,
      "step": 649
    },
    {
      "epoch": 8.78,
      "learning_rate": 0.00012979999999999998,
      "loss": 0.3926,
      "step": 650
    },
    {
      "epoch": 8.8,
      "learning_rate": 0.00013,
      "loss": 0.7427,
      "step": 651
    },
    {
      "epoch": 8.81,
      "learning_rate": 0.0001302,
      "loss": 0.4233,
      "step": 652
    },
    {
      "epoch": 8.82,
      "learning_rate": 0.00013039999999999997,
      "loss": 0.578,
      "step": 653
    },
    {
      "epoch": 8.84,
      "learning_rate": 0.0001306,
      "loss": 0.7197,
      "step": 654
    },
    {
      "epoch": 8.85,
      "learning_rate": 0.00013079999999999998,
      "loss": 0.3764,
      "step": 655
    },
    {
      "epoch": 8.86,
      "learning_rate": 0.00013099999999999999,
      "loss": 0.383,
      "step": 656
    },
    {
      "epoch": 8.88,
      "learning_rate": 0.0001312,
      "loss": 0.6592,
      "step": 657
    },
    {
      "epoch": 8.89,
      "learning_rate": 0.0001314,
      "loss": 0.4666,
      "step": 658
    },
    {
      "epoch": 8.91,
      "learning_rate": 0.00013159999999999997,
      "loss": 0.4817,
      "step": 659
    },
    {
      "epoch": 8.92,
      "learning_rate": 0.0001318,
      "loss": 0.3728,
      "step": 660
    },
    {
      "epoch": 8.93,
      "learning_rate": 0.00013199999999999998,
      "loss": 0.532,
      "step": 661
    },
    {
      "epoch": 8.95,
      "learning_rate": 0.0001322,
      "loss": 0.387,
      "step": 662
    },
    {
      "epoch": 8.96,
      "learning_rate": 0.0001324,
      "loss": 0.4721,
      "step": 663
    },
    {
      "epoch": 8.97,
      "learning_rate": 0.0001326,
      "loss": 0.4894,
      "step": 664
    },
    {
      "epoch": 8.99,
      "learning_rate": 0.00013279999999999998,
      "loss": 0.4863,
      "step": 665
    },
    {
      "epoch": 9.0,
      "learning_rate": 0.000133,
      "loss": 1.126,
      "step": 666
    },
    {
      "epoch": 9.01,
      "learning_rate": 0.00013319999999999999,
      "loss": 0.7618,
      "step": 667
    },
    {
      "epoch": 9.03,
      "learning_rate": 0.0001334,
      "loss": 0.3804,
      "step": 668
    },
    {
      "epoch": 9.04,
      "learning_rate": 0.0001336,
      "loss": 0.189,
      "step": 669
    },
    {
      "epoch": 9.05,
      "learning_rate": 0.0001338,
      "loss": 0.4924,
      "step": 670
    },
    {
      "epoch": 9.07,
      "learning_rate": 0.00013399999999999998,
      "loss": 0.5211,
      "step": 671
    },
    {
      "epoch": 9.08,
      "learning_rate": 0.0001342,
      "loss": 0.6293,
      "step": 672
    },
    {
      "epoch": 9.09,
      "learning_rate": 0.0001344,
      "loss": 0.4885,
      "step": 673
    },
    {
      "epoch": 9.11,
      "learning_rate": 0.0001346,
      "loss": 0.3761,
      "step": 674
    },
    {
      "epoch": 9.12,
      "learning_rate": 0.00013479999999999997,
      "loss": 0.5912,
      "step": 675
    },
    {
      "epoch": 9.14,
      "learning_rate": 0.000135,
      "loss": 0.4683,
      "step": 676
    },
    {
      "epoch": 9.15,
      "learning_rate": 0.00013519999999999998,
      "loss": 0.2803,
      "step": 677
    },
    {
      "epoch": 9.16,
      "learning_rate": 0.00013539999999999998,
      "loss": 0.2866,
      "step": 678
    },
    {
      "epoch": 9.18,
      "learning_rate": 0.0001356,
      "loss": 0.2145,
      "step": 679
    },
    {
      "epoch": 9.19,
      "learning_rate": 0.0001358,
      "loss": 0.5625,
      "step": 680
    },
    {
      "epoch": 9.2,
      "learning_rate": 0.00013599999999999997,
      "loss": 0.2651,
      "step": 681
    },
    {
      "epoch": 9.22,
      "learning_rate": 0.0001362,
      "loss": 0.2906,
      "step": 682
    },
    {
      "epoch": 9.23,
      "learning_rate": 0.00013639999999999998,
      "loss": 0.4286,
      "step": 683
    },
    {
      "epoch": 9.24,
      "learning_rate": 0.00013659999999999999,
      "loss": 0.5018,
      "step": 684
    },
    {
      "epoch": 9.26,
      "learning_rate": 0.0001368,
      "loss": 0.6861,
      "step": 685
    },
    {
      "epoch": 9.27,
      "learning_rate": 0.000137,
      "loss": 0.3622,
      "step": 686
    },
    {
      "epoch": 9.28,
      "learning_rate": 0.00013719999999999997,
      "loss": 0.5301,
      "step": 687
    },
    {
      "epoch": 9.3,
      "learning_rate": 0.0001374,
      "loss": 0.3624,
      "step": 688
    },
    {
      "epoch": 9.31,
      "learning_rate": 0.00013759999999999998,
      "loss": 0.4899,
      "step": 689
    },
    {
      "epoch": 9.32,
      "learning_rate": 0.0001378,
      "loss": 0.4123,
      "step": 690
    },
    {
      "epoch": 9.34,
      "learning_rate": 0.000138,
      "loss": 0.8012,
      "step": 691
    },
    {
      "epoch": 9.35,
      "learning_rate": 0.0001382,
      "loss": 0.5121,
      "step": 692
    },
    {
      "epoch": 9.36,
      "learning_rate": 0.00013839999999999998,
      "loss": 0.7474,
      "step": 693
    },
    {
      "epoch": 9.38,
      "learning_rate": 0.0001386,
      "loss": 0.4101,
      "step": 694
    },
    {
      "epoch": 9.39,
      "learning_rate": 0.00013879999999999999,
      "loss": 0.7177,
      "step": 695
    },
    {
      "epoch": 9.41,
      "learning_rate": 0.000139,
      "loss": 0.5872,
      "step": 696
    },
    {
      "epoch": 9.42,
      "learning_rate": 0.0001392,
      "loss": 0.5828,
      "step": 697
    },
    {
      "epoch": 9.43,
      "learning_rate": 0.0001394,
      "loss": 0.5493,
      "step": 698
    },
    {
      "epoch": 9.45,
      "learning_rate": 0.00013959999999999998,
      "loss": 0.396,
      "step": 699
    },
    {
      "epoch": 9.46,
      "learning_rate": 0.00013979999999999998,
      "loss": 0.4257,
      "step": 700
    },
    {
      "epoch": 9.47,
      "learning_rate": 0.00014,
      "loss": 0.5564,
      "step": 701
    },
    {
      "epoch": 9.49,
      "learning_rate": 0.0001402,
      "loss": 0.5198,
      "step": 702
    },
    {
      "epoch": 9.5,
      "learning_rate": 0.0001404,
      "loss": 0.4734,
      "step": 703
    },
    {
      "epoch": 9.51,
      "learning_rate": 0.0001406,
      "loss": 0.4548,
      "step": 704
    },
    {
      "epoch": 9.53,
      "learning_rate": 0.00014079999999999998,
      "loss": 0.3187,
      "step": 705
    },
    {
      "epoch": 9.54,
      "learning_rate": 0.00014099999999999998,
      "loss": 0.6428,
      "step": 706
    },
    {
      "epoch": 9.55,
      "learning_rate": 0.0001412,
      "loss": 0.274,
      "step": 707
    },
    {
      "epoch": 9.57,
      "learning_rate": 0.0001414,
      "loss": 0.9343,
      "step": 708
    },
    {
      "epoch": 9.58,
      "learning_rate": 0.00014159999999999997,
      "loss": 0.7951,
      "step": 709
    },
    {
      "epoch": 9.59,
      "learning_rate": 0.0001418,
      "loss": 0.8463,
      "step": 710
    },
    {
      "epoch": 9.61,
      "learning_rate": 0.00014199999999999998,
      "loss": 0.4867,
      "step": 711
    },
    {
      "epoch": 9.62,
      "learning_rate": 0.0001422,
      "loss": 0.5664,
      "step": 712
    },
    {
      "epoch": 9.64,
      "learning_rate": 0.0001424,
      "loss": 0.595,
      "step": 713
    },
    {
      "epoch": 9.65,
      "learning_rate": 0.0001426,
      "loss": 0.7758,
      "step": 714
    },
    {
      "epoch": 9.66,
      "learning_rate": 0.00014279999999999997,
      "loss": 0.3557,
      "step": 715
    },
    {
      "epoch": 9.68,
      "learning_rate": 0.00014299999999999998,
      "loss": 0.561,
      "step": 716
    },
    {
      "epoch": 9.69,
      "learning_rate": 0.00014319999999999998,
      "loss": 0.4548,
      "step": 717
    },
    {
      "epoch": 9.7,
      "learning_rate": 0.0001434,
      "loss": 0.3968,
      "step": 718
    },
    {
      "epoch": 9.72,
      "learning_rate": 0.0001436,
      "loss": 0.2919,
      "step": 719
    },
    {
      "epoch": 9.73,
      "learning_rate": 0.0001438,
      "loss": 0.4018,
      "step": 720
    },
    {
      "epoch": 9.74,
      "learning_rate": 0.00014399999999999998,
      "loss": 0.4195,
      "step": 721
    },
    {
      "epoch": 9.76,
      "learning_rate": 0.00014419999999999998,
      "loss": 0.7186,
      "step": 722
    },
    {
      "epoch": 9.77,
      "learning_rate": 0.00014439999999999999,
      "loss": 0.4516,
      "step": 723
    },
    {
      "epoch": 9.78,
      "learning_rate": 0.0001446,
      "loss": 0.6922,
      "step": 724
    },
    {
      "epoch": 9.8,
      "learning_rate": 0.0001448,
      "loss": 0.4468,
      "step": 725
    },
    {
      "epoch": 9.81,
      "learning_rate": 0.000145,
      "loss": 0.341,
      "step": 726
    },
    {
      "epoch": 9.82,
      "learning_rate": 0.00014519999999999998,
      "loss": 0.3343,
      "step": 727
    },
    {
      "epoch": 9.84,
      "learning_rate": 0.00014539999999999998,
      "loss": 0.369,
      "step": 728
    },
    {
      "epoch": 9.85,
      "learning_rate": 0.0001456,
      "loss": 0.4698,
      "step": 729
    },
    {
      "epoch": 9.86,
      "learning_rate": 0.0001458,
      "loss": 0.2977,
      "step": 730
    },
    {
      "epoch": 9.88,
      "learning_rate": 0.000146,
      "loss": 0.4826,
      "step": 731
    },
    {
      "epoch": 9.89,
      "learning_rate": 0.0001462,
      "loss": 0.1673,
      "step": 732
    },
    {
      "epoch": 9.91,
      "learning_rate": 0.00014639999999999998,
      "loss": 0.4151,
      "step": 733
    },
    {
      "epoch": 9.92,
      "learning_rate": 0.00014659999999999999,
      "loss": 0.4255,
      "step": 734
    },
    {
      "epoch": 9.93,
      "learning_rate": 0.0001468,
      "loss": 0.166,
      "step": 735
    },
    {
      "epoch": 9.95,
      "learning_rate": 0.000147,
      "loss": 0.5173,
      "step": 736
    },
    {
      "epoch": 9.96,
      "learning_rate": 0.00014719999999999997,
      "loss": 0.2576,
      "step": 737
    },
    {
      "epoch": 9.97,
      "learning_rate": 0.00014739999999999998,
      "loss": 0.4011,
      "step": 738
    },
    {
      "epoch": 9.99,
      "learning_rate": 0.00014759999999999998,
      "loss": 0.5601,
      "step": 739
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.0001478,
      "loss": 1.0177,
      "step": 740
    },
    {
      "epoch": 10.01,
      "learning_rate": 0.000148,
      "loss": 0.6317,
      "step": 741
    },
    {
      "epoch": 10.03,
      "learning_rate": 0.0001482,
      "loss": 0.4215,
      "step": 742
    },
    {
      "epoch": 10.04,
      "learning_rate": 0.00014839999999999998,
      "loss": 0.5085,
      "step": 743
    },
    {
      "epoch": 10.05,
      "learning_rate": 0.00014859999999999998,
      "loss": 0.3684,
      "step": 744
    },
    {
      "epoch": 10.07,
      "learning_rate": 0.00014879999999999998,
      "loss": 0.2348,
      "step": 745
    },
    {
      "epoch": 10.08,
      "learning_rate": 0.000149,
      "loss": 0.3223,
      "step": 746
    },
    {
      "epoch": 10.09,
      "learning_rate": 0.0001492,
      "loss": 0.6774,
      "step": 747
    },
    {
      "epoch": 10.11,
      "learning_rate": 0.0001494,
      "loss": 0.2788,
      "step": 748
    },
    {
      "epoch": 10.12,
      "learning_rate": 0.00014959999999999998,
      "loss": 0.3955,
      "step": 749
    },
    {
      "epoch": 10.14,
      "learning_rate": 0.00014979999999999998,
      "loss": 0.3231,
      "step": 750
    },
    {
      "epoch": 10.15,
      "learning_rate": 0.00015,
      "loss": 0.5253,
      "step": 751
    },
    {
      "epoch": 10.16,
      "learning_rate": 0.0001502,
      "loss": 0.2835,
      "step": 752
    },
    {
      "epoch": 10.18,
      "learning_rate": 0.00015039999999999997,
      "loss": 0.4606,
      "step": 753
    },
    {
      "epoch": 10.19,
      "learning_rate": 0.00015059999999999997,
      "loss": 0.3202,
      "step": 754
    },
    {
      "epoch": 10.2,
      "learning_rate": 0.0001508,
      "loss": 0.4767,
      "step": 755
    },
    {
      "epoch": 10.22,
      "learning_rate": 0.00015099999999999998,
      "loss": 0.4594,
      "step": 756
    },
    {
      "epoch": 10.23,
      "learning_rate": 0.0001512,
      "loss": 0.3462,
      "step": 757
    },
    {
      "epoch": 10.24,
      "learning_rate": 0.0001514,
      "loss": 0.4671,
      "step": 758
    },
    {
      "epoch": 10.26,
      "learning_rate": 0.00015159999999999997,
      "loss": 0.5477,
      "step": 759
    },
    {
      "epoch": 10.27,
      "learning_rate": 0.00015179999999999998,
      "loss": 0.5803,
      "step": 760
    },
    {
      "epoch": 10.28,
      "learning_rate": 0.000152,
      "loss": 0.1185,
      "step": 761
    },
    {
      "epoch": 10.3,
      "learning_rate": 0.00015219999999999999,
      "loss": 0.1844,
      "step": 762
    },
    {
      "epoch": 10.31,
      "learning_rate": 0.0001524,
      "loss": 0.2591,
      "step": 763
    },
    {
      "epoch": 10.32,
      "learning_rate": 0.0001526,
      "loss": 0.2455,
      "step": 764
    },
    {
      "epoch": 10.34,
      "learning_rate": 0.00015279999999999997,
      "loss": 0.2625,
      "step": 765
    },
    {
      "epoch": 10.35,
      "learning_rate": 0.00015299999999999998,
      "loss": 0.5752,
      "step": 766
    },
    {
      "epoch": 10.36,
      "learning_rate": 0.0001532,
      "loss": 0.4907,
      "step": 767
    },
    {
      "epoch": 10.38,
      "learning_rate": 0.0001534,
      "loss": 0.4674,
      "step": 768
    },
    {
      "epoch": 10.39,
      "learning_rate": 0.0001536,
      "loss": 0.479,
      "step": 769
    },
    {
      "epoch": 10.41,
      "learning_rate": 0.0001538,
      "loss": 0.6095,
      "step": 770
    },
    {
      "epoch": 10.42,
      "learning_rate": 0.00015399999999999998,
      "loss": 0.4979,
      "step": 771
    },
    {
      "epoch": 10.43,
      "learning_rate": 0.00015419999999999998,
      "loss": 0.4728,
      "step": 772
    },
    {
      "epoch": 10.45,
      "learning_rate": 0.0001544,
      "loss": 0.5265,
      "step": 773
    },
    {
      "epoch": 10.46,
      "learning_rate": 0.0001546,
      "loss": 0.6304,
      "step": 774
    },
    {
      "epoch": 10.47,
      "learning_rate": 0.0001548,
      "loss": 0.5397,
      "step": 775
    },
    {
      "epoch": 10.49,
      "learning_rate": 0.000155,
      "loss": 0.5455,
      "step": 776
    },
    {
      "epoch": 10.5,
      "learning_rate": 0.00015519999999999998,
      "loss": 0.2712,
      "step": 777
    },
    {
      "epoch": 10.51,
      "learning_rate": 0.00015539999999999998,
      "loss": 0.2948,
      "step": 778
    },
    {
      "epoch": 10.53,
      "learning_rate": 0.00015560000000000001,
      "loss": 0.2207,
      "step": 779
    },
    {
      "epoch": 10.54,
      "learning_rate": 0.0001558,
      "loss": 0.3967,
      "step": 780
    },
    {
      "epoch": 10.55,
      "learning_rate": 0.000156,
      "loss": 0.298,
      "step": 781
    },
    {
      "epoch": 10.57,
      "learning_rate": 0.0001562,
      "loss": 0.4501,
      "step": 782
    },
    {
      "epoch": 10.58,
      "learning_rate": 0.00015639999999999998,
      "loss": 0.6336,
      "step": 783
    },
    {
      "epoch": 10.59,
      "learning_rate": 0.00015659999999999998,
      "loss": 0.431,
      "step": 784
    },
    {
      "epoch": 10.61,
      "learning_rate": 0.00015679999999999996,
      "loss": 0.376,
      "step": 785
    },
    {
      "epoch": 10.62,
      "learning_rate": 0.000157,
      "loss": 0.3214,
      "step": 786
    },
    {
      "epoch": 10.64,
      "learning_rate": 0.0001572,
      "loss": 0.3098,
      "step": 787
    },
    {
      "epoch": 10.65,
      "learning_rate": 0.00015739999999999998,
      "loss": 0.4673,
      "step": 788
    },
    {
      "epoch": 10.66,
      "learning_rate": 0.00015759999999999998,
      "loss": 0.4017,
      "step": 789
    },
    {
      "epoch": 10.68,
      "learning_rate": 0.0001578,
      "loss": 0.3766,
      "step": 790
    },
    {
      "epoch": 10.69,
      "learning_rate": 0.00015799999999999996,
      "loss": 0.5101,
      "step": 791
    },
    {
      "epoch": 10.7,
      "learning_rate": 0.00015819999999999997,
      "loss": 0.3636,
      "step": 792
    },
    {
      "epoch": 10.72,
      "learning_rate": 0.0001584,
      "loss": 0.2796,
      "step": 793
    },
    {
      "epoch": 10.73,
      "learning_rate": 0.00015859999999999998,
      "loss": 0.6011,
      "step": 794
    },
    {
      "epoch": 10.74,
      "learning_rate": 0.00015879999999999998,
      "loss": 0.5597,
      "step": 795
    },
    {
      "epoch": 10.76,
      "learning_rate": 0.000159,
      "loss": 0.5279,
      "step": 796
    },
    {
      "epoch": 10.77,
      "learning_rate": 0.00015919999999999997,
      "loss": 0.2779,
      "step": 797
    },
    {
      "epoch": 10.78,
      "learning_rate": 0.00015939999999999997,
      "loss": 0.5145,
      "step": 798
    },
    {
      "epoch": 10.8,
      "learning_rate": 0.0001596,
      "loss": 0.5517,
      "step": 799
    },
    {
      "epoch": 10.81,
      "learning_rate": 0.00015979999999999998,
      "loss": 0.3896,
      "step": 800
    },
    {
      "epoch": 10.81,
      "eval_accuracy": 0.8890243902439025,
      "eval_f1": 0.8904715912027061,
      "eval_loss": 0.44023460149765015,
      "eval_runtime": 125.3604,
      "eval_samples_per_second": 13.082,
      "eval_steps_per_second": 0.822,
      "step": 800
    },
    {
      "epoch": 10.82,
      "learning_rate": 0.00015999999999999999,
      "loss": 0.3689,
      "step": 801
    },
    {
      "epoch": 10.84,
      "learning_rate": 0.0001602,
      "loss": 0.3823,
      "step": 802
    },
    {
      "epoch": 10.85,
      "learning_rate": 0.00016039999999999997,
      "loss": 0.3426,
      "step": 803
    },
    {
      "epoch": 10.86,
      "learning_rate": 0.00016059999999999997,
      "loss": 0.28,
      "step": 804
    },
    {
      "epoch": 10.88,
      "learning_rate": 0.0001608,
      "loss": 0.487,
      "step": 805
    },
    {
      "epoch": 10.89,
      "learning_rate": 0.00016099999999999998,
      "loss": 0.4399,
      "step": 806
    },
    {
      "epoch": 10.91,
      "learning_rate": 0.0001612,
      "loss": 0.3,
      "step": 807
    },
    {
      "epoch": 10.92,
      "learning_rate": 0.0001614,
      "loss": 0.725,
      "step": 808
    },
    {
      "epoch": 10.93,
      "learning_rate": 0.00016159999999999997,
      "loss": 0.4034,
      "step": 809
    },
    {
      "epoch": 10.95,
      "learning_rate": 0.00016179999999999998,
      "loss": 0.4548,
      "step": 810
    },
    {
      "epoch": 10.96,
      "learning_rate": 0.000162,
      "loss": 0.2968,
      "step": 811
    },
    {
      "epoch": 10.97,
      "learning_rate": 0.00016219999999999999,
      "loss": 0.4476,
      "step": 812
    },
    {
      "epoch": 10.99,
      "learning_rate": 0.0001624,
      "loss": 0.5409,
      "step": 813
    },
    {
      "epoch": 11.0,
      "learning_rate": 0.0001626,
      "loss": 0.8629,
      "step": 814
    },
    {
      "epoch": 11.01,
      "learning_rate": 0.00016279999999999997,
      "loss": 0.3788,
      "step": 815
    },
    {
      "epoch": 11.03,
      "learning_rate": 0.00016299999999999998,
      "loss": 0.248,
      "step": 816
    },
    {
      "epoch": 11.04,
      "learning_rate": 0.0001632,
      "loss": 0.237,
      "step": 817
    },
    {
      "epoch": 11.05,
      "learning_rate": 0.0001634,
      "loss": 0.3325,
      "step": 818
    },
    {
      "epoch": 11.07,
      "learning_rate": 0.0001636,
      "loss": 0.2004,
      "step": 819
    },
    {
      "epoch": 11.08,
      "learning_rate": 0.0001638,
      "loss": 0.2927,
      "step": 820
    },
    {
      "epoch": 11.09,
      "learning_rate": 0.00016399999999999997,
      "loss": 0.3305,
      "step": 821
    },
    {
      "epoch": 11.11,
      "learning_rate": 0.00016419999999999998,
      "loss": 0.2182,
      "step": 822
    },
    {
      "epoch": 11.12,
      "learning_rate": 0.0001644,
      "loss": 0.2784,
      "step": 823
    },
    {
      "epoch": 11.14,
      "learning_rate": 0.0001646,
      "loss": 0.3769,
      "step": 824
    },
    {
      "epoch": 11.15,
      "learning_rate": 0.0001648,
      "loss": 0.2562,
      "step": 825
    },
    {
      "epoch": 11.16,
      "learning_rate": 0.000165,
      "loss": 0.3715,
      "step": 826
    },
    {
      "epoch": 11.18,
      "learning_rate": 0.00016519999999999998,
      "loss": 0.1913,
      "step": 827
    },
    {
      "epoch": 11.19,
      "learning_rate": 0.00016539999999999998,
      "loss": 0.4582,
      "step": 828
    },
    {
      "epoch": 11.2,
      "learning_rate": 0.0001656,
      "loss": 0.2838,
      "step": 829
    },
    {
      "epoch": 11.22,
      "learning_rate": 0.00016579999999999996,
      "loss": 0.3113,
      "step": 830
    },
    {
      "epoch": 11.23,
      "learning_rate": 0.000166,
      "loss": 0.2097,
      "step": 831
    },
    {
      "epoch": 11.24,
      "learning_rate": 0.0001662,
      "loss": 0.4201,
      "step": 832
    },
    {
      "epoch": 11.26,
      "learning_rate": 0.00016639999999999998,
      "loss": 0.8733,
      "step": 833
    },
    {
      "epoch": 11.27,
      "learning_rate": 0.00016659999999999998,
      "loss": 0.6876,
      "step": 834
    },
    {
      "epoch": 11.28,
      "learning_rate": 0.0001668,
      "loss": 0.2757,
      "step": 835
    },
    {
      "epoch": 11.3,
      "learning_rate": 0.00016699999999999997,
      "loss": 0.3713,
      "step": 836
    },
    {
      "epoch": 11.31,
      "learning_rate": 0.0001672,
      "loss": 0.2666,
      "step": 837
    },
    {
      "epoch": 11.32,
      "learning_rate": 0.0001674,
      "loss": 0.5656,
      "step": 838
    },
    {
      "epoch": 11.34,
      "learning_rate": 0.00016759999999999998,
      "loss": 0.3265,
      "step": 839
    },
    {
      "epoch": 11.35,
      "learning_rate": 0.00016779999999999999,
      "loss": 0.2791,
      "step": 840
    },
    {
      "epoch": 11.36,
      "learning_rate": 0.000168,
      "loss": 0.4903,
      "step": 841
    },
    {
      "epoch": 11.38,
      "learning_rate": 0.00016819999999999997,
      "loss": 0.3422,
      "step": 842
    },
    {
      "epoch": 11.39,
      "learning_rate": 0.0001684,
      "loss": 0.2877,
      "step": 843
    },
    {
      "epoch": 11.41,
      "learning_rate": 0.0001686,
      "loss": 0.3822,
      "step": 844
    },
    {
      "epoch": 11.42,
      "learning_rate": 0.00016879999999999998,
      "loss": 0.5591,
      "step": 845
    },
    {
      "epoch": 11.43,
      "learning_rate": 0.000169,
      "loss": 0.4508,
      "step": 846
    },
    {
      "epoch": 11.45,
      "learning_rate": 0.00016919999999999997,
      "loss": 0.4009,
      "step": 847
    },
    {
      "epoch": 11.46,
      "learning_rate": 0.00016939999999999997,
      "loss": 0.3507,
      "step": 848
    },
    {
      "epoch": 11.47,
      "learning_rate": 0.0001696,
      "loss": 0.4247,
      "step": 849
    },
    {
      "epoch": 11.49,
      "learning_rate": 0.00016979999999999998,
      "loss": 0.134,
      "step": 850
    },
    {
      "epoch": 11.5,
      "learning_rate": 0.00016999999999999999,
      "loss": 0.5958,
      "step": 851
    },
    {
      "epoch": 11.51,
      "learning_rate": 0.0001702,
      "loss": 0.2694,
      "step": 852
    },
    {
      "epoch": 11.53,
      "learning_rate": 0.00017039999999999997,
      "loss": 0.2421,
      "step": 853
    },
    {
      "epoch": 11.54,
      "learning_rate": 0.00017059999999999997,
      "loss": 0.3944,
      "step": 854
    },
    {
      "epoch": 11.55,
      "learning_rate": 0.0001708,
      "loss": 0.4214,
      "step": 855
    },
    {
      "epoch": 11.57,
      "learning_rate": 0.00017099999999999998,
      "loss": 0.6751,
      "step": 856
    },
    {
      "epoch": 11.58,
      "learning_rate": 0.0001712,
      "loss": 0.2831,
      "step": 857
    },
    {
      "epoch": 11.59,
      "learning_rate": 0.0001714,
      "loss": 0.3382,
      "step": 858
    },
    {
      "epoch": 11.61,
      "learning_rate": 0.00017159999999999997,
      "loss": 0.3508,
      "step": 859
    },
    {
      "epoch": 11.62,
      "learning_rate": 0.00017179999999999997,
      "loss": 0.2955,
      "step": 860
    },
    {
      "epoch": 11.64,
      "learning_rate": 0.000172,
      "loss": 0.4587,
      "step": 861
    },
    {
      "epoch": 11.65,
      "learning_rate": 0.00017219999999999998,
      "loss": 0.3305,
      "step": 862
    },
    {
      "epoch": 11.66,
      "learning_rate": 0.0001724,
      "loss": 0.2766,
      "step": 863
    },
    {
      "epoch": 11.68,
      "learning_rate": 0.0001726,
      "loss": 0.6875,
      "step": 864
    },
    {
      "epoch": 11.69,
      "learning_rate": 0.00017279999999999997,
      "loss": 0.1867,
      "step": 865
    },
    {
      "epoch": 11.7,
      "learning_rate": 0.00017299999999999998,
      "loss": 0.3185,
      "step": 866
    },
    {
      "epoch": 11.72,
      "learning_rate": 0.00017319999999999998,
      "loss": 0.1229,
      "step": 867
    },
    {
      "epoch": 11.73,
      "learning_rate": 0.00017339999999999996,
      "loss": 0.4529,
      "step": 868
    },
    {
      "epoch": 11.74,
      "learning_rate": 0.0001736,
      "loss": 0.649,
      "step": 869
    },
    {
      "epoch": 11.76,
      "learning_rate": 0.0001738,
      "loss": 0.5674,
      "step": 870
    },
    {
      "epoch": 11.77,
      "learning_rate": 0.00017399999999999997,
      "loss": 0.436,
      "step": 871
    },
    {
      "epoch": 11.78,
      "learning_rate": 0.00017419999999999998,
      "loss": 0.3787,
      "step": 872
    },
    {
      "epoch": 11.8,
      "learning_rate": 0.00017439999999999998,
      "loss": 0.4965,
      "step": 873
    },
    {
      "epoch": 11.81,
      "learning_rate": 0.00017459999999999996,
      "loss": 0.4163,
      "step": 874
    },
    {
      "epoch": 11.82,
      "learning_rate": 0.0001748,
      "loss": 0.3618,
      "step": 875
    },
    {
      "epoch": 11.84,
      "learning_rate": 0.000175,
      "loss": 0.6179,
      "step": 876
    },
    {
      "epoch": 11.85,
      "learning_rate": 0.00017519999999999998,
      "loss": 0.2493,
      "step": 877
    },
    {
      "epoch": 11.86,
      "learning_rate": 0.00017539999999999998,
      "loss": 0.2251,
      "step": 878
    },
    {
      "epoch": 11.88,
      "learning_rate": 0.00017559999999999999,
      "loss": 0.347,
      "step": 879
    },
    {
      "epoch": 11.89,
      "learning_rate": 0.00017579999999999996,
      "loss": 0.3893,
      "step": 880
    },
    {
      "epoch": 11.91,
      "learning_rate": 0.000176,
      "loss": 0.355,
      "step": 881
    },
    {
      "epoch": 11.92,
      "learning_rate": 0.0001762,
      "loss": 0.1993,
      "step": 882
    },
    {
      "epoch": 11.93,
      "learning_rate": 0.00017639999999999998,
      "loss": 0.4811,
      "step": 883
    },
    {
      "epoch": 11.95,
      "learning_rate": 0.00017659999999999998,
      "loss": 0.202,
      "step": 884
    },
    {
      "epoch": 11.96,
      "learning_rate": 0.0001768,
      "loss": 0.3182,
      "step": 885
    },
    {
      "epoch": 11.97,
      "learning_rate": 0.00017699999999999997,
      "loss": 0.1344,
      "step": 886
    },
    {
      "epoch": 11.99,
      "learning_rate": 0.0001772,
      "loss": 0.2897,
      "step": 887
    },
    {
      "epoch": 12.0,
      "learning_rate": 0.0001774,
      "loss": 0.0793,
      "step": 888
    },
    {
      "epoch": 12.01,
      "learning_rate": 0.00017759999999999998,
      "loss": 0.2755,
      "step": 889
    },
    {
      "epoch": 12.03,
      "learning_rate": 0.00017779999999999998,
      "loss": 0.3367,
      "step": 890
    },
    {
      "epoch": 12.04,
      "learning_rate": 0.000178,
      "loss": 0.2513,
      "step": 891
    },
    {
      "epoch": 12.05,
      "learning_rate": 0.00017819999999999997,
      "loss": 0.3077,
      "step": 892
    },
    {
      "epoch": 12.07,
      "learning_rate": 0.0001784,
      "loss": 0.1405,
      "step": 893
    },
    {
      "epoch": 12.08,
      "learning_rate": 0.0001786,
      "loss": 0.2833,
      "step": 894
    },
    {
      "epoch": 12.09,
      "learning_rate": 0.00017879999999999998,
      "loss": 0.3358,
      "step": 895
    },
    {
      "epoch": 12.11,
      "learning_rate": 0.000179,
      "loss": 0.3534,
      "step": 896
    },
    {
      "epoch": 12.12,
      "learning_rate": 0.0001792,
      "loss": 0.1715,
      "step": 897
    },
    {
      "epoch": 12.14,
      "learning_rate": 0.00017939999999999997,
      "loss": 0.2949,
      "step": 898
    },
    {
      "epoch": 12.15,
      "learning_rate": 0.0001796,
      "loss": 0.5721,
      "step": 899
    },
    {
      "epoch": 12.16,
      "learning_rate": 0.0001798,
      "loss": 0.3286,
      "step": 900
    },
    {
      "epoch": 12.18,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.3115,
      "step": 901
    },
    {
      "epoch": 12.19,
      "learning_rate": 0.0001802,
      "loss": 0.2943,
      "step": 902
    },
    {
      "epoch": 12.2,
      "learning_rate": 0.0001804,
      "loss": 0.3089,
      "step": 903
    },
    {
      "epoch": 12.22,
      "learning_rate": 0.00018059999999999997,
      "loss": 0.5787,
      "step": 904
    },
    {
      "epoch": 12.23,
      "learning_rate": 0.00018079999999999998,
      "loss": 0.2188,
      "step": 905
    },
    {
      "epoch": 12.24,
      "learning_rate": 0.000181,
      "loss": 0.3985,
      "step": 906
    },
    {
      "epoch": 12.26,
      "learning_rate": 0.00018119999999999999,
      "loss": 1.2399,
      "step": 907
    },
    {
      "epoch": 12.27,
      "learning_rate": 0.0001814,
      "loss": 0.3419,
      "step": 908
    },
    {
      "epoch": 12.28,
      "learning_rate": 0.00018159999999999997,
      "loss": 0.557,
      "step": 909
    },
    {
      "epoch": 12.3,
      "learning_rate": 0.00018179999999999997,
      "loss": 0.4813,
      "step": 910
    },
    {
      "epoch": 12.31,
      "learning_rate": 0.00018199999999999998,
      "loss": 0.2294,
      "step": 911
    },
    {
      "epoch": 12.32,
      "learning_rate": 0.00018219999999999996,
      "loss": 0.1271,
      "step": 912
    },
    {
      "epoch": 12.34,
      "learning_rate": 0.0001824,
      "loss": 0.482,
      "step": 913
    },
    {
      "epoch": 12.35,
      "learning_rate": 0.0001826,
      "loss": 0.3189,
      "step": 914
    },
    {
      "epoch": 12.36,
      "learning_rate": 0.00018279999999999997,
      "loss": 0.3893,
      "step": 915
    },
    {
      "epoch": 12.38,
      "learning_rate": 0.00018299999999999998,
      "loss": 0.1099,
      "step": 916
    },
    {
      "epoch": 12.39,
      "learning_rate": 0.00018319999999999998,
      "loss": 0.3328,
      "step": 917
    },
    {
      "epoch": 12.41,
      "learning_rate": 0.00018339999999999996,
      "loss": 0.413,
      "step": 918
    },
    {
      "epoch": 12.42,
      "learning_rate": 0.0001836,
      "loss": 0.4575,
      "step": 919
    },
    {
      "epoch": 12.43,
      "learning_rate": 0.0001838,
      "loss": 0.2529,
      "step": 920
    },
    {
      "epoch": 12.45,
      "learning_rate": 0.00018399999999999997,
      "loss": 0.2385,
      "step": 921
    },
    {
      "epoch": 12.46,
      "learning_rate": 0.00018419999999999998,
      "loss": 0.3634,
      "step": 922
    },
    {
      "epoch": 12.47,
      "learning_rate": 0.00018439999999999998,
      "loss": 0.3706,
      "step": 923
    },
    {
      "epoch": 12.49,
      "learning_rate": 0.00018459999999999996,
      "loss": 0.284,
      "step": 924
    },
    {
      "epoch": 12.5,
      "learning_rate": 0.0001848,
      "loss": 0.5566,
      "step": 925
    },
    {
      "epoch": 12.51,
      "learning_rate": 0.000185,
      "loss": 0.4146,
      "step": 926
    },
    {
      "epoch": 12.53,
      "learning_rate": 0.00018519999999999998,
      "loss": 0.2445,
      "step": 927
    },
    {
      "epoch": 12.54,
      "learning_rate": 0.00018539999999999998,
      "loss": 0.241,
      "step": 928
    },
    {
      "epoch": 12.55,
      "learning_rate": 0.00018559999999999998,
      "loss": 0.2227,
      "step": 929
    },
    {
      "epoch": 12.57,
      "learning_rate": 0.00018579999999999996,
      "loss": 0.2727,
      "step": 930
    },
    {
      "epoch": 12.58,
      "learning_rate": 0.000186,
      "loss": 0.295,
      "step": 931
    },
    {
      "epoch": 12.59,
      "learning_rate": 0.0001862,
      "loss": 0.2323,
      "step": 932
    },
    {
      "epoch": 12.61,
      "learning_rate": 0.00018639999999999998,
      "loss": 0.2695,
      "step": 933
    },
    {
      "epoch": 12.62,
      "learning_rate": 0.00018659999999999998,
      "loss": 0.3341,
      "step": 934
    },
    {
      "epoch": 12.64,
      "learning_rate": 0.0001868,
      "loss": 0.4686,
      "step": 935
    },
    {
      "epoch": 12.65,
      "learning_rate": 0.00018699999999999996,
      "loss": 0.4248,
      "step": 936
    },
    {
      "epoch": 12.66,
      "learning_rate": 0.0001872,
      "loss": 0.3269,
      "step": 937
    },
    {
      "epoch": 12.68,
      "learning_rate": 0.0001874,
      "loss": 0.3956,
      "step": 938
    },
    {
      "epoch": 12.69,
      "learning_rate": 0.00018759999999999998,
      "loss": 0.3659,
      "step": 939
    },
    {
      "epoch": 12.7,
      "learning_rate": 0.00018779999999999998,
      "loss": 0.3046,
      "step": 940
    },
    {
      "epoch": 12.72,
      "learning_rate": 0.000188,
      "loss": 0.2415,
      "step": 941
    },
    {
      "epoch": 12.73,
      "learning_rate": 0.00018819999999999997,
      "loss": 0.3719,
      "step": 942
    },
    {
      "epoch": 12.74,
      "learning_rate": 0.00018839999999999997,
      "loss": 0.3319,
      "step": 943
    },
    {
      "epoch": 12.76,
      "learning_rate": 0.0001886,
      "loss": 0.2757,
      "step": 944
    },
    {
      "epoch": 12.77,
      "learning_rate": 0.00018879999999999998,
      "loss": 0.276,
      "step": 945
    },
    {
      "epoch": 12.78,
      "learning_rate": 0.00018899999999999999,
      "loss": 0.3849,
      "step": 946
    },
    {
      "epoch": 12.8,
      "learning_rate": 0.0001892,
      "loss": 0.283,
      "step": 947
    },
    {
      "epoch": 12.81,
      "learning_rate": 0.00018939999999999997,
      "loss": 0.3924,
      "step": 948
    },
    {
      "epoch": 12.82,
      "learning_rate": 0.00018959999999999997,
      "loss": 0.2229,
      "step": 949
    },
    {
      "epoch": 12.84,
      "learning_rate": 0.0001898,
      "loss": 0.1429,
      "step": 950
    },
    {
      "epoch": 12.85,
      "learning_rate": 0.00018999999999999998,
      "loss": 0.1475,
      "step": 951
    },
    {
      "epoch": 12.86,
      "learning_rate": 0.0001902,
      "loss": 0.2689,
      "step": 952
    },
    {
      "epoch": 12.88,
      "learning_rate": 0.0001904,
      "loss": 0.3573,
      "step": 953
    },
    {
      "epoch": 12.89,
      "learning_rate": 0.00019059999999999997,
      "loss": 0.1952,
      "step": 954
    },
    {
      "epoch": 12.91,
      "learning_rate": 0.00019079999999999998,
      "loss": 0.5259,
      "step": 955
    },
    {
      "epoch": 12.92,
      "learning_rate": 0.000191,
      "loss": 0.7877,
      "step": 956
    },
    {
      "epoch": 12.93,
      "learning_rate": 0.00019119999999999999,
      "loss": 0.3804,
      "step": 957
    },
    {
      "epoch": 12.95,
      "learning_rate": 0.0001914,
      "loss": 0.4601,
      "step": 958
    },
    {
      "epoch": 12.96,
      "learning_rate": 0.0001916,
      "loss": 0.2674,
      "step": 959
    },
    {
      "epoch": 12.97,
      "learning_rate": 0.00019179999999999997,
      "loss": 0.359,
      "step": 960
    },
    {
      "epoch": 12.99,
      "learning_rate": 0.00019199999999999998,
      "loss": 0.4247,
      "step": 961
    },
    {
      "epoch": 13.0,
      "learning_rate": 0.0001922,
      "loss": 0.0471,
      "step": 962
    },
    {
      "epoch": 13.01,
      "learning_rate": 0.0001924,
      "loss": 0.4423,
      "step": 963
    },
    {
      "epoch": 13.03,
      "learning_rate": 0.0001926,
      "loss": 0.2549,
      "step": 964
    },
    {
      "epoch": 13.04,
      "learning_rate": 0.0001928,
      "loss": 0.4503,
      "step": 965
    },
    {
      "epoch": 13.05,
      "learning_rate": 0.00019299999999999997,
      "loss": 0.2096,
      "step": 966
    },
    {
      "epoch": 13.07,
      "learning_rate": 0.00019319999999999998,
      "loss": 0.2175,
      "step": 967
    },
    {
      "epoch": 13.08,
      "learning_rate": 0.0001934,
      "loss": 0.2819,
      "step": 968
    },
    {
      "epoch": 13.09,
      "learning_rate": 0.0001936,
      "loss": 0.2348,
      "step": 969
    },
    {
      "epoch": 13.11,
      "learning_rate": 0.0001938,
      "loss": 0.0683,
      "step": 970
    },
    {
      "epoch": 13.12,
      "learning_rate": 0.00019399999999999997,
      "loss": 0.1704,
      "step": 971
    },
    {
      "epoch": 13.14,
      "learning_rate": 0.00019419999999999998,
      "loss": 0.1762,
      "step": 972
    },
    {
      "epoch": 13.15,
      "learning_rate": 0.00019439999999999998,
      "loss": 0.3678,
      "step": 973
    },
    {
      "epoch": 13.16,
      "learning_rate": 0.00019459999999999996,
      "loss": 0.6126,
      "step": 974
    },
    {
      "epoch": 13.18,
      "learning_rate": 0.0001948,
      "loss": 0.3294,
      "step": 975
    },
    {
      "epoch": 13.19,
      "learning_rate": 0.000195,
      "loss": 0.2787,
      "step": 976
    },
    {
      "epoch": 13.2,
      "learning_rate": 0.00019519999999999997,
      "loss": 0.2343,
      "step": 977
    },
    {
      "epoch": 13.22,
      "learning_rate": 0.00019539999999999998,
      "loss": 0.3347,
      "step": 978
    },
    {
      "epoch": 13.23,
      "learning_rate": 0.00019559999999999998,
      "loss": 0.166,
      "step": 979
    },
    {
      "epoch": 13.24,
      "learning_rate": 0.00019579999999999996,
      "loss": 0.2731,
      "step": 980
    },
    {
      "epoch": 13.26,
      "learning_rate": 0.00019599999999999997,
      "loss": 0.3828,
      "step": 981
    },
    {
      "epoch": 13.27,
      "learning_rate": 0.0001962,
      "loss": 0.3111,
      "step": 982
    },
    {
      "epoch": 13.28,
      "learning_rate": 0.00019639999999999998,
      "loss": 0.1187,
      "step": 983
    },
    {
      "epoch": 13.3,
      "learning_rate": 0.00019659999999999998,
      "loss": 0.1651,
      "step": 984
    },
    {
      "epoch": 13.31,
      "learning_rate": 0.00019679999999999999,
      "loss": 0.5256,
      "step": 985
    },
    {
      "epoch": 13.32,
      "learning_rate": 0.00019699999999999996,
      "loss": 0.3372,
      "step": 986
    },
    {
      "epoch": 13.34,
      "learning_rate": 0.00019719999999999997,
      "loss": 0.5123,
      "step": 987
    },
    {
      "epoch": 13.35,
      "learning_rate": 0.0001974,
      "loss": 0.3644,
      "step": 988
    },
    {
      "epoch": 13.36,
      "learning_rate": 0.00019759999999999998,
      "loss": 0.3128,
      "step": 989
    },
    {
      "epoch": 13.38,
      "learning_rate": 0.00019779999999999998,
      "loss": 0.229,
      "step": 990
    },
    {
      "epoch": 13.39,
      "learning_rate": 0.000198,
      "loss": 0.2792,
      "step": 991
    },
    {
      "epoch": 13.41,
      "learning_rate": 0.00019819999999999997,
      "loss": 0.4767,
      "step": 992
    },
    {
      "epoch": 13.42,
      "learning_rate": 0.00019839999999999997,
      "loss": 0.281,
      "step": 993
    },
    {
      "epoch": 13.43,
      "learning_rate": 0.0001986,
      "loss": 0.3698,
      "step": 994
    },
    {
      "epoch": 13.45,
      "learning_rate": 0.00019879999999999998,
      "loss": 0.2052,
      "step": 995
    },
    {
      "epoch": 13.46,
      "learning_rate": 0.00019899999999999999,
      "loss": 0.5402,
      "step": 996
    },
    {
      "epoch": 13.47,
      "learning_rate": 0.0001992,
      "loss": 0.4236,
      "step": 997
    },
    {
      "epoch": 13.49,
      "learning_rate": 0.00019939999999999997,
      "loss": 0.5331,
      "step": 998
    },
    {
      "epoch": 13.5,
      "learning_rate": 0.00019959999999999997,
      "loss": 0.2445,
      "step": 999
    },
    {
      "epoch": 13.51,
      "learning_rate": 0.0001998,
      "loss": 0.5052,
      "step": 1000
    },
    {
      "epoch": 13.51,
      "eval_accuracy": 0.8829268292682927,
      "eval_f1": 0.8836828628013275,
      "eval_loss": 0.44825276732444763,
      "eval_runtime": 125.8333,
      "eval_samples_per_second": 13.033,
      "eval_steps_per_second": 0.819,
      "step": 1000
    },
    {
      "epoch": 13.53,
      "learning_rate": 0.00019999999999999998,
      "loss": 0.2012,
      "step": 1001
    },
    {
      "epoch": 13.54,
      "learning_rate": 0.0002002,
      "loss": 0.3914,
      "step": 1002
    },
    {
      "epoch": 13.55,
      "learning_rate": 0.0002004,
      "loss": 0.2981,
      "step": 1003
    },
    {
      "epoch": 13.57,
      "learning_rate": 0.00020059999999999997,
      "loss": 0.4039,
      "step": 1004
    },
    {
      "epoch": 13.58,
      "learning_rate": 0.00020079999999999997,
      "loss": 0.752,
      "step": 1005
    },
    {
      "epoch": 13.59,
      "learning_rate": 0.000201,
      "loss": 0.2834,
      "step": 1006
    },
    {
      "epoch": 13.61,
      "learning_rate": 0.00020119999999999998,
      "loss": 0.3535,
      "step": 1007
    },
    {
      "epoch": 13.62,
      "learning_rate": 0.0002014,
      "loss": 0.3489,
      "step": 1008
    },
    {
      "epoch": 13.64,
      "learning_rate": 0.0002016,
      "loss": 0.4348,
      "step": 1009
    },
    {
      "epoch": 13.65,
      "learning_rate": 0.00020179999999999997,
      "loss": 0.2463,
      "step": 1010
    },
    {
      "epoch": 13.66,
      "learning_rate": 0.00020199999999999998,
      "loss": 0.3341,
      "step": 1011
    },
    {
      "epoch": 13.68,
      "learning_rate": 0.0002022,
      "loss": 0.2501,
      "step": 1012
    },
    {
      "epoch": 13.69,
      "learning_rate": 0.0002024,
      "loss": 0.4123,
      "step": 1013
    },
    {
      "epoch": 13.7,
      "learning_rate": 0.0002026,
      "loss": 0.1957,
      "step": 1014
    },
    {
      "epoch": 13.72,
      "learning_rate": 0.0002028,
      "loss": 0.3476,
      "step": 1015
    },
    {
      "epoch": 13.73,
      "learning_rate": 0.00020299999999999997,
      "loss": 0.1455,
      "step": 1016
    },
    {
      "epoch": 13.74,
      "learning_rate": 0.00020319999999999998,
      "loss": 0.3194,
      "step": 1017
    },
    {
      "epoch": 13.76,
      "learning_rate": 0.00020339999999999998,
      "loss": 0.2485,
      "step": 1018
    },
    {
      "epoch": 13.77,
      "learning_rate": 0.00020359999999999996,
      "loss": 0.1819,
      "step": 1019
    },
    {
      "epoch": 13.78,
      "learning_rate": 0.0002038,
      "loss": 0.35,
      "step": 1020
    },
    {
      "epoch": 13.8,
      "learning_rate": 0.000204,
      "loss": 0.3816,
      "step": 1021
    },
    {
      "epoch": 13.81,
      "learning_rate": 0.00020419999999999998,
      "loss": 0.4049,
      "step": 1022
    },
    {
      "epoch": 13.82,
      "learning_rate": 0.00020439999999999998,
      "loss": 0.2847,
      "step": 1023
    },
    {
      "epoch": 13.84,
      "learning_rate": 0.00020459999999999999,
      "loss": 0.4564,
      "step": 1024
    },
    {
      "epoch": 13.85,
      "learning_rate": 0.00020479999999999996,
      "loss": 0.2521,
      "step": 1025
    },
    {
      "epoch": 13.86,
      "learning_rate": 0.000205,
      "loss": 0.165,
      "step": 1026
    },
    {
      "epoch": 13.88,
      "learning_rate": 0.0002052,
      "loss": 0.0882,
      "step": 1027
    },
    {
      "epoch": 13.89,
      "learning_rate": 0.00020539999999999998,
      "loss": 0.3402,
      "step": 1028
    },
    {
      "epoch": 13.91,
      "learning_rate": 0.00020559999999999998,
      "loss": 0.4013,
      "step": 1029
    },
    {
      "epoch": 13.92,
      "learning_rate": 0.0002058,
      "loss": 0.1632,
      "step": 1030
    },
    {
      "epoch": 13.93,
      "learning_rate": 0.00020599999999999997,
      "loss": 0.4254,
      "step": 1031
    },
    {
      "epoch": 13.95,
      "learning_rate": 0.0002062,
      "loss": 0.4055,
      "step": 1032
    },
    {
      "epoch": 13.96,
      "learning_rate": 0.00020639999999999998,
      "loss": 0.4918,
      "step": 1033
    },
    {
      "epoch": 13.97,
      "learning_rate": 0.00020659999999999998,
      "loss": 0.2221,
      "step": 1034
    },
    {
      "epoch": 13.99,
      "learning_rate": 0.00020679999999999999,
      "loss": 0.4653,
      "step": 1035
    },
    {
      "epoch": 14.0,
      "learning_rate": 0.00020699999999999996,
      "loss": 0.0819,
      "step": 1036
    },
    {
      "epoch": 14.01,
      "learning_rate": 0.00020719999999999997,
      "loss": 0.2083,
      "step": 1037
    },
    {
      "epoch": 14.03,
      "learning_rate": 0.0002074,
      "loss": 0.2339,
      "step": 1038
    },
    {
      "epoch": 14.04,
      "learning_rate": 0.00020759999999999998,
      "loss": 0.2366,
      "step": 1039
    },
    {
      "epoch": 14.05,
      "learning_rate": 0.00020779999999999998,
      "loss": 0.3703,
      "step": 1040
    },
    {
      "epoch": 14.07,
      "learning_rate": 0.000208,
      "loss": 0.3229,
      "step": 1041
    },
    {
      "epoch": 14.08,
      "learning_rate": 0.00020819999999999996,
      "loss": 0.6243,
      "step": 1042
    },
    {
      "epoch": 14.09,
      "learning_rate": 0.00020839999999999997,
      "loss": 0.3471,
      "step": 1043
    },
    {
      "epoch": 14.11,
      "learning_rate": 0.0002086,
      "loss": 0.2212,
      "step": 1044
    },
    {
      "epoch": 14.12,
      "learning_rate": 0.00020879999999999998,
      "loss": 0.3031,
      "step": 1045
    },
    {
      "epoch": 14.14,
      "learning_rate": 0.00020899999999999998,
      "loss": 0.3312,
      "step": 1046
    },
    {
      "epoch": 14.15,
      "learning_rate": 0.0002092,
      "loss": 0.222,
      "step": 1047
    },
    {
      "epoch": 14.16,
      "learning_rate": 0.00020939999999999997,
      "loss": 0.3275,
      "step": 1048
    },
    {
      "epoch": 14.18,
      "learning_rate": 0.00020959999999999997,
      "loss": 0.1459,
      "step": 1049
    },
    {
      "epoch": 14.19,
      "learning_rate": 0.0002098,
      "loss": 0.3815,
      "step": 1050
    },
    {
      "epoch": 14.2,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.1899,
      "step": 1051
    },
    {
      "epoch": 14.22,
      "learning_rate": 0.0002102,
      "loss": 0.2299,
      "step": 1052
    },
    {
      "epoch": 14.23,
      "learning_rate": 0.0002104,
      "loss": 0.3879,
      "step": 1053
    },
    {
      "epoch": 14.24,
      "learning_rate": 0.00021059999999999997,
      "loss": 0.0571,
      "step": 1054
    },
    {
      "epoch": 14.26,
      "learning_rate": 0.00021079999999999997,
      "loss": 0.5104,
      "step": 1055
    },
    {
      "epoch": 14.27,
      "learning_rate": 0.00021099999999999998,
      "loss": 0.1428,
      "step": 1056
    },
    {
      "epoch": 14.28,
      "learning_rate": 0.00021119999999999996,
      "loss": 0.257,
      "step": 1057
    },
    {
      "epoch": 14.3,
      "learning_rate": 0.0002114,
      "loss": 0.2537,
      "step": 1058
    },
    {
      "epoch": 14.31,
      "learning_rate": 0.0002116,
      "loss": 0.4106,
      "step": 1059
    },
    {
      "epoch": 14.32,
      "learning_rate": 0.00021179999999999997,
      "loss": 0.2948,
      "step": 1060
    },
    {
      "epoch": 14.34,
      "learning_rate": 0.00021199999999999998,
      "loss": 0.0874,
      "step": 1061
    },
    {
      "epoch": 14.35,
      "learning_rate": 0.00021219999999999998,
      "loss": 0.0932,
      "step": 1062
    },
    {
      "epoch": 14.36,
      "learning_rate": 0.00021239999999999996,
      "loss": 0.1386,
      "step": 1063
    },
    {
      "epoch": 14.38,
      "learning_rate": 0.0002126,
      "loss": 0.3554,
      "step": 1064
    },
    {
      "epoch": 14.39,
      "learning_rate": 0.0002128,
      "loss": 0.335,
      "step": 1065
    },
    {
      "epoch": 14.41,
      "learning_rate": 0.00021299999999999997,
      "loss": 0.7293,
      "step": 1066
    },
    {
      "epoch": 14.42,
      "learning_rate": 0.00021319999999999998,
      "loss": 0.566,
      "step": 1067
    },
    {
      "epoch": 14.43,
      "learning_rate": 0.00021339999999999998,
      "loss": 0.3303,
      "step": 1068
    },
    {
      "epoch": 14.45,
      "learning_rate": 0.00021359999999999996,
      "loss": 0.3264,
      "step": 1069
    },
    {
      "epoch": 14.46,
      "learning_rate": 0.0002138,
      "loss": 0.3679,
      "step": 1070
    },
    {
      "epoch": 14.47,
      "learning_rate": 0.000214,
      "loss": 0.2717,
      "step": 1071
    },
    {
      "epoch": 14.49,
      "learning_rate": 0.00021419999999999998,
      "loss": 0.3882,
      "step": 1072
    },
    {
      "epoch": 14.5,
      "learning_rate": 0.00021439999999999998,
      "loss": 0.3344,
      "step": 1073
    },
    {
      "epoch": 14.51,
      "learning_rate": 0.00021459999999999998,
      "loss": 0.6592,
      "step": 1074
    },
    {
      "epoch": 14.53,
      "learning_rate": 0.00021479999999999996,
      "loss": 0.199,
      "step": 1075
    },
    {
      "epoch": 14.54,
      "learning_rate": 0.000215,
      "loss": 0.2164,
      "step": 1076
    },
    {
      "epoch": 14.55,
      "learning_rate": 0.0002152,
      "loss": 0.3898,
      "step": 1077
    },
    {
      "epoch": 14.57,
      "learning_rate": 0.00021539999999999998,
      "loss": 0.143,
      "step": 1078
    },
    {
      "epoch": 14.58,
      "learning_rate": 0.00021559999999999998,
      "loss": 0.2811,
      "step": 1079
    },
    {
      "epoch": 14.59,
      "learning_rate": 0.0002158,
      "loss": 0.4553,
      "step": 1080
    },
    {
      "epoch": 14.61,
      "learning_rate": 0.00021599999999999996,
      "loss": 0.2813,
      "step": 1081
    },
    {
      "epoch": 14.62,
      "learning_rate": 0.0002162,
      "loss": 0.2797,
      "step": 1082
    },
    {
      "epoch": 14.64,
      "learning_rate": 0.0002164,
      "loss": 0.3564,
      "step": 1083
    },
    {
      "epoch": 14.65,
      "learning_rate": 0.00021659999999999998,
      "loss": 0.3781,
      "step": 1084
    },
    {
      "epoch": 14.66,
      "learning_rate": 0.00021679999999999998,
      "loss": 0.4588,
      "step": 1085
    },
    {
      "epoch": 14.68,
      "learning_rate": 0.000217,
      "loss": 0.4004,
      "step": 1086
    },
    {
      "epoch": 14.69,
      "learning_rate": 0.00021719999999999997,
      "loss": 0.1897,
      "step": 1087
    },
    {
      "epoch": 14.7,
      "learning_rate": 0.0002174,
      "loss": 0.459,
      "step": 1088
    },
    {
      "epoch": 14.72,
      "learning_rate": 0.0002176,
      "loss": 0.188,
      "step": 1089
    },
    {
      "epoch": 14.73,
      "learning_rate": 0.00021779999999999998,
      "loss": 0.3457,
      "step": 1090
    },
    {
      "epoch": 14.74,
      "learning_rate": 0.00021799999999999999,
      "loss": 0.2979,
      "step": 1091
    },
    {
      "epoch": 14.76,
      "learning_rate": 0.0002182,
      "loss": 0.1578,
      "step": 1092
    },
    {
      "epoch": 14.77,
      "learning_rate": 0.00021839999999999997,
      "loss": 0.1184,
      "step": 1093
    },
    {
      "epoch": 14.78,
      "learning_rate": 0.00021859999999999997,
      "loss": 0.3905,
      "step": 1094
    },
    {
      "epoch": 14.8,
      "learning_rate": 0.00021879999999999995,
      "loss": 0.0972,
      "step": 1095
    },
    {
      "epoch": 14.81,
      "learning_rate": 0.00021899999999999998,
      "loss": 0.2593,
      "step": 1096
    },
    {
      "epoch": 14.82,
      "learning_rate": 0.0002192,
      "loss": 0.3834,
      "step": 1097
    },
    {
      "epoch": 14.84,
      "learning_rate": 0.00021939999999999997,
      "loss": 0.2939,
      "step": 1098
    },
    {
      "epoch": 14.85,
      "learning_rate": 0.00021959999999999997,
      "loss": 0.2584,
      "step": 1099
    },
    {
      "epoch": 14.86,
      "learning_rate": 0.00021979999999999998,
      "loss": 0.0747,
      "step": 1100
    },
    {
      "epoch": 14.88,
      "learning_rate": 0.00021999999999999995,
      "loss": 0.3543,
      "step": 1101
    },
    {
      "epoch": 14.89,
      "learning_rate": 0.00022019999999999999,
      "loss": 0.2122,
      "step": 1102
    },
    {
      "epoch": 14.91,
      "learning_rate": 0.0002204,
      "loss": 0.3469,
      "step": 1103
    },
    {
      "epoch": 14.92,
      "learning_rate": 0.00022059999999999997,
      "loss": 0.2568,
      "step": 1104
    },
    {
      "epoch": 14.93,
      "learning_rate": 0.00022079999999999997,
      "loss": 0.2669,
      "step": 1105
    },
    {
      "epoch": 14.95,
      "learning_rate": 0.00022099999999999998,
      "loss": 0.2511,
      "step": 1106
    },
    {
      "epoch": 14.96,
      "learning_rate": 0.00022119999999999996,
      "loss": 0.2956,
      "step": 1107
    },
    {
      "epoch": 14.97,
      "learning_rate": 0.0002214,
      "loss": 0.3712,
      "step": 1108
    },
    {
      "epoch": 14.99,
      "learning_rate": 0.0002216,
      "loss": 0.3847,
      "step": 1109
    },
    {
      "epoch": 15.0,
      "learning_rate": 0.00022179999999999997,
      "loss": 0.0722,
      "step": 1110
    },
    {
      "epoch": 15.01,
      "learning_rate": 0.00022199999999999998,
      "loss": 0.5841,
      "step": 1111
    },
    {
      "epoch": 15.03,
      "learning_rate": 0.00022219999999999998,
      "loss": 0.1387,
      "step": 1112
    },
    {
      "epoch": 15.04,
      "learning_rate": 0.00022239999999999996,
      "loss": 0.378,
      "step": 1113
    },
    {
      "epoch": 15.05,
      "learning_rate": 0.0002226,
      "loss": 0.6368,
      "step": 1114
    },
    {
      "epoch": 15.07,
      "learning_rate": 0.0002228,
      "loss": 0.1306,
      "step": 1115
    },
    {
      "epoch": 15.08,
      "learning_rate": 0.00022299999999999997,
      "loss": 0.1314,
      "step": 1116
    },
    {
      "epoch": 15.09,
      "learning_rate": 0.00022319999999999998,
      "loss": 0.2115,
      "step": 1117
    },
    {
      "epoch": 15.11,
      "learning_rate": 0.00022339999999999998,
      "loss": 0.283,
      "step": 1118
    },
    {
      "epoch": 15.12,
      "learning_rate": 0.00022359999999999996,
      "loss": 0.2049,
      "step": 1119
    },
    {
      "epoch": 15.14,
      "learning_rate": 0.0002238,
      "loss": 0.3485,
      "step": 1120
    },
    {
      "epoch": 15.15,
      "learning_rate": 0.000224,
      "loss": 0.2479,
      "step": 1121
    },
    {
      "epoch": 15.16,
      "learning_rate": 0.00022419999999999997,
      "loss": 0.2157,
      "step": 1122
    },
    {
      "epoch": 15.18,
      "learning_rate": 0.00022439999999999998,
      "loss": 0.3228,
      "step": 1123
    },
    {
      "epoch": 15.19,
      "learning_rate": 0.00022459999999999998,
      "loss": 0.4792,
      "step": 1124
    },
    {
      "epoch": 15.2,
      "learning_rate": 0.00022479999999999996,
      "loss": 0.4126,
      "step": 1125
    },
    {
      "epoch": 15.22,
      "learning_rate": 0.000225,
      "loss": 0.5049,
      "step": 1126
    },
    {
      "epoch": 15.23,
      "learning_rate": 0.0002252,
      "loss": 0.2743,
      "step": 1127
    },
    {
      "epoch": 15.24,
      "learning_rate": 0.00022539999999999998,
      "loss": 0.2337,
      "step": 1128
    },
    {
      "epoch": 15.26,
      "learning_rate": 0.00022559999999999998,
      "loss": 0.37,
      "step": 1129
    },
    {
      "epoch": 15.27,
      "learning_rate": 0.00022579999999999999,
      "loss": 0.6755,
      "step": 1130
    },
    {
      "epoch": 15.28,
      "learning_rate": 0.00022599999999999996,
      "loss": 0.2819,
      "step": 1131
    },
    {
      "epoch": 15.3,
      "learning_rate": 0.00022619999999999997,
      "loss": 0.2521,
      "step": 1132
    },
    {
      "epoch": 15.31,
      "learning_rate": 0.0002264,
      "loss": 0.5303,
      "step": 1133
    },
    {
      "epoch": 15.32,
      "learning_rate": 0.00022659999999999998,
      "loss": 0.3225,
      "step": 1134
    },
    {
      "epoch": 15.34,
      "learning_rate": 0.00022679999999999998,
      "loss": 0.3284,
      "step": 1135
    },
    {
      "epoch": 15.35,
      "learning_rate": 0.000227,
      "loss": 0.3309,
      "step": 1136
    },
    {
      "epoch": 15.36,
      "learning_rate": 0.00022719999999999997,
      "loss": 0.4213,
      "step": 1137
    },
    {
      "epoch": 15.38,
      "learning_rate": 0.00022739999999999997,
      "loss": 0.2267,
      "step": 1138
    },
    {
      "epoch": 15.39,
      "learning_rate": 0.0002276,
      "loss": 0.3373,
      "step": 1139
    },
    {
      "epoch": 15.41,
      "learning_rate": 0.00022779999999999998,
      "loss": 0.5039,
      "step": 1140
    },
    {
      "epoch": 15.42,
      "learning_rate": 0.00022799999999999999,
      "loss": 0.2451,
      "step": 1141
    },
    {
      "epoch": 15.43,
      "learning_rate": 0.0002282,
      "loss": 0.1464,
      "step": 1142
    },
    {
      "epoch": 15.45,
      "learning_rate": 0.00022839999999999997,
      "loss": 0.4184,
      "step": 1143
    },
    {
      "epoch": 15.46,
      "learning_rate": 0.00022859999999999997,
      "loss": 0.3909,
      "step": 1144
    },
    {
      "epoch": 15.47,
      "learning_rate": 0.0002288,
      "loss": 0.5066,
      "step": 1145
    },
    {
      "epoch": 15.49,
      "learning_rate": 0.00022899999999999998,
      "loss": 0.2531,
      "step": 1146
    },
    {
      "epoch": 15.5,
      "learning_rate": 0.0002292,
      "loss": 0.3147,
      "step": 1147
    },
    {
      "epoch": 15.51,
      "learning_rate": 0.0002294,
      "loss": 0.2858,
      "step": 1148
    },
    {
      "epoch": 15.53,
      "learning_rate": 0.00022959999999999997,
      "loss": 0.0563,
      "step": 1149
    },
    {
      "epoch": 15.54,
      "learning_rate": 0.00022979999999999997,
      "loss": 0.2699,
      "step": 1150
    },
    {
      "epoch": 15.55,
      "learning_rate": 0.00023,
      "loss": 0.3324,
      "step": 1151
    },
    {
      "epoch": 15.57,
      "learning_rate": 0.00023019999999999998,
      "loss": 0.2507,
      "step": 1152
    },
    {
      "epoch": 15.58,
      "learning_rate": 0.0002304,
      "loss": 0.1709,
      "step": 1153
    },
    {
      "epoch": 15.59,
      "learning_rate": 0.0002306,
      "loss": 0.1634,
      "step": 1154
    },
    {
      "epoch": 15.61,
      "learning_rate": 0.00023079999999999997,
      "loss": 0.1479,
      "step": 1155
    },
    {
      "epoch": 15.62,
      "learning_rate": 0.00023099999999999998,
      "loss": 0.2849,
      "step": 1156
    },
    {
      "epoch": 15.64,
      "learning_rate": 0.0002312,
      "loss": 0.1701,
      "step": 1157
    },
    {
      "epoch": 15.65,
      "learning_rate": 0.0002314,
      "loss": 0.1956,
      "step": 1158
    },
    {
      "epoch": 15.66,
      "learning_rate": 0.0002316,
      "loss": 0.2546,
      "step": 1159
    },
    {
      "epoch": 15.68,
      "learning_rate": 0.00023179999999999997,
      "loss": 0.5456,
      "step": 1160
    },
    {
      "epoch": 15.69,
      "learning_rate": 0.00023199999999999997,
      "loss": 0.1569,
      "step": 1161
    },
    {
      "epoch": 15.7,
      "learning_rate": 0.00023219999999999998,
      "loss": 0.1738,
      "step": 1162
    },
    {
      "epoch": 15.72,
      "learning_rate": 0.00023239999999999996,
      "loss": 0.3728,
      "step": 1163
    },
    {
      "epoch": 15.73,
      "learning_rate": 0.00023259999999999996,
      "loss": 0.3801,
      "step": 1164
    },
    {
      "epoch": 15.74,
      "learning_rate": 0.0002328,
      "loss": 0.6128,
      "step": 1165
    },
    {
      "epoch": 15.76,
      "learning_rate": 0.00023299999999999997,
      "loss": 0.1603,
      "step": 1166
    },
    {
      "epoch": 15.77,
      "learning_rate": 0.00023319999999999998,
      "loss": 0.2088,
      "step": 1167
    },
    {
      "epoch": 15.78,
      "learning_rate": 0.00023339999999999998,
      "loss": 0.4385,
      "step": 1168
    },
    {
      "epoch": 15.8,
      "learning_rate": 0.00023359999999999996,
      "loss": 0.3069,
      "step": 1169
    },
    {
      "epoch": 15.81,
      "learning_rate": 0.00023379999999999996,
      "loss": 0.2775,
      "step": 1170
    },
    {
      "epoch": 15.82,
      "learning_rate": 0.000234,
      "loss": 0.2033,
      "step": 1171
    },
    {
      "epoch": 15.84,
      "learning_rate": 0.00023419999999999997,
      "loss": 0.232,
      "step": 1172
    },
    {
      "epoch": 15.85,
      "learning_rate": 0.00023439999999999998,
      "loss": 0.3359,
      "step": 1173
    },
    {
      "epoch": 15.86,
      "learning_rate": 0.00023459999999999998,
      "loss": 0.296,
      "step": 1174
    },
    {
      "epoch": 15.88,
      "learning_rate": 0.00023479999999999996,
      "loss": 0.184,
      "step": 1175
    },
    {
      "epoch": 15.89,
      "learning_rate": 0.00023499999999999997,
      "loss": 0.3075,
      "step": 1176
    },
    {
      "epoch": 15.91,
      "learning_rate": 0.0002352,
      "loss": 0.4092,
      "step": 1177
    },
    {
      "epoch": 15.92,
      "learning_rate": 0.00023539999999999998,
      "loss": 0.2754,
      "step": 1178
    },
    {
      "epoch": 15.93,
      "learning_rate": 0.00023559999999999998,
      "loss": 0.3209,
      "step": 1179
    },
    {
      "epoch": 15.95,
      "learning_rate": 0.00023579999999999999,
      "loss": 0.3336,
      "step": 1180
    },
    {
      "epoch": 15.96,
      "learning_rate": 0.00023599999999999996,
      "loss": 0.3614,
      "step": 1181
    },
    {
      "epoch": 15.97,
      "learning_rate": 0.00023619999999999997,
      "loss": 0.2762,
      "step": 1182
    },
    {
      "epoch": 15.99,
      "learning_rate": 0.0002364,
      "loss": 0.0869,
      "step": 1183
    },
    {
      "epoch": 16.0,
      "learning_rate": 0.00023659999999999998,
      "loss": 0.225,
      "step": 1184
    },
    {
      "epoch": 16.01,
      "learning_rate": 0.00023679999999999998,
      "loss": 0.2133,
      "step": 1185
    },
    {
      "epoch": 16.03,
      "learning_rate": 0.000237,
      "loss": 0.3941,
      "step": 1186
    },
    {
      "epoch": 16.04,
      "learning_rate": 0.00023719999999999997,
      "loss": 0.1468,
      "step": 1187
    },
    {
      "epoch": 16.05,
      "learning_rate": 0.00023739999999999997,
      "loss": 0.247,
      "step": 1188
    },
    {
      "epoch": 16.07,
      "learning_rate": 0.0002376,
      "loss": 0.1894,
      "step": 1189
    },
    {
      "epoch": 16.08,
      "learning_rate": 0.00023779999999999998,
      "loss": 0.2545,
      "step": 1190
    },
    {
      "epoch": 16.09,
      "learning_rate": 0.00023799999999999998,
      "loss": 0.5769,
      "step": 1191
    },
    {
      "epoch": 16.11,
      "learning_rate": 0.0002382,
      "loss": 0.1261,
      "step": 1192
    },
    {
      "epoch": 16.12,
      "learning_rate": 0.00023839999999999997,
      "loss": 0.1674,
      "step": 1193
    },
    {
      "epoch": 16.14,
      "learning_rate": 0.00023859999999999997,
      "loss": 0.2802,
      "step": 1194
    },
    {
      "epoch": 16.15,
      "learning_rate": 0.0002388,
      "loss": 0.3018,
      "step": 1195
    },
    {
      "epoch": 16.16,
      "learning_rate": 0.00023899999999999998,
      "loss": 0.5797,
      "step": 1196
    },
    {
      "epoch": 16.18,
      "learning_rate": 0.0002392,
      "loss": 0.2643,
      "step": 1197
    },
    {
      "epoch": 16.19,
      "learning_rate": 0.0002394,
      "loss": 0.5432,
      "step": 1198
    },
    {
      "epoch": 16.2,
      "learning_rate": 0.00023959999999999997,
      "loss": 0.3085,
      "step": 1199
    },
    {
      "epoch": 16.22,
      "learning_rate": 0.00023979999999999997,
      "loss": 0.4806,
      "step": 1200
    },
    {
      "epoch": 16.22,
      "eval_accuracy": 0.8786585365853659,
      "eval_f1": 0.8784462019085745,
      "eval_loss": 0.49812883138656616,
      "eval_runtime": 125.819,
      "eval_samples_per_second": 13.035,
      "eval_steps_per_second": 0.819,
      "step": 1200
    },
    {
      "epoch": 16.23,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.6398,
      "step": 1201
    },
    {
      "epoch": 16.24,
      "learning_rate": 0.00024019999999999996,
      "loss": 0.381,
      "step": 1202
    },
    {
      "epoch": 16.26,
      "learning_rate": 0.0002404,
      "loss": 0.5224,
      "step": 1203
    },
    {
      "epoch": 16.27,
      "learning_rate": 0.0002406,
      "loss": 0.4102,
      "step": 1204
    },
    {
      "epoch": 16.28,
      "learning_rate": 0.00024079999999999997,
      "loss": 0.1748,
      "step": 1205
    },
    {
      "epoch": 16.3,
      "learning_rate": 0.00024099999999999998,
      "loss": 0.147,
      "step": 1206
    },
    {
      "epoch": 16.31,
      "learning_rate": 0.00024119999999999998,
      "loss": 0.1146,
      "step": 1207
    },
    {
      "epoch": 16.32,
      "learning_rate": 0.00024139999999999996,
      "loss": 0.2012,
      "step": 1208
    },
    {
      "epoch": 16.34,
      "learning_rate": 0.0002416,
      "loss": 0.3077,
      "step": 1209
    },
    {
      "epoch": 16.35,
      "learning_rate": 0.0002418,
      "loss": 0.3532,
      "step": 1210
    },
    {
      "epoch": 16.36,
      "learning_rate": 0.00024199999999999997,
      "loss": 0.1483,
      "step": 1211
    },
    {
      "epoch": 16.38,
      "learning_rate": 0.00024219999999999998,
      "loss": 0.3194,
      "step": 1212
    },
    {
      "epoch": 16.39,
      "learning_rate": 0.00024239999999999998,
      "loss": 0.3533,
      "step": 1213
    },
    {
      "epoch": 16.41,
      "learning_rate": 0.00024259999999999996,
      "loss": 0.4186,
      "step": 1214
    },
    {
      "epoch": 16.42,
      "learning_rate": 0.0002428,
      "loss": 0.3087,
      "step": 1215
    },
    {
      "epoch": 16.43,
      "learning_rate": 0.000243,
      "loss": 0.3661,
      "step": 1216
    },
    {
      "epoch": 16.45,
      "learning_rate": 0.00024319999999999998,
      "loss": 0.4381,
      "step": 1217
    },
    {
      "epoch": 16.46,
      "learning_rate": 0.00024339999999999998,
      "loss": 0.8433,
      "step": 1218
    },
    {
      "epoch": 16.47,
      "learning_rate": 0.00024359999999999999,
      "loss": 0.2987,
      "step": 1219
    },
    {
      "epoch": 16.49,
      "learning_rate": 0.00024379999999999996,
      "loss": 0.3431,
      "step": 1220
    },
    {
      "epoch": 16.5,
      "learning_rate": 0.000244,
      "loss": 0.4714,
      "step": 1221
    },
    {
      "epoch": 16.51,
      "learning_rate": 0.00024419999999999997,
      "loss": 0.2592,
      "step": 1222
    },
    {
      "epoch": 16.53,
      "learning_rate": 0.0002444,
      "loss": 0.1472,
      "step": 1223
    },
    {
      "epoch": 16.54,
      "learning_rate": 0.0002446,
      "loss": 0.3633,
      "step": 1224
    },
    {
      "epoch": 16.55,
      "learning_rate": 0.0002448,
      "loss": 0.3477,
      "step": 1225
    },
    {
      "epoch": 16.57,
      "learning_rate": 0.000245,
      "loss": 0.2391,
      "step": 1226
    },
    {
      "epoch": 16.58,
      "learning_rate": 0.0002452,
      "loss": 0.2062,
      "step": 1227
    },
    {
      "epoch": 16.59,
      "learning_rate": 0.00024539999999999995,
      "loss": 0.149,
      "step": 1228
    },
    {
      "epoch": 16.61,
      "learning_rate": 0.00024559999999999995,
      "loss": 0.2565,
      "step": 1229
    },
    {
      "epoch": 16.62,
      "learning_rate": 0.0002458,
      "loss": 0.1375,
      "step": 1230
    },
    {
      "epoch": 16.64,
      "learning_rate": 0.00024599999999999996,
      "loss": 0.4086,
      "step": 1231
    },
    {
      "epoch": 16.65,
      "learning_rate": 0.00024619999999999997,
      "loss": 0.3422,
      "step": 1232
    },
    {
      "epoch": 16.66,
      "learning_rate": 0.00024639999999999997,
      "loss": 0.3616,
      "step": 1233
    },
    {
      "epoch": 16.68,
      "learning_rate": 0.0002466,
      "loss": 0.131,
      "step": 1234
    },
    {
      "epoch": 16.69,
      "learning_rate": 0.0002468,
      "loss": 0.2289,
      "step": 1235
    },
    {
      "epoch": 16.7,
      "learning_rate": 0.000247,
      "loss": 0.0955,
      "step": 1236
    },
    {
      "epoch": 16.72,
      "learning_rate": 0.0002472,
      "loss": 0.3041,
      "step": 1237
    },
    {
      "epoch": 16.73,
      "learning_rate": 0.0002474,
      "loss": 0.1215,
      "step": 1238
    },
    {
      "epoch": 16.74,
      "learning_rate": 0.0002476,
      "loss": 0.4083,
      "step": 1239
    },
    {
      "epoch": 16.76,
      "learning_rate": 0.00024779999999999995,
      "loss": 0.2564,
      "step": 1240
    },
    {
      "epoch": 16.77,
      "learning_rate": 0.00024799999999999996,
      "loss": 0.545,
      "step": 1241
    },
    {
      "epoch": 16.78,
      "learning_rate": 0.00024819999999999996,
      "loss": 0.2072,
      "step": 1242
    },
    {
      "epoch": 16.8,
      "learning_rate": 0.00024839999999999997,
      "loss": 0.451,
      "step": 1243
    },
    {
      "epoch": 16.81,
      "learning_rate": 0.00024859999999999997,
      "loss": 0.1766,
      "step": 1244
    },
    {
      "epoch": 16.82,
      "learning_rate": 0.0002488,
      "loss": 0.2121,
      "step": 1245
    },
    {
      "epoch": 16.84,
      "learning_rate": 0.000249,
      "loss": 0.3126,
      "step": 1246
    },
    {
      "epoch": 16.85,
      "learning_rate": 0.0002492,
      "loss": 0.1877,
      "step": 1247
    },
    {
      "epoch": 16.86,
      "learning_rate": 0.0002494,
      "loss": 0.5007,
      "step": 1248
    },
    {
      "epoch": 16.88,
      "learning_rate": 0.00024959999999999994,
      "loss": 0.2483,
      "step": 1249
    },
    {
      "epoch": 16.89,
      "learning_rate": 0.0002498,
      "loss": 0.2738,
      "step": 1250
    },
    {
      "epoch": 16.91,
      "learning_rate": 0.00025,
      "loss": 0.351,
      "step": 1251
    },
    {
      "epoch": 16.92,
      "learning_rate": 0.00025019999999999996,
      "loss": 0.3176,
      "step": 1252
    },
    {
      "epoch": 16.93,
      "learning_rate": 0.00025039999999999996,
      "loss": 0.204,
      "step": 1253
    },
    {
      "epoch": 16.95,
      "learning_rate": 0.00025059999999999997,
      "loss": 0.3051,
      "step": 1254
    },
    {
      "epoch": 16.96,
      "learning_rate": 0.00025079999999999997,
      "loss": 0.1357,
      "step": 1255
    },
    {
      "epoch": 16.97,
      "learning_rate": 0.000251,
      "loss": 0.4072,
      "step": 1256
    },
    {
      "epoch": 16.99,
      "learning_rate": 0.0002512,
      "loss": 0.3807,
      "step": 1257
    },
    {
      "epoch": 17.0,
      "learning_rate": 0.0002514,
      "loss": 0.4054,
      "step": 1258
    },
    {
      "epoch": 17.01,
      "learning_rate": 0.0002516,
      "loss": 0.5336,
      "step": 1259
    },
    {
      "epoch": 17.03,
      "learning_rate": 0.0002518,
      "loss": 0.308,
      "step": 1260
    },
    {
      "epoch": 17.04,
      "learning_rate": 0.00025199999999999995,
      "loss": 0.2131,
      "step": 1261
    },
    {
      "epoch": 17.05,
      "learning_rate": 0.0002522,
      "loss": 0.0305,
      "step": 1262
    },
    {
      "epoch": 17.07,
      "learning_rate": 0.0002524,
      "loss": 0.0908,
      "step": 1263
    },
    {
      "epoch": 17.08,
      "learning_rate": 0.00025259999999999996,
      "loss": 0.2512,
      "step": 1264
    },
    {
      "epoch": 17.09,
      "learning_rate": 0.00025279999999999996,
      "loss": 0.1997,
      "step": 1265
    },
    {
      "epoch": 17.11,
      "learning_rate": 0.00025299999999999997,
      "loss": 0.4145,
      "step": 1266
    },
    {
      "epoch": 17.12,
      "learning_rate": 0.0002532,
      "loss": 0.5771,
      "step": 1267
    },
    {
      "epoch": 17.14,
      "learning_rate": 0.0002534,
      "loss": 0.2776,
      "step": 1268
    },
    {
      "epoch": 17.15,
      "learning_rate": 0.0002536,
      "loss": 0.653,
      "step": 1269
    },
    {
      "epoch": 17.16,
      "learning_rate": 0.0002538,
      "loss": 0.3912,
      "step": 1270
    },
    {
      "epoch": 17.18,
      "learning_rate": 0.000254,
      "loss": 0.4761,
      "step": 1271
    },
    {
      "epoch": 17.19,
      "learning_rate": 0.0002542,
      "loss": 0.3934,
      "step": 1272
    },
    {
      "epoch": 17.2,
      "learning_rate": 0.00025439999999999995,
      "loss": 0.2676,
      "step": 1273
    },
    {
      "epoch": 17.22,
      "learning_rate": 0.0002546,
      "loss": 0.2582,
      "step": 1274
    },
    {
      "epoch": 17.23,
      "learning_rate": 0.0002548,
      "loss": 0.2249,
      "step": 1275
    },
    {
      "epoch": 17.24,
      "learning_rate": 0.00025499999999999996,
      "loss": 0.1659,
      "step": 1276
    },
    {
      "epoch": 17.26,
      "learning_rate": 0.00025519999999999997,
      "loss": 0.1882,
      "step": 1277
    },
    {
      "epoch": 17.27,
      "learning_rate": 0.0002554,
      "loss": 0.2045,
      "step": 1278
    },
    {
      "epoch": 17.28,
      "learning_rate": 0.0002556,
      "loss": 0.2392,
      "step": 1279
    },
    {
      "epoch": 17.3,
      "learning_rate": 0.0002558,
      "loss": 0.3478,
      "step": 1280
    },
    {
      "epoch": 17.31,
      "learning_rate": 0.000256,
      "loss": 0.2754,
      "step": 1281
    },
    {
      "epoch": 17.32,
      "learning_rate": 0.0002562,
      "loss": 0.259,
      "step": 1282
    },
    {
      "epoch": 17.34,
      "learning_rate": 0.0002564,
      "loss": 0.3611,
      "step": 1283
    },
    {
      "epoch": 17.35,
      "learning_rate": 0.00025659999999999995,
      "loss": 0.146,
      "step": 1284
    },
    {
      "epoch": 17.36,
      "learning_rate": 0.00025679999999999995,
      "loss": 0.3597,
      "step": 1285
    },
    {
      "epoch": 17.38,
      "learning_rate": 0.00025699999999999996,
      "loss": 0.4947,
      "step": 1286
    },
    {
      "epoch": 17.39,
      "learning_rate": 0.00025719999999999996,
      "loss": 0.4217,
      "step": 1287
    },
    {
      "epoch": 17.41,
      "learning_rate": 0.00025739999999999997,
      "loss": 0.2974,
      "step": 1288
    },
    {
      "epoch": 17.42,
      "learning_rate": 0.0002576,
      "loss": 0.1555,
      "step": 1289
    },
    {
      "epoch": 17.43,
      "learning_rate": 0.0002578,
      "loss": 0.1611,
      "step": 1290
    },
    {
      "epoch": 17.45,
      "learning_rate": 0.000258,
      "loss": 0.2897,
      "step": 1291
    },
    {
      "epoch": 17.46,
      "learning_rate": 0.0002582,
      "loss": 0.3173,
      "step": 1292
    },
    {
      "epoch": 17.47,
      "learning_rate": 0.00025839999999999994,
      "loss": 0.2727,
      "step": 1293
    },
    {
      "epoch": 17.49,
      "learning_rate": 0.0002586,
      "loss": 0.2257,
      "step": 1294
    },
    {
      "epoch": 17.5,
      "learning_rate": 0.0002588,
      "loss": 0.4893,
      "step": 1295
    },
    {
      "epoch": 17.51,
      "learning_rate": 0.00025899999999999995,
      "loss": 0.3561,
      "step": 1296
    },
    {
      "epoch": 17.53,
      "learning_rate": 0.00025919999999999996,
      "loss": 0.1397,
      "step": 1297
    },
    {
      "epoch": 17.54,
      "learning_rate": 0.00025939999999999996,
      "loss": 0.2824,
      "step": 1298
    },
    {
      "epoch": 17.55,
      "learning_rate": 0.00025959999999999997,
      "loss": 0.3632,
      "step": 1299
    },
    {
      "epoch": 17.57,
      "learning_rate": 0.00025979999999999997,
      "loss": 0.2974,
      "step": 1300
    },
    {
      "epoch": 17.58,
      "learning_rate": 0.00026,
      "loss": 0.4531,
      "step": 1301
    },
    {
      "epoch": 17.59,
      "learning_rate": 0.0002602,
      "loss": 0.4489,
      "step": 1302
    },
    {
      "epoch": 17.61,
      "learning_rate": 0.0002604,
      "loss": 0.3786,
      "step": 1303
    },
    {
      "epoch": 17.62,
      "learning_rate": 0.0002606,
      "loss": 0.2942,
      "step": 1304
    },
    {
      "epoch": 17.64,
      "learning_rate": 0.00026079999999999994,
      "loss": 0.6698,
      "step": 1305
    },
    {
      "epoch": 17.65,
      "learning_rate": 0.000261,
      "loss": 0.1663,
      "step": 1306
    },
    {
      "epoch": 17.66,
      "learning_rate": 0.0002612,
      "loss": 0.3783,
      "step": 1307
    },
    {
      "epoch": 17.68,
      "learning_rate": 0.00026139999999999996,
      "loss": 0.312,
      "step": 1308
    },
    {
      "epoch": 17.69,
      "learning_rate": 0.00026159999999999996,
      "loss": 0.3783,
      "step": 1309
    },
    {
      "epoch": 17.7,
      "learning_rate": 0.00026179999999999997,
      "loss": 0.1481,
      "step": 1310
    },
    {
      "epoch": 17.72,
      "learning_rate": 0.00026199999999999997,
      "loss": 0.1699,
      "step": 1311
    },
    {
      "epoch": 17.73,
      "learning_rate": 0.0002622,
      "loss": 0.4477,
      "step": 1312
    },
    {
      "epoch": 17.74,
      "learning_rate": 0.0002624,
      "loss": 0.2032,
      "step": 1313
    },
    {
      "epoch": 17.76,
      "learning_rate": 0.0002626,
      "loss": 0.1765,
      "step": 1314
    },
    {
      "epoch": 17.77,
      "learning_rate": 0.0002626,
      "loss": 0.3896,
      "step": 1315
    },
    {
      "epoch": 17.78,
      "learning_rate": 0.0002628,
      "loss": 0.3777,
      "step": 1316
    },
    {
      "epoch": 17.8,
      "learning_rate": 0.000263,
      "loss": 0.1653,
      "step": 1317
    },
    {
      "epoch": 17.81,
      "learning_rate": 0.00026319999999999995,
      "loss": 0.2024,
      "step": 1318
    },
    {
      "epoch": 17.82,
      "learning_rate": 0.00026339999999999995,
      "loss": 0.2158,
      "step": 1319
    },
    {
      "epoch": 17.84,
      "learning_rate": 0.0002636,
      "loss": 0.3709,
      "step": 1320
    },
    {
      "epoch": 17.85,
      "learning_rate": 0.00026379999999999996,
      "loss": 0.1483,
      "step": 1321
    },
    {
      "epoch": 17.86,
      "learning_rate": 0.00026399999999999997,
      "loss": 0.1921,
      "step": 1322
    },
    {
      "epoch": 17.88,
      "learning_rate": 0.00026419999999999997,
      "loss": 0.1324,
      "step": 1323
    },
    {
      "epoch": 17.89,
      "learning_rate": 0.0002644,
      "loss": 0.3457,
      "step": 1324
    },
    {
      "epoch": 17.91,
      "learning_rate": 0.0002646,
      "loss": 0.1695,
      "step": 1325
    },
    {
      "epoch": 17.92,
      "learning_rate": 0.0002648,
      "loss": 0.3445,
      "step": 1326
    },
    {
      "epoch": 17.93,
      "learning_rate": 0.000265,
      "loss": 0.3744,
      "step": 1327
    },
    {
      "epoch": 17.95,
      "learning_rate": 0.0002652,
      "loss": 0.1488,
      "step": 1328
    },
    {
      "epoch": 17.96,
      "learning_rate": 0.0002654,
      "loss": 0.2231,
      "step": 1329
    },
    {
      "epoch": 17.97,
      "learning_rate": 0.00026559999999999995,
      "loss": 0.1888,
      "step": 1330
    },
    {
      "epoch": 17.99,
      "learning_rate": 0.00026579999999999996,
      "loss": 0.6292,
      "step": 1331
    },
    {
      "epoch": 18.0,
      "learning_rate": 0.000266,
      "loss": 0.067,
      "step": 1332
    },
    {
      "epoch": 18.01,
      "learning_rate": 0.00026619999999999997,
      "loss": 0.3133,
      "step": 1333
    },
    {
      "epoch": 18.03,
      "learning_rate": 0.00026639999999999997,
      "loss": 0.4072,
      "step": 1334
    },
    {
      "epoch": 18.04,
      "learning_rate": 0.0002666,
      "loss": 0.4405,
      "step": 1335
    },
    {
      "epoch": 18.05,
      "learning_rate": 0.0002668,
      "loss": 0.5238,
      "step": 1336
    },
    {
      "epoch": 18.07,
      "learning_rate": 0.000267,
      "loss": 0.5586,
      "step": 1337
    },
    {
      "epoch": 18.08,
      "learning_rate": 0.0002672,
      "loss": 0.1807,
      "step": 1338
    },
    {
      "epoch": 18.09,
      "learning_rate": 0.0002674,
      "loss": 0.2143,
      "step": 1339
    },
    {
      "epoch": 18.11,
      "learning_rate": 0.0002676,
      "loss": 0.2594,
      "step": 1340
    },
    {
      "epoch": 18.12,
      "learning_rate": 0.0002678,
      "loss": 0.2064,
      "step": 1341
    },
    {
      "epoch": 18.14,
      "learning_rate": 0.00026799999999999995,
      "loss": 0.0694,
      "step": 1342
    },
    {
      "epoch": 18.15,
      "learning_rate": 0.00026819999999999996,
      "loss": 0.0922,
      "step": 1343
    },
    {
      "epoch": 18.16,
      "learning_rate": 0.0002684,
      "loss": 0.5782,
      "step": 1344
    },
    {
      "epoch": 18.18,
      "learning_rate": 0.00026859999999999997,
      "loss": 0.3298,
      "step": 1345
    },
    {
      "epoch": 18.19,
      "learning_rate": 0.0002688,
      "loss": 0.08,
      "step": 1346
    },
    {
      "epoch": 18.2,
      "learning_rate": 0.000269,
      "loss": 0.3685,
      "step": 1347
    },
    {
      "epoch": 18.22,
      "learning_rate": 0.0002692,
      "loss": 0.1472,
      "step": 1348
    },
    {
      "epoch": 18.23,
      "learning_rate": 0.0002694,
      "loss": 0.3624,
      "step": 1349
    },
    {
      "epoch": 18.24,
      "learning_rate": 0.00026959999999999994,
      "loss": 0.3671,
      "step": 1350
    },
    {
      "epoch": 18.26,
      "learning_rate": 0.0002698,
      "loss": 0.6211,
      "step": 1351
    },
    {
      "epoch": 18.27,
      "learning_rate": 0.00027,
      "loss": 0.5175,
      "step": 1352
    },
    {
      "epoch": 18.28,
      "learning_rate": 0.00027019999999999995,
      "loss": 0.147,
      "step": 1353
    },
    {
      "epoch": 18.3,
      "learning_rate": 0.00027039999999999996,
      "loss": 0.2328,
      "step": 1354
    },
    {
      "epoch": 18.31,
      "learning_rate": 0.00027059999999999996,
      "loss": 0.1154,
      "step": 1355
    },
    {
      "epoch": 18.32,
      "learning_rate": 0.00027079999999999997,
      "loss": 0.0677,
      "step": 1356
    },
    {
      "epoch": 18.34,
      "learning_rate": 0.000271,
      "loss": 0.294,
      "step": 1357
    },
    {
      "epoch": 18.35,
      "learning_rate": 0.0002712,
      "loss": 0.4698,
      "step": 1358
    },
    {
      "epoch": 18.36,
      "learning_rate": 0.0002714,
      "loss": 0.1337,
      "step": 1359
    },
    {
      "epoch": 18.38,
      "learning_rate": 0.0002716,
      "loss": 0.2759,
      "step": 1360
    },
    {
      "epoch": 18.39,
      "learning_rate": 0.0002718,
      "loss": 0.3038,
      "step": 1361
    },
    {
      "epoch": 18.41,
      "learning_rate": 0.00027199999999999994,
      "loss": 0.1626,
      "step": 1362
    },
    {
      "epoch": 18.42,
      "learning_rate": 0.00027219999999999995,
      "loss": 0.1619,
      "step": 1363
    },
    {
      "epoch": 18.43,
      "learning_rate": 0.0002724,
      "loss": 0.2661,
      "step": 1364
    },
    {
      "epoch": 18.45,
      "learning_rate": 0.00027259999999999996,
      "loss": 0.2124,
      "step": 1365
    },
    {
      "epoch": 18.46,
      "learning_rate": 0.00027279999999999996,
      "loss": 0.4175,
      "step": 1366
    },
    {
      "epoch": 18.47,
      "learning_rate": 0.00027299999999999997,
      "loss": 0.3095,
      "step": 1367
    },
    {
      "epoch": 18.49,
      "learning_rate": 0.00027319999999999997,
      "loss": 0.3653,
      "step": 1368
    },
    {
      "epoch": 18.5,
      "learning_rate": 0.0002734,
      "loss": 0.7272,
      "step": 1369
    },
    {
      "epoch": 18.51,
      "learning_rate": 0.0002736,
      "loss": 0.3239,
      "step": 1370
    },
    {
      "epoch": 18.53,
      "learning_rate": 0.0002738,
      "loss": 0.2081,
      "step": 1371
    },
    {
      "epoch": 18.54,
      "learning_rate": 0.000274,
      "loss": 0.1915,
      "step": 1372
    },
    {
      "epoch": 18.55,
      "learning_rate": 0.0002742,
      "loss": 0.1812,
      "step": 1373
    },
    {
      "epoch": 18.57,
      "learning_rate": 0.00027439999999999995,
      "loss": 0.2143,
      "step": 1374
    },
    {
      "epoch": 18.58,
      "learning_rate": 0.00027459999999999995,
      "loss": 0.3531,
      "step": 1375
    },
    {
      "epoch": 18.59,
      "learning_rate": 0.0002748,
      "loss": 0.5,
      "step": 1376
    },
    {
      "epoch": 18.61,
      "learning_rate": 0.00027499999999999996,
      "loss": 0.2664,
      "step": 1377
    },
    {
      "epoch": 18.62,
      "learning_rate": 0.00027519999999999997,
      "loss": 0.1786,
      "step": 1378
    },
    {
      "epoch": 18.64,
      "learning_rate": 0.00027539999999999997,
      "loss": 0.2092,
      "step": 1379
    },
    {
      "epoch": 18.65,
      "learning_rate": 0.0002756,
      "loss": 0.1366,
      "step": 1380
    },
    {
      "epoch": 18.66,
      "learning_rate": 0.0002758,
      "loss": 0.3179,
      "step": 1381
    },
    {
      "epoch": 18.68,
      "learning_rate": 0.000276,
      "loss": 0.2658,
      "step": 1382
    },
    {
      "epoch": 18.69,
      "learning_rate": 0.0002762,
      "loss": 0.1435,
      "step": 1383
    },
    {
      "epoch": 18.7,
      "learning_rate": 0.0002764,
      "loss": 0.2156,
      "step": 1384
    },
    {
      "epoch": 18.72,
      "learning_rate": 0.0002766,
      "loss": 0.2627,
      "step": 1385
    },
    {
      "epoch": 18.73,
      "learning_rate": 0.00027679999999999995,
      "loss": 0.4329,
      "step": 1386
    },
    {
      "epoch": 18.74,
      "learning_rate": 0.00027699999999999996,
      "loss": 0.2768,
      "step": 1387
    },
    {
      "epoch": 18.76,
      "learning_rate": 0.0002772,
      "loss": 0.2883,
      "step": 1388
    },
    {
      "epoch": 18.77,
      "learning_rate": 0.00027739999999999997,
      "loss": 0.2949,
      "step": 1389
    },
    {
      "epoch": 18.78,
      "learning_rate": 0.00027759999999999997,
      "loss": 0.1977,
      "step": 1390
    },
    {
      "epoch": 18.8,
      "learning_rate": 0.0002778,
      "loss": 0.2669,
      "step": 1391
    },
    {
      "epoch": 18.81,
      "learning_rate": 0.000278,
      "loss": 0.0968,
      "step": 1392
    },
    {
      "epoch": 18.82,
      "learning_rate": 0.0002782,
      "loss": 0.3387,
      "step": 1393
    },
    {
      "epoch": 18.84,
      "learning_rate": 0.0002784,
      "loss": 0.2539,
      "step": 1394
    },
    {
      "epoch": 18.85,
      "learning_rate": 0.00027859999999999994,
      "loss": 0.2499,
      "step": 1395
    },
    {
      "epoch": 18.86,
      "learning_rate": 0.0002788,
      "loss": 0.1083,
      "step": 1396
    },
    {
      "epoch": 18.88,
      "learning_rate": 0.000279,
      "loss": 0.1436,
      "step": 1397
    },
    {
      "epoch": 18.89,
      "learning_rate": 0.00027919999999999996,
      "loss": 0.2188,
      "step": 1398
    },
    {
      "epoch": 18.91,
      "learning_rate": 0.00027939999999999996,
      "loss": 0.2322,
      "step": 1399
    },
    {
      "epoch": 18.92,
      "learning_rate": 0.00027959999999999997,
      "loss": 0.2103,
      "step": 1400
    },
    {
      "epoch": 18.92,
      "eval_accuracy": 0.8817073170731707,
      "eval_f1": 0.8810075269224288,
      "eval_loss": 0.4956743121147156,
      "eval_runtime": 125.5403,
      "eval_samples_per_second": 13.064,
      "eval_steps_per_second": 0.82,
      "step": 1400
    },
    {
      "epoch": 18.93,
      "learning_rate": 0.00027979999999999997,
      "loss": 0.4907,
      "step": 1401
    },
    {
      "epoch": 18.95,
      "learning_rate": 0.00028,
      "loss": 0.4818,
      "step": 1402
    },
    {
      "epoch": 18.96,
      "learning_rate": 0.0002802,
      "loss": 0.2171,
      "step": 1403
    },
    {
      "epoch": 18.97,
      "learning_rate": 0.0002804,
      "loss": 0.3013,
      "step": 1404
    },
    {
      "epoch": 18.99,
      "learning_rate": 0.0002806,
      "loss": 0.4552,
      "step": 1405
    },
    {
      "epoch": 19.0,
      "learning_rate": 0.0002808,
      "loss": 0.3314,
      "step": 1406
    },
    {
      "epoch": 19.01,
      "learning_rate": 0.00028099999999999995,
      "loss": 0.4121,
      "step": 1407
    },
    {
      "epoch": 19.03,
      "learning_rate": 0.0002812,
      "loss": 0.1902,
      "step": 1408
    },
    {
      "epoch": 19.04,
      "learning_rate": 0.00028139999999999996,
      "loss": 0.1849,
      "step": 1409
    },
    {
      "epoch": 19.05,
      "learning_rate": 0.00028159999999999996,
      "loss": 0.2525,
      "step": 1410
    },
    {
      "epoch": 19.07,
      "learning_rate": 0.00028179999999999997,
      "loss": 0.1704,
      "step": 1411
    },
    {
      "epoch": 19.08,
      "learning_rate": 0.00028199999999999997,
      "loss": 0.2586,
      "step": 1412
    },
    {
      "epoch": 19.09,
      "learning_rate": 0.0002822,
      "loss": 0.2289,
      "step": 1413
    },
    {
      "epoch": 19.11,
      "learning_rate": 0.0002824,
      "loss": 0.3416,
      "step": 1414
    },
    {
      "epoch": 19.12,
      "learning_rate": 0.0002826,
      "loss": 0.6191,
      "step": 1415
    },
    {
      "epoch": 19.14,
      "learning_rate": 0.0002828,
      "loss": 0.2435,
      "step": 1416
    },
    {
      "epoch": 19.15,
      "learning_rate": 0.000283,
      "loss": 0.3784,
      "step": 1417
    },
    {
      "epoch": 19.16,
      "learning_rate": 0.00028319999999999994,
      "loss": 0.2906,
      "step": 1418
    },
    {
      "epoch": 19.18,
      "learning_rate": 0.00028339999999999995,
      "loss": 0.2081,
      "step": 1419
    },
    {
      "epoch": 19.19,
      "learning_rate": 0.0002836,
      "loss": 0.4352,
      "step": 1420
    },
    {
      "epoch": 19.2,
      "learning_rate": 0.00028379999999999996,
      "loss": 0.4684,
      "step": 1421
    },
    {
      "epoch": 19.22,
      "learning_rate": 0.00028399999999999996,
      "loss": 0.1617,
      "step": 1422
    },
    {
      "epoch": 19.23,
      "learning_rate": 0.00028419999999999997,
      "loss": 0.2626,
      "step": 1423
    },
    {
      "epoch": 19.24,
      "learning_rate": 0.0002844,
      "loss": 0.1178,
      "step": 1424
    },
    {
      "epoch": 19.26,
      "learning_rate": 0.0002846,
      "loss": 0.2969,
      "step": 1425
    },
    {
      "epoch": 19.27,
      "learning_rate": 0.0002848,
      "loss": 0.5409,
      "step": 1426
    },
    {
      "epoch": 19.28,
      "learning_rate": 0.000285,
      "loss": 0.2508,
      "step": 1427
    },
    {
      "epoch": 19.3,
      "learning_rate": 0.0002852,
      "loss": 0.1179,
      "step": 1428
    },
    {
      "epoch": 19.31,
      "learning_rate": 0.0002854,
      "loss": 0.4968,
      "step": 1429
    },
    {
      "epoch": 19.32,
      "learning_rate": 0.00028559999999999995,
      "loss": 0.4972,
      "step": 1430
    },
    {
      "epoch": 19.34,
      "learning_rate": 0.00028579999999999995,
      "loss": 0.7266,
      "step": 1431
    },
    {
      "epoch": 19.35,
      "learning_rate": 0.00028599999999999996,
      "loss": 0.1823,
      "step": 1432
    },
    {
      "epoch": 19.36,
      "learning_rate": 0.00028619999999999996,
      "loss": 0.2619,
      "step": 1433
    },
    {
      "epoch": 19.38,
      "learning_rate": 0.00028639999999999997,
      "loss": 0.181,
      "step": 1434
    },
    {
      "epoch": 19.39,
      "learning_rate": 0.0002866,
      "loss": 0.407,
      "step": 1435
    },
    {
      "epoch": 19.41,
      "learning_rate": 0.0002868,
      "loss": 0.3326,
      "step": 1436
    },
    {
      "epoch": 19.42,
      "learning_rate": 0.000287,
      "loss": 0.4503,
      "step": 1437
    },
    {
      "epoch": 19.43,
      "learning_rate": 0.0002872,
      "loss": 0.419,
      "step": 1438
    },
    {
      "epoch": 19.45,
      "learning_rate": 0.00028739999999999994,
      "loss": 0.2988,
      "step": 1439
    },
    {
      "epoch": 19.46,
      "learning_rate": 0.0002876,
      "loss": 0.1889,
      "step": 1440
    },
    {
      "epoch": 19.47,
      "learning_rate": 0.0002878,
      "loss": 0.4885,
      "step": 1441
    },
    {
      "epoch": 19.49,
      "learning_rate": 0.00028799999999999995,
      "loss": 0.1875,
      "step": 1442
    },
    {
      "epoch": 19.5,
      "learning_rate": 0.00028819999999999996,
      "loss": 0.5513,
      "step": 1443
    },
    {
      "epoch": 19.51,
      "learning_rate": 0.00028839999999999996,
      "loss": 0.2832,
      "step": 1444
    },
    {
      "epoch": 19.53,
      "learning_rate": 0.00028859999999999997,
      "loss": 0.2068,
      "step": 1445
    },
    {
      "epoch": 19.54,
      "learning_rate": 0.00028879999999999997,
      "loss": 0.3752,
      "step": 1446
    },
    {
      "epoch": 19.55,
      "learning_rate": 0.000289,
      "loss": 0.264,
      "step": 1447
    },
    {
      "epoch": 19.57,
      "learning_rate": 0.0002892,
      "loss": 0.2028,
      "step": 1448
    },
    {
      "epoch": 19.58,
      "learning_rate": 0.0002894,
      "loss": 0.2439,
      "step": 1449
    },
    {
      "epoch": 19.59,
      "learning_rate": 0.0002896,
      "loss": 0.4324,
      "step": 1450
    },
    {
      "epoch": 19.61,
      "learning_rate": 0.00028979999999999994,
      "loss": 0.3333,
      "step": 1451
    },
    {
      "epoch": 19.62,
      "learning_rate": 0.00029,
      "loss": 0.5027,
      "step": 1452
    },
    {
      "epoch": 19.64,
      "learning_rate": 0.0002902,
      "loss": 0.349,
      "step": 1453
    },
    {
      "epoch": 19.65,
      "learning_rate": 0.00029039999999999996,
      "loss": 0.4158,
      "step": 1454
    },
    {
      "epoch": 19.66,
      "learning_rate": 0.00029059999999999996,
      "loss": 0.2422,
      "step": 1455
    },
    {
      "epoch": 19.68,
      "learning_rate": 0.00029079999999999997,
      "loss": 0.4739,
      "step": 1456
    },
    {
      "epoch": 19.69,
      "learning_rate": 0.00029099999999999997,
      "loss": 0.2941,
      "step": 1457
    },
    {
      "epoch": 19.7,
      "learning_rate": 0.0002912,
      "loss": 0.1082,
      "step": 1458
    },
    {
      "epoch": 19.72,
      "learning_rate": 0.0002914,
      "loss": 0.3418,
      "step": 1459
    },
    {
      "epoch": 19.73,
      "learning_rate": 0.0002916,
      "loss": 0.1156,
      "step": 1460
    },
    {
      "epoch": 19.74,
      "learning_rate": 0.0002918,
      "loss": 0.3589,
      "step": 1461
    },
    {
      "epoch": 19.76,
      "learning_rate": 0.000292,
      "loss": 0.1954,
      "step": 1462
    },
    {
      "epoch": 19.77,
      "learning_rate": 0.00029219999999999995,
      "loss": 0.2009,
      "step": 1463
    },
    {
      "epoch": 19.78,
      "learning_rate": 0.0002924,
      "loss": 0.3068,
      "step": 1464
    },
    {
      "epoch": 19.8,
      "learning_rate": 0.0002926,
      "loss": 0.2784,
      "step": 1465
    },
    {
      "epoch": 19.81,
      "learning_rate": 0.00029279999999999996,
      "loss": 0.2938,
      "step": 1466
    },
    {
      "epoch": 19.82,
      "learning_rate": 0.00029299999999999997,
      "loss": 0.1915,
      "step": 1467
    },
    {
      "epoch": 19.84,
      "learning_rate": 0.00029319999999999997,
      "loss": 0.2096,
      "step": 1468
    },
    {
      "epoch": 19.85,
      "learning_rate": 0.0002934,
      "loss": 0.2208,
      "step": 1469
    },
    {
      "epoch": 19.86,
      "learning_rate": 0.0002936,
      "loss": 0.1089,
      "step": 1470
    },
    {
      "epoch": 19.88,
      "learning_rate": 0.00029379999999999993,
      "loss": 0.3738,
      "step": 1471
    },
    {
      "epoch": 19.89,
      "learning_rate": 0.000294,
      "loss": 0.2546,
      "step": 1472
    },
    {
      "epoch": 19.91,
      "learning_rate": 0.0002942,
      "loss": 0.4319,
      "step": 1473
    },
    {
      "epoch": 19.92,
      "learning_rate": 0.00029439999999999995,
      "loss": 0.2622,
      "step": 1474
    },
    {
      "epoch": 19.93,
      "learning_rate": 0.00029459999999999995,
      "loss": 0.1809,
      "step": 1475
    },
    {
      "epoch": 19.95,
      "learning_rate": 0.00029479999999999996,
      "loss": 0.2845,
      "step": 1476
    },
    {
      "epoch": 19.96,
      "learning_rate": 0.00029499999999999996,
      "loss": 0.4761,
      "step": 1477
    },
    {
      "epoch": 19.97,
      "learning_rate": 0.00029519999999999997,
      "loss": 0.1853,
      "step": 1478
    },
    {
      "epoch": 19.99,
      "learning_rate": 0.00029539999999999997,
      "loss": 0.2662,
      "step": 1479
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0002956,
      "loss": 0.0137,
      "step": 1480
    },
    {
      "epoch": 20.01,
      "learning_rate": 0.0002958,
      "loss": 0.3405,
      "step": 1481
    },
    {
      "epoch": 20.03,
      "learning_rate": 0.000296,
      "loss": 0.5129,
      "step": 1482
    },
    {
      "epoch": 20.04,
      "learning_rate": 0.00029619999999999994,
      "loss": 0.1391,
      "step": 1483
    },
    {
      "epoch": 20.05,
      "learning_rate": 0.0002964,
      "loss": 0.141,
      "step": 1484
    },
    {
      "epoch": 20.07,
      "learning_rate": 0.0002966,
      "loss": 0.2597,
      "step": 1485
    },
    {
      "epoch": 20.08,
      "learning_rate": 0.00029679999999999995,
      "loss": 0.3591,
      "step": 1486
    },
    {
      "epoch": 20.09,
      "learning_rate": 0.00029699999999999996,
      "loss": 0.28,
      "step": 1487
    },
    {
      "epoch": 20.11,
      "learning_rate": 0.00029719999999999996,
      "loss": 0.1378,
      "step": 1488
    },
    {
      "epoch": 20.12,
      "learning_rate": 0.00029739999999999996,
      "loss": 0.3172,
      "step": 1489
    },
    {
      "epoch": 20.14,
      "learning_rate": 0.00029759999999999997,
      "loss": 0.0962,
      "step": 1490
    },
    {
      "epoch": 20.15,
      "learning_rate": 0.0002978,
      "loss": 0.2351,
      "step": 1491
    },
    {
      "epoch": 20.16,
      "learning_rate": 0.000298,
      "loss": 0.2254,
      "step": 1492
    },
    {
      "epoch": 20.18,
      "learning_rate": 0.0002982,
      "loss": 0.0793,
      "step": 1493
    },
    {
      "epoch": 20.19,
      "learning_rate": 0.0002984,
      "loss": 0.289,
      "step": 1494
    },
    {
      "epoch": 20.2,
      "learning_rate": 0.00029859999999999994,
      "loss": 0.3008,
      "step": 1495
    },
    {
      "epoch": 20.22,
      "learning_rate": 0.0002988,
      "loss": 0.445,
      "step": 1496
    },
    {
      "epoch": 20.23,
      "learning_rate": 0.000299,
      "loss": 0.2335,
      "step": 1497
    },
    {
      "epoch": 20.24,
      "learning_rate": 0.00029919999999999995,
      "loss": 0.2395,
      "step": 1498
    },
    {
      "epoch": 20.26,
      "learning_rate": 0.00029939999999999996,
      "loss": 0.3997,
      "step": 1499
    },
    {
      "epoch": 20.27,
      "learning_rate": 0.00029959999999999996,
      "loss": 0.5179,
      "step": 1500
    },
    {
      "epoch": 20.28,
      "learning_rate": 0.00029979999999999997,
      "loss": 0.2444,
      "step": 1501
    },
    {
      "epoch": 20.3,
      "learning_rate": 0.0003,
      "loss": 0.4619,
      "step": 1502
    },
    {
      "epoch": 20.31,
      "learning_rate": 0.0002998636363636363,
      "loss": 0.224,
      "step": 1503
    },
    {
      "epoch": 20.32,
      "learning_rate": 0.0002997272727272727,
      "loss": 0.164,
      "step": 1504
    },
    {
      "epoch": 20.34,
      "learning_rate": 0.0002995909090909091,
      "loss": 0.3859,
      "step": 1505
    },
    {
      "epoch": 20.35,
      "learning_rate": 0.0002994545454545454,
      "loss": 0.3783,
      "step": 1506
    },
    {
      "epoch": 20.36,
      "learning_rate": 0.0002993181818181818,
      "loss": 0.4678,
      "step": 1507
    },
    {
      "epoch": 20.38,
      "learning_rate": 0.0002991818181818182,
      "loss": 0.2121,
      "step": 1508
    },
    {
      "epoch": 20.39,
      "learning_rate": 0.0002990454545454545,
      "loss": 0.526,
      "step": 1509
    },
    {
      "epoch": 20.41,
      "learning_rate": 0.0002989090909090909,
      "loss": 0.331,
      "step": 1510
    },
    {
      "epoch": 20.42,
      "learning_rate": 0.0002987727272727272,
      "loss": 0.3109,
      "step": 1511
    },
    {
      "epoch": 20.43,
      "learning_rate": 0.0002986363636363636,
      "loss": 0.5405,
      "step": 1512
    },
    {
      "epoch": 20.45,
      "learning_rate": 0.0002985,
      "loss": 0.271,
      "step": 1513
    },
    {
      "epoch": 20.46,
      "learning_rate": 0.0002983636363636363,
      "loss": 0.2466,
      "step": 1514
    },
    {
      "epoch": 20.47,
      "learning_rate": 0.0002982272727272727,
      "loss": 0.3996,
      "step": 1515
    },
    {
      "epoch": 20.49,
      "learning_rate": 0.0002980909090909091,
      "loss": 0.389,
      "step": 1516
    },
    {
      "epoch": 20.5,
      "learning_rate": 0.0002979545454545454,
      "loss": 0.4764,
      "step": 1517
    },
    {
      "epoch": 20.51,
      "learning_rate": 0.00029781818181818175,
      "loss": 0.2163,
      "step": 1518
    },
    {
      "epoch": 20.53,
      "learning_rate": 0.0002976818181818182,
      "loss": 0.2643,
      "step": 1519
    },
    {
      "epoch": 20.54,
      "learning_rate": 0.0002975454545454545,
      "loss": 0.3524,
      "step": 1520
    },
    {
      "epoch": 20.55,
      "learning_rate": 0.00029740909090909085,
      "loss": 0.2639,
      "step": 1521
    },
    {
      "epoch": 20.57,
      "learning_rate": 0.00029727272727272724,
      "loss": 0.3574,
      "step": 1522
    },
    {
      "epoch": 20.58,
      "learning_rate": 0.0002971363636363636,
      "loss": 0.1491,
      "step": 1523
    },
    {
      "epoch": 20.59,
      "learning_rate": 0.00029699999999999996,
      "loss": 0.2792,
      "step": 1524
    },
    {
      "epoch": 20.61,
      "learning_rate": 0.00029686363636363634,
      "loss": 0.2062,
      "step": 1525
    },
    {
      "epoch": 20.62,
      "learning_rate": 0.0002967272727272727,
      "loss": 0.141,
      "step": 1526
    },
    {
      "epoch": 20.64,
      "learning_rate": 0.00029659090909090906,
      "loss": 0.1464,
      "step": 1527
    },
    {
      "epoch": 20.65,
      "learning_rate": 0.00029645454545454544,
      "loss": 0.4313,
      "step": 1528
    },
    {
      "epoch": 20.66,
      "learning_rate": 0.00029631818181818177,
      "loss": 0.2937,
      "step": 1529
    },
    {
      "epoch": 20.68,
      "learning_rate": 0.00029618181818181816,
      "loss": 0.3701,
      "step": 1530
    },
    {
      "epoch": 20.69,
      "learning_rate": 0.00029604545454545454,
      "loss": 0.4596,
      "step": 1531
    },
    {
      "epoch": 20.7,
      "learning_rate": 0.00029590909090909087,
      "loss": 0.3119,
      "step": 1532
    },
    {
      "epoch": 20.72,
      "learning_rate": 0.00029577272727272726,
      "loss": 0.1145,
      "step": 1533
    },
    {
      "epoch": 20.73,
      "learning_rate": 0.00029563636363636364,
      "loss": 0.4607,
      "step": 1534
    },
    {
      "epoch": 20.74,
      "learning_rate": 0.00029549999999999997,
      "loss": 0.3676,
      "step": 1535
    },
    {
      "epoch": 20.76,
      "learning_rate": 0.0002953636363636363,
      "loss": 0.3686,
      "step": 1536
    },
    {
      "epoch": 20.77,
      "learning_rate": 0.0002952272727272727,
      "loss": 0.3785,
      "step": 1537
    },
    {
      "epoch": 20.78,
      "learning_rate": 0.0002950909090909091,
      "loss": 0.4068,
      "step": 1538
    },
    {
      "epoch": 20.8,
      "learning_rate": 0.0002949545454545454,
      "loss": 0.7279,
      "step": 1539
    },
    {
      "epoch": 20.81,
      "learning_rate": 0.0002948181818181818,
      "loss": 0.3187,
      "step": 1540
    },
    {
      "epoch": 20.82,
      "learning_rate": 0.0002946818181818182,
      "loss": 0.162,
      "step": 1541
    },
    {
      "epoch": 20.84,
      "learning_rate": 0.0002945454545454545,
      "loss": 0.1778,
      "step": 1542
    },
    {
      "epoch": 20.85,
      "learning_rate": 0.0002944090909090909,
      "loss": 0.257,
      "step": 1543
    },
    {
      "epoch": 20.86,
      "learning_rate": 0.0002942727272727273,
      "loss": 0.1947,
      "step": 1544
    },
    {
      "epoch": 20.88,
      "learning_rate": 0.0002941363636363636,
      "loss": 0.5775,
      "step": 1545
    },
    {
      "epoch": 20.89,
      "learning_rate": 0.000294,
      "loss": 0.3027,
      "step": 1546
    },
    {
      "epoch": 20.91,
      "learning_rate": 0.0002938636363636363,
      "loss": 0.2331,
      "step": 1547
    },
    {
      "epoch": 20.92,
      "learning_rate": 0.0002937272727272727,
      "loss": 0.4825,
      "step": 1548
    },
    {
      "epoch": 20.93,
      "learning_rate": 0.0002935909090909091,
      "loss": 0.3809,
      "step": 1549
    },
    {
      "epoch": 20.95,
      "learning_rate": 0.0002934545454545454,
      "loss": 0.245,
      "step": 1550
    },
    {
      "epoch": 20.96,
      "learning_rate": 0.0002933181818181818,
      "loss": 0.2066,
      "step": 1551
    },
    {
      "epoch": 20.97,
      "learning_rate": 0.00029318181818181814,
      "loss": 0.2879,
      "step": 1552
    },
    {
      "epoch": 20.99,
      "learning_rate": 0.0002930454545454545,
      "loss": 0.5975,
      "step": 1553
    },
    {
      "epoch": 21.0,
      "learning_rate": 0.00029290909090909085,
      "loss": 0.0093,
      "step": 1554
    },
    {
      "epoch": 21.01,
      "learning_rate": 0.00029277272727272724,
      "loss": 0.4224,
      "step": 1555
    },
    {
      "epoch": 21.03,
      "learning_rate": 0.0002926363636363636,
      "loss": 0.197,
      "step": 1556
    },
    {
      "epoch": 21.04,
      "learning_rate": 0.00029249999999999995,
      "loss": 0.0739,
      "step": 1557
    },
    {
      "epoch": 21.05,
      "learning_rate": 0.00029236363636363634,
      "loss": 0.1717,
      "step": 1558
    },
    {
      "epoch": 21.07,
      "learning_rate": 0.0002922272727272727,
      "loss": 0.2575,
      "step": 1559
    },
    {
      "epoch": 21.08,
      "learning_rate": 0.00029209090909090905,
      "loss": 0.476,
      "step": 1560
    },
    {
      "epoch": 21.09,
      "learning_rate": 0.00029195454545454544,
      "loss": 0.17,
      "step": 1561
    },
    {
      "epoch": 21.11,
      "learning_rate": 0.0002918181818181818,
      "loss": 0.2702,
      "step": 1562
    },
    {
      "epoch": 21.12,
      "learning_rate": 0.00029168181818181816,
      "loss": 0.1037,
      "step": 1563
    },
    {
      "epoch": 21.14,
      "learning_rate": 0.00029154545454545454,
      "loss": 0.1993,
      "step": 1564
    },
    {
      "epoch": 21.15,
      "learning_rate": 0.00029140909090909087,
      "loss": 0.4193,
      "step": 1565
    },
    {
      "epoch": 21.16,
      "learning_rate": 0.00029127272727272726,
      "loss": 0.2972,
      "step": 1566
    },
    {
      "epoch": 21.18,
      "learning_rate": 0.0002911363636363636,
      "loss": 0.0503,
      "step": 1567
    },
    {
      "epoch": 21.19,
      "learning_rate": 0.00029099999999999997,
      "loss": 0.3607,
      "step": 1568
    },
    {
      "epoch": 21.2,
      "learning_rate": 0.00029086363636363636,
      "loss": 0.2318,
      "step": 1569
    },
    {
      "epoch": 21.22,
      "learning_rate": 0.0002907272727272727,
      "loss": 0.1534,
      "step": 1570
    },
    {
      "epoch": 21.23,
      "learning_rate": 0.00029059090909090907,
      "loss": 0.1734,
      "step": 1571
    },
    {
      "epoch": 21.24,
      "learning_rate": 0.0002904545454545454,
      "loss": 0.2455,
      "step": 1572
    },
    {
      "epoch": 21.26,
      "learning_rate": 0.0002903181818181818,
      "loss": 0.3929,
      "step": 1573
    },
    {
      "epoch": 21.27,
      "learning_rate": 0.0002901818181818182,
      "loss": 0.5712,
      "step": 1574
    },
    {
      "epoch": 21.28,
      "learning_rate": 0.0002900454545454545,
      "loss": 0.3195,
      "step": 1575
    },
    {
      "epoch": 21.3,
      "learning_rate": 0.0002899090909090909,
      "loss": 0.2438,
      "step": 1576
    },
    {
      "epoch": 21.31,
      "learning_rate": 0.0002897727272727273,
      "loss": 0.3007,
      "step": 1577
    },
    {
      "epoch": 21.32,
      "learning_rate": 0.0002896363636363636,
      "loss": 0.3912,
      "step": 1578
    },
    {
      "epoch": 21.34,
      "learning_rate": 0.0002895,
      "loss": 0.1686,
      "step": 1579
    },
    {
      "epoch": 21.35,
      "learning_rate": 0.0002893636363636364,
      "loss": 0.3925,
      "step": 1580
    },
    {
      "epoch": 21.36,
      "learning_rate": 0.0002892272727272727,
      "loss": 0.2356,
      "step": 1581
    },
    {
      "epoch": 21.38,
      "learning_rate": 0.00028909090909090904,
      "loss": 0.3012,
      "step": 1582
    },
    {
      "epoch": 21.39,
      "learning_rate": 0.0002889545454545454,
      "loss": 0.1322,
      "step": 1583
    },
    {
      "epoch": 21.41,
      "learning_rate": 0.0002888181818181818,
      "loss": 0.3258,
      "step": 1584
    },
    {
      "epoch": 21.42,
      "learning_rate": 0.00028868181818181814,
      "loss": 0.4799,
      "step": 1585
    },
    {
      "epoch": 21.43,
      "learning_rate": 0.0002885454545454545,
      "loss": 0.2358,
      "step": 1586
    },
    {
      "epoch": 21.45,
      "learning_rate": 0.00028840909090909085,
      "loss": 0.0452,
      "step": 1587
    },
    {
      "epoch": 21.46,
      "learning_rate": 0.00028827272727272724,
      "loss": 0.2533,
      "step": 1588
    },
    {
      "epoch": 21.47,
      "learning_rate": 0.0002881363636363636,
      "loss": 0.0345,
      "step": 1589
    },
    {
      "epoch": 21.49,
      "learning_rate": 0.00028799999999999995,
      "loss": 0.0819,
      "step": 1590
    },
    {
      "epoch": 21.5,
      "learning_rate": 0.00028786363636363634,
      "loss": 0.3334,
      "step": 1591
    },
    {
      "epoch": 21.51,
      "learning_rate": 0.0002877272727272727,
      "loss": 0.1502,
      "step": 1592
    },
    {
      "epoch": 21.53,
      "learning_rate": 0.00028759090909090905,
      "loss": 0.3242,
      "step": 1593
    },
    {
      "epoch": 21.54,
      "learning_rate": 0.00028745454545454544,
      "loss": 0.1348,
      "step": 1594
    },
    {
      "epoch": 21.55,
      "learning_rate": 0.0002873181818181818,
      "loss": 0.1893,
      "step": 1595
    },
    {
      "epoch": 21.57,
      "learning_rate": 0.00028718181818181815,
      "loss": 0.1705,
      "step": 1596
    },
    {
      "epoch": 21.58,
      "learning_rate": 0.0002870454545454545,
      "loss": 0.2497,
      "step": 1597
    },
    {
      "epoch": 21.59,
      "learning_rate": 0.0002869090909090909,
      "loss": 0.3114,
      "step": 1598
    },
    {
      "epoch": 21.61,
      "learning_rate": 0.00028677272727272726,
      "loss": 0.3253,
      "step": 1599
    },
    {
      "epoch": 21.62,
      "learning_rate": 0.0002866363636363636,
      "loss": 0.4198,
      "step": 1600
    },
    {
      "epoch": 21.62,
      "eval_accuracy": 0.8920731707317073,
      "eval_f1": 0.8927475772424293,
      "eval_loss": 0.5161228179931641,
      "eval_runtime": 126.2965,
      "eval_samples_per_second": 12.985,
      "eval_steps_per_second": 0.816,
      "step": 1600
    },
    {
      "epoch": 21.64,
      "learning_rate": 0.00028649999999999997,
      "loss": 0.3042,
      "step": 1601
    },
    {
      "epoch": 21.65,
      "learning_rate": 0.00028636363636363636,
      "loss": 0.2985,
      "step": 1602
    },
    {
      "epoch": 21.66,
      "learning_rate": 0.0002862272727272727,
      "loss": 0.4288,
      "step": 1603
    },
    {
      "epoch": 21.68,
      "learning_rate": 0.00028609090909090907,
      "loss": 0.3174,
      "step": 1604
    },
    {
      "epoch": 21.69,
      "learning_rate": 0.0002859545454545454,
      "loss": 0.2812,
      "step": 1605
    },
    {
      "epoch": 21.7,
      "learning_rate": 0.0002858181818181818,
      "loss": 0.2943,
      "step": 1606
    },
    {
      "epoch": 21.72,
      "learning_rate": 0.00028568181818181817,
      "loss": 0.3087,
      "step": 1607
    },
    {
      "epoch": 21.73,
      "learning_rate": 0.0002855454545454545,
      "loss": 0.2508,
      "step": 1608
    },
    {
      "epoch": 21.74,
      "learning_rate": 0.0002854090909090909,
      "loss": 0.3396,
      "step": 1609
    },
    {
      "epoch": 21.76,
      "learning_rate": 0.00028527272727272727,
      "loss": 0.2491,
      "step": 1610
    },
    {
      "epoch": 21.77,
      "learning_rate": 0.0002851363636363636,
      "loss": 0.1654,
      "step": 1611
    },
    {
      "epoch": 21.78,
      "learning_rate": 0.000285,
      "loss": 0.2094,
      "step": 1612
    },
    {
      "epoch": 21.8,
      "learning_rate": 0.0002848636363636364,
      "loss": 0.2235,
      "step": 1613
    },
    {
      "epoch": 21.81,
      "learning_rate": 0.0002847272727272727,
      "loss": 0.4307,
      "step": 1614
    },
    {
      "epoch": 21.82,
      "learning_rate": 0.00028459090909090904,
      "loss": 0.2795,
      "step": 1615
    },
    {
      "epoch": 21.84,
      "learning_rate": 0.0002844545454545454,
      "loss": 0.4182,
      "step": 1616
    },
    {
      "epoch": 21.85,
      "learning_rate": 0.0002843181818181818,
      "loss": 0.2589,
      "step": 1617
    },
    {
      "epoch": 21.86,
      "learning_rate": 0.00028418181818181814,
      "loss": 0.3623,
      "step": 1618
    },
    {
      "epoch": 21.88,
      "learning_rate": 0.0002840454545454545,
      "loss": 0.0805,
      "step": 1619
    },
    {
      "epoch": 21.89,
      "learning_rate": 0.0002839090909090909,
      "loss": 0.3676,
      "step": 1620
    },
    {
      "epoch": 21.91,
      "learning_rate": 0.00028377272727272724,
      "loss": 0.1785,
      "step": 1621
    },
    {
      "epoch": 21.92,
      "learning_rate": 0.0002836363636363636,
      "loss": 0.1595,
      "step": 1622
    },
    {
      "epoch": 21.93,
      "learning_rate": 0.00028349999999999995,
      "loss": 0.1825,
      "step": 1623
    },
    {
      "epoch": 21.95,
      "learning_rate": 0.00028336363636363634,
      "loss": 0.2676,
      "step": 1624
    },
    {
      "epoch": 21.96,
      "learning_rate": 0.0002832272727272727,
      "loss": 0.2298,
      "step": 1625
    },
    {
      "epoch": 21.97,
      "learning_rate": 0.00028309090909090905,
      "loss": 0.198,
      "step": 1626
    },
    {
      "epoch": 21.99,
      "learning_rate": 0.00028295454545454544,
      "loss": 0.273,
      "step": 1627
    },
    {
      "epoch": 22.0,
      "learning_rate": 0.0002828181818181818,
      "loss": 0.0056,
      "step": 1628
    },
    {
      "epoch": 22.01,
      "learning_rate": 0.00028268181818181815,
      "loss": 0.1965,
      "step": 1629
    },
    {
      "epoch": 22.03,
      "learning_rate": 0.0002825454545454545,
      "loss": 0.0628,
      "step": 1630
    },
    {
      "epoch": 22.04,
      "learning_rate": 0.00028240909090909087,
      "loss": 0.1768,
      "step": 1631
    },
    {
      "epoch": 22.05,
      "learning_rate": 0.00028227272727272725,
      "loss": 0.3128,
      "step": 1632
    },
    {
      "epoch": 22.07,
      "learning_rate": 0.0002821363636363636,
      "loss": 0.4677,
      "step": 1633
    },
    {
      "epoch": 22.08,
      "learning_rate": 0.00028199999999999997,
      "loss": 0.0113,
      "step": 1634
    },
    {
      "epoch": 22.09,
      "learning_rate": 0.00028186363636363635,
      "loss": 0.1787,
      "step": 1635
    },
    {
      "epoch": 22.11,
      "learning_rate": 0.0002817272727272727,
      "loss": 0.1061,
      "step": 1636
    },
    {
      "epoch": 22.12,
      "learning_rate": 0.00028159090909090907,
      "loss": 0.0728,
      "step": 1637
    },
    {
      "epoch": 22.14,
      "learning_rate": 0.00028145454545454546,
      "loss": 0.28,
      "step": 1638
    },
    {
      "epoch": 22.15,
      "learning_rate": 0.0002813181818181818,
      "loss": 0.2459,
      "step": 1639
    },
    {
      "epoch": 22.16,
      "learning_rate": 0.00028118181818181817,
      "loss": 0.3317,
      "step": 1640
    },
    {
      "epoch": 22.18,
      "learning_rate": 0.0002810454545454545,
      "loss": 0.2559,
      "step": 1641
    },
    {
      "epoch": 22.19,
      "learning_rate": 0.0002809090909090909,
      "loss": 0.1789,
      "step": 1642
    },
    {
      "epoch": 22.2,
      "learning_rate": 0.00028077272727272727,
      "loss": 0.3074,
      "step": 1643
    },
    {
      "epoch": 22.22,
      "learning_rate": 0.0002806363636363636,
      "loss": 0.1215,
      "step": 1644
    },
    {
      "epoch": 22.23,
      "learning_rate": 0.0002805,
      "loss": 0.1202,
      "step": 1645
    },
    {
      "epoch": 22.24,
      "learning_rate": 0.00028036363636363637,
      "loss": 0.2177,
      "step": 1646
    },
    {
      "epoch": 22.26,
      "learning_rate": 0.0002802272727272727,
      "loss": 0.084,
      "step": 1647
    },
    {
      "epoch": 22.27,
      "learning_rate": 0.00028009090909090903,
      "loss": 0.0748,
      "step": 1648
    },
    {
      "epoch": 22.28,
      "learning_rate": 0.0002799545454545454,
      "loss": 0.0977,
      "step": 1649
    },
    {
      "epoch": 22.3,
      "learning_rate": 0.0002798181818181818,
      "loss": 0.1012,
      "step": 1650
    },
    {
      "epoch": 22.31,
      "learning_rate": 0.00027968181818181813,
      "loss": 0.2113,
      "step": 1651
    },
    {
      "epoch": 22.32,
      "learning_rate": 0.0002795454545454545,
      "loss": 0.1774,
      "step": 1652
    },
    {
      "epoch": 22.34,
      "learning_rate": 0.0002794090909090909,
      "loss": 0.137,
      "step": 1653
    },
    {
      "epoch": 22.35,
      "learning_rate": 0.00027927272727272724,
      "loss": 0.2654,
      "step": 1654
    },
    {
      "epoch": 22.36,
      "learning_rate": 0.0002791363636363636,
      "loss": 0.3057,
      "step": 1655
    },
    {
      "epoch": 22.38,
      "learning_rate": 0.000279,
      "loss": 0.178,
      "step": 1656
    },
    {
      "epoch": 22.39,
      "learning_rate": 0.00027886363636363634,
      "loss": 0.1978,
      "step": 1657
    },
    {
      "epoch": 22.41,
      "learning_rate": 0.0002787272727272727,
      "loss": 0.1352,
      "step": 1658
    },
    {
      "epoch": 22.42,
      "learning_rate": 0.00027859090909090905,
      "loss": 0.1218,
      "step": 1659
    },
    {
      "epoch": 22.43,
      "learning_rate": 0.00027845454545454544,
      "loss": 0.3554,
      "step": 1660
    },
    {
      "epoch": 22.45,
      "learning_rate": 0.0002783181818181818,
      "loss": 0.33,
      "step": 1661
    },
    {
      "epoch": 22.46,
      "learning_rate": 0.00027818181818181815,
      "loss": 0.2422,
      "step": 1662
    },
    {
      "epoch": 22.47,
      "learning_rate": 0.0002780454545454545,
      "loss": 0.0491,
      "step": 1663
    },
    {
      "epoch": 22.49,
      "learning_rate": 0.00027790909090909087,
      "loss": 0.2101,
      "step": 1664
    },
    {
      "epoch": 22.5,
      "learning_rate": 0.00027777272727272725,
      "loss": 0.0659,
      "step": 1665
    },
    {
      "epoch": 22.51,
      "learning_rate": 0.0002776363636363636,
      "loss": 0.1878,
      "step": 1666
    },
    {
      "epoch": 22.53,
      "learning_rate": 0.00027749999999999997,
      "loss": 0.3798,
      "step": 1667
    },
    {
      "epoch": 22.54,
      "learning_rate": 0.00027736363636363635,
      "loss": 0.4378,
      "step": 1668
    },
    {
      "epoch": 22.55,
      "learning_rate": 0.0002772272727272727,
      "loss": 0.4857,
      "step": 1669
    },
    {
      "epoch": 22.57,
      "learning_rate": 0.00027709090909090907,
      "loss": 0.1317,
      "step": 1670
    },
    {
      "epoch": 22.58,
      "learning_rate": 0.00027695454545454545,
      "loss": 0.1876,
      "step": 1671
    },
    {
      "epoch": 22.59,
      "learning_rate": 0.0002768181818181818,
      "loss": 0.3141,
      "step": 1672
    },
    {
      "epoch": 22.61,
      "learning_rate": 0.00027668181818181817,
      "loss": 0.1698,
      "step": 1673
    },
    {
      "epoch": 22.62,
      "learning_rate": 0.00027654545454545456,
      "loss": 0.1952,
      "step": 1674
    },
    {
      "epoch": 22.64,
      "learning_rate": 0.0002764090909090909,
      "loss": 0.2714,
      "step": 1675
    },
    {
      "epoch": 22.65,
      "learning_rate": 0.00027627272727272727,
      "loss": 0.2635,
      "step": 1676
    },
    {
      "epoch": 22.66,
      "learning_rate": 0.0002761363636363636,
      "loss": 0.3304,
      "step": 1677
    },
    {
      "epoch": 22.68,
      "learning_rate": 0.000276,
      "loss": 0.2477,
      "step": 1678
    },
    {
      "epoch": 22.69,
      "learning_rate": 0.0002758636363636363,
      "loss": 0.3064,
      "step": 1679
    },
    {
      "epoch": 22.7,
      "learning_rate": 0.0002757272727272727,
      "loss": 0.2813,
      "step": 1680
    },
    {
      "epoch": 22.72,
      "learning_rate": 0.00027559090909090903,
      "loss": 0.2512,
      "step": 1681
    },
    {
      "epoch": 22.73,
      "learning_rate": 0.0002754545454545454,
      "loss": 0.2418,
      "step": 1682
    },
    {
      "epoch": 22.74,
      "learning_rate": 0.0002753181818181818,
      "loss": 0.2615,
      "step": 1683
    },
    {
      "epoch": 22.76,
      "learning_rate": 0.00027518181818181813,
      "loss": 0.2313,
      "step": 1684
    },
    {
      "epoch": 22.77,
      "learning_rate": 0.0002750454545454545,
      "loss": 0.092,
      "step": 1685
    },
    {
      "epoch": 22.78,
      "learning_rate": 0.0002749090909090909,
      "loss": 0.1665,
      "step": 1686
    },
    {
      "epoch": 22.8,
      "learning_rate": 0.00027477272727272723,
      "loss": 0.1313,
      "step": 1687
    },
    {
      "epoch": 22.81,
      "learning_rate": 0.0002746363636363636,
      "loss": 0.2268,
      "step": 1688
    },
    {
      "epoch": 22.82,
      "learning_rate": 0.0002745,
      "loss": 0.0834,
      "step": 1689
    },
    {
      "epoch": 22.84,
      "learning_rate": 0.00027436363636363634,
      "loss": 0.2072,
      "step": 1690
    },
    {
      "epoch": 22.85,
      "learning_rate": 0.0002742272727272727,
      "loss": 0.109,
      "step": 1691
    },
    {
      "epoch": 22.86,
      "learning_rate": 0.0002740909090909091,
      "loss": 0.5027,
      "step": 1692
    },
    {
      "epoch": 22.88,
      "learning_rate": 0.00027395454545454544,
      "loss": 0.1952,
      "step": 1693
    },
    {
      "epoch": 22.89,
      "learning_rate": 0.00027381818181818177,
      "loss": 0.2268,
      "step": 1694
    },
    {
      "epoch": 22.91,
      "learning_rate": 0.00027368181818181815,
      "loss": 0.3494,
      "step": 1695
    },
    {
      "epoch": 22.92,
      "learning_rate": 0.00027354545454545454,
      "loss": 0.2211,
      "step": 1696
    },
    {
      "epoch": 22.93,
      "learning_rate": 0.00027340909090909087,
      "loss": 0.3437,
      "step": 1697
    },
    {
      "epoch": 22.95,
      "learning_rate": 0.00027327272727272725,
      "loss": 0.134,
      "step": 1698
    },
    {
      "epoch": 22.96,
      "learning_rate": 0.0002731363636363636,
      "loss": 0.3108,
      "step": 1699
    },
    {
      "epoch": 22.97,
      "learning_rate": 0.00027299999999999997,
      "loss": 0.0967,
      "step": 1700
    },
    {
      "epoch": 22.99,
      "learning_rate": 0.00027286363636363635,
      "loss": 0.2448,
      "step": 1701
    },
    {
      "epoch": 23.0,
      "learning_rate": 0.0002727272727272727,
      "loss": 0.0185,
      "step": 1702
    },
    {
      "epoch": 23.01,
      "learning_rate": 0.00027259090909090907,
      "loss": 0.2684,
      "step": 1703
    },
    {
      "epoch": 23.03,
      "learning_rate": 0.00027245454545454545,
      "loss": 0.154,
      "step": 1704
    },
    {
      "epoch": 23.04,
      "learning_rate": 0.0002723181818181818,
      "loss": 0.2449,
      "step": 1705
    },
    {
      "epoch": 23.05,
      "learning_rate": 0.00027218181818181817,
      "loss": 0.109,
      "step": 1706
    },
    {
      "epoch": 23.07,
      "learning_rate": 0.00027204545454545455,
      "loss": 0.2459,
      "step": 1707
    },
    {
      "epoch": 23.08,
      "learning_rate": 0.0002719090909090909,
      "loss": 0.1213,
      "step": 1708
    },
    {
      "epoch": 23.09,
      "learning_rate": 0.0002717727272727272,
      "loss": 0.155,
      "step": 1709
    },
    {
      "epoch": 23.11,
      "learning_rate": 0.0002716363636363636,
      "loss": 0.0256,
      "step": 1710
    },
    {
      "epoch": 23.12,
      "learning_rate": 0.0002715,
      "loss": 0.1641,
      "step": 1711
    },
    {
      "epoch": 23.14,
      "learning_rate": 0.0002713636363636363,
      "loss": 0.2562,
      "step": 1712
    },
    {
      "epoch": 23.15,
      "learning_rate": 0.0002712272727272727,
      "loss": 0.1885,
      "step": 1713
    },
    {
      "epoch": 23.16,
      "learning_rate": 0.0002710909090909091,
      "loss": 0.3949,
      "step": 1714
    },
    {
      "epoch": 23.18,
      "learning_rate": 0.0002709545454545454,
      "loss": 0.1054,
      "step": 1715
    },
    {
      "epoch": 23.19,
      "learning_rate": 0.0002708181818181818,
      "loss": 0.1519,
      "step": 1716
    },
    {
      "epoch": 23.2,
      "learning_rate": 0.00027068181818181813,
      "loss": 0.0803,
      "step": 1717
    },
    {
      "epoch": 23.22,
      "learning_rate": 0.0002705454545454545,
      "loss": 0.0906,
      "step": 1718
    },
    {
      "epoch": 23.23,
      "learning_rate": 0.0002704090909090909,
      "loss": 0.2292,
      "step": 1719
    },
    {
      "epoch": 23.24,
      "learning_rate": 0.00027027272727272723,
      "loss": 0.4874,
      "step": 1720
    },
    {
      "epoch": 23.26,
      "learning_rate": 0.0002701363636363636,
      "loss": 0.0722,
      "step": 1721
    },
    {
      "epoch": 23.27,
      "learning_rate": 0.00027,
      "loss": 0.3082,
      "step": 1722
    },
    {
      "epoch": 23.28,
      "learning_rate": 0.00026986363636363633,
      "loss": 0.2147,
      "step": 1723
    },
    {
      "epoch": 23.3,
      "learning_rate": 0.00026972727272727266,
      "loss": 0.317,
      "step": 1724
    },
    {
      "epoch": 23.31,
      "learning_rate": 0.0002695909090909091,
      "loss": 0.0885,
      "step": 1725
    },
    {
      "epoch": 23.32,
      "learning_rate": 0.00026945454545454543,
      "loss": 0.2791,
      "step": 1726
    },
    {
      "epoch": 23.34,
      "learning_rate": 0.00026931818181818177,
      "loss": 0.085,
      "step": 1727
    },
    {
      "epoch": 23.35,
      "learning_rate": 0.00026918181818181815,
      "loss": 0.2282,
      "step": 1728
    },
    {
      "epoch": 23.36,
      "learning_rate": 0.00026904545454545454,
      "loss": 0.0594,
      "step": 1729
    },
    {
      "epoch": 23.38,
      "learning_rate": 0.00026890909090909087,
      "loss": 0.2585,
      "step": 1730
    },
    {
      "epoch": 23.39,
      "learning_rate": 0.00026877272727272725,
      "loss": 0.2417,
      "step": 1731
    },
    {
      "epoch": 23.41,
      "learning_rate": 0.00026863636363636364,
      "loss": 0.2306,
      "step": 1732
    },
    {
      "epoch": 23.42,
      "learning_rate": 0.00026849999999999997,
      "loss": 0.1729,
      "step": 1733
    },
    {
      "epoch": 23.43,
      "learning_rate": 0.00026836363636363635,
      "loss": 0.1845,
      "step": 1734
    },
    {
      "epoch": 23.45,
      "learning_rate": 0.0002682272727272727,
      "loss": 0.3174,
      "step": 1735
    },
    {
      "epoch": 23.46,
      "learning_rate": 0.00026809090909090907,
      "loss": 0.2382,
      "step": 1736
    },
    {
      "epoch": 23.47,
      "learning_rate": 0.00026795454545454545,
      "loss": 0.3121,
      "step": 1737
    },
    {
      "epoch": 23.49,
      "learning_rate": 0.0002678181818181818,
      "loss": 0.1611,
      "step": 1738
    },
    {
      "epoch": 23.5,
      "learning_rate": 0.00026768181818181817,
      "loss": 0.1945,
      "step": 1739
    },
    {
      "epoch": 23.51,
      "learning_rate": 0.00026754545454545455,
      "loss": 0.2888,
      "step": 1740
    },
    {
      "epoch": 23.53,
      "learning_rate": 0.0002674090909090909,
      "loss": 0.2245,
      "step": 1741
    },
    {
      "epoch": 23.54,
      "learning_rate": 0.0002672727272727272,
      "loss": 0.1795,
      "step": 1742
    },
    {
      "epoch": 23.55,
      "learning_rate": 0.0002671363636363636,
      "loss": 0.1523,
      "step": 1743
    },
    {
      "epoch": 23.57,
      "learning_rate": 0.000267,
      "loss": 0.3644,
      "step": 1744
    },
    {
      "epoch": 23.58,
      "learning_rate": 0.0002668636363636363,
      "loss": 0.2604,
      "step": 1745
    },
    {
      "epoch": 23.59,
      "learning_rate": 0.0002667272727272727,
      "loss": 0.231,
      "step": 1746
    },
    {
      "epoch": 23.61,
      "learning_rate": 0.0002665909090909091,
      "loss": 0.1812,
      "step": 1747
    },
    {
      "epoch": 23.62,
      "learning_rate": 0.0002664545454545454,
      "loss": 0.1782,
      "step": 1748
    },
    {
      "epoch": 23.64,
      "learning_rate": 0.0002663181818181818,
      "loss": 0.1734,
      "step": 1749
    },
    {
      "epoch": 23.65,
      "learning_rate": 0.0002661818181818182,
      "loss": 0.1888,
      "step": 1750
    },
    {
      "epoch": 23.66,
      "learning_rate": 0.0002660454545454545,
      "loss": 0.0729,
      "step": 1751
    },
    {
      "epoch": 23.68,
      "learning_rate": 0.0002659090909090909,
      "loss": 0.3066,
      "step": 1752
    },
    {
      "epoch": 23.69,
      "learning_rate": 0.00026577272727272723,
      "loss": 0.1339,
      "step": 1753
    },
    {
      "epoch": 23.7,
      "learning_rate": 0.0002656363636363636,
      "loss": 0.1564,
      "step": 1754
    },
    {
      "epoch": 23.72,
      "learning_rate": 0.0002655,
      "loss": 0.2072,
      "step": 1755
    },
    {
      "epoch": 23.73,
      "learning_rate": 0.00026536363636363633,
      "loss": 0.0992,
      "step": 1756
    },
    {
      "epoch": 23.74,
      "learning_rate": 0.00026522727272727266,
      "loss": 0.324,
      "step": 1757
    },
    {
      "epoch": 23.76,
      "learning_rate": 0.00026509090909090905,
      "loss": 0.2179,
      "step": 1758
    },
    {
      "epoch": 23.77,
      "learning_rate": 0.00026495454545454543,
      "loss": 0.3787,
      "step": 1759
    },
    {
      "epoch": 23.78,
      "learning_rate": 0.00026481818181818176,
      "loss": 0.2366,
      "step": 1760
    },
    {
      "epoch": 23.8,
      "learning_rate": 0.00026468181818181815,
      "loss": 0.31,
      "step": 1761
    },
    {
      "epoch": 23.81,
      "learning_rate": 0.00026454545454545453,
      "loss": 0.1698,
      "step": 1762
    },
    {
      "epoch": 23.82,
      "learning_rate": 0.00026440909090909087,
      "loss": 0.1387,
      "step": 1763
    },
    {
      "epoch": 23.84,
      "learning_rate": 0.00026427272727272725,
      "loss": 0.2816,
      "step": 1764
    },
    {
      "epoch": 23.85,
      "learning_rate": 0.00026413636363636364,
      "loss": 0.0943,
      "step": 1765
    },
    {
      "epoch": 23.86,
      "learning_rate": 0.00026399999999999997,
      "loss": 0.2615,
      "step": 1766
    },
    {
      "epoch": 23.88,
      "learning_rate": 0.00026386363636363635,
      "loss": 0.118,
      "step": 1767
    },
    {
      "epoch": 23.89,
      "learning_rate": 0.00026372727272727274,
      "loss": 0.1107,
      "step": 1768
    },
    {
      "epoch": 23.91,
      "learning_rate": 0.00026359090909090907,
      "loss": 0.3253,
      "step": 1769
    },
    {
      "epoch": 23.92,
      "learning_rate": 0.00026345454545454545,
      "loss": 0.2935,
      "step": 1770
    },
    {
      "epoch": 23.93,
      "learning_rate": 0.0002633181818181818,
      "loss": 0.1319,
      "step": 1771
    },
    {
      "epoch": 23.95,
      "learning_rate": 0.00026318181818181817,
      "loss": 0.1092,
      "step": 1772
    },
    {
      "epoch": 23.96,
      "learning_rate": 0.00026304545454545455,
      "loss": 0.2581,
      "step": 1773
    },
    {
      "epoch": 23.97,
      "learning_rate": 0.0002629090909090909,
      "loss": 0.1175,
      "step": 1774
    },
    {
      "epoch": 23.99,
      "learning_rate": 0.0002627727272727272,
      "loss": 0.259,
      "step": 1775
    },
    {
      "epoch": 24.0,
      "learning_rate": 0.0002626363636363636,
      "loss": 0.1706,
      "step": 1776
    },
    {
      "epoch": 24.01,
      "learning_rate": 0.0002625,
      "loss": 0.288,
      "step": 1777
    },
    {
      "epoch": 24.03,
      "learning_rate": 0.0002623636363636363,
      "loss": 0.295,
      "step": 1778
    },
    {
      "epoch": 24.04,
      "learning_rate": 0.0002622272727272727,
      "loss": 0.3335,
      "step": 1779
    },
    {
      "epoch": 24.05,
      "learning_rate": 0.0002620909090909091,
      "loss": 0.1457,
      "step": 1780
    },
    {
      "epoch": 24.07,
      "learning_rate": 0.0002619545454545454,
      "loss": 0.1947,
      "step": 1781
    },
    {
      "epoch": 24.08,
      "learning_rate": 0.0002618181818181818,
      "loss": 0.194,
      "step": 1782
    },
    {
      "epoch": 24.09,
      "learning_rate": 0.0002616818181818182,
      "loss": 0.2611,
      "step": 1783
    },
    {
      "epoch": 24.11,
      "learning_rate": 0.0002615454545454545,
      "loss": 0.0556,
      "step": 1784
    },
    {
      "epoch": 24.12,
      "learning_rate": 0.0002614090909090909,
      "loss": 0.044,
      "step": 1785
    },
    {
      "epoch": 24.14,
      "learning_rate": 0.00026127272727272723,
      "loss": 0.1713,
      "step": 1786
    },
    {
      "epoch": 24.15,
      "learning_rate": 0.0002611363636363636,
      "loss": 0.1638,
      "step": 1787
    },
    {
      "epoch": 24.16,
      "learning_rate": 0.000261,
      "loss": 0.2565,
      "step": 1788
    },
    {
      "epoch": 24.18,
      "learning_rate": 0.00026086363636363633,
      "loss": 0.13,
      "step": 1789
    },
    {
      "epoch": 24.19,
      "learning_rate": 0.0002607272727272727,
      "loss": 0.1342,
      "step": 1790
    },
    {
      "epoch": 24.2,
      "learning_rate": 0.00026059090909090905,
      "loss": 0.1664,
      "step": 1791
    },
    {
      "epoch": 24.22,
      "learning_rate": 0.00026045454545454543,
      "loss": 0.1111,
      "step": 1792
    },
    {
      "epoch": 24.23,
      "learning_rate": 0.00026031818181818176,
      "loss": 0.0487,
      "step": 1793
    },
    {
      "epoch": 24.24,
      "learning_rate": 0.00026018181818181815,
      "loss": 0.0339,
      "step": 1794
    },
    {
      "epoch": 24.26,
      "learning_rate": 0.00026004545454545453,
      "loss": 0.2588,
      "step": 1795
    },
    {
      "epoch": 24.27,
      "learning_rate": 0.00025990909090909086,
      "loss": 0.0167,
      "step": 1796
    },
    {
      "epoch": 24.28,
      "learning_rate": 0.00025977272727272725,
      "loss": 0.0864,
      "step": 1797
    },
    {
      "epoch": 24.3,
      "learning_rate": 0.00025963636363636363,
      "loss": 0.0533,
      "step": 1798
    },
    {
      "epoch": 24.31,
      "learning_rate": 0.00025949999999999997,
      "loss": 0.1137,
      "step": 1799
    },
    {
      "epoch": 24.32,
      "learning_rate": 0.00025936363636363635,
      "loss": 0.11,
      "step": 1800
    },
    {
      "epoch": 24.32,
      "eval_accuracy": 0.8902439024390244,
      "eval_f1": 0.8922991745266968,
      "eval_loss": 0.44561275839805603,
      "eval_runtime": 125.7247,
      "eval_samples_per_second": 13.044,
      "eval_steps_per_second": 0.819,
      "step": 1800
    },
    {
      "epoch": 24.34,
      "learning_rate": 0.00025922727272727273,
      "loss": 0.4028,
      "step": 1801
    },
    {
      "epoch": 24.35,
      "learning_rate": 0.00025909090909090907,
      "loss": 0.262,
      "step": 1802
    },
    {
      "epoch": 24.36,
      "learning_rate": 0.00025895454545454545,
      "loss": 0.0625,
      "step": 1803
    },
    {
      "epoch": 24.38,
      "learning_rate": 0.0002588181818181818,
      "loss": 0.27,
      "step": 1804
    },
    {
      "epoch": 24.39,
      "learning_rate": 0.00025868181818181817,
      "loss": 0.239,
      "step": 1805
    },
    {
      "epoch": 24.41,
      "learning_rate": 0.0002585454545454545,
      "loss": 0.3534,
      "step": 1806
    },
    {
      "epoch": 24.42,
      "learning_rate": 0.0002584090909090909,
      "loss": 0.0038,
      "step": 1807
    },
    {
      "epoch": 24.43,
      "learning_rate": 0.00025827272727272727,
      "loss": 0.3399,
      "step": 1808
    },
    {
      "epoch": 24.45,
      "learning_rate": 0.0002581363636363636,
      "loss": 0.1625,
      "step": 1809
    },
    {
      "epoch": 24.46,
      "learning_rate": 0.000258,
      "loss": 0.2527,
      "step": 1810
    },
    {
      "epoch": 24.47,
      "learning_rate": 0.0002578636363636363,
      "loss": 0.0594,
      "step": 1811
    },
    {
      "epoch": 24.49,
      "learning_rate": 0.0002577272727272727,
      "loss": 0.3317,
      "step": 1812
    },
    {
      "epoch": 24.5,
      "learning_rate": 0.0002575909090909091,
      "loss": 0.2897,
      "step": 1813
    },
    {
      "epoch": 24.51,
      "learning_rate": 0.0002574545454545454,
      "loss": 0.0629,
      "step": 1814
    },
    {
      "epoch": 24.53,
      "learning_rate": 0.0002573181818181818,
      "loss": 0.2923,
      "step": 1815
    },
    {
      "epoch": 24.54,
      "learning_rate": 0.0002571818181818182,
      "loss": 0.1588,
      "step": 1816
    },
    {
      "epoch": 24.55,
      "learning_rate": 0.0002570454545454545,
      "loss": 0.3346,
      "step": 1817
    },
    {
      "epoch": 24.57,
      "learning_rate": 0.0002569090909090909,
      "loss": 0.1453,
      "step": 1818
    },
    {
      "epoch": 24.58,
      "learning_rate": 0.0002567727272727273,
      "loss": 0.3337,
      "step": 1819
    },
    {
      "epoch": 24.59,
      "learning_rate": 0.0002566363636363636,
      "loss": 0.0774,
      "step": 1820
    },
    {
      "epoch": 24.61,
      "learning_rate": 0.00025649999999999995,
      "loss": 0.298,
      "step": 1821
    },
    {
      "epoch": 24.62,
      "learning_rate": 0.00025636363636363633,
      "loss": 0.0667,
      "step": 1822
    },
    {
      "epoch": 24.64,
      "learning_rate": 0.0002562272727272727,
      "loss": 0.1372,
      "step": 1823
    },
    {
      "epoch": 24.65,
      "learning_rate": 0.00025609090909090905,
      "loss": 0.1455,
      "step": 1824
    },
    {
      "epoch": 24.66,
      "learning_rate": 0.00025595454545454543,
      "loss": 0.0909,
      "step": 1825
    },
    {
      "epoch": 24.68,
      "learning_rate": 0.0002558181818181818,
      "loss": 0.3342,
      "step": 1826
    },
    {
      "epoch": 24.69,
      "learning_rate": 0.00025568181818181815,
      "loss": 0.0812,
      "step": 1827
    },
    {
      "epoch": 24.7,
      "learning_rate": 0.00025554545454545453,
      "loss": 0.1847,
      "step": 1828
    },
    {
      "epoch": 24.72,
      "learning_rate": 0.00025540909090909086,
      "loss": 0.1712,
      "step": 1829
    },
    {
      "epoch": 24.73,
      "learning_rate": 0.00025527272727272725,
      "loss": 0.1051,
      "step": 1830
    },
    {
      "epoch": 24.74,
      "learning_rate": 0.00025513636363636363,
      "loss": 0.19,
      "step": 1831
    },
    {
      "epoch": 24.76,
      "learning_rate": 0.00025499999999999996,
      "loss": 0.0458,
      "step": 1832
    },
    {
      "epoch": 24.77,
      "learning_rate": 0.00025486363636363635,
      "loss": 0.1272,
      "step": 1833
    },
    {
      "epoch": 24.78,
      "learning_rate": 0.00025472727272727273,
      "loss": 0.0687,
      "step": 1834
    },
    {
      "epoch": 24.8,
      "learning_rate": 0.00025459090909090906,
      "loss": 0.2198,
      "step": 1835
    },
    {
      "epoch": 24.81,
      "learning_rate": 0.0002544545454545454,
      "loss": 0.1435,
      "step": 1836
    },
    {
      "epoch": 24.82,
      "learning_rate": 0.00025431818181818183,
      "loss": 0.2627,
      "step": 1837
    },
    {
      "epoch": 24.84,
      "learning_rate": 0.00025418181818181817,
      "loss": 0.151,
      "step": 1838
    },
    {
      "epoch": 24.85,
      "learning_rate": 0.0002540454545454545,
      "loss": 0.1113,
      "step": 1839
    },
    {
      "epoch": 24.86,
      "learning_rate": 0.0002539090909090909,
      "loss": 0.0626,
      "step": 1840
    },
    {
      "epoch": 24.88,
      "learning_rate": 0.00025377272727272727,
      "loss": 0.0508,
      "step": 1841
    },
    {
      "epoch": 24.89,
      "learning_rate": 0.0002536363636363636,
      "loss": 0.1027,
      "step": 1842
    },
    {
      "epoch": 24.91,
      "learning_rate": 0.0002535,
      "loss": 0.108,
      "step": 1843
    },
    {
      "epoch": 24.92,
      "learning_rate": 0.00025336363636363637,
      "loss": 0.3143,
      "step": 1844
    },
    {
      "epoch": 24.93,
      "learning_rate": 0.0002532272727272727,
      "loss": 0.1871,
      "step": 1845
    },
    {
      "epoch": 24.95,
      "learning_rate": 0.0002530909090909091,
      "loss": 0.101,
      "step": 1846
    },
    {
      "epoch": 24.96,
      "learning_rate": 0.0002529545454545454,
      "loss": 0.0933,
      "step": 1847
    },
    {
      "epoch": 24.97,
      "learning_rate": 0.0002528181818181818,
      "loss": 0.0516,
      "step": 1848
    },
    {
      "epoch": 24.99,
      "learning_rate": 0.0002526818181818182,
      "loss": 0.2128,
      "step": 1849
    },
    {
      "epoch": 25.0,
      "learning_rate": 0.0002525454545454545,
      "loss": 0.0077,
      "step": 1850
    },
    {
      "epoch": 25.01,
      "learning_rate": 0.0002524090909090909,
      "loss": 0.1135,
      "step": 1851
    },
    {
      "epoch": 25.03,
      "learning_rate": 0.0002522727272727273,
      "loss": 0.1197,
      "step": 1852
    },
    {
      "epoch": 25.04,
      "learning_rate": 0.0002521363636363636,
      "loss": 0.0363,
      "step": 1853
    },
    {
      "epoch": 25.05,
      "learning_rate": 0.00025199999999999995,
      "loss": 0.2009,
      "step": 1854
    },
    {
      "epoch": 25.07,
      "learning_rate": 0.00025186363636363633,
      "loss": 0.1142,
      "step": 1855
    },
    {
      "epoch": 25.08,
      "learning_rate": 0.0002517272727272727,
      "loss": 0.0806,
      "step": 1856
    },
    {
      "epoch": 25.09,
      "learning_rate": 0.00025159090909090905,
      "loss": 0.2615,
      "step": 1857
    },
    {
      "epoch": 25.11,
      "learning_rate": 0.00025145454545454543,
      "loss": 0.314,
      "step": 1858
    },
    {
      "epoch": 25.12,
      "learning_rate": 0.0002513181818181818,
      "loss": 0.343,
      "step": 1859
    },
    {
      "epoch": 25.14,
      "learning_rate": 0.00025118181818181815,
      "loss": 0.2554,
      "step": 1860
    },
    {
      "epoch": 25.15,
      "learning_rate": 0.00025104545454545453,
      "loss": 0.0406,
      "step": 1861
    },
    {
      "epoch": 25.16,
      "learning_rate": 0.00025090909090909086,
      "loss": 0.2055,
      "step": 1862
    },
    {
      "epoch": 25.18,
      "learning_rate": 0.00025077272727272725,
      "loss": 0.0505,
      "step": 1863
    },
    {
      "epoch": 25.19,
      "learning_rate": 0.00025063636363636363,
      "loss": 0.0705,
      "step": 1864
    },
    {
      "epoch": 25.2,
      "learning_rate": 0.00025049999999999996,
      "loss": 0.1113,
      "step": 1865
    },
    {
      "epoch": 25.22,
      "learning_rate": 0.00025036363636363635,
      "loss": 0.0884,
      "step": 1866
    },
    {
      "epoch": 25.23,
      "learning_rate": 0.00025022727272727273,
      "loss": 0.2396,
      "step": 1867
    },
    {
      "epoch": 25.24,
      "learning_rate": 0.00025009090909090906,
      "loss": 0.2443,
      "step": 1868
    },
    {
      "epoch": 25.26,
      "learning_rate": 0.0002499545454545454,
      "loss": 0.3722,
      "step": 1869
    },
    {
      "epoch": 25.27,
      "learning_rate": 0.0002498181818181818,
      "loss": 0.1946,
      "step": 1870
    },
    {
      "epoch": 25.28,
      "learning_rate": 0.00024968181818181816,
      "loss": 0.216,
      "step": 1871
    },
    {
      "epoch": 25.3,
      "learning_rate": 0.0002495454545454545,
      "loss": 0.0253,
      "step": 1872
    },
    {
      "epoch": 25.31,
      "learning_rate": 0.0002494090909090909,
      "loss": 0.1381,
      "step": 1873
    },
    {
      "epoch": 25.32,
      "learning_rate": 0.00024927272727272727,
      "loss": 0.1171,
      "step": 1874
    },
    {
      "epoch": 25.34,
      "learning_rate": 0.0002491363636363636,
      "loss": 0.0857,
      "step": 1875
    },
    {
      "epoch": 25.35,
      "learning_rate": 0.000249,
      "loss": 0.2689,
      "step": 1876
    },
    {
      "epoch": 25.36,
      "learning_rate": 0.00024886363636363637,
      "loss": 0.2315,
      "step": 1877
    },
    {
      "epoch": 25.38,
      "learning_rate": 0.0002487272727272727,
      "loss": 0.1348,
      "step": 1878
    },
    {
      "epoch": 25.39,
      "learning_rate": 0.0002485909090909091,
      "loss": 0.2247,
      "step": 1879
    },
    {
      "epoch": 25.41,
      "learning_rate": 0.0002484545454545454,
      "loss": 0.0571,
      "step": 1880
    },
    {
      "epoch": 25.42,
      "learning_rate": 0.0002483181818181818,
      "loss": 0.3023,
      "step": 1881
    },
    {
      "epoch": 25.43,
      "learning_rate": 0.0002481818181818182,
      "loss": 0.115,
      "step": 1882
    },
    {
      "epoch": 25.45,
      "learning_rate": 0.0002480454545454545,
      "loss": 0.0876,
      "step": 1883
    },
    {
      "epoch": 25.46,
      "learning_rate": 0.0002479090909090909,
      "loss": 0.0945,
      "step": 1884
    },
    {
      "epoch": 25.47,
      "learning_rate": 0.0002477727272727273,
      "loss": 0.2239,
      "step": 1885
    },
    {
      "epoch": 25.49,
      "learning_rate": 0.0002476363636363636,
      "loss": 0.2139,
      "step": 1886
    },
    {
      "epoch": 25.5,
      "learning_rate": 0.00024749999999999994,
      "loss": 0.3384,
      "step": 1887
    },
    {
      "epoch": 25.51,
      "learning_rate": 0.00024736363636363633,
      "loss": 0.1876,
      "step": 1888
    },
    {
      "epoch": 25.53,
      "learning_rate": 0.0002472272727272727,
      "loss": 0.1824,
      "step": 1889
    },
    {
      "epoch": 25.54,
      "learning_rate": 0.00024709090909090905,
      "loss": 0.2401,
      "step": 1890
    },
    {
      "epoch": 25.55,
      "learning_rate": 0.00024695454545454543,
      "loss": 0.1896,
      "step": 1891
    },
    {
      "epoch": 25.57,
      "learning_rate": 0.0002468181818181818,
      "loss": 0.2098,
      "step": 1892
    },
    {
      "epoch": 25.58,
      "learning_rate": 0.00024668181818181815,
      "loss": 0.4642,
      "step": 1893
    },
    {
      "epoch": 25.59,
      "learning_rate": 0.00024654545454545453,
      "loss": 0.175,
      "step": 1894
    },
    {
      "epoch": 25.61,
      "learning_rate": 0.0002464090909090909,
      "loss": 0.1565,
      "step": 1895
    },
    {
      "epoch": 25.62,
      "learning_rate": 0.00024627272727272725,
      "loss": 0.2635,
      "step": 1896
    },
    {
      "epoch": 25.64,
      "learning_rate": 0.00024613636363636363,
      "loss": 0.2436,
      "step": 1897
    },
    {
      "epoch": 25.65,
      "learning_rate": 0.00024599999999999996,
      "loss": 0.0712,
      "step": 1898
    },
    {
      "epoch": 25.66,
      "learning_rate": 0.00024586363636363635,
      "loss": 0.035,
      "step": 1899
    },
    {
      "epoch": 25.68,
      "learning_rate": 0.00024572727272727273,
      "loss": 0.0505,
      "step": 1900
    },
    {
      "epoch": 25.69,
      "learning_rate": 0.00024559090909090906,
      "loss": 0.1159,
      "step": 1901
    },
    {
      "epoch": 25.7,
      "learning_rate": 0.00024545454545454545,
      "loss": 0.0439,
      "step": 1902
    },
    {
      "epoch": 25.72,
      "learning_rate": 0.0002453181818181818,
      "loss": 0.0375,
      "step": 1903
    },
    {
      "epoch": 25.73,
      "learning_rate": 0.00024518181818181816,
      "loss": 0.0477,
      "step": 1904
    },
    {
      "epoch": 25.74,
      "learning_rate": 0.0002450454545454545,
      "loss": 0.1999,
      "step": 1905
    },
    {
      "epoch": 25.76,
      "learning_rate": 0.0002449090909090909,
      "loss": 0.1514,
      "step": 1906
    },
    {
      "epoch": 25.77,
      "learning_rate": 0.00024477272727272726,
      "loss": 0.2802,
      "step": 1907
    },
    {
      "epoch": 25.78,
      "learning_rate": 0.0002446363636363636,
      "loss": 0.1237,
      "step": 1908
    },
    {
      "epoch": 25.8,
      "learning_rate": 0.0002445,
      "loss": 0.2066,
      "step": 1909
    },
    {
      "epoch": 25.81,
      "learning_rate": 0.00024436363636363636,
      "loss": 0.0945,
      "step": 1910
    },
    {
      "epoch": 25.82,
      "learning_rate": 0.0002442272727272727,
      "loss": 0.1026,
      "step": 1911
    },
    {
      "epoch": 25.84,
      "learning_rate": 0.00024409090909090905,
      "loss": 0.235,
      "step": 1912
    },
    {
      "epoch": 25.85,
      "learning_rate": 0.00024395454545454544,
      "loss": 0.2406,
      "step": 1913
    },
    {
      "epoch": 25.86,
      "learning_rate": 0.0002438181818181818,
      "loss": 0.1285,
      "step": 1914
    },
    {
      "epoch": 25.88,
      "learning_rate": 0.00024368181818181815,
      "loss": 0.1139,
      "step": 1915
    },
    {
      "epoch": 25.89,
      "learning_rate": 0.0002435454545454545,
      "loss": 0.142,
      "step": 1916
    },
    {
      "epoch": 25.91,
      "learning_rate": 0.0002434090909090909,
      "loss": 0.1142,
      "step": 1917
    },
    {
      "epoch": 25.92,
      "learning_rate": 0.00024327272727272725,
      "loss": 0.2274,
      "step": 1918
    },
    {
      "epoch": 25.93,
      "learning_rate": 0.0002431363636363636,
      "loss": 0.1682,
      "step": 1919
    },
    {
      "epoch": 25.95,
      "learning_rate": 0.000243,
      "loss": 0.3041,
      "step": 1920
    },
    {
      "epoch": 25.96,
      "learning_rate": 0.00024286363636363636,
      "loss": 0.1586,
      "step": 1921
    },
    {
      "epoch": 25.97,
      "learning_rate": 0.0002427272727272727,
      "loss": 0.3,
      "step": 1922
    },
    {
      "epoch": 25.99,
      "learning_rate": 0.00024259090909090907,
      "loss": 0.0882,
      "step": 1923
    },
    {
      "epoch": 26.0,
      "learning_rate": 0.00024245454545454546,
      "loss": 0.0113,
      "step": 1924
    },
    {
      "epoch": 26.01,
      "learning_rate": 0.0002423181818181818,
      "loss": 0.1909,
      "step": 1925
    },
    {
      "epoch": 26.03,
      "learning_rate": 0.00024218181818181814,
      "loss": 0.2588,
      "step": 1926
    },
    {
      "epoch": 26.04,
      "learning_rate": 0.0002420454545454545,
      "loss": 0.0747,
      "step": 1927
    },
    {
      "epoch": 26.05,
      "learning_rate": 0.0002419090909090909,
      "loss": 0.0834,
      "step": 1928
    },
    {
      "epoch": 26.07,
      "learning_rate": 0.00024177272727272725,
      "loss": 0.1769,
      "step": 1929
    },
    {
      "epoch": 26.08,
      "learning_rate": 0.0002416363636363636,
      "loss": 0.3137,
      "step": 1930
    },
    {
      "epoch": 26.09,
      "learning_rate": 0.0002415,
      "loss": 0.0906,
      "step": 1931
    },
    {
      "epoch": 26.11,
      "learning_rate": 0.00024136363636363635,
      "loss": 0.1523,
      "step": 1932
    },
    {
      "epoch": 26.12,
      "learning_rate": 0.0002412272727272727,
      "loss": 0.0865,
      "step": 1933
    },
    {
      "epoch": 26.14,
      "learning_rate": 0.00024109090909090906,
      "loss": 0.1033,
      "step": 1934
    },
    {
      "epoch": 26.15,
      "learning_rate": 0.00024095454545454545,
      "loss": 0.1658,
      "step": 1935
    },
    {
      "epoch": 26.16,
      "learning_rate": 0.0002408181818181818,
      "loss": 0.0765,
      "step": 1936
    },
    {
      "epoch": 26.18,
      "learning_rate": 0.00024068181818181816,
      "loss": 0.0276,
      "step": 1937
    },
    {
      "epoch": 26.19,
      "learning_rate": 0.00024054545454545452,
      "loss": 0.0518,
      "step": 1938
    },
    {
      "epoch": 26.2,
      "learning_rate": 0.0002404090909090909,
      "loss": 0.1242,
      "step": 1939
    },
    {
      "epoch": 26.22,
      "learning_rate": 0.00024027272727272726,
      "loss": 0.097,
      "step": 1940
    },
    {
      "epoch": 26.23,
      "learning_rate": 0.0002401363636363636,
      "loss": 0.2588,
      "step": 1941
    },
    {
      "epoch": 26.24,
      "learning_rate": 0.00023999999999999998,
      "loss": 0.0907,
      "step": 1942
    },
    {
      "epoch": 26.26,
      "learning_rate": 0.00023986363636363634,
      "loss": 0.0854,
      "step": 1943
    },
    {
      "epoch": 26.27,
      "learning_rate": 0.0002397272727272727,
      "loss": 0.1622,
      "step": 1944
    },
    {
      "epoch": 26.28,
      "learning_rate": 0.00023959090909090905,
      "loss": 0.0664,
      "step": 1945
    },
    {
      "epoch": 26.3,
      "learning_rate": 0.00023945454545454544,
      "loss": 0.0781,
      "step": 1946
    },
    {
      "epoch": 26.31,
      "learning_rate": 0.0002393181818181818,
      "loss": 0.0851,
      "step": 1947
    },
    {
      "epoch": 26.32,
      "learning_rate": 0.00023918181818181815,
      "loss": 0.2294,
      "step": 1948
    },
    {
      "epoch": 26.34,
      "learning_rate": 0.00023904545454545454,
      "loss": 0.0974,
      "step": 1949
    },
    {
      "epoch": 26.35,
      "learning_rate": 0.0002389090909090909,
      "loss": 0.2368,
      "step": 1950
    },
    {
      "epoch": 26.36,
      "learning_rate": 0.00023877272727272725,
      "loss": 0.2725,
      "step": 1951
    },
    {
      "epoch": 26.38,
      "learning_rate": 0.0002386363636363636,
      "loss": 0.2426,
      "step": 1952
    },
    {
      "epoch": 26.39,
      "learning_rate": 0.0002385,
      "loss": 0.2089,
      "step": 1953
    },
    {
      "epoch": 26.41,
      "learning_rate": 0.00023836363636363635,
      "loss": 0.0094,
      "step": 1954
    },
    {
      "epoch": 26.42,
      "learning_rate": 0.0002382272727272727,
      "loss": 0.4318,
      "step": 1955
    },
    {
      "epoch": 26.43,
      "learning_rate": 0.00023809090909090904,
      "loss": 0.0826,
      "step": 1956
    },
    {
      "epoch": 26.45,
      "learning_rate": 0.00023795454545454546,
      "loss": 0.0989,
      "step": 1957
    },
    {
      "epoch": 26.46,
      "learning_rate": 0.00023781818181818179,
      "loss": 0.2138,
      "step": 1958
    },
    {
      "epoch": 26.47,
      "learning_rate": 0.00023768181818181814,
      "loss": 0.0829,
      "step": 1959
    },
    {
      "epoch": 26.49,
      "learning_rate": 0.00023754545454545453,
      "loss": 0.1673,
      "step": 1960
    },
    {
      "epoch": 26.5,
      "learning_rate": 0.0002374090909090909,
      "loss": 0.1015,
      "step": 1961
    },
    {
      "epoch": 26.51,
      "learning_rate": 0.00023727272727272724,
      "loss": 0.0993,
      "step": 1962
    },
    {
      "epoch": 26.53,
      "learning_rate": 0.0002371363636363636,
      "loss": 0.118,
      "step": 1963
    },
    {
      "epoch": 26.54,
      "learning_rate": 0.000237,
      "loss": 0.1631,
      "step": 1964
    },
    {
      "epoch": 26.55,
      "learning_rate": 0.00023686363636363635,
      "loss": 0.3296,
      "step": 1965
    },
    {
      "epoch": 26.57,
      "learning_rate": 0.0002367272727272727,
      "loss": 0.3113,
      "step": 1966
    },
    {
      "epoch": 26.58,
      "learning_rate": 0.0002365909090909091,
      "loss": 0.215,
      "step": 1967
    },
    {
      "epoch": 26.59,
      "learning_rate": 0.00023645454545454545,
      "loss": 0.21,
      "step": 1968
    },
    {
      "epoch": 26.61,
      "learning_rate": 0.0002363181818181818,
      "loss": 0.1172,
      "step": 1969
    },
    {
      "epoch": 26.62,
      "learning_rate": 0.00023618181818181816,
      "loss": 0.0408,
      "step": 1970
    },
    {
      "epoch": 26.64,
      "learning_rate": 0.00023604545454545455,
      "loss": 0.0413,
      "step": 1971
    },
    {
      "epoch": 26.65,
      "learning_rate": 0.0002359090909090909,
      "loss": 0.1335,
      "step": 1972
    },
    {
      "epoch": 26.66,
      "learning_rate": 0.00023577272727272724,
      "loss": 0.2736,
      "step": 1973
    },
    {
      "epoch": 26.68,
      "learning_rate": 0.0002356363636363636,
      "loss": 0.0171,
      "step": 1974
    },
    {
      "epoch": 26.69,
      "learning_rate": 0.00023549999999999998,
      "loss": 0.0248,
      "step": 1975
    },
    {
      "epoch": 26.7,
      "learning_rate": 0.00023536363636363634,
      "loss": 0.1061,
      "step": 1976
    },
    {
      "epoch": 26.72,
      "learning_rate": 0.0002352272727272727,
      "loss": 0.0724,
      "step": 1977
    },
    {
      "epoch": 26.73,
      "learning_rate": 0.00023509090909090908,
      "loss": 0.0256,
      "step": 1978
    },
    {
      "epoch": 26.74,
      "learning_rate": 0.00023495454545454544,
      "loss": 0.099,
      "step": 1979
    },
    {
      "epoch": 26.76,
      "learning_rate": 0.0002348181818181818,
      "loss": 0.0848,
      "step": 1980
    },
    {
      "epoch": 26.77,
      "learning_rate": 0.00023468181818181815,
      "loss": 0.0607,
      "step": 1981
    },
    {
      "epoch": 26.78,
      "learning_rate": 0.00023454545454545454,
      "loss": 0.2041,
      "step": 1982
    },
    {
      "epoch": 26.8,
      "learning_rate": 0.0002344090909090909,
      "loss": 0.273,
      "step": 1983
    },
    {
      "epoch": 26.81,
      "learning_rate": 0.00023427272727272725,
      "loss": 0.0848,
      "step": 1984
    },
    {
      "epoch": 26.82,
      "learning_rate": 0.0002341363636363636,
      "loss": 0.1738,
      "step": 1985
    },
    {
      "epoch": 26.84,
      "learning_rate": 0.000234,
      "loss": 0.2546,
      "step": 1986
    },
    {
      "epoch": 26.85,
      "learning_rate": 0.00023386363636363635,
      "loss": 0.1166,
      "step": 1987
    },
    {
      "epoch": 26.86,
      "learning_rate": 0.00023372727272727268,
      "loss": 0.3557,
      "step": 1988
    },
    {
      "epoch": 26.88,
      "learning_rate": 0.0002335909090909091,
      "loss": 0.0796,
      "step": 1989
    },
    {
      "epoch": 26.89,
      "learning_rate": 0.00023345454545454543,
      "loss": 0.0395,
      "step": 1990
    },
    {
      "epoch": 26.91,
      "learning_rate": 0.00023331818181818179,
      "loss": 0.2891,
      "step": 1991
    },
    {
      "epoch": 26.92,
      "learning_rate": 0.00023318181818181814,
      "loss": 0.0233,
      "step": 1992
    },
    {
      "epoch": 26.93,
      "learning_rate": 0.00023304545454545453,
      "loss": 0.0482,
      "step": 1993
    },
    {
      "epoch": 26.95,
      "learning_rate": 0.00023290909090909089,
      "loss": 0.169,
      "step": 1994
    },
    {
      "epoch": 26.96,
      "learning_rate": 0.00023277272727272724,
      "loss": 0.1048,
      "step": 1995
    },
    {
      "epoch": 26.97,
      "learning_rate": 0.00023263636363636363,
      "loss": 0.0844,
      "step": 1996
    },
    {
      "epoch": 26.99,
      "learning_rate": 0.00023249999999999999,
      "loss": 0.1168,
      "step": 1997
    },
    {
      "epoch": 27.0,
      "learning_rate": 0.00023236363636363634,
      "loss": 0.1283,
      "step": 1998
    },
    {
      "epoch": 27.01,
      "learning_rate": 0.0002322272727272727,
      "loss": 0.0246,
      "step": 1999
    },
    {
      "epoch": 27.03,
      "learning_rate": 0.0002320909090909091,
      "loss": 0.1233,
      "step": 2000
    },
    {
      "epoch": 27.03,
      "eval_accuracy": 0.901219512195122,
      "eval_f1": 0.901558848734687,
      "eval_loss": 0.3857743442058563,
      "eval_runtime": 125.9887,
      "eval_samples_per_second": 13.017,
      "eval_steps_per_second": 0.818,
      "step": 2000
    },
    {
      "epoch": 27.04,
      "learning_rate": 0.00023195454545454544,
      "loss": 0.0161,
      "step": 2001
    },
    {
      "epoch": 27.05,
      "learning_rate": 0.0002318181818181818,
      "loss": 0.1077,
      "step": 2002
    },
    {
      "epoch": 27.07,
      "learning_rate": 0.00023168181818181813,
      "loss": 0.2394,
      "step": 2003
    },
    {
      "epoch": 27.08,
      "learning_rate": 0.00023154545454545455,
      "loss": 0.1017,
      "step": 2004
    },
    {
      "epoch": 27.09,
      "learning_rate": 0.00023140909090909088,
      "loss": 0.0878,
      "step": 2005
    },
    {
      "epoch": 27.11,
      "learning_rate": 0.00023127272727272723,
      "loss": 0.3061,
      "step": 2006
    },
    {
      "epoch": 27.12,
      "learning_rate": 0.00023113636363636362,
      "loss": 0.0429,
      "step": 2007
    },
    {
      "epoch": 27.14,
      "learning_rate": 0.00023099999999999998,
      "loss": 0.2579,
      "step": 2008
    },
    {
      "epoch": 27.15,
      "learning_rate": 0.00023086363636363633,
      "loss": 0.0705,
      "step": 2009
    },
    {
      "epoch": 27.16,
      "learning_rate": 0.0002307272727272727,
      "loss": 0.092,
      "step": 2010
    },
    {
      "epoch": 27.18,
      "learning_rate": 0.00023059090909090908,
      "loss": 0.0152,
      "step": 2011
    },
    {
      "epoch": 27.19,
      "learning_rate": 0.00023045454545454544,
      "loss": 0.0917,
      "step": 2012
    },
    {
      "epoch": 27.2,
      "learning_rate": 0.0002303181818181818,
      "loss": 0.1082,
      "step": 2013
    },
    {
      "epoch": 27.22,
      "learning_rate": 0.00023018181818181815,
      "loss": 0.1,
      "step": 2014
    },
    {
      "epoch": 27.23,
      "learning_rate": 0.00023004545454545454,
      "loss": 0.3105,
      "step": 2015
    },
    {
      "epoch": 27.24,
      "learning_rate": 0.0002299090909090909,
      "loss": 0.0836,
      "step": 2016
    },
    {
      "epoch": 27.26,
      "learning_rate": 0.00022977272727272725,
      "loss": 0.3182,
      "step": 2017
    },
    {
      "epoch": 27.27,
      "learning_rate": 0.00022963636363636364,
      "loss": 0.3598,
      "step": 2018
    },
    {
      "epoch": 27.28,
      "learning_rate": 0.0002295,
      "loss": 0.1413,
      "step": 2019
    },
    {
      "epoch": 27.3,
      "learning_rate": 0.00022936363636363633,
      "loss": 0.1641,
      "step": 2020
    },
    {
      "epoch": 27.31,
      "learning_rate": 0.00022922727272727268,
      "loss": 0.2478,
      "step": 2021
    },
    {
      "epoch": 27.32,
      "learning_rate": 0.00022909090909090907,
      "loss": 0.2928,
      "step": 2022
    },
    {
      "epoch": 27.34,
      "learning_rate": 0.00022895454545454543,
      "loss": 0.105,
      "step": 2023
    },
    {
      "epoch": 27.35,
      "learning_rate": 0.00022881818181818178,
      "loss": 0.3172,
      "step": 2024
    },
    {
      "epoch": 27.36,
      "learning_rate": 0.00022868181818181817,
      "loss": 0.1671,
      "step": 2025
    },
    {
      "epoch": 27.38,
      "learning_rate": 0.00022854545454545453,
      "loss": 0.1107,
      "step": 2026
    },
    {
      "epoch": 27.39,
      "learning_rate": 0.00022840909090909088,
      "loss": 0.2299,
      "step": 2027
    },
    {
      "epoch": 27.41,
      "learning_rate": 0.00022827272727272724,
      "loss": 0.0854,
      "step": 2028
    },
    {
      "epoch": 27.42,
      "learning_rate": 0.00022813636363636363,
      "loss": 0.2436,
      "step": 2029
    },
    {
      "epoch": 27.43,
      "learning_rate": 0.00022799999999999999,
      "loss": 0.1392,
      "step": 2030
    },
    {
      "epoch": 27.45,
      "learning_rate": 0.00022786363636363634,
      "loss": 0.1177,
      "step": 2031
    },
    {
      "epoch": 27.46,
      "learning_rate": 0.0002277272727272727,
      "loss": 0.0889,
      "step": 2032
    },
    {
      "epoch": 27.47,
      "learning_rate": 0.00022759090909090909,
      "loss": 0.2346,
      "step": 2033
    },
    {
      "epoch": 27.49,
      "learning_rate": 0.00022745454545454544,
      "loss": 0.1249,
      "step": 2034
    },
    {
      "epoch": 27.5,
      "learning_rate": 0.00022731818181818177,
      "loss": 0.2483,
      "step": 2035
    },
    {
      "epoch": 27.51,
      "learning_rate": 0.0002271818181818182,
      "loss": 0.0163,
      "step": 2036
    },
    {
      "epoch": 27.53,
      "learning_rate": 0.00022704545454545452,
      "loss": 0.2514,
      "step": 2037
    },
    {
      "epoch": 27.54,
      "learning_rate": 0.00022690909090909088,
      "loss": 0.3235,
      "step": 2038
    },
    {
      "epoch": 27.55,
      "learning_rate": 0.00022677272727272723,
      "loss": 0.1002,
      "step": 2039
    },
    {
      "epoch": 27.57,
      "learning_rate": 0.00022663636363636362,
      "loss": 0.3972,
      "step": 2040
    },
    {
      "epoch": 27.58,
      "learning_rate": 0.00022649999999999998,
      "loss": 0.0415,
      "step": 2041
    },
    {
      "epoch": 27.59,
      "learning_rate": 0.00022636363636363633,
      "loss": 0.1361,
      "step": 2042
    },
    {
      "epoch": 27.61,
      "learning_rate": 0.00022622727272727272,
      "loss": 0.1071,
      "step": 2043
    },
    {
      "epoch": 27.62,
      "learning_rate": 0.00022609090909090908,
      "loss": 0.0681,
      "step": 2044
    },
    {
      "epoch": 27.64,
      "learning_rate": 0.00022595454545454543,
      "loss": 0.0864,
      "step": 2045
    },
    {
      "epoch": 27.65,
      "learning_rate": 0.0002258181818181818,
      "loss": 0.222,
      "step": 2046
    },
    {
      "epoch": 27.66,
      "learning_rate": 0.00022568181818181818,
      "loss": 0.2676,
      "step": 2047
    },
    {
      "epoch": 27.68,
      "learning_rate": 0.00022554545454545454,
      "loss": 0.1847,
      "step": 2048
    },
    {
      "epoch": 27.69,
      "learning_rate": 0.0002254090909090909,
      "loss": 0.1731,
      "step": 2049
    },
    {
      "epoch": 27.7,
      "learning_rate": 0.00022527272727272725,
      "loss": 0.1061,
      "step": 2050
    },
    {
      "epoch": 27.72,
      "learning_rate": 0.00022513636363636364,
      "loss": 0.1717,
      "step": 2051
    },
    {
      "epoch": 27.73,
      "learning_rate": 0.000225,
      "loss": 0.1303,
      "step": 2052
    },
    {
      "epoch": 27.74,
      "learning_rate": 0.00022486363636363632,
      "loss": 0.3189,
      "step": 2053
    },
    {
      "epoch": 27.76,
      "learning_rate": 0.0002247272727272727,
      "loss": 0.1302,
      "step": 2054
    },
    {
      "epoch": 27.77,
      "learning_rate": 0.00022459090909090907,
      "loss": 0.2625,
      "step": 2055
    },
    {
      "epoch": 27.78,
      "learning_rate": 0.00022445454545454543,
      "loss": 0.1367,
      "step": 2056
    },
    {
      "epoch": 27.8,
      "learning_rate": 0.00022431818181818178,
      "loss": 0.0412,
      "step": 2057
    },
    {
      "epoch": 27.81,
      "learning_rate": 0.00022418181818181817,
      "loss": 0.1383,
      "step": 2058
    },
    {
      "epoch": 27.82,
      "learning_rate": 0.00022404545454545453,
      "loss": 0.122,
      "step": 2059
    },
    {
      "epoch": 27.84,
      "learning_rate": 0.00022390909090909088,
      "loss": 0.1917,
      "step": 2060
    },
    {
      "epoch": 27.85,
      "learning_rate": 0.00022377272727272724,
      "loss": 0.0237,
      "step": 2061
    },
    {
      "epoch": 27.86,
      "learning_rate": 0.00022363636363636363,
      "loss": 0.2369,
      "step": 2062
    },
    {
      "epoch": 27.88,
      "learning_rate": 0.00022349999999999998,
      "loss": 0.0903,
      "step": 2063
    },
    {
      "epoch": 27.89,
      "learning_rate": 0.00022336363636363634,
      "loss": 0.2801,
      "step": 2064
    },
    {
      "epoch": 27.91,
      "learning_rate": 0.00022322727272727273,
      "loss": 0.0391,
      "step": 2065
    },
    {
      "epoch": 27.92,
      "learning_rate": 0.00022309090909090909,
      "loss": 0.1689,
      "step": 2066
    },
    {
      "epoch": 27.93,
      "learning_rate": 0.00022295454545454544,
      "loss": 0.0328,
      "step": 2067
    },
    {
      "epoch": 27.95,
      "learning_rate": 0.00022281818181818177,
      "loss": 0.0095,
      "step": 2068
    },
    {
      "epoch": 27.96,
      "learning_rate": 0.00022268181818181819,
      "loss": 0.2245,
      "step": 2069
    },
    {
      "epoch": 27.97,
      "learning_rate": 0.00022254545454545452,
      "loss": 0.1462,
      "step": 2070
    },
    {
      "epoch": 27.99,
      "learning_rate": 0.00022240909090909087,
      "loss": 0.0761,
      "step": 2071
    },
    {
      "epoch": 28.0,
      "learning_rate": 0.00022227272727272726,
      "loss": 0.4568,
      "step": 2072
    },
    {
      "epoch": 28.01,
      "learning_rate": 0.00022213636363636362,
      "loss": 0.0994,
      "step": 2073
    },
    {
      "epoch": 28.03,
      "learning_rate": 0.00022199999999999998,
      "loss": 0.0959,
      "step": 2074
    },
    {
      "epoch": 28.04,
      "learning_rate": 0.00022186363636363633,
      "loss": 0.0744,
      "step": 2075
    },
    {
      "epoch": 28.05,
      "learning_rate": 0.00022172727272727272,
      "loss": 0.0387,
      "step": 2076
    },
    {
      "epoch": 28.07,
      "learning_rate": 0.00022159090909090908,
      "loss": 0.0198,
      "step": 2077
    },
    {
      "epoch": 28.08,
      "learning_rate": 0.00022145454545454543,
      "loss": 0.0474,
      "step": 2078
    },
    {
      "epoch": 28.09,
      "learning_rate": 0.0002213181818181818,
      "loss": 0.1095,
      "step": 2079
    },
    {
      "epoch": 28.11,
      "learning_rate": 0.00022118181818181818,
      "loss": 0.307,
      "step": 2080
    },
    {
      "epoch": 28.12,
      "learning_rate": 0.00022104545454545453,
      "loss": 0.1064,
      "step": 2081
    },
    {
      "epoch": 28.14,
      "learning_rate": 0.0002209090909090909,
      "loss": 0.1047,
      "step": 2082
    },
    {
      "epoch": 28.15,
      "learning_rate": 0.00022077272727272728,
      "loss": 0.0324,
      "step": 2083
    },
    {
      "epoch": 28.16,
      "learning_rate": 0.00022063636363636364,
      "loss": 0.0716,
      "step": 2084
    },
    {
      "epoch": 28.18,
      "learning_rate": 0.00022049999999999997,
      "loss": 0.046,
      "step": 2085
    },
    {
      "epoch": 28.19,
      "learning_rate": 0.00022036363636363632,
      "loss": 0.2854,
      "step": 2086
    },
    {
      "epoch": 28.2,
      "learning_rate": 0.0002202272727272727,
      "loss": 0.2158,
      "step": 2087
    },
    {
      "epoch": 28.22,
      "learning_rate": 0.00022009090909090907,
      "loss": 0.0162,
      "step": 2088
    },
    {
      "epoch": 28.23,
      "learning_rate": 0.00021995454545454542,
      "loss": 0.1148,
      "step": 2089
    },
    {
      "epoch": 28.24,
      "learning_rate": 0.0002198181818181818,
      "loss": 0.0333,
      "step": 2090
    },
    {
      "epoch": 28.26,
      "learning_rate": 0.00021968181818181817,
      "loss": 0.1136,
      "step": 2091
    },
    {
      "epoch": 28.27,
      "learning_rate": 0.00021954545454545452,
      "loss": 0.1825,
      "step": 2092
    },
    {
      "epoch": 28.28,
      "learning_rate": 0.00021940909090909088,
      "loss": 0.0397,
      "step": 2093
    },
    {
      "epoch": 28.3,
      "learning_rate": 0.00021927272727272727,
      "loss": 0.1092,
      "step": 2094
    },
    {
      "epoch": 28.31,
      "learning_rate": 0.00021913636363636363,
      "loss": 0.2171,
      "step": 2095
    },
    {
      "epoch": 28.32,
      "learning_rate": 0.00021899999999999998,
      "loss": 0.0045,
      "step": 2096
    },
    {
      "epoch": 28.34,
      "learning_rate": 0.00021886363636363634,
      "loss": 0.0505,
      "step": 2097
    },
    {
      "epoch": 28.35,
      "learning_rate": 0.00021872727272727273,
      "loss": 0.1625,
      "step": 2098
    },
    {
      "epoch": 28.36,
      "learning_rate": 0.00021859090909090908,
      "loss": 0.0394,
      "step": 2099
    },
    {
      "epoch": 28.38,
      "learning_rate": 0.00021845454545454541,
      "loss": 0.2265,
      "step": 2100
    },
    {
      "epoch": 28.39,
      "learning_rate": 0.00021831818181818183,
      "loss": 0.0457,
      "step": 2101
    },
    {
      "epoch": 28.41,
      "learning_rate": 0.00021818181818181816,
      "loss": 0.0873,
      "step": 2102
    },
    {
      "epoch": 28.42,
      "learning_rate": 0.00021804545454545452,
      "loss": 0.1428,
      "step": 2103
    },
    {
      "epoch": 28.43,
      "learning_rate": 0.00021790909090909087,
      "loss": 0.189,
      "step": 2104
    },
    {
      "epoch": 28.45,
      "learning_rate": 0.00021777272727272726,
      "loss": 0.0923,
      "step": 2105
    },
    {
      "epoch": 28.46,
      "learning_rate": 0.00021763636363636362,
      "loss": 0.3614,
      "step": 2106
    },
    {
      "epoch": 28.47,
      "learning_rate": 0.00021749999999999997,
      "loss": 0.2029,
      "step": 2107
    },
    {
      "epoch": 28.49,
      "learning_rate": 0.00021736363636363633,
      "loss": 0.0653,
      "step": 2108
    },
    {
      "epoch": 28.5,
      "learning_rate": 0.00021722727272727272,
      "loss": 0.2142,
      "step": 2109
    },
    {
      "epoch": 28.51,
      "learning_rate": 0.00021709090909090907,
      "loss": 0.11,
      "step": 2110
    },
    {
      "epoch": 28.53,
      "learning_rate": 0.00021695454545454543,
      "loss": 0.0422,
      "step": 2111
    },
    {
      "epoch": 28.54,
      "learning_rate": 0.00021681818181818182,
      "loss": 0.1004,
      "step": 2112
    },
    {
      "epoch": 28.55,
      "learning_rate": 0.00021668181818181818,
      "loss": 0.054,
      "step": 2113
    },
    {
      "epoch": 28.57,
      "learning_rate": 0.00021654545454545453,
      "loss": 0.0184,
      "step": 2114
    },
    {
      "epoch": 28.58,
      "learning_rate": 0.00021640909090909086,
      "loss": 0.0182,
      "step": 2115
    },
    {
      "epoch": 28.59,
      "learning_rate": 0.00021627272727272728,
      "loss": 0.0243,
      "step": 2116
    },
    {
      "epoch": 28.61,
      "learning_rate": 0.0002161363636363636,
      "loss": 0.102,
      "step": 2117
    },
    {
      "epoch": 28.62,
      "learning_rate": 0.00021599999999999996,
      "loss": 0.0386,
      "step": 2118
    },
    {
      "epoch": 28.64,
      "learning_rate": 0.00021586363636363635,
      "loss": 0.1234,
      "step": 2119
    },
    {
      "epoch": 28.65,
      "learning_rate": 0.0002157272727272727,
      "loss": 0.1237,
      "step": 2120
    },
    {
      "epoch": 28.66,
      "learning_rate": 0.00021559090909090907,
      "loss": 0.0427,
      "step": 2121
    },
    {
      "epoch": 28.68,
      "learning_rate": 0.00021545454545454542,
      "loss": 0.2283,
      "step": 2122
    },
    {
      "epoch": 28.69,
      "learning_rate": 0.0002153181818181818,
      "loss": 0.2248,
      "step": 2123
    },
    {
      "epoch": 28.7,
      "learning_rate": 0.00021518181818181817,
      "loss": 0.0086,
      "step": 2124
    },
    {
      "epoch": 28.72,
      "learning_rate": 0.00021504545454545452,
      "loss": 0.1013,
      "step": 2125
    },
    {
      "epoch": 28.73,
      "learning_rate": 0.00021490909090909088,
      "loss": 0.2585,
      "step": 2126
    },
    {
      "epoch": 28.74,
      "learning_rate": 0.00021477272727272727,
      "loss": 0.1371,
      "step": 2127
    },
    {
      "epoch": 28.76,
      "learning_rate": 0.00021463636363636362,
      "loss": 0.1463,
      "step": 2128
    },
    {
      "epoch": 28.77,
      "learning_rate": 0.00021449999999999998,
      "loss": 0.2079,
      "step": 2129
    },
    {
      "epoch": 28.78,
      "learning_rate": 0.00021436363636363637,
      "loss": 0.191,
      "step": 2130
    },
    {
      "epoch": 28.8,
      "learning_rate": 0.00021422727272727273,
      "loss": 0.2167,
      "step": 2131
    },
    {
      "epoch": 28.81,
      "learning_rate": 0.00021409090909090906,
      "loss": 0.0533,
      "step": 2132
    },
    {
      "epoch": 28.82,
      "learning_rate": 0.00021395454545454541,
      "loss": 0.0634,
      "step": 2133
    },
    {
      "epoch": 28.84,
      "learning_rate": 0.0002138181818181818,
      "loss": 0.049,
      "step": 2134
    },
    {
      "epoch": 28.85,
      "learning_rate": 0.00021368181818181816,
      "loss": 0.0292,
      "step": 2135
    },
    {
      "epoch": 28.86,
      "learning_rate": 0.00021354545454545451,
      "loss": 0.123,
      "step": 2136
    },
    {
      "epoch": 28.88,
      "learning_rate": 0.00021340909090909087,
      "loss": 0.0965,
      "step": 2137
    },
    {
      "epoch": 28.89,
      "learning_rate": 0.00021327272727272726,
      "loss": 0.0467,
      "step": 2138
    },
    {
      "epoch": 28.91,
      "learning_rate": 0.00021313636363636362,
      "loss": 0.0513,
      "step": 2139
    },
    {
      "epoch": 28.92,
      "learning_rate": 0.00021299999999999997,
      "loss": 0.1383,
      "step": 2140
    },
    {
      "epoch": 28.93,
      "learning_rate": 0.00021286363636363636,
      "loss": 0.0189,
      "step": 2141
    },
    {
      "epoch": 28.95,
      "learning_rate": 0.00021272727272727272,
      "loss": 0.089,
      "step": 2142
    },
    {
      "epoch": 28.96,
      "learning_rate": 0.00021259090909090907,
      "loss": 0.021,
      "step": 2143
    },
    {
      "epoch": 28.97,
      "learning_rate": 0.00021245454545454543,
      "loss": 0.0157,
      "step": 2144
    },
    {
      "epoch": 28.99,
      "learning_rate": 0.00021231818181818182,
      "loss": 0.1666,
      "step": 2145
    },
    {
      "epoch": 29.0,
      "learning_rate": 0.00021218181818181817,
      "loss": 0.037,
      "step": 2146
    },
    {
      "epoch": 29.01,
      "learning_rate": 0.0002120454545454545,
      "loss": 0.1405,
      "step": 2147
    },
    {
      "epoch": 29.03,
      "learning_rate": 0.00021190909090909092,
      "loss": 0.0838,
      "step": 2148
    },
    {
      "epoch": 29.04,
      "learning_rate": 0.00021177272727272725,
      "loss": 0.1355,
      "step": 2149
    },
    {
      "epoch": 29.05,
      "learning_rate": 0.0002116363636363636,
      "loss": 0.0412,
      "step": 2150
    },
    {
      "epoch": 29.07,
      "learning_rate": 0.00021149999999999996,
      "loss": 0.1226,
      "step": 2151
    },
    {
      "epoch": 29.08,
      "learning_rate": 0.00021136363636363635,
      "loss": 0.0068,
      "step": 2152
    },
    {
      "epoch": 29.09,
      "learning_rate": 0.0002112272727272727,
      "loss": 0.0049,
      "step": 2153
    },
    {
      "epoch": 29.11,
      "learning_rate": 0.00021109090909090906,
      "loss": 0.3355,
      "step": 2154
    },
    {
      "epoch": 29.12,
      "learning_rate": 0.00021095454545454542,
      "loss": 0.0275,
      "step": 2155
    },
    {
      "epoch": 29.14,
      "learning_rate": 0.0002108181818181818,
      "loss": 0.3329,
      "step": 2156
    },
    {
      "epoch": 29.15,
      "learning_rate": 0.00021068181818181817,
      "loss": 0.2534,
      "step": 2157
    },
    {
      "epoch": 29.16,
      "learning_rate": 0.00021054545454545452,
      "loss": 0.2138,
      "step": 2158
    },
    {
      "epoch": 29.18,
      "learning_rate": 0.0002104090909090909,
      "loss": 0.2221,
      "step": 2159
    },
    {
      "epoch": 29.19,
      "learning_rate": 0.00021027272727272727,
      "loss": 0.1441,
      "step": 2160
    },
    {
      "epoch": 29.2,
      "learning_rate": 0.00021013636363636362,
      "loss": 0.1837,
      "step": 2161
    },
    {
      "epoch": 29.22,
      "learning_rate": 0.00020999999999999998,
      "loss": 0.3599,
      "step": 2162
    },
    {
      "epoch": 29.23,
      "learning_rate": 0.00020986363636363637,
      "loss": 0.5427,
      "step": 2163
    },
    {
      "epoch": 29.24,
      "learning_rate": 0.0002097272727272727,
      "loss": 0.1108,
      "step": 2164
    },
    {
      "epoch": 29.26,
      "learning_rate": 0.00020959090909090906,
      "loss": 0.1108,
      "step": 2165
    },
    {
      "epoch": 29.27,
      "learning_rate": 0.00020945454545454544,
      "loss": 0.0395,
      "step": 2166
    },
    {
      "epoch": 29.28,
      "learning_rate": 0.0002093181818181818,
      "loss": 0.0123,
      "step": 2167
    },
    {
      "epoch": 29.3,
      "learning_rate": 0.00020918181818181816,
      "loss": 0.2293,
      "step": 2168
    },
    {
      "epoch": 29.31,
      "learning_rate": 0.00020904545454545451,
      "loss": 0.0657,
      "step": 2169
    },
    {
      "epoch": 29.32,
      "learning_rate": 0.0002089090909090909,
      "loss": 0.1432,
      "step": 2170
    },
    {
      "epoch": 29.34,
      "learning_rate": 0.00020877272727272726,
      "loss": 0.058,
      "step": 2171
    },
    {
      "epoch": 29.35,
      "learning_rate": 0.00020863636363636361,
      "loss": 0.0418,
      "step": 2172
    },
    {
      "epoch": 29.36,
      "learning_rate": 0.00020849999999999997,
      "loss": 0.0339,
      "step": 2173
    },
    {
      "epoch": 29.38,
      "learning_rate": 0.00020836363636363636,
      "loss": 0.0799,
      "step": 2174
    },
    {
      "epoch": 29.39,
      "learning_rate": 0.00020822727272727272,
      "loss": 0.1186,
      "step": 2175
    },
    {
      "epoch": 29.41,
      "learning_rate": 0.00020809090909090907,
      "loss": 0.1528,
      "step": 2176
    },
    {
      "epoch": 29.42,
      "learning_rate": 0.00020795454545454546,
      "loss": 0.0445,
      "step": 2177
    },
    {
      "epoch": 29.43,
      "learning_rate": 0.00020781818181818182,
      "loss": 0.0244,
      "step": 2178
    },
    {
      "epoch": 29.45,
      "learning_rate": 0.00020768181818181817,
      "loss": 0.1037,
      "step": 2179
    },
    {
      "epoch": 29.46,
      "learning_rate": 0.0002075454545454545,
      "loss": 0.0417,
      "step": 2180
    },
    {
      "epoch": 29.47,
      "learning_rate": 0.0002074090909090909,
      "loss": 0.0536,
      "step": 2181
    },
    {
      "epoch": 29.49,
      "learning_rate": 0.00020727272727272725,
      "loss": 0.1513,
      "step": 2182
    },
    {
      "epoch": 29.5,
      "learning_rate": 0.0002071363636363636,
      "loss": 0.1578,
      "step": 2183
    },
    {
      "epoch": 29.51,
      "learning_rate": 0.00020699999999999996,
      "loss": 0.1211,
      "step": 2184
    },
    {
      "epoch": 29.53,
      "learning_rate": 0.00020686363636363635,
      "loss": 0.2802,
      "step": 2185
    },
    {
      "epoch": 29.54,
      "learning_rate": 0.0002067272727272727,
      "loss": 0.0997,
      "step": 2186
    },
    {
      "epoch": 29.55,
      "learning_rate": 0.00020659090909090906,
      "loss": 0.0569,
      "step": 2187
    },
    {
      "epoch": 29.57,
      "learning_rate": 0.00020645454545454545,
      "loss": 0.075,
      "step": 2188
    },
    {
      "epoch": 29.58,
      "learning_rate": 0.0002063181818181818,
      "loss": 0.0081,
      "step": 2189
    },
    {
      "epoch": 29.59,
      "learning_rate": 0.00020618181818181816,
      "loss": 0.2103,
      "step": 2190
    },
    {
      "epoch": 29.61,
      "learning_rate": 0.00020604545454545452,
      "loss": 0.0515,
      "step": 2191
    },
    {
      "epoch": 29.62,
      "learning_rate": 0.0002059090909090909,
      "loss": 0.0396,
      "step": 2192
    },
    {
      "epoch": 29.64,
      "learning_rate": 0.00020577272727272726,
      "loss": 0.061,
      "step": 2193
    },
    {
      "epoch": 29.65,
      "learning_rate": 0.00020563636363636362,
      "loss": 0.0766,
      "step": 2194
    },
    {
      "epoch": 29.66,
      "learning_rate": 0.0002055,
      "loss": 0.0443,
      "step": 2195
    },
    {
      "epoch": 29.68,
      "learning_rate": 0.00020536363636363637,
      "loss": 0.0702,
      "step": 2196
    },
    {
      "epoch": 29.69,
      "learning_rate": 0.0002052272727272727,
      "loss": 0.0077,
      "step": 2197
    },
    {
      "epoch": 29.7,
      "learning_rate": 0.00020509090909090905,
      "loss": 0.1492,
      "step": 2198
    },
    {
      "epoch": 29.72,
      "learning_rate": 0.00020495454545454544,
      "loss": 0.2001,
      "step": 2199
    },
    {
      "epoch": 29.73,
      "learning_rate": 0.0002048181818181818,
      "loss": 0.1827,
      "step": 2200
    },
    {
      "epoch": 29.73,
      "eval_accuracy": 0.9158536585365854,
      "eval_f1": 0.9162126400447651,
      "eval_loss": 0.37654536962509155,
      "eval_runtime": 127.0981,
      "eval_samples_per_second": 12.903,
      "eval_steps_per_second": 0.81,
      "step": 2200
    },
    {
      "epoch": 29.74,
      "learning_rate": 0.00020468181818181815,
      "loss": 0.3433,
      "step": 2201
    },
    {
      "epoch": 29.76,
      "learning_rate": 0.0002045454545454545,
      "loss": 0.3333,
      "step": 2202
    },
    {
      "epoch": 29.77,
      "learning_rate": 0.0002044090909090909,
      "loss": 0.1939,
      "step": 2203
    },
    {
      "epoch": 29.78,
      "learning_rate": 0.00020427272727272726,
      "loss": 0.1106,
      "step": 2204
    },
    {
      "epoch": 29.8,
      "learning_rate": 0.0002041363636363636,
      "loss": 0.2558,
      "step": 2205
    },
    {
      "epoch": 29.81,
      "learning_rate": 0.000204,
      "loss": 0.0633,
      "step": 2206
    },
    {
      "epoch": 29.82,
      "learning_rate": 0.00020386363636363636,
      "loss": 0.1534,
      "step": 2207
    },
    {
      "epoch": 29.84,
      "learning_rate": 0.00020372727272727271,
      "loss": 0.1143,
      "step": 2208
    },
    {
      "epoch": 29.85,
      "learning_rate": 0.00020359090909090907,
      "loss": 0.0927,
      "step": 2209
    },
    {
      "epoch": 29.86,
      "learning_rate": 0.00020345454545454546,
      "loss": 0.1398,
      "step": 2210
    },
    {
      "epoch": 29.88,
      "learning_rate": 0.00020331818181818181,
      "loss": 0.1189,
      "step": 2211
    },
    {
      "epoch": 29.89,
      "learning_rate": 0.00020318181818181815,
      "loss": 0.0276,
      "step": 2212
    },
    {
      "epoch": 29.91,
      "learning_rate": 0.0002030454545454545,
      "loss": 0.058,
      "step": 2213
    },
    {
      "epoch": 29.92,
      "learning_rate": 0.0002029090909090909,
      "loss": 0.0782,
      "step": 2214
    },
    {
      "epoch": 29.93,
      "learning_rate": 0.00020277272727272725,
      "loss": 0.0339,
      "step": 2215
    },
    {
      "epoch": 29.95,
      "learning_rate": 0.0002026363636363636,
      "loss": 0.0492,
      "step": 2216
    },
    {
      "epoch": 29.96,
      "learning_rate": 0.0002025,
      "loss": 0.1743,
      "step": 2217
    },
    {
      "epoch": 29.97,
      "learning_rate": 0.00020236363636363635,
      "loss": 0.0371,
      "step": 2218
    },
    {
      "epoch": 29.99,
      "learning_rate": 0.0002022272727272727,
      "loss": 0.0249,
      "step": 2219
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.00020209090909090906,
      "loss": 0.0625,
      "step": 2220
    },
    {
      "epoch": 30.01,
      "learning_rate": 0.00020195454545454545,
      "loss": 0.0385,
      "step": 2221
    },
    {
      "epoch": 30.03,
      "learning_rate": 0.0002018181818181818,
      "loss": 0.1195,
      "step": 2222
    },
    {
      "epoch": 30.04,
      "learning_rate": 0.00020168181818181816,
      "loss": 0.1306,
      "step": 2223
    },
    {
      "epoch": 30.05,
      "learning_rate": 0.00020154545454545455,
      "loss": 0.0609,
      "step": 2224
    },
    {
      "epoch": 30.07,
      "learning_rate": 0.0002014090909090909,
      "loss": 0.0213,
      "step": 2225
    },
    {
      "epoch": 30.08,
      "learning_rate": 0.00020127272727272726,
      "loss": 0.1538,
      "step": 2226
    },
    {
      "epoch": 30.09,
      "learning_rate": 0.0002011363636363636,
      "loss": 0.0153,
      "step": 2227
    },
    {
      "epoch": 30.11,
      "learning_rate": 0.000201,
      "loss": 0.1658,
      "step": 2228
    },
    {
      "epoch": 30.12,
      "learning_rate": 0.00020086363636363634,
      "loss": 0.0626,
      "step": 2229
    },
    {
      "epoch": 30.14,
      "learning_rate": 0.0002007272727272727,
      "loss": 0.4123,
      "step": 2230
    },
    {
      "epoch": 30.15,
      "learning_rate": 0.00020059090909090905,
      "loss": 0.3604,
      "step": 2231
    },
    {
      "epoch": 30.16,
      "learning_rate": 0.00020045454545454544,
      "loss": 0.1101,
      "step": 2232
    },
    {
      "epoch": 30.18,
      "learning_rate": 0.0002003181818181818,
      "loss": 0.0748,
      "step": 2233
    },
    {
      "epoch": 30.19,
      "learning_rate": 0.00020018181818181815,
      "loss": 0.067,
      "step": 2234
    },
    {
      "epoch": 30.2,
      "learning_rate": 0.00020004545454545454,
      "loss": 0.0219,
      "step": 2235
    },
    {
      "epoch": 30.22,
      "learning_rate": 0.0001999090909090909,
      "loss": 0.0844,
      "step": 2236
    },
    {
      "epoch": 30.23,
      "learning_rate": 0.00019977272727272725,
      "loss": 0.0426,
      "step": 2237
    },
    {
      "epoch": 30.24,
      "learning_rate": 0.0001996363636363636,
      "loss": 0.2019,
      "step": 2238
    },
    {
      "epoch": 30.26,
      "learning_rate": 0.0001995,
      "loss": 0.1129,
      "step": 2239
    },
    {
      "epoch": 30.27,
      "learning_rate": 0.00019936363636363636,
      "loss": 0.0342,
      "step": 2240
    },
    {
      "epoch": 30.28,
      "learning_rate": 0.0001992272727272727,
      "loss": 0.181,
      "step": 2241
    },
    {
      "epoch": 30.3,
      "learning_rate": 0.0001990909090909091,
      "loss": 0.0527,
      "step": 2242
    },
    {
      "epoch": 30.31,
      "learning_rate": 0.00019895454545454546,
      "loss": 0.0068,
      "step": 2243
    },
    {
      "epoch": 30.32,
      "learning_rate": 0.0001988181818181818,
      "loss": 0.1686,
      "step": 2244
    },
    {
      "epoch": 30.34,
      "learning_rate": 0.00019868181818181814,
      "loss": 0.2276,
      "step": 2245
    },
    {
      "epoch": 30.35,
      "learning_rate": 0.00019854545454545453,
      "loss": 0.2094,
      "step": 2246
    },
    {
      "epoch": 30.36,
      "learning_rate": 0.0001984090909090909,
      "loss": 0.0678,
      "step": 2247
    },
    {
      "epoch": 30.38,
      "learning_rate": 0.00019827272727272725,
      "loss": 0.1396,
      "step": 2248
    },
    {
      "epoch": 30.39,
      "learning_rate": 0.0001981363636363636,
      "loss": 0.0502,
      "step": 2249
    },
    {
      "epoch": 30.41,
      "learning_rate": 0.000198,
      "loss": 0.3148,
      "step": 2250
    },
    {
      "epoch": 30.42,
      "learning_rate": 0.00019786363636363635,
      "loss": 0.0987,
      "step": 2251
    },
    {
      "epoch": 30.43,
      "learning_rate": 0.0001977272727272727,
      "loss": 0.136,
      "step": 2252
    },
    {
      "epoch": 30.45,
      "learning_rate": 0.0001975909090909091,
      "loss": 0.1138,
      "step": 2253
    },
    {
      "epoch": 30.46,
      "learning_rate": 0.00019745454545454545,
      "loss": 0.0163,
      "step": 2254
    },
    {
      "epoch": 30.47,
      "learning_rate": 0.0001973181818181818,
      "loss": 0.0478,
      "step": 2255
    },
    {
      "epoch": 30.49,
      "learning_rate": 0.00019718181818181816,
      "loss": 0.036,
      "step": 2256
    },
    {
      "epoch": 30.5,
      "learning_rate": 0.00019704545454545455,
      "loss": 0.2147,
      "step": 2257
    },
    {
      "epoch": 30.51,
      "learning_rate": 0.0001969090909090909,
      "loss": 0.1457,
      "step": 2258
    },
    {
      "epoch": 30.53,
      "learning_rate": 0.00019677272727272724,
      "loss": 0.0079,
      "step": 2259
    },
    {
      "epoch": 30.54,
      "learning_rate": 0.0001966363636363636,
      "loss": 0.1162,
      "step": 2260
    },
    {
      "epoch": 30.55,
      "learning_rate": 0.00019649999999999998,
      "loss": 0.0253,
      "step": 2261
    },
    {
      "epoch": 30.57,
      "learning_rate": 0.00019636363636363634,
      "loss": 0.0142,
      "step": 2262
    },
    {
      "epoch": 30.58,
      "learning_rate": 0.0001962272727272727,
      "loss": 0.0977,
      "step": 2263
    },
    {
      "epoch": 30.59,
      "learning_rate": 0.00019609090909090908,
      "loss": 0.1311,
      "step": 2264
    },
    {
      "epoch": 30.61,
      "learning_rate": 0.00019595454545454544,
      "loss": 0.0286,
      "step": 2265
    },
    {
      "epoch": 30.62,
      "learning_rate": 0.0001958181818181818,
      "loss": 0.0928,
      "step": 2266
    },
    {
      "epoch": 30.64,
      "learning_rate": 0.00019568181818181815,
      "loss": 0.1386,
      "step": 2267
    },
    {
      "epoch": 30.65,
      "learning_rate": 0.00019554545454545454,
      "loss": 0.1163,
      "step": 2268
    },
    {
      "epoch": 30.66,
      "learning_rate": 0.0001954090909090909,
      "loss": 0.1445,
      "step": 2269
    },
    {
      "epoch": 30.68,
      "learning_rate": 0.00019527272727272725,
      "loss": 0.0866,
      "step": 2270
    },
    {
      "epoch": 30.69,
      "learning_rate": 0.00019513636363636364,
      "loss": 0.0335,
      "step": 2271
    },
    {
      "epoch": 30.7,
      "learning_rate": 0.000195,
      "loss": 0.0257,
      "step": 2272
    },
    {
      "epoch": 30.72,
      "learning_rate": 0.00019486363636363635,
      "loss": 0.3548,
      "step": 2273
    },
    {
      "epoch": 30.73,
      "learning_rate": 0.00019472727272727269,
      "loss": 0.013,
      "step": 2274
    },
    {
      "epoch": 30.74,
      "learning_rate": 0.0001945909090909091,
      "loss": 0.0873,
      "step": 2275
    },
    {
      "epoch": 30.76,
      "learning_rate": 0.00019445454545454543,
      "loss": 0.0453,
      "step": 2276
    },
    {
      "epoch": 30.77,
      "learning_rate": 0.00019431818181818179,
      "loss": 0.1085,
      "step": 2277
    },
    {
      "epoch": 30.78,
      "learning_rate": 0.00019418181818181814,
      "loss": 0.1187,
      "step": 2278
    },
    {
      "epoch": 30.8,
      "learning_rate": 0.00019404545454545453,
      "loss": 0.0747,
      "step": 2279
    },
    {
      "epoch": 30.81,
      "learning_rate": 0.00019390909090909089,
      "loss": 0.0696,
      "step": 2280
    },
    {
      "epoch": 30.82,
      "learning_rate": 0.00019377272727272724,
      "loss": 0.0029,
      "step": 2281
    },
    {
      "epoch": 30.84,
      "learning_rate": 0.00019363636363636363,
      "loss": 0.0457,
      "step": 2282
    },
    {
      "epoch": 30.85,
      "learning_rate": 0.0001935,
      "loss": 0.1246,
      "step": 2283
    },
    {
      "epoch": 30.86,
      "learning_rate": 0.00019336363636363634,
      "loss": 0.17,
      "step": 2284
    },
    {
      "epoch": 30.88,
      "learning_rate": 0.0001932272727272727,
      "loss": 0.0366,
      "step": 2285
    },
    {
      "epoch": 30.89,
      "learning_rate": 0.0001930909090909091,
      "loss": 0.042,
      "step": 2286
    },
    {
      "epoch": 30.91,
      "learning_rate": 0.00019295454545454545,
      "loss": 0.0217,
      "step": 2287
    },
    {
      "epoch": 30.92,
      "learning_rate": 0.0001928181818181818,
      "loss": 0.2218,
      "step": 2288
    },
    {
      "epoch": 30.93,
      "learning_rate": 0.00019268181818181816,
      "loss": 0.0442,
      "step": 2289
    },
    {
      "epoch": 30.95,
      "learning_rate": 0.00019254545454545455,
      "loss": 0.072,
      "step": 2290
    },
    {
      "epoch": 30.96,
      "learning_rate": 0.00019240909090909088,
      "loss": 0.0442,
      "step": 2291
    },
    {
      "epoch": 30.97,
      "learning_rate": 0.00019227272727272723,
      "loss": 0.1432,
      "step": 2292
    },
    {
      "epoch": 30.99,
      "learning_rate": 0.00019213636363636362,
      "loss": 0.1086,
      "step": 2293
    },
    {
      "epoch": 31.0,
      "learning_rate": 0.00019199999999999998,
      "loss": 0.0019,
      "step": 2294
    },
    {
      "epoch": 31.01,
      "learning_rate": 0.00019186363636363634,
      "loss": 0.072,
      "step": 2295
    },
    {
      "epoch": 31.03,
      "learning_rate": 0.0001917272727272727,
      "loss": 0.0661,
      "step": 2296
    },
    {
      "epoch": 31.04,
      "learning_rate": 0.00019159090909090908,
      "loss": 0.0205,
      "step": 2297
    },
    {
      "epoch": 31.05,
      "learning_rate": 0.00019145454545454544,
      "loss": 0.0637,
      "step": 2298
    },
    {
      "epoch": 31.07,
      "learning_rate": 0.0001913181818181818,
      "loss": 0.172,
      "step": 2299
    },
    {
      "epoch": 31.08,
      "learning_rate": 0.00019118181818181818,
      "loss": 0.0142,
      "step": 2300
    },
    {
      "epoch": 31.09,
      "learning_rate": 0.00019104545454545454,
      "loss": 0.0886,
      "step": 2301
    },
    {
      "epoch": 31.11,
      "learning_rate": 0.0001909090909090909,
      "loss": 0.005,
      "step": 2302
    },
    {
      "epoch": 31.12,
      "learning_rate": 0.00019077272727272725,
      "loss": 0.0372,
      "step": 2303
    },
    {
      "epoch": 31.14,
      "learning_rate": 0.00019063636363636364,
      "loss": 0.0723,
      "step": 2304
    },
    {
      "epoch": 31.15,
      "learning_rate": 0.0001905,
      "loss": 0.0081,
      "step": 2305
    },
    {
      "epoch": 31.16,
      "learning_rate": 0.00019036363636363635,
      "loss": 0.153,
      "step": 2306
    },
    {
      "epoch": 31.18,
      "learning_rate": 0.00019022727272727268,
      "loss": 0.0307,
      "step": 2307
    },
    {
      "epoch": 31.19,
      "learning_rate": 0.00019009090909090907,
      "loss": 0.0497,
      "step": 2308
    },
    {
      "epoch": 31.2,
      "learning_rate": 0.00018995454545454543,
      "loss": 0.0206,
      "step": 2309
    },
    {
      "epoch": 31.22,
      "learning_rate": 0.00018981818181818178,
      "loss": 0.0551,
      "step": 2310
    },
    {
      "epoch": 31.23,
      "learning_rate": 0.00018968181818181817,
      "loss": 0.1578,
      "step": 2311
    },
    {
      "epoch": 31.24,
      "learning_rate": 0.00018954545454545453,
      "loss": 0.0136,
      "step": 2312
    },
    {
      "epoch": 31.26,
      "learning_rate": 0.00018940909090909089,
      "loss": 0.1762,
      "step": 2313
    },
    {
      "epoch": 31.27,
      "learning_rate": 0.00018927272727272724,
      "loss": 0.0604,
      "step": 2314
    },
    {
      "epoch": 31.28,
      "learning_rate": 0.00018913636363636363,
      "loss": 0.0339,
      "step": 2315
    },
    {
      "epoch": 31.3,
      "learning_rate": 0.00018899999999999999,
      "loss": 0.0167,
      "step": 2316
    },
    {
      "epoch": 31.31,
      "learning_rate": 0.00018886363636363634,
      "loss": 0.0381,
      "step": 2317
    },
    {
      "epoch": 31.32,
      "learning_rate": 0.00018872727272727273,
      "loss": 0.0776,
      "step": 2318
    },
    {
      "epoch": 31.34,
      "learning_rate": 0.0001885909090909091,
      "loss": 0.0073,
      "step": 2319
    },
    {
      "epoch": 31.35,
      "learning_rate": 0.00018845454545454544,
      "loss": 0.0777,
      "step": 2320
    },
    {
      "epoch": 31.36,
      "learning_rate": 0.0001883181818181818,
      "loss": 0.0057,
      "step": 2321
    },
    {
      "epoch": 31.38,
      "learning_rate": 0.0001881818181818182,
      "loss": 0.0247,
      "step": 2322
    },
    {
      "epoch": 31.39,
      "learning_rate": 0.00018804545454545455,
      "loss": 0.0091,
      "step": 2323
    },
    {
      "epoch": 31.41,
      "learning_rate": 0.00018790909090909088,
      "loss": 0.0631,
      "step": 2324
    },
    {
      "epoch": 31.42,
      "learning_rate": 0.00018777272727272723,
      "loss": 0.0979,
      "step": 2325
    },
    {
      "epoch": 31.43,
      "learning_rate": 0.00018763636363636362,
      "loss": 0.0909,
      "step": 2326
    },
    {
      "epoch": 31.45,
      "learning_rate": 0.00018749999999999998,
      "loss": 0.0232,
      "step": 2327
    },
    {
      "epoch": 31.46,
      "learning_rate": 0.00018736363636363633,
      "loss": 0.05,
      "step": 2328
    },
    {
      "epoch": 31.47,
      "learning_rate": 0.00018722727272727272,
      "loss": 0.1449,
      "step": 2329
    },
    {
      "epoch": 31.49,
      "learning_rate": 0.00018709090909090908,
      "loss": 0.0249,
      "step": 2330
    },
    {
      "epoch": 31.5,
      "learning_rate": 0.00018695454545454544,
      "loss": 0.1014,
      "step": 2331
    },
    {
      "epoch": 31.51,
      "learning_rate": 0.0001868181818181818,
      "loss": 0.4306,
      "step": 2332
    },
    {
      "epoch": 31.53,
      "learning_rate": 0.00018668181818181818,
      "loss": 0.1899,
      "step": 2333
    },
    {
      "epoch": 31.54,
      "learning_rate": 0.00018654545454545454,
      "loss": 0.0857,
      "step": 2334
    },
    {
      "epoch": 31.55,
      "learning_rate": 0.0001864090909090909,
      "loss": 0.1644,
      "step": 2335
    },
    {
      "epoch": 31.57,
      "learning_rate": 0.00018627272727272725,
      "loss": 0.0915,
      "step": 2336
    },
    {
      "epoch": 31.58,
      "learning_rate": 0.00018613636363636364,
      "loss": 0.008,
      "step": 2337
    },
    {
      "epoch": 31.59,
      "learning_rate": 0.000186,
      "loss": 0.2789,
      "step": 2338
    },
    {
      "epoch": 31.61,
      "learning_rate": 0.00018586363636363633,
      "loss": 0.0183,
      "step": 2339
    },
    {
      "epoch": 31.62,
      "learning_rate": 0.00018572727272727274,
      "loss": 0.1705,
      "step": 2340
    },
    {
      "epoch": 31.64,
      "learning_rate": 0.00018559090909090907,
      "loss": 0.0803,
      "step": 2341
    },
    {
      "epoch": 31.65,
      "learning_rate": 0.00018545454545454543,
      "loss": 0.0715,
      "step": 2342
    },
    {
      "epoch": 31.66,
      "learning_rate": 0.00018531818181818178,
      "loss": 0.2476,
      "step": 2343
    },
    {
      "epoch": 31.68,
      "learning_rate": 0.00018518181818181817,
      "loss": 0.0195,
      "step": 2344
    },
    {
      "epoch": 31.69,
      "learning_rate": 0.00018504545454545453,
      "loss": 0.0474,
      "step": 2345
    },
    {
      "epoch": 31.7,
      "learning_rate": 0.00018490909090909088,
      "loss": 0.123,
      "step": 2346
    },
    {
      "epoch": 31.72,
      "learning_rate": 0.00018477272727272727,
      "loss": 0.0374,
      "step": 2347
    },
    {
      "epoch": 31.73,
      "learning_rate": 0.00018463636363636363,
      "loss": 0.0332,
      "step": 2348
    },
    {
      "epoch": 31.74,
      "learning_rate": 0.00018449999999999999,
      "loss": 0.1695,
      "step": 2349
    },
    {
      "epoch": 31.76,
      "learning_rate": 0.00018436363636363634,
      "loss": 0.1992,
      "step": 2350
    },
    {
      "epoch": 31.77,
      "learning_rate": 0.00018422727272727273,
      "loss": 0.0404,
      "step": 2351
    },
    {
      "epoch": 31.78,
      "learning_rate": 0.00018409090909090909,
      "loss": 0.2296,
      "step": 2352
    },
    {
      "epoch": 31.8,
      "learning_rate": 0.00018395454545454544,
      "loss": 0.2091,
      "step": 2353
    },
    {
      "epoch": 31.81,
      "learning_rate": 0.00018381818181818177,
      "loss": 0.0078,
      "step": 2354
    },
    {
      "epoch": 31.82,
      "learning_rate": 0.0001836818181818182,
      "loss": 0.1723,
      "step": 2355
    },
    {
      "epoch": 31.84,
      "learning_rate": 0.00018354545454545452,
      "loss": 0.0207,
      "step": 2356
    },
    {
      "epoch": 31.85,
      "learning_rate": 0.00018340909090909088,
      "loss": 0.0068,
      "step": 2357
    },
    {
      "epoch": 31.86,
      "learning_rate": 0.00018327272727272726,
      "loss": 0.0713,
      "step": 2358
    },
    {
      "epoch": 31.88,
      "learning_rate": 0.00018313636363636362,
      "loss": 0.1246,
      "step": 2359
    },
    {
      "epoch": 31.89,
      "learning_rate": 0.00018299999999999998,
      "loss": 0.0106,
      "step": 2360
    },
    {
      "epoch": 31.91,
      "learning_rate": 0.00018286363636363633,
      "loss": 0.2002,
      "step": 2361
    },
    {
      "epoch": 31.92,
      "learning_rate": 0.00018272727272727272,
      "loss": 0.0435,
      "step": 2362
    },
    {
      "epoch": 31.93,
      "learning_rate": 0.00018259090909090908,
      "loss": 0.0437,
      "step": 2363
    },
    {
      "epoch": 31.95,
      "learning_rate": 0.00018245454545454543,
      "loss": 0.3254,
      "step": 2364
    },
    {
      "epoch": 31.96,
      "learning_rate": 0.00018231818181818182,
      "loss": 0.0303,
      "step": 2365
    },
    {
      "epoch": 31.97,
      "learning_rate": 0.00018218181818181818,
      "loss": 0.0979,
      "step": 2366
    },
    {
      "epoch": 31.99,
      "learning_rate": 0.00018204545454545454,
      "loss": 0.1353,
      "step": 2367
    },
    {
      "epoch": 32.0,
      "learning_rate": 0.0001819090909090909,
      "loss": 0.0028,
      "step": 2368
    },
    {
      "epoch": 32.01,
      "learning_rate": 0.00018177272727272728,
      "loss": 0.0662,
      "step": 2369
    },
    {
      "epoch": 32.03,
      "learning_rate": 0.00018163636363636364,
      "loss": 0.0215,
      "step": 2370
    },
    {
      "epoch": 32.04,
      "learning_rate": 0.00018149999999999997,
      "loss": 0.1003,
      "step": 2371
    },
    {
      "epoch": 32.05,
      "learning_rate": 0.00018136363636363632,
      "loss": 0.0085,
      "step": 2372
    },
    {
      "epoch": 32.07,
      "learning_rate": 0.0001812272727272727,
      "loss": 0.015,
      "step": 2373
    },
    {
      "epoch": 32.08,
      "learning_rate": 0.00018109090909090907,
      "loss": 0.0067,
      "step": 2374
    },
    {
      "epoch": 32.09,
      "learning_rate": 0.00018095454545454543,
      "loss": 0.0093,
      "step": 2375
    },
    {
      "epoch": 32.11,
      "learning_rate": 0.0001808181818181818,
      "loss": 0.1371,
      "step": 2376
    },
    {
      "epoch": 32.12,
      "learning_rate": 0.00018068181818181817,
      "loss": 0.0841,
      "step": 2377
    },
    {
      "epoch": 32.14,
      "learning_rate": 0.00018054545454545453,
      "loss": 0.0772,
      "step": 2378
    },
    {
      "epoch": 32.15,
      "learning_rate": 0.00018040909090909088,
      "loss": 0.0053,
      "step": 2379
    },
    {
      "epoch": 32.16,
      "learning_rate": 0.00018027272727272727,
      "loss": 0.1558,
      "step": 2380
    },
    {
      "epoch": 32.18,
      "learning_rate": 0.00018013636363636363,
      "loss": 0.1868,
      "step": 2381
    },
    {
      "epoch": 32.19,
      "learning_rate": 0.00017999999999999998,
      "loss": 0.0822,
      "step": 2382
    },
    {
      "epoch": 32.2,
      "learning_rate": 0.00017986363636363634,
      "loss": 0.0046,
      "step": 2383
    },
    {
      "epoch": 32.22,
      "learning_rate": 0.00017972727272727273,
      "loss": 0.112,
      "step": 2384
    },
    {
      "epoch": 32.23,
      "learning_rate": 0.00017959090909090908,
      "loss": 0.005,
      "step": 2385
    },
    {
      "epoch": 32.24,
      "learning_rate": 0.00017945454545454542,
      "loss": 0.0444,
      "step": 2386
    },
    {
      "epoch": 32.26,
      "learning_rate": 0.00017931818181818183,
      "loss": 0.0294,
      "step": 2387
    },
    {
      "epoch": 32.27,
      "learning_rate": 0.00017918181818181816,
      "loss": 0.056,
      "step": 2388
    },
    {
      "epoch": 32.28,
      "learning_rate": 0.00017904545454545452,
      "loss": 0.007,
      "step": 2389
    },
    {
      "epoch": 32.3,
      "learning_rate": 0.00017890909090909087,
      "loss": 0.0045,
      "step": 2390
    },
    {
      "epoch": 32.31,
      "learning_rate": 0.00017877272727272726,
      "loss": 0.0891,
      "step": 2391
    },
    {
      "epoch": 32.32,
      "learning_rate": 0.00017863636363636362,
      "loss": 0.1266,
      "step": 2392
    },
    {
      "epoch": 32.34,
      "learning_rate": 0.00017849999999999997,
      "loss": 0.0984,
      "step": 2393
    },
    {
      "epoch": 32.35,
      "learning_rate": 0.00017836363636363636,
      "loss": 0.0168,
      "step": 2394
    },
    {
      "epoch": 32.36,
      "learning_rate": 0.00017822727272727272,
      "loss": 0.0991,
      "step": 2395
    },
    {
      "epoch": 32.38,
      "learning_rate": 0.00017809090909090908,
      "loss": 0.1488,
      "step": 2396
    },
    {
      "epoch": 32.39,
      "learning_rate": 0.00017795454545454543,
      "loss": 0.0066,
      "step": 2397
    },
    {
      "epoch": 32.41,
      "learning_rate": 0.00017781818181818182,
      "loss": 0.1299,
      "step": 2398
    },
    {
      "epoch": 32.42,
      "learning_rate": 0.00017768181818181818,
      "loss": 0.0182,
      "step": 2399
    },
    {
      "epoch": 32.43,
      "learning_rate": 0.00017754545454545453,
      "loss": 0.1235,
      "step": 2400
    },
    {
      "epoch": 32.43,
      "eval_accuracy": 0.9128048780487805,
      "eval_f1": 0.9133774522906959,
      "eval_loss": 0.37155774235725403,
      "eval_runtime": 125.9919,
      "eval_samples_per_second": 13.017,
      "eval_steps_per_second": 0.818,
      "step": 2400
    },
    {
      "epoch": 32.45,
      "learning_rate": 0.0001774090909090909,
      "loss": 0.1091,
      "step": 2401
    },
    {
      "epoch": 32.46,
      "learning_rate": 0.00017727272727272728,
      "loss": 0.2826,
      "step": 2402
    },
    {
      "epoch": 32.47,
      "learning_rate": 0.0001771363636363636,
      "loss": 0.0216,
      "step": 2403
    },
    {
      "epoch": 32.49,
      "learning_rate": 0.00017699999999999997,
      "loss": 0.1259,
      "step": 2404
    },
    {
      "epoch": 32.5,
      "learning_rate": 0.00017686363636363635,
      "loss": 0.0489,
      "step": 2405
    },
    {
      "epoch": 32.51,
      "learning_rate": 0.0001767272727272727,
      "loss": 0.0587,
      "step": 2406
    },
    {
      "epoch": 32.53,
      "learning_rate": 0.00017659090909090907,
      "loss": 0.0285,
      "step": 2407
    },
    {
      "epoch": 32.54,
      "learning_rate": 0.00017645454545454542,
      "loss": 0.0904,
      "step": 2408
    },
    {
      "epoch": 32.55,
      "learning_rate": 0.0001763181818181818,
      "loss": 0.0946,
      "step": 2409
    },
    {
      "epoch": 32.57,
      "learning_rate": 0.00017618181818181817,
      "loss": 0.1204,
      "step": 2410
    },
    {
      "epoch": 32.58,
      "learning_rate": 0.00017604545454545452,
      "loss": 0.2206,
      "step": 2411
    },
    {
      "epoch": 32.59,
      "learning_rate": 0.00017590909090909088,
      "loss": 0.0835,
      "step": 2412
    },
    {
      "epoch": 32.61,
      "learning_rate": 0.00017577272727272727,
      "loss": 0.0455,
      "step": 2413
    },
    {
      "epoch": 32.62,
      "learning_rate": 0.00017563636363636363,
      "loss": 0.0782,
      "step": 2414
    },
    {
      "epoch": 32.64,
      "learning_rate": 0.00017549999999999998,
      "loss": 0.0275,
      "step": 2415
    },
    {
      "epoch": 32.65,
      "learning_rate": 0.00017536363636363637,
      "loss": 0.132,
      "step": 2416
    },
    {
      "epoch": 32.66,
      "learning_rate": 0.00017522727272727273,
      "loss": 0.0122,
      "step": 2417
    },
    {
      "epoch": 32.68,
      "learning_rate": 0.00017509090909090908,
      "loss": 0.0099,
      "step": 2418
    },
    {
      "epoch": 32.69,
      "learning_rate": 0.00017495454545454541,
      "loss": 0.1056,
      "step": 2419
    },
    {
      "epoch": 32.7,
      "learning_rate": 0.0001748181818181818,
      "loss": 0.2305,
      "step": 2420
    },
    {
      "epoch": 32.72,
      "learning_rate": 0.00017468181818181816,
      "loss": 0.0101,
      "step": 2421
    },
    {
      "epoch": 32.73,
      "learning_rate": 0.00017454545454545452,
      "loss": 0.0671,
      "step": 2422
    },
    {
      "epoch": 32.74,
      "learning_rate": 0.0001744090909090909,
      "loss": 0.0407,
      "step": 2423
    },
    {
      "epoch": 32.76,
      "learning_rate": 0.00017427272727272726,
      "loss": 0.0077,
      "step": 2424
    },
    {
      "epoch": 32.77,
      "learning_rate": 0.00017413636363636362,
      "loss": 0.1073,
      "step": 2425
    },
    {
      "epoch": 32.78,
      "learning_rate": 0.00017399999999999997,
      "loss": 0.11,
      "step": 2426
    },
    {
      "epoch": 32.8,
      "learning_rate": 0.00017386363636363636,
      "loss": 0.017,
      "step": 2427
    },
    {
      "epoch": 32.81,
      "learning_rate": 0.00017372727272727272,
      "loss": 0.1183,
      "step": 2428
    },
    {
      "epoch": 32.82,
      "learning_rate": 0.00017359090909090907,
      "loss": 0.1131,
      "step": 2429
    },
    {
      "epoch": 32.84,
      "learning_rate": 0.00017345454545454543,
      "loss": 0.1418,
      "step": 2430
    },
    {
      "epoch": 32.85,
      "learning_rate": 0.00017331818181818182,
      "loss": 0.0942,
      "step": 2431
    },
    {
      "epoch": 32.86,
      "learning_rate": 0.00017318181818181818,
      "loss": 0.2103,
      "step": 2432
    },
    {
      "epoch": 32.88,
      "learning_rate": 0.00017304545454545453,
      "loss": 0.0732,
      "step": 2433
    },
    {
      "epoch": 32.89,
      "learning_rate": 0.00017290909090909092,
      "loss": 0.0621,
      "step": 2434
    },
    {
      "epoch": 32.91,
      "learning_rate": 0.00017277272727272728,
      "loss": 0.1071,
      "step": 2435
    },
    {
      "epoch": 32.92,
      "learning_rate": 0.0001726363636363636,
      "loss": 0.0593,
      "step": 2436
    },
    {
      "epoch": 32.93,
      "learning_rate": 0.00017249999999999996,
      "loss": 0.0031,
      "step": 2437
    },
    {
      "epoch": 32.95,
      "learning_rate": 0.00017236363636363635,
      "loss": 0.0433,
      "step": 2438
    },
    {
      "epoch": 32.96,
      "learning_rate": 0.0001722272727272727,
      "loss": 0.0504,
      "step": 2439
    },
    {
      "epoch": 32.97,
      "learning_rate": 0.00017209090909090907,
      "loss": 0.0026,
      "step": 2440
    },
    {
      "epoch": 32.99,
      "learning_rate": 0.00017195454545454545,
      "loss": 0.0791,
      "step": 2441
    },
    {
      "epoch": 33.0,
      "learning_rate": 0.0001718181818181818,
      "loss": 0.0031,
      "step": 2442
    },
    {
      "epoch": 33.01,
      "learning_rate": 0.00017168181818181817,
      "loss": 0.0197,
      "step": 2443
    },
    {
      "epoch": 33.03,
      "learning_rate": 0.00017154545454545452,
      "loss": 0.0215,
      "step": 2444
    },
    {
      "epoch": 33.04,
      "learning_rate": 0.0001714090909090909,
      "loss": 0.1159,
      "step": 2445
    },
    {
      "epoch": 33.05,
      "learning_rate": 0.00017127272727272727,
      "loss": 0.015,
      "step": 2446
    },
    {
      "epoch": 33.07,
      "learning_rate": 0.00017113636363636362,
      "loss": 0.1175,
      "step": 2447
    },
    {
      "epoch": 33.08,
      "learning_rate": 0.00017099999999999998,
      "loss": 0.0054,
      "step": 2448
    },
    {
      "epoch": 33.09,
      "learning_rate": 0.00017086363636363637,
      "loss": 0.1216,
      "step": 2449
    },
    {
      "epoch": 33.11,
      "learning_rate": 0.00017072727272727273,
      "loss": 0.0086,
      "step": 2450
    },
    {
      "epoch": 33.12,
      "learning_rate": 0.00017059090909090906,
      "loss": 0.0595,
      "step": 2451
    },
    {
      "epoch": 33.14,
      "learning_rate": 0.00017045454545454547,
      "loss": 0.2143,
      "step": 2452
    },
    {
      "epoch": 33.15,
      "learning_rate": 0.0001703181818181818,
      "loss": 0.064,
      "step": 2453
    },
    {
      "epoch": 33.16,
      "learning_rate": 0.00017018181818181816,
      "loss": 0.1046,
      "step": 2454
    },
    {
      "epoch": 33.18,
      "learning_rate": 0.00017004545454545451,
      "loss": 0.1097,
      "step": 2455
    },
    {
      "epoch": 33.19,
      "learning_rate": 0.0001699090909090909,
      "loss": 0.0508,
      "step": 2456
    },
    {
      "epoch": 33.2,
      "learning_rate": 0.00016977272727272726,
      "loss": 0.0168,
      "step": 2457
    },
    {
      "epoch": 33.22,
      "learning_rate": 0.00016963636363636362,
      "loss": 0.0964,
      "step": 2458
    },
    {
      "epoch": 33.23,
      "learning_rate": 0.00016949999999999997,
      "loss": 0.0804,
      "step": 2459
    },
    {
      "epoch": 33.24,
      "learning_rate": 0.00016936363636363636,
      "loss": 0.0068,
      "step": 2460
    },
    {
      "epoch": 33.26,
      "learning_rate": 0.00016922727272727272,
      "loss": 0.0279,
      "step": 2461
    },
    {
      "epoch": 33.27,
      "learning_rate": 0.00016909090909090907,
      "loss": 0.0192,
      "step": 2462
    },
    {
      "epoch": 33.28,
      "learning_rate": 0.00016895454545454546,
      "loss": 0.0682,
      "step": 2463
    },
    {
      "epoch": 33.3,
      "learning_rate": 0.00016881818181818182,
      "loss": 0.0362,
      "step": 2464
    },
    {
      "epoch": 33.31,
      "learning_rate": 0.00016868181818181817,
      "loss": 0.0325,
      "step": 2465
    },
    {
      "epoch": 33.32,
      "learning_rate": 0.0001685454545454545,
      "loss": 0.019,
      "step": 2466
    },
    {
      "epoch": 33.34,
      "learning_rate": 0.00016840909090909092,
      "loss": 0.1534,
      "step": 2467
    },
    {
      "epoch": 33.35,
      "learning_rate": 0.00016827272727272725,
      "loss": 0.066,
      "step": 2468
    },
    {
      "epoch": 33.36,
      "learning_rate": 0.0001681363636363636,
      "loss": 0.0037,
      "step": 2469
    },
    {
      "epoch": 33.38,
      "learning_rate": 0.000168,
      "loss": 0.1713,
      "step": 2470
    },
    {
      "epoch": 33.39,
      "learning_rate": 0.00016786363636363635,
      "loss": 0.0579,
      "step": 2471
    },
    {
      "epoch": 33.41,
      "learning_rate": 0.0001677272727272727,
      "loss": 0.005,
      "step": 2472
    },
    {
      "epoch": 33.42,
      "learning_rate": 0.00016759090909090906,
      "loss": 0.0021,
      "step": 2473
    },
    {
      "epoch": 33.43,
      "learning_rate": 0.00016745454545454545,
      "loss": 0.0028,
      "step": 2474
    },
    {
      "epoch": 33.45,
      "learning_rate": 0.0001673181818181818,
      "loss": 0.0731,
      "step": 2475
    },
    {
      "epoch": 33.46,
      "learning_rate": 0.00016718181818181816,
      "loss": 0.277,
      "step": 2476
    },
    {
      "epoch": 33.47,
      "learning_rate": 0.00016704545454545452,
      "loss": 0.0663,
      "step": 2477
    },
    {
      "epoch": 33.49,
      "learning_rate": 0.0001669090909090909,
      "loss": 0.0278,
      "step": 2478
    },
    {
      "epoch": 33.5,
      "learning_rate": 0.00016677272727272727,
      "loss": 0.0295,
      "step": 2479
    },
    {
      "epoch": 33.51,
      "learning_rate": 0.00016663636363636362,
      "loss": 0.005,
      "step": 2480
    },
    {
      "epoch": 33.53,
      "learning_rate": 0.0001665,
      "loss": 0.0225,
      "step": 2481
    },
    {
      "epoch": 33.54,
      "learning_rate": 0.00016636363636363637,
      "loss": 0.0539,
      "step": 2482
    },
    {
      "epoch": 33.55,
      "learning_rate": 0.0001662272727272727,
      "loss": 0.009,
      "step": 2483
    },
    {
      "epoch": 33.57,
      "learning_rate": 0.00016609090909090905,
      "loss": 0.0287,
      "step": 2484
    },
    {
      "epoch": 33.58,
      "learning_rate": 0.00016595454545454544,
      "loss": 0.0346,
      "step": 2485
    },
    {
      "epoch": 33.59,
      "learning_rate": 0.0001658181818181818,
      "loss": 0.1232,
      "step": 2486
    },
    {
      "epoch": 33.61,
      "learning_rate": 0.00016568181818181816,
      "loss": 0.1098,
      "step": 2487
    },
    {
      "epoch": 33.62,
      "learning_rate": 0.0001655454545454545,
      "loss": 0.1968,
      "step": 2488
    },
    {
      "epoch": 33.64,
      "learning_rate": 0.0001654090909090909,
      "loss": 0.0241,
      "step": 2489
    },
    {
      "epoch": 33.65,
      "learning_rate": 0.00016527272727272726,
      "loss": 0.1126,
      "step": 2490
    },
    {
      "epoch": 33.66,
      "learning_rate": 0.00016513636363636361,
      "loss": 0.0348,
      "step": 2491
    },
    {
      "epoch": 33.68,
      "learning_rate": 0.000165,
      "loss": 0.0291,
      "step": 2492
    },
    {
      "epoch": 33.69,
      "learning_rate": 0.00016486363636363636,
      "loss": 0.0783,
      "step": 2493
    },
    {
      "epoch": 33.7,
      "learning_rate": 0.00016472727272727271,
      "loss": 0.0501,
      "step": 2494
    },
    {
      "epoch": 33.72,
      "learning_rate": 0.00016459090909090907,
      "loss": 0.0823,
      "step": 2495
    },
    {
      "epoch": 33.73,
      "learning_rate": 0.00016445454545454546,
      "loss": 0.1524,
      "step": 2496
    },
    {
      "epoch": 33.74,
      "learning_rate": 0.00016431818181818182,
      "loss": 0.0479,
      "step": 2497
    },
    {
      "epoch": 33.76,
      "learning_rate": 0.00016418181818181815,
      "loss": 0.1178,
      "step": 2498
    },
    {
      "epoch": 33.77,
      "learning_rate": 0.00016404545454545456,
      "loss": 0.0088,
      "step": 2499
    },
    {
      "epoch": 33.78,
      "learning_rate": 0.0001639090909090909,
      "loss": 0.192,
      "step": 2500
    },
    {
      "epoch": 33.8,
      "learning_rate": 0.00016377272727272725,
      "loss": 0.1047,
      "step": 2501
    },
    {
      "epoch": 33.81,
      "learning_rate": 0.0001636363636363636,
      "loss": 0.0913,
      "step": 2502
    },
    {
      "epoch": 33.82,
      "learning_rate": 0.0001635,
      "loss": 0.128,
      "step": 2503
    },
    {
      "epoch": 33.84,
      "learning_rate": 0.00016336363636363635,
      "loss": 0.0631,
      "step": 2504
    },
    {
      "epoch": 33.85,
      "learning_rate": 0.0001632272727272727,
      "loss": 0.1836,
      "step": 2505
    },
    {
      "epoch": 33.86,
      "learning_rate": 0.00016309090909090906,
      "loss": 0.0353,
      "step": 2506
    },
    {
      "epoch": 33.88,
      "learning_rate": 0.00016295454545454545,
      "loss": 0.1644,
      "step": 2507
    },
    {
      "epoch": 33.89,
      "learning_rate": 0.0001628181818181818,
      "loss": 0.105,
      "step": 2508
    },
    {
      "epoch": 33.91,
      "learning_rate": 0.00016268181818181816,
      "loss": 0.0829,
      "step": 2509
    },
    {
      "epoch": 33.92,
      "learning_rate": 0.00016254545454545455,
      "loss": 0.2486,
      "step": 2510
    },
    {
      "epoch": 33.93,
      "learning_rate": 0.0001624090909090909,
      "loss": 0.0591,
      "step": 2511
    },
    {
      "epoch": 33.95,
      "learning_rate": 0.00016227272727272726,
      "loss": 0.2114,
      "step": 2512
    },
    {
      "epoch": 33.96,
      "learning_rate": 0.0001621363636363636,
      "loss": 0.0693,
      "step": 2513
    },
    {
      "epoch": 33.97,
      "learning_rate": 0.000162,
      "loss": 0.1027,
      "step": 2514
    },
    {
      "epoch": 33.99,
      "learning_rate": 0.00016186363636363634,
      "loss": 0.0076,
      "step": 2515
    },
    {
      "epoch": 34.0,
      "learning_rate": 0.0001617272727272727,
      "loss": 0.0038,
      "step": 2516
    },
    {
      "epoch": 34.01,
      "learning_rate": 0.00016159090909090908,
      "loss": 0.0086,
      "step": 2517
    },
    {
      "epoch": 34.03,
      "learning_rate": 0.00016145454545454544,
      "loss": 0.0077,
      "step": 2518
    },
    {
      "epoch": 34.04,
      "learning_rate": 0.0001613181818181818,
      "loss": 0.0554,
      "step": 2519
    },
    {
      "epoch": 34.05,
      "learning_rate": 0.00016118181818181815,
      "loss": 0.0158,
      "step": 2520
    },
    {
      "epoch": 34.07,
      "learning_rate": 0.00016104545454545454,
      "loss": 0.0548,
      "step": 2521
    },
    {
      "epoch": 34.08,
      "learning_rate": 0.0001609090909090909,
      "loss": 0.0872,
      "step": 2522
    },
    {
      "epoch": 34.09,
      "learning_rate": 0.00016077272727272726,
      "loss": 0.0134,
      "step": 2523
    },
    {
      "epoch": 34.11,
      "learning_rate": 0.0001606363636363636,
      "loss": 0.0136,
      "step": 2524
    },
    {
      "epoch": 34.12,
      "learning_rate": 0.0001605,
      "loss": 0.0333,
      "step": 2525
    },
    {
      "epoch": 34.14,
      "learning_rate": 0.00016036363636363636,
      "loss": 0.0924,
      "step": 2526
    },
    {
      "epoch": 34.15,
      "learning_rate": 0.00016022727272727271,
      "loss": 0.0673,
      "step": 2527
    },
    {
      "epoch": 34.16,
      "learning_rate": 0.0001600909090909091,
      "loss": 0.1784,
      "step": 2528
    },
    {
      "epoch": 34.18,
      "learning_rate": 0.00015995454545454546,
      "loss": 0.2059,
      "step": 2529
    },
    {
      "epoch": 34.19,
      "learning_rate": 0.0001598181818181818,
      "loss": 0.0432,
      "step": 2530
    },
    {
      "epoch": 34.2,
      "learning_rate": 0.00015968181818181815,
      "loss": 0.0385,
      "step": 2531
    },
    {
      "epoch": 34.22,
      "learning_rate": 0.00015954545454545453,
      "loss": 0.0569,
      "step": 2532
    },
    {
      "epoch": 34.23,
      "learning_rate": 0.0001594090909090909,
      "loss": 0.0156,
      "step": 2533
    },
    {
      "epoch": 34.24,
      "learning_rate": 0.00015927272727272725,
      "loss": 0.0272,
      "step": 2534
    },
    {
      "epoch": 34.26,
      "learning_rate": 0.0001591363636363636,
      "loss": 0.0732,
      "step": 2535
    },
    {
      "epoch": 34.27,
      "learning_rate": 0.000159,
      "loss": 0.0282,
      "step": 2536
    },
    {
      "epoch": 34.28,
      "learning_rate": 0.00015886363636363635,
      "loss": 0.0053,
      "step": 2537
    },
    {
      "epoch": 34.3,
      "learning_rate": 0.0001587272727272727,
      "loss": 0.0201,
      "step": 2538
    },
    {
      "epoch": 34.31,
      "learning_rate": 0.0001585909090909091,
      "loss": 0.0674,
      "step": 2539
    },
    {
      "epoch": 34.32,
      "learning_rate": 0.00015845454545454545,
      "loss": 0.0681,
      "step": 2540
    },
    {
      "epoch": 34.34,
      "learning_rate": 0.0001583181818181818,
      "loss": 0.2394,
      "step": 2541
    },
    {
      "epoch": 34.35,
      "learning_rate": 0.00015818181818181816,
      "loss": 0.0918,
      "step": 2542
    },
    {
      "epoch": 34.36,
      "learning_rate": 0.00015804545454545455,
      "loss": 0.0156,
      "step": 2543
    },
    {
      "epoch": 34.38,
      "learning_rate": 0.0001579090909090909,
      "loss": 0.1176,
      "step": 2544
    },
    {
      "epoch": 34.39,
      "learning_rate": 0.00015777272727272726,
      "loss": 0.0345,
      "step": 2545
    },
    {
      "epoch": 34.41,
      "learning_rate": 0.00015763636363636365,
      "loss": 0.009,
      "step": 2546
    },
    {
      "epoch": 34.42,
      "learning_rate": 0.00015749999999999998,
      "loss": 0.0275,
      "step": 2547
    },
    {
      "epoch": 34.43,
      "learning_rate": 0.00015736363636363634,
      "loss": 0.0095,
      "step": 2548
    },
    {
      "epoch": 34.45,
      "learning_rate": 0.0001572272727272727,
      "loss": 0.0032,
      "step": 2549
    },
    {
      "epoch": 34.46,
      "learning_rate": 0.00015709090909090908,
      "loss": 0.066,
      "step": 2550
    },
    {
      "epoch": 34.47,
      "learning_rate": 0.00015695454545454544,
      "loss": 0.1519,
      "step": 2551
    },
    {
      "epoch": 34.49,
      "learning_rate": 0.0001568181818181818,
      "loss": 0.145,
      "step": 2552
    },
    {
      "epoch": 34.5,
      "learning_rate": 0.00015668181818181815,
      "loss": 0.0483,
      "step": 2553
    },
    {
      "epoch": 34.51,
      "learning_rate": 0.00015654545454545454,
      "loss": 0.021,
      "step": 2554
    },
    {
      "epoch": 34.53,
      "learning_rate": 0.0001564090909090909,
      "loss": 0.0022,
      "step": 2555
    },
    {
      "epoch": 34.54,
      "learning_rate": 0.00015627272727272725,
      "loss": 0.0147,
      "step": 2556
    },
    {
      "epoch": 34.55,
      "learning_rate": 0.00015613636363636364,
      "loss": 0.017,
      "step": 2557
    },
    {
      "epoch": 34.57,
      "learning_rate": 0.000156,
      "loss": 0.0284,
      "step": 2558
    },
    {
      "epoch": 34.58,
      "learning_rate": 0.00015586363636363636,
      "loss": 0.0038,
      "step": 2559
    },
    {
      "epoch": 34.59,
      "learning_rate": 0.0001557272727272727,
      "loss": 0.0837,
      "step": 2560
    },
    {
      "epoch": 34.61,
      "learning_rate": 0.0001555909090909091,
      "loss": 0.006,
      "step": 2561
    },
    {
      "epoch": 34.62,
      "learning_rate": 0.00015545454545454546,
      "loss": 0.0904,
      "step": 2562
    },
    {
      "epoch": 34.64,
      "learning_rate": 0.0001553181818181818,
      "loss": 0.0054,
      "step": 2563
    },
    {
      "epoch": 34.65,
      "learning_rate": 0.00015518181818181814,
      "loss": 0.1394,
      "step": 2564
    },
    {
      "epoch": 34.66,
      "learning_rate": 0.00015504545454545453,
      "loss": 0.005,
      "step": 2565
    },
    {
      "epoch": 34.68,
      "learning_rate": 0.0001549090909090909,
      "loss": 0.1438,
      "step": 2566
    },
    {
      "epoch": 34.69,
      "learning_rate": 0.00015477272727272725,
      "loss": 0.1457,
      "step": 2567
    },
    {
      "epoch": 34.7,
      "learning_rate": 0.00015463636363636363,
      "loss": 0.0933,
      "step": 2568
    },
    {
      "epoch": 34.72,
      "learning_rate": 0.0001545,
      "loss": 0.0026,
      "step": 2569
    },
    {
      "epoch": 34.73,
      "learning_rate": 0.00015436363636363635,
      "loss": 0.1825,
      "step": 2570
    },
    {
      "epoch": 34.74,
      "learning_rate": 0.0001542272727272727,
      "loss": 0.0025,
      "step": 2571
    },
    {
      "epoch": 34.76,
      "learning_rate": 0.0001540909090909091,
      "loss": 0.0554,
      "step": 2572
    },
    {
      "epoch": 34.77,
      "learning_rate": 0.00015395454545454545,
      "loss": 0.0079,
      "step": 2573
    },
    {
      "epoch": 34.78,
      "learning_rate": 0.0001538181818181818,
      "loss": 0.0031,
      "step": 2574
    },
    {
      "epoch": 34.8,
      "learning_rate": 0.0001536818181818182,
      "loss": 0.0816,
      "step": 2575
    },
    {
      "epoch": 34.81,
      "learning_rate": 0.00015354545454545455,
      "loss": 0.1835,
      "step": 2576
    },
    {
      "epoch": 34.82,
      "learning_rate": 0.0001534090909090909,
      "loss": 0.0871,
      "step": 2577
    },
    {
      "epoch": 34.84,
      "learning_rate": 0.00015327272727272724,
      "loss": 0.0525,
      "step": 2578
    },
    {
      "epoch": 34.85,
      "learning_rate": 0.00015313636363636365,
      "loss": 0.024,
      "step": 2579
    },
    {
      "epoch": 34.86,
      "learning_rate": 0.00015299999999999998,
      "loss": 0.0448,
      "step": 2580
    },
    {
      "epoch": 34.88,
      "learning_rate": 0.00015286363636363634,
      "loss": 0.1355,
      "step": 2581
    },
    {
      "epoch": 34.89,
      "learning_rate": 0.0001527272727272727,
      "loss": 0.1749,
      "step": 2582
    },
    {
      "epoch": 34.91,
      "learning_rate": 0.00015259090909090908,
      "loss": 0.018,
      "step": 2583
    },
    {
      "epoch": 34.92,
      "learning_rate": 0.00015245454545454544,
      "loss": 0.0069,
      "step": 2584
    },
    {
      "epoch": 34.93,
      "learning_rate": 0.0001523181818181818,
      "loss": 0.0274,
      "step": 2585
    },
    {
      "epoch": 34.95,
      "learning_rate": 0.00015218181818181818,
      "loss": 0.0415,
      "step": 2586
    },
    {
      "epoch": 34.96,
      "learning_rate": 0.00015204545454545454,
      "loss": 0.0064,
      "step": 2587
    },
    {
      "epoch": 34.97,
      "learning_rate": 0.0001519090909090909,
      "loss": 0.073,
      "step": 2588
    },
    {
      "epoch": 34.99,
      "learning_rate": 0.00015177272727272725,
      "loss": 0.0522,
      "step": 2589
    },
    {
      "epoch": 35.0,
      "learning_rate": 0.00015163636363636364,
      "loss": 0.0026,
      "step": 2590
    },
    {
      "epoch": 35.01,
      "learning_rate": 0.0001515,
      "loss": 0.0451,
      "step": 2591
    },
    {
      "epoch": 35.03,
      "learning_rate": 0.00015136363636363635,
      "loss": 0.0343,
      "step": 2592
    },
    {
      "epoch": 35.04,
      "learning_rate": 0.00015122727272727274,
      "loss": 0.1028,
      "step": 2593
    },
    {
      "epoch": 35.05,
      "learning_rate": 0.0001510909090909091,
      "loss": 0.0907,
      "step": 2594
    },
    {
      "epoch": 35.07,
      "learning_rate": 0.00015095454545454543,
      "loss": 0.0104,
      "step": 2595
    },
    {
      "epoch": 35.08,
      "learning_rate": 0.00015081818181818179,
      "loss": 0.0383,
      "step": 2596
    },
    {
      "epoch": 35.09,
      "learning_rate": 0.00015068181818181817,
      "loss": 0.0644,
      "step": 2597
    },
    {
      "epoch": 35.11,
      "learning_rate": 0.00015054545454545453,
      "loss": 0.0225,
      "step": 2598
    },
    {
      "epoch": 35.12,
      "learning_rate": 0.00015040909090909089,
      "loss": 0.0056,
      "step": 2599
    },
    {
      "epoch": 35.14,
      "learning_rate": 0.00015027272727272724,
      "loss": 0.1873,
      "step": 2600
    },
    {
      "epoch": 35.14,
      "eval_accuracy": 0.9310975609756098,
      "eval_f1": 0.9314209347501917,
      "eval_loss": 0.3080303370952606,
      "eval_runtime": 126.4881,
      "eval_samples_per_second": 12.966,
      "eval_steps_per_second": 0.814,
      "step": 2600
    },
    {
      "epoch": 35.15,
      "learning_rate": 0.00015013636363636363,
      "loss": 0.0101,
      "step": 2601
    },
    {
      "epoch": 35.16,
      "learning_rate": 0.00015,
      "loss": 0.0161,
      "step": 2602
    },
    {
      "epoch": 35.18,
      "learning_rate": 0.00014986363636363634,
      "loss": 0.0085,
      "step": 2603
    },
    {
      "epoch": 35.19,
      "learning_rate": 0.0001497272727272727,
      "loss": 0.0068,
      "step": 2604
    },
    {
      "epoch": 35.2,
      "learning_rate": 0.0001495909090909091,
      "loss": 0.0053,
      "step": 2605
    },
    {
      "epoch": 35.22,
      "learning_rate": 0.00014945454545454545,
      "loss": 0.0088,
      "step": 2606
    },
    {
      "epoch": 35.23,
      "learning_rate": 0.0001493181818181818,
      "loss": 0.0029,
      "step": 2607
    },
    {
      "epoch": 35.24,
      "learning_rate": 0.00014918181818181816,
      "loss": 0.0986,
      "step": 2608
    },
    {
      "epoch": 35.26,
      "learning_rate": 0.00014904545454545455,
      "loss": 0.0037,
      "step": 2609
    },
    {
      "epoch": 35.27,
      "learning_rate": 0.00014890909090909088,
      "loss": 0.0892,
      "step": 2610
    },
    {
      "epoch": 35.28,
      "learning_rate": 0.00014877272727272726,
      "loss": 0.142,
      "step": 2611
    },
    {
      "epoch": 35.3,
      "learning_rate": 0.00014863636363636362,
      "loss": 0.0403,
      "step": 2612
    },
    {
      "epoch": 35.31,
      "learning_rate": 0.00014849999999999998,
      "loss": 0.0039,
      "step": 2613
    },
    {
      "epoch": 35.32,
      "learning_rate": 0.00014836363636363636,
      "loss": 0.0029,
      "step": 2614
    },
    {
      "epoch": 35.34,
      "learning_rate": 0.00014822727272727272,
      "loss": 0.153,
      "step": 2615
    },
    {
      "epoch": 35.35,
      "learning_rate": 0.00014809090909090908,
      "loss": 0.0109,
      "step": 2616
    },
    {
      "epoch": 35.36,
      "learning_rate": 0.00014795454545454544,
      "loss": 0.0449,
      "step": 2617
    },
    {
      "epoch": 35.38,
      "learning_rate": 0.00014781818181818182,
      "loss": 0.0056,
      "step": 2618
    },
    {
      "epoch": 35.39,
      "learning_rate": 0.00014768181818181815,
      "loss": 0.0062,
      "step": 2619
    },
    {
      "epoch": 35.41,
      "learning_rate": 0.00014754545454545454,
      "loss": 0.0179,
      "step": 2620
    },
    {
      "epoch": 35.42,
      "learning_rate": 0.0001474090909090909,
      "loss": 0.0018,
      "step": 2621
    },
    {
      "epoch": 35.43,
      "learning_rate": 0.00014727272727272725,
      "loss": 0.0205,
      "step": 2622
    },
    {
      "epoch": 35.45,
      "learning_rate": 0.00014713636363636364,
      "loss": 0.0025,
      "step": 2623
    },
    {
      "epoch": 35.46,
      "learning_rate": 0.000147,
      "loss": 0.11,
      "step": 2624
    },
    {
      "epoch": 35.47,
      "learning_rate": 0.00014686363636363635,
      "loss": 0.0156,
      "step": 2625
    },
    {
      "epoch": 35.49,
      "learning_rate": 0.0001467272727272727,
      "loss": 0.026,
      "step": 2626
    },
    {
      "epoch": 35.5,
      "learning_rate": 0.00014659090909090907,
      "loss": 0.0075,
      "step": 2627
    },
    {
      "epoch": 35.51,
      "learning_rate": 0.00014645454545454543,
      "loss": 0.0091,
      "step": 2628
    },
    {
      "epoch": 35.53,
      "learning_rate": 0.0001463181818181818,
      "loss": 0.0464,
      "step": 2629
    },
    {
      "epoch": 35.54,
      "learning_rate": 0.00014618181818181817,
      "loss": 0.052,
      "step": 2630
    },
    {
      "epoch": 35.55,
      "learning_rate": 0.00014604545454545453,
      "loss": 0.1407,
      "step": 2631
    },
    {
      "epoch": 35.57,
      "learning_rate": 0.0001459090909090909,
      "loss": 0.005,
      "step": 2632
    },
    {
      "epoch": 35.58,
      "learning_rate": 0.00014577272727272727,
      "loss": 0.1875,
      "step": 2633
    },
    {
      "epoch": 35.59,
      "learning_rate": 0.00014563636363636363,
      "loss": 0.0017,
      "step": 2634
    },
    {
      "epoch": 35.61,
      "learning_rate": 0.00014549999999999999,
      "loss": 0.0027,
      "step": 2635
    },
    {
      "epoch": 35.62,
      "learning_rate": 0.00014536363636363634,
      "loss": 0.029,
      "step": 2636
    },
    {
      "epoch": 35.64,
      "learning_rate": 0.0001452272727272727,
      "loss": 0.0048,
      "step": 2637
    },
    {
      "epoch": 35.65,
      "learning_rate": 0.0001450909090909091,
      "loss": 0.0753,
      "step": 2638
    },
    {
      "epoch": 35.66,
      "learning_rate": 0.00014495454545454544,
      "loss": 0.032,
      "step": 2639
    },
    {
      "epoch": 35.68,
      "learning_rate": 0.0001448181818181818,
      "loss": 0.0928,
      "step": 2640
    },
    {
      "epoch": 35.69,
      "learning_rate": 0.0001446818181818182,
      "loss": 0.0011,
      "step": 2641
    },
    {
      "epoch": 35.7,
      "learning_rate": 0.00014454545454545452,
      "loss": 0.0538,
      "step": 2642
    },
    {
      "epoch": 35.72,
      "learning_rate": 0.0001444090909090909,
      "loss": 0.0019,
      "step": 2643
    },
    {
      "epoch": 35.73,
      "learning_rate": 0.00014427272727272726,
      "loss": 0.0494,
      "step": 2644
    },
    {
      "epoch": 35.74,
      "learning_rate": 0.00014413636363636362,
      "loss": 0.1645,
      "step": 2645
    },
    {
      "epoch": 35.76,
      "learning_rate": 0.00014399999999999998,
      "loss": 0.0207,
      "step": 2646
    },
    {
      "epoch": 35.77,
      "learning_rate": 0.00014386363636363636,
      "loss": 0.1069,
      "step": 2647
    },
    {
      "epoch": 35.78,
      "learning_rate": 0.00014372727272727272,
      "loss": 0.0044,
      "step": 2648
    },
    {
      "epoch": 35.8,
      "learning_rate": 0.00014359090909090908,
      "loss": 0.0294,
      "step": 2649
    },
    {
      "epoch": 35.81,
      "learning_rate": 0.00014345454545454546,
      "loss": 0.0083,
      "step": 2650
    },
    {
      "epoch": 35.82,
      "learning_rate": 0.0001433181818181818,
      "loss": 0.0493,
      "step": 2651
    },
    {
      "epoch": 35.84,
      "learning_rate": 0.00014318181818181818,
      "loss": 0.0017,
      "step": 2652
    },
    {
      "epoch": 35.85,
      "learning_rate": 0.00014304545454545454,
      "loss": 0.0428,
      "step": 2653
    },
    {
      "epoch": 35.86,
      "learning_rate": 0.0001429090909090909,
      "loss": 0.1183,
      "step": 2654
    },
    {
      "epoch": 35.88,
      "learning_rate": 0.00014277272727272725,
      "loss": 0.0257,
      "step": 2655
    },
    {
      "epoch": 35.89,
      "learning_rate": 0.00014263636363636364,
      "loss": 0.1095,
      "step": 2656
    },
    {
      "epoch": 35.91,
      "learning_rate": 0.0001425,
      "loss": 0.0236,
      "step": 2657
    },
    {
      "epoch": 35.92,
      "learning_rate": 0.00014236363636363635,
      "loss": 0.025,
      "step": 2658
    },
    {
      "epoch": 35.93,
      "learning_rate": 0.0001422272727272727,
      "loss": 0.0126,
      "step": 2659
    },
    {
      "epoch": 35.95,
      "learning_rate": 0.00014209090909090907,
      "loss": 0.0531,
      "step": 2660
    },
    {
      "epoch": 35.96,
      "learning_rate": 0.00014195454545454545,
      "loss": 0.1377,
      "step": 2661
    },
    {
      "epoch": 35.97,
      "learning_rate": 0.0001418181818181818,
      "loss": 0.0308,
      "step": 2662
    },
    {
      "epoch": 35.99,
      "learning_rate": 0.00014168181818181817,
      "loss": 0.021,
      "step": 2663
    },
    {
      "epoch": 36.0,
      "learning_rate": 0.00014154545454545453,
      "loss": 0.0106,
      "step": 2664
    },
    {
      "epoch": 36.01,
      "learning_rate": 0.0001414090909090909,
      "loss": 0.0331,
      "step": 2665
    },
    {
      "epoch": 36.03,
      "learning_rate": 0.00014127272727272724,
      "loss": 0.005,
      "step": 2666
    },
    {
      "epoch": 36.04,
      "learning_rate": 0.00014113636363636363,
      "loss": 0.061,
      "step": 2667
    },
    {
      "epoch": 36.05,
      "learning_rate": 0.00014099999999999998,
      "loss": 0.0259,
      "step": 2668
    },
    {
      "epoch": 36.07,
      "learning_rate": 0.00014086363636363634,
      "loss": 0.0015,
      "step": 2669
    },
    {
      "epoch": 36.08,
      "learning_rate": 0.00014072727272727273,
      "loss": 0.0709,
      "step": 2670
    },
    {
      "epoch": 36.09,
      "learning_rate": 0.00014059090909090909,
      "loss": 0.0149,
      "step": 2671
    },
    {
      "epoch": 36.11,
      "learning_rate": 0.00014045454545454544,
      "loss": 0.1043,
      "step": 2672
    },
    {
      "epoch": 36.12,
      "learning_rate": 0.0001403181818181818,
      "loss": 0.0347,
      "step": 2673
    },
    {
      "epoch": 36.14,
      "learning_rate": 0.00014018181818181819,
      "loss": 0.0146,
      "step": 2674
    },
    {
      "epoch": 36.15,
      "learning_rate": 0.00014004545454545452,
      "loss": 0.0617,
      "step": 2675
    },
    {
      "epoch": 36.16,
      "learning_rate": 0.0001399090909090909,
      "loss": 0.1,
      "step": 2676
    },
    {
      "epoch": 36.18,
      "learning_rate": 0.00013977272727272726,
      "loss": 0.0484,
      "step": 2677
    },
    {
      "epoch": 36.19,
      "learning_rate": 0.00013963636363636362,
      "loss": 0.0857,
      "step": 2678
    },
    {
      "epoch": 36.2,
      "learning_rate": 0.0001395,
      "loss": 0.0771,
      "step": 2679
    },
    {
      "epoch": 36.22,
      "learning_rate": 0.00013936363636363636,
      "loss": 0.002,
      "step": 2680
    },
    {
      "epoch": 36.23,
      "learning_rate": 0.00013922727272727272,
      "loss": 0.0009,
      "step": 2681
    },
    {
      "epoch": 36.24,
      "learning_rate": 0.00013909090909090908,
      "loss": 0.0855,
      "step": 2682
    },
    {
      "epoch": 36.26,
      "learning_rate": 0.00013895454545454543,
      "loss": 0.0025,
      "step": 2683
    },
    {
      "epoch": 36.27,
      "learning_rate": 0.0001388181818181818,
      "loss": 0.0027,
      "step": 2684
    },
    {
      "epoch": 36.28,
      "learning_rate": 0.00013868181818181818,
      "loss": 0.0039,
      "step": 2685
    },
    {
      "epoch": 36.3,
      "learning_rate": 0.00013854545454545453,
      "loss": 0.1325,
      "step": 2686
    },
    {
      "epoch": 36.31,
      "learning_rate": 0.0001384090909090909,
      "loss": 0.0763,
      "step": 2687
    },
    {
      "epoch": 36.32,
      "learning_rate": 0.00013827272727272728,
      "loss": 0.0018,
      "step": 2688
    },
    {
      "epoch": 36.34,
      "learning_rate": 0.00013813636363636364,
      "loss": 0.0819,
      "step": 2689
    },
    {
      "epoch": 36.35,
      "learning_rate": 0.000138,
      "loss": 0.1725,
      "step": 2690
    },
    {
      "epoch": 36.36,
      "learning_rate": 0.00013786363636363635,
      "loss": 0.0022,
      "step": 2691
    },
    {
      "epoch": 36.38,
      "learning_rate": 0.0001377272727272727,
      "loss": 0.0883,
      "step": 2692
    },
    {
      "epoch": 36.39,
      "learning_rate": 0.00013759090909090907,
      "loss": 0.0047,
      "step": 2693
    },
    {
      "epoch": 36.41,
      "learning_rate": 0.00013745454545454545,
      "loss": 0.0334,
      "step": 2694
    },
    {
      "epoch": 36.42,
      "learning_rate": 0.0001373181818181818,
      "loss": 0.0535,
      "step": 2695
    },
    {
      "epoch": 36.43,
      "learning_rate": 0.00013718181818181817,
      "loss": 0.0256,
      "step": 2696
    },
    {
      "epoch": 36.45,
      "learning_rate": 0.00013704545454545455,
      "loss": 0.04,
      "step": 2697
    },
    {
      "epoch": 36.46,
      "learning_rate": 0.00013690909090909088,
      "loss": 0.0152,
      "step": 2698
    },
    {
      "epoch": 36.47,
      "learning_rate": 0.00013677272727272727,
      "loss": 0.1128,
      "step": 2699
    },
    {
      "epoch": 36.49,
      "learning_rate": 0.00013663636363636363,
      "loss": 0.0071,
      "step": 2700
    },
    {
      "epoch": 36.5,
      "learning_rate": 0.00013649999999999998,
      "loss": 0.1079,
      "step": 2701
    },
    {
      "epoch": 36.51,
      "learning_rate": 0.00013636363636363634,
      "loss": 0.1359,
      "step": 2702
    },
    {
      "epoch": 36.53,
      "learning_rate": 0.00013622727272727273,
      "loss": 0.0543,
      "step": 2703
    },
    {
      "epoch": 36.54,
      "learning_rate": 0.00013609090909090908,
      "loss": 0.0062,
      "step": 2704
    },
    {
      "epoch": 36.55,
      "learning_rate": 0.00013595454545454544,
      "loss": 0.0805,
      "step": 2705
    },
    {
      "epoch": 36.57,
      "learning_rate": 0.0001358181818181818,
      "loss": 0.0044,
      "step": 2706
    },
    {
      "epoch": 36.58,
      "learning_rate": 0.00013568181818181816,
      "loss": 0.0024,
      "step": 2707
    },
    {
      "epoch": 36.59,
      "learning_rate": 0.00013554545454545454,
      "loss": 0.0583,
      "step": 2708
    },
    {
      "epoch": 36.61,
      "learning_rate": 0.0001354090909090909,
      "loss": 0.014,
      "step": 2709
    },
    {
      "epoch": 36.62,
      "learning_rate": 0.00013527272727272726,
      "loss": 0.0262,
      "step": 2710
    },
    {
      "epoch": 36.64,
      "learning_rate": 0.00013513636363636362,
      "loss": 0.1524,
      "step": 2711
    },
    {
      "epoch": 36.65,
      "learning_rate": 0.000135,
      "loss": 0.0324,
      "step": 2712
    },
    {
      "epoch": 36.66,
      "learning_rate": 0.00013486363636363633,
      "loss": 0.0627,
      "step": 2713
    },
    {
      "epoch": 36.68,
      "learning_rate": 0.00013472727272727272,
      "loss": 0.0912,
      "step": 2714
    },
    {
      "epoch": 36.69,
      "learning_rate": 0.00013459090909090908,
      "loss": 0.0396,
      "step": 2715
    },
    {
      "epoch": 36.7,
      "learning_rate": 0.00013445454545454543,
      "loss": 0.017,
      "step": 2716
    },
    {
      "epoch": 36.72,
      "learning_rate": 0.00013431818181818182,
      "loss": 0.0023,
      "step": 2717
    },
    {
      "epoch": 36.73,
      "learning_rate": 0.00013418181818181818,
      "loss": 0.1085,
      "step": 2718
    },
    {
      "epoch": 36.74,
      "learning_rate": 0.00013404545454545453,
      "loss": 0.0027,
      "step": 2719
    },
    {
      "epoch": 36.76,
      "learning_rate": 0.0001339090909090909,
      "loss": 0.0397,
      "step": 2720
    },
    {
      "epoch": 36.77,
      "learning_rate": 0.00013377272727272728,
      "loss": 0.0046,
      "step": 2721
    },
    {
      "epoch": 36.78,
      "learning_rate": 0.0001336363636363636,
      "loss": 0.0036,
      "step": 2722
    },
    {
      "epoch": 36.8,
      "learning_rate": 0.0001335,
      "loss": 0.0049,
      "step": 2723
    },
    {
      "epoch": 36.81,
      "learning_rate": 0.00013336363636363635,
      "loss": 0.0635,
      "step": 2724
    },
    {
      "epoch": 36.82,
      "learning_rate": 0.0001332272727272727,
      "loss": 0.0054,
      "step": 2725
    },
    {
      "epoch": 36.84,
      "learning_rate": 0.0001330909090909091,
      "loss": 0.1117,
      "step": 2726
    },
    {
      "epoch": 36.85,
      "learning_rate": 0.00013295454545454545,
      "loss": 0.008,
      "step": 2727
    },
    {
      "epoch": 36.86,
      "learning_rate": 0.0001328181818181818,
      "loss": 0.0427,
      "step": 2728
    },
    {
      "epoch": 36.88,
      "learning_rate": 0.00013268181818181817,
      "loss": 0.002,
      "step": 2729
    },
    {
      "epoch": 36.89,
      "learning_rate": 0.00013254545454545452,
      "loss": 0.1127,
      "step": 2730
    },
    {
      "epoch": 36.91,
      "learning_rate": 0.00013240909090909088,
      "loss": 0.0558,
      "step": 2731
    },
    {
      "epoch": 36.92,
      "learning_rate": 0.00013227272727272727,
      "loss": 0.1041,
      "step": 2732
    },
    {
      "epoch": 36.93,
      "learning_rate": 0.00013213636363636363,
      "loss": 0.002,
      "step": 2733
    },
    {
      "epoch": 36.95,
      "learning_rate": 0.00013199999999999998,
      "loss": 0.0011,
      "step": 2734
    },
    {
      "epoch": 36.96,
      "learning_rate": 0.00013186363636363637,
      "loss": 0.0265,
      "step": 2735
    },
    {
      "epoch": 36.97,
      "learning_rate": 0.00013172727272727273,
      "loss": 0.0989,
      "step": 2736
    },
    {
      "epoch": 36.99,
      "learning_rate": 0.00013159090909090908,
      "loss": 0.0313,
      "step": 2737
    },
    {
      "epoch": 37.0,
      "learning_rate": 0.00013145454545454544,
      "loss": 0.0048,
      "step": 2738
    },
    {
      "epoch": 37.01,
      "learning_rate": 0.0001313181818181818,
      "loss": 0.0303,
      "step": 2739
    },
    {
      "epoch": 37.03,
      "learning_rate": 0.00013118181818181816,
      "loss": 0.0157,
      "step": 2740
    },
    {
      "epoch": 37.04,
      "learning_rate": 0.00013104545454545454,
      "loss": 0.0107,
      "step": 2741
    },
    {
      "epoch": 37.05,
      "learning_rate": 0.0001309090909090909,
      "loss": 0.0699,
      "step": 2742
    },
    {
      "epoch": 37.07,
      "learning_rate": 0.00013077272727272726,
      "loss": 0.0165,
      "step": 2743
    },
    {
      "epoch": 37.08,
      "learning_rate": 0.00013063636363636362,
      "loss": 0.0026,
      "step": 2744
    },
    {
      "epoch": 37.09,
      "learning_rate": 0.0001305,
      "loss": 0.0037,
      "step": 2745
    },
    {
      "epoch": 37.11,
      "learning_rate": 0.00013036363636363636,
      "loss": 0.1017,
      "step": 2746
    },
    {
      "epoch": 37.12,
      "learning_rate": 0.00013022727272727272,
      "loss": 0.0894,
      "step": 2747
    },
    {
      "epoch": 37.14,
      "learning_rate": 0.00013009090909090907,
      "loss": 0.1263,
      "step": 2748
    },
    {
      "epoch": 37.15,
      "learning_rate": 0.00012995454545454543,
      "loss": 0.0014,
      "step": 2749
    },
    {
      "epoch": 37.16,
      "learning_rate": 0.00012981818181818182,
      "loss": 0.0294,
      "step": 2750
    },
    {
      "epoch": 37.18,
      "learning_rate": 0.00012968181818181818,
      "loss": 0.0584,
      "step": 2751
    },
    {
      "epoch": 37.19,
      "learning_rate": 0.00012954545454545453,
      "loss": 0.0061,
      "step": 2752
    },
    {
      "epoch": 37.2,
      "learning_rate": 0.0001294090909090909,
      "loss": 0.0031,
      "step": 2753
    },
    {
      "epoch": 37.22,
      "learning_rate": 0.00012927272727272725,
      "loss": 0.0292,
      "step": 2754
    },
    {
      "epoch": 37.23,
      "learning_rate": 0.00012913636363636363,
      "loss": 0.0579,
      "step": 2755
    },
    {
      "epoch": 37.24,
      "learning_rate": 0.000129,
      "loss": 0.0277,
      "step": 2756
    },
    {
      "epoch": 37.26,
      "learning_rate": 0.00012886363636363635,
      "loss": 0.0552,
      "step": 2757
    },
    {
      "epoch": 37.27,
      "learning_rate": 0.0001287272727272727,
      "loss": 0.0131,
      "step": 2758
    },
    {
      "epoch": 37.28,
      "learning_rate": 0.0001285909090909091,
      "loss": 0.0065,
      "step": 2759
    },
    {
      "epoch": 37.3,
      "learning_rate": 0.00012845454545454545,
      "loss": 0.0493,
      "step": 2760
    },
    {
      "epoch": 37.31,
      "learning_rate": 0.0001283181818181818,
      "loss": 0.0181,
      "step": 2761
    },
    {
      "epoch": 37.32,
      "learning_rate": 0.00012818181818181817,
      "loss": 0.0901,
      "step": 2762
    },
    {
      "epoch": 37.34,
      "learning_rate": 0.00012804545454545452,
      "loss": 0.0126,
      "step": 2763
    },
    {
      "epoch": 37.35,
      "learning_rate": 0.0001279090909090909,
      "loss": 0.0048,
      "step": 2764
    },
    {
      "epoch": 37.36,
      "learning_rate": 0.00012777272727272727,
      "loss": 0.0052,
      "step": 2765
    },
    {
      "epoch": 37.38,
      "learning_rate": 0.00012763636363636362,
      "loss": 0.0046,
      "step": 2766
    },
    {
      "epoch": 37.39,
      "learning_rate": 0.00012749999999999998,
      "loss": 0.0083,
      "step": 2767
    },
    {
      "epoch": 37.41,
      "learning_rate": 0.00012736363636363637,
      "loss": 0.0033,
      "step": 2768
    },
    {
      "epoch": 37.42,
      "learning_rate": 0.0001272272727272727,
      "loss": 0.0257,
      "step": 2769
    },
    {
      "epoch": 37.43,
      "learning_rate": 0.00012709090909090908,
      "loss": 0.0119,
      "step": 2770
    },
    {
      "epoch": 37.45,
      "learning_rate": 0.00012695454545454544,
      "loss": 0.0579,
      "step": 2771
    },
    {
      "epoch": 37.46,
      "learning_rate": 0.0001268181818181818,
      "loss": 0.0861,
      "step": 2772
    },
    {
      "epoch": 37.47,
      "learning_rate": 0.00012668181818181818,
      "loss": 0.0629,
      "step": 2773
    },
    {
      "epoch": 37.49,
      "learning_rate": 0.00012654545454545454,
      "loss": 0.2032,
      "step": 2774
    },
    {
      "epoch": 37.5,
      "learning_rate": 0.0001264090909090909,
      "loss": 0.0391,
      "step": 2775
    },
    {
      "epoch": 37.51,
      "learning_rate": 0.00012627272727272726,
      "loss": 0.0026,
      "step": 2776
    },
    {
      "epoch": 37.53,
      "learning_rate": 0.00012613636363636364,
      "loss": 0.0466,
      "step": 2777
    },
    {
      "epoch": 37.54,
      "learning_rate": 0.00012599999999999997,
      "loss": 0.0039,
      "step": 2778
    },
    {
      "epoch": 37.55,
      "learning_rate": 0.00012586363636363636,
      "loss": 0.1719,
      "step": 2779
    },
    {
      "epoch": 37.57,
      "learning_rate": 0.00012572727272727272,
      "loss": 0.021,
      "step": 2780
    },
    {
      "epoch": 37.58,
      "learning_rate": 0.00012559090909090907,
      "loss": 0.031,
      "step": 2781
    },
    {
      "epoch": 37.59,
      "learning_rate": 0.00012545454545454543,
      "loss": 0.0614,
      "step": 2782
    },
    {
      "epoch": 37.61,
      "learning_rate": 0.00012531818181818182,
      "loss": 0.0042,
      "step": 2783
    },
    {
      "epoch": 37.62,
      "learning_rate": 0.00012518181818181817,
      "loss": 0.0224,
      "step": 2784
    },
    {
      "epoch": 37.64,
      "learning_rate": 0.00012504545454545453,
      "loss": 0.0014,
      "step": 2785
    },
    {
      "epoch": 37.65,
      "learning_rate": 0.0001249090909090909,
      "loss": 0.0017,
      "step": 2786
    },
    {
      "epoch": 37.66,
      "learning_rate": 0.00012477272727272725,
      "loss": 0.002,
      "step": 2787
    },
    {
      "epoch": 37.68,
      "learning_rate": 0.00012463636363636363,
      "loss": 0.1094,
      "step": 2788
    },
    {
      "epoch": 37.69,
      "learning_rate": 0.0001245,
      "loss": 0.0458,
      "step": 2789
    },
    {
      "epoch": 37.7,
      "learning_rate": 0.00012436363636363635,
      "loss": 0.0556,
      "step": 2790
    },
    {
      "epoch": 37.72,
      "learning_rate": 0.0001242272727272727,
      "loss": 0.0065,
      "step": 2791
    },
    {
      "epoch": 37.73,
      "learning_rate": 0.0001240909090909091,
      "loss": 0.1657,
      "step": 2792
    },
    {
      "epoch": 37.74,
      "learning_rate": 0.00012395454545454545,
      "loss": 0.0273,
      "step": 2793
    },
    {
      "epoch": 37.76,
      "learning_rate": 0.0001238181818181818,
      "loss": 0.0019,
      "step": 2794
    },
    {
      "epoch": 37.77,
      "learning_rate": 0.00012368181818181816,
      "loss": 0.151,
      "step": 2795
    },
    {
      "epoch": 37.78,
      "learning_rate": 0.00012354545454545452,
      "loss": 0.0017,
      "step": 2796
    },
    {
      "epoch": 37.8,
      "learning_rate": 0.0001234090909090909,
      "loss": 0.142,
      "step": 2797
    },
    {
      "epoch": 37.81,
      "learning_rate": 0.00012327272727272727,
      "loss": 0.0518,
      "step": 2798
    },
    {
      "epoch": 37.82,
      "learning_rate": 0.00012313636363636362,
      "loss": 0.0741,
      "step": 2799
    },
    {
      "epoch": 37.84,
      "learning_rate": 0.00012299999999999998,
      "loss": 0.017,
      "step": 2800
    },
    {
      "epoch": 37.84,
      "eval_accuracy": 0.9408536585365853,
      "eval_f1": 0.941470393860948,
      "eval_loss": 0.2628859579563141,
      "eval_runtime": 126.368,
      "eval_samples_per_second": 12.978,
      "eval_steps_per_second": 0.815,
      "step": 2800
    },
    {
      "epoch": 37.85,
      "learning_rate": 0.00012286363636363637,
      "loss": 0.0282,
      "step": 2801
    },
    {
      "epoch": 37.86,
      "learning_rate": 0.00012272727272727272,
      "loss": 0.0011,
      "step": 2802
    },
    {
      "epoch": 37.88,
      "learning_rate": 0.00012259090909090908,
      "loss": 0.002,
      "step": 2803
    },
    {
      "epoch": 37.89,
      "learning_rate": 0.00012245454545454544,
      "loss": 0.0631,
      "step": 2804
    },
    {
      "epoch": 37.91,
      "learning_rate": 0.0001223181818181818,
      "loss": 0.1144,
      "step": 2805
    },
    {
      "epoch": 37.92,
      "learning_rate": 0.00012218181818181818,
      "loss": 0.0874,
      "step": 2806
    },
    {
      "epoch": 37.93,
      "learning_rate": 0.00012204545454545453,
      "loss": 0.063,
      "step": 2807
    },
    {
      "epoch": 37.95,
      "learning_rate": 0.0001219090909090909,
      "loss": 0.0594,
      "step": 2808
    },
    {
      "epoch": 37.96,
      "learning_rate": 0.00012177272727272726,
      "loss": 0.0024,
      "step": 2809
    },
    {
      "epoch": 37.97,
      "learning_rate": 0.00012163636363636363,
      "loss": 0.0031,
      "step": 2810
    },
    {
      "epoch": 37.99,
      "learning_rate": 0.0001215,
      "loss": 0.2026,
      "step": 2811
    },
    {
      "epoch": 38.0,
      "learning_rate": 0.00012136363636363636,
      "loss": 0.016,
      "step": 2812
    },
    {
      "epoch": 38.01,
      "learning_rate": 0.00012122727272727273,
      "loss": 0.0427,
      "step": 2813
    },
    {
      "epoch": 38.03,
      "learning_rate": 0.00012109090909090907,
      "loss": 0.0092,
      "step": 2814
    },
    {
      "epoch": 38.04,
      "learning_rate": 0.00012095454545454544,
      "loss": 0.0159,
      "step": 2815
    },
    {
      "epoch": 38.05,
      "learning_rate": 0.0001208181818181818,
      "loss": 0.0455,
      "step": 2816
    },
    {
      "epoch": 38.07,
      "learning_rate": 0.00012068181818181817,
      "loss": 0.0861,
      "step": 2817
    },
    {
      "epoch": 38.08,
      "learning_rate": 0.00012054545454545453,
      "loss": 0.0013,
      "step": 2818
    },
    {
      "epoch": 38.09,
      "learning_rate": 0.0001204090909090909,
      "loss": 0.011,
      "step": 2819
    },
    {
      "epoch": 38.11,
      "learning_rate": 0.00012027272727272726,
      "loss": 0.0169,
      "step": 2820
    },
    {
      "epoch": 38.12,
      "learning_rate": 0.00012013636363636363,
      "loss": 0.0201,
      "step": 2821
    },
    {
      "epoch": 38.14,
      "learning_rate": 0.00011999999999999999,
      "loss": 0.0091,
      "step": 2822
    },
    {
      "epoch": 38.15,
      "learning_rate": 0.00011986363636363635,
      "loss": 0.0893,
      "step": 2823
    },
    {
      "epoch": 38.16,
      "learning_rate": 0.00011972727272727272,
      "loss": 0.0028,
      "step": 2824
    },
    {
      "epoch": 38.18,
      "learning_rate": 0.00011959090909090908,
      "loss": 0.0365,
      "step": 2825
    },
    {
      "epoch": 38.19,
      "learning_rate": 0.00011945454545454545,
      "loss": 0.0047,
      "step": 2826
    },
    {
      "epoch": 38.2,
      "learning_rate": 0.0001193181818181818,
      "loss": 0.002,
      "step": 2827
    },
    {
      "epoch": 38.22,
      "learning_rate": 0.00011918181818181818,
      "loss": 0.0048,
      "step": 2828
    },
    {
      "epoch": 38.23,
      "learning_rate": 0.00011904545454545452,
      "loss": 0.0124,
      "step": 2829
    },
    {
      "epoch": 38.24,
      "learning_rate": 0.00011890909090909089,
      "loss": 0.0048,
      "step": 2830
    },
    {
      "epoch": 38.26,
      "learning_rate": 0.00011877272727272726,
      "loss": 0.0023,
      "step": 2831
    },
    {
      "epoch": 38.27,
      "learning_rate": 0.00011863636363636362,
      "loss": 0.0033,
      "step": 2832
    },
    {
      "epoch": 38.28,
      "learning_rate": 0.0001185,
      "loss": 0.0023,
      "step": 2833
    },
    {
      "epoch": 38.3,
      "learning_rate": 0.00011836363636363635,
      "loss": 0.0341,
      "step": 2834
    },
    {
      "epoch": 38.31,
      "learning_rate": 0.00011822727272727272,
      "loss": 0.0206,
      "step": 2835
    },
    {
      "epoch": 38.32,
      "learning_rate": 0.00011809090909090908,
      "loss": 0.0133,
      "step": 2836
    },
    {
      "epoch": 38.34,
      "learning_rate": 0.00011795454545454545,
      "loss": 0.0025,
      "step": 2837
    },
    {
      "epoch": 38.35,
      "learning_rate": 0.0001178181818181818,
      "loss": 0.0199,
      "step": 2838
    },
    {
      "epoch": 38.36,
      "learning_rate": 0.00011768181818181817,
      "loss": 0.0051,
      "step": 2839
    },
    {
      "epoch": 38.38,
      "learning_rate": 0.00011754545454545454,
      "loss": 0.0661,
      "step": 2840
    },
    {
      "epoch": 38.39,
      "learning_rate": 0.0001174090909090909,
      "loss": 0.0871,
      "step": 2841
    },
    {
      "epoch": 38.41,
      "learning_rate": 0.00011727272727272727,
      "loss": 0.0412,
      "step": 2842
    },
    {
      "epoch": 38.42,
      "learning_rate": 0.00011713636363636363,
      "loss": 0.0027,
      "step": 2843
    },
    {
      "epoch": 38.43,
      "learning_rate": 0.000117,
      "loss": 0.0008,
      "step": 2844
    },
    {
      "epoch": 38.45,
      "learning_rate": 0.00011686363636363634,
      "loss": 0.0277,
      "step": 2845
    },
    {
      "epoch": 38.46,
      "learning_rate": 0.00011672727272727271,
      "loss": 0.0596,
      "step": 2846
    },
    {
      "epoch": 38.47,
      "learning_rate": 0.00011659090909090907,
      "loss": 0.0011,
      "step": 2847
    },
    {
      "epoch": 38.49,
      "learning_rate": 0.00011645454545454544,
      "loss": 0.0024,
      "step": 2848
    },
    {
      "epoch": 38.5,
      "learning_rate": 0.00011631818181818181,
      "loss": 0.0017,
      "step": 2849
    },
    {
      "epoch": 38.51,
      "learning_rate": 0.00011618181818181817,
      "loss": 0.0036,
      "step": 2850
    },
    {
      "epoch": 38.53,
      "learning_rate": 0.00011604545454545454,
      "loss": 0.1438,
      "step": 2851
    },
    {
      "epoch": 38.54,
      "learning_rate": 0.0001159090909090909,
      "loss": 0.0021,
      "step": 2852
    },
    {
      "epoch": 38.55,
      "learning_rate": 0.00011577272727272727,
      "loss": 0.0062,
      "step": 2853
    },
    {
      "epoch": 38.57,
      "learning_rate": 0.00011563636363636362,
      "loss": 0.0144,
      "step": 2854
    },
    {
      "epoch": 38.58,
      "learning_rate": 0.00011549999999999999,
      "loss": 0.0064,
      "step": 2855
    },
    {
      "epoch": 38.59,
      "learning_rate": 0.00011536363636363635,
      "loss": 0.0863,
      "step": 2856
    },
    {
      "epoch": 38.61,
      "learning_rate": 0.00011522727272727272,
      "loss": 0.0063,
      "step": 2857
    },
    {
      "epoch": 38.62,
      "learning_rate": 0.00011509090909090908,
      "loss": 0.0032,
      "step": 2858
    },
    {
      "epoch": 38.64,
      "learning_rate": 0.00011495454545454545,
      "loss": 0.0037,
      "step": 2859
    },
    {
      "epoch": 38.65,
      "learning_rate": 0.00011481818181818182,
      "loss": 0.0074,
      "step": 2860
    },
    {
      "epoch": 38.66,
      "learning_rate": 0.00011468181818181816,
      "loss": 0.0803,
      "step": 2861
    },
    {
      "epoch": 38.68,
      "learning_rate": 0.00011454545454545453,
      "loss": 0.0022,
      "step": 2862
    },
    {
      "epoch": 38.69,
      "learning_rate": 0.00011440909090909089,
      "loss": 0.0018,
      "step": 2863
    },
    {
      "epoch": 38.7,
      "learning_rate": 0.00011427272727272726,
      "loss": 0.0281,
      "step": 2864
    },
    {
      "epoch": 38.72,
      "learning_rate": 0.00011413636363636362,
      "loss": 0.1033,
      "step": 2865
    },
    {
      "epoch": 38.73,
      "learning_rate": 0.00011399999999999999,
      "loss": 0.0139,
      "step": 2866
    },
    {
      "epoch": 38.74,
      "learning_rate": 0.00011386363636363635,
      "loss": 0.0156,
      "step": 2867
    },
    {
      "epoch": 38.76,
      "learning_rate": 0.00011372727272727272,
      "loss": 0.002,
      "step": 2868
    },
    {
      "epoch": 38.77,
      "learning_rate": 0.0001135909090909091,
      "loss": 0.0336,
      "step": 2869
    },
    {
      "epoch": 38.78,
      "learning_rate": 0.00011345454545454544,
      "loss": 0.029,
      "step": 2870
    },
    {
      "epoch": 38.8,
      "learning_rate": 0.00011331818181818181,
      "loss": 0.0087,
      "step": 2871
    },
    {
      "epoch": 38.81,
      "learning_rate": 0.00011318181818181817,
      "loss": 0.1186,
      "step": 2872
    },
    {
      "epoch": 38.82,
      "learning_rate": 0.00011304545454545454,
      "loss": 0.0065,
      "step": 2873
    },
    {
      "epoch": 38.84,
      "learning_rate": 0.0001129090909090909,
      "loss": 0.0988,
      "step": 2874
    },
    {
      "epoch": 38.85,
      "learning_rate": 0.00011277272727272727,
      "loss": 0.005,
      "step": 2875
    },
    {
      "epoch": 38.86,
      "learning_rate": 0.00011263636363636363,
      "loss": 0.0012,
      "step": 2876
    },
    {
      "epoch": 38.88,
      "learning_rate": 0.0001125,
      "loss": 0.0017,
      "step": 2877
    },
    {
      "epoch": 38.89,
      "learning_rate": 0.00011236363636363635,
      "loss": 0.0104,
      "step": 2878
    },
    {
      "epoch": 38.91,
      "learning_rate": 0.00011222727272727271,
      "loss": 0.1734,
      "step": 2879
    },
    {
      "epoch": 38.92,
      "learning_rate": 0.00011209090909090908,
      "loss": 0.1448,
      "step": 2880
    },
    {
      "epoch": 38.93,
      "learning_rate": 0.00011195454545454544,
      "loss": 0.0523,
      "step": 2881
    },
    {
      "epoch": 38.95,
      "learning_rate": 0.00011181818181818181,
      "loss": 0.042,
      "step": 2882
    },
    {
      "epoch": 38.96,
      "learning_rate": 0.00011168181818181817,
      "loss": 0.0027,
      "step": 2883
    },
    {
      "epoch": 38.97,
      "learning_rate": 0.00011154545454545454,
      "loss": 0.0248,
      "step": 2884
    },
    {
      "epoch": 38.99,
      "learning_rate": 0.00011140909090909089,
      "loss": 0.0033,
      "step": 2885
    },
    {
      "epoch": 39.0,
      "learning_rate": 0.00011127272727272726,
      "loss": 0.0009,
      "step": 2886
    },
    {
      "epoch": 39.01,
      "learning_rate": 0.00011113636363636363,
      "loss": 0.0063,
      "step": 2887
    },
    {
      "epoch": 39.03,
      "learning_rate": 0.00011099999999999999,
      "loss": 0.0017,
      "step": 2888
    },
    {
      "epoch": 39.04,
      "learning_rate": 0.00011086363636363636,
      "loss": 0.0124,
      "step": 2889
    },
    {
      "epoch": 39.05,
      "learning_rate": 0.00011072727272727272,
      "loss": 0.0165,
      "step": 2890
    },
    {
      "epoch": 39.07,
      "learning_rate": 0.00011059090909090909,
      "loss": 0.0673,
      "step": 2891
    },
    {
      "epoch": 39.08,
      "learning_rate": 0.00011045454545454545,
      "loss": 0.0409,
      "step": 2892
    },
    {
      "epoch": 39.09,
      "learning_rate": 0.00011031818181818182,
      "loss": 0.0206,
      "step": 2893
    },
    {
      "epoch": 39.11,
      "learning_rate": 0.00011018181818181816,
      "loss": 0.0028,
      "step": 2894
    },
    {
      "epoch": 39.12,
      "learning_rate": 0.00011004545454545453,
      "loss": 0.0026,
      "step": 2895
    },
    {
      "epoch": 39.14,
      "learning_rate": 0.0001099090909090909,
      "loss": 0.0006,
      "step": 2896
    },
    {
      "epoch": 39.15,
      "learning_rate": 0.00010977272727272726,
      "loss": 0.0548,
      "step": 2897
    },
    {
      "epoch": 39.16,
      "learning_rate": 0.00010963636363636363,
      "loss": 0.0009,
      "step": 2898
    },
    {
      "epoch": 39.18,
      "learning_rate": 0.00010949999999999999,
      "loss": 0.0182,
      "step": 2899
    },
    {
      "epoch": 39.19,
      "learning_rate": 0.00010936363636363636,
      "loss": 0.0158,
      "step": 2900
    },
    {
      "epoch": 39.2,
      "learning_rate": 0.00010922727272727271,
      "loss": 0.0008,
      "step": 2901
    },
    {
      "epoch": 39.22,
      "learning_rate": 0.00010909090909090908,
      "loss": 0.0007,
      "step": 2902
    },
    {
      "epoch": 39.23,
      "learning_rate": 0.00010895454545454544,
      "loss": 0.103,
      "step": 2903
    },
    {
      "epoch": 39.24,
      "learning_rate": 0.00010881818181818181,
      "loss": 0.0092,
      "step": 2904
    },
    {
      "epoch": 39.26,
      "learning_rate": 0.00010868181818181817,
      "loss": 0.0495,
      "step": 2905
    },
    {
      "epoch": 39.27,
      "learning_rate": 0.00010854545454545454,
      "loss": 0.0011,
      "step": 2906
    },
    {
      "epoch": 39.28,
      "learning_rate": 0.00010840909090909091,
      "loss": 0.0142,
      "step": 2907
    },
    {
      "epoch": 39.3,
      "learning_rate": 0.00010827272727272727,
      "loss": 0.0066,
      "step": 2908
    },
    {
      "epoch": 39.31,
      "learning_rate": 0.00010813636363636364,
      "loss": 0.0054,
      "step": 2909
    },
    {
      "epoch": 39.32,
      "learning_rate": 0.00010799999999999998,
      "loss": 0.0201,
      "step": 2910
    },
    {
      "epoch": 39.34,
      "learning_rate": 0.00010786363636363635,
      "loss": 0.0015,
      "step": 2911
    },
    {
      "epoch": 39.35,
      "learning_rate": 0.00010772727272727271,
      "loss": 0.0889,
      "step": 2912
    },
    {
      "epoch": 39.36,
      "learning_rate": 0.00010759090909090908,
      "loss": 0.1231,
      "step": 2913
    },
    {
      "epoch": 39.38,
      "learning_rate": 0.00010745454545454544,
      "loss": 0.0012,
      "step": 2914
    },
    {
      "epoch": 39.39,
      "learning_rate": 0.00010731818181818181,
      "loss": 0.0537,
      "step": 2915
    },
    {
      "epoch": 39.41,
      "learning_rate": 0.00010718181818181818,
      "loss": 0.0135,
      "step": 2916
    },
    {
      "epoch": 39.42,
      "learning_rate": 0.00010704545454545453,
      "loss": 0.0595,
      "step": 2917
    },
    {
      "epoch": 39.43,
      "learning_rate": 0.0001069090909090909,
      "loss": 0.0269,
      "step": 2918
    },
    {
      "epoch": 39.45,
      "learning_rate": 0.00010677272727272726,
      "loss": 0.0105,
      "step": 2919
    },
    {
      "epoch": 39.46,
      "learning_rate": 0.00010663636363636363,
      "loss": 0.0011,
      "step": 2920
    },
    {
      "epoch": 39.47,
      "learning_rate": 0.00010649999999999999,
      "loss": 0.0024,
      "step": 2921
    },
    {
      "epoch": 39.49,
      "learning_rate": 0.00010636363636363636,
      "loss": 0.1045,
      "step": 2922
    },
    {
      "epoch": 39.5,
      "learning_rate": 0.00010622727272727272,
      "loss": 0.0589,
      "step": 2923
    },
    {
      "epoch": 39.51,
      "learning_rate": 0.00010609090909090909,
      "loss": 0.0009,
      "step": 2924
    },
    {
      "epoch": 39.53,
      "learning_rate": 0.00010595454545454546,
      "loss": 0.0573,
      "step": 2925
    },
    {
      "epoch": 39.54,
      "learning_rate": 0.0001058181818181818,
      "loss": 0.0043,
      "step": 2926
    },
    {
      "epoch": 39.55,
      "learning_rate": 0.00010568181818181817,
      "loss": 0.0316,
      "step": 2927
    },
    {
      "epoch": 39.57,
      "learning_rate": 0.00010554545454545453,
      "loss": 0.0694,
      "step": 2928
    },
    {
      "epoch": 39.58,
      "learning_rate": 0.0001054090909090909,
      "loss": 0.0018,
      "step": 2929
    },
    {
      "epoch": 39.59,
      "learning_rate": 0.00010527272727272726,
      "loss": 0.0018,
      "step": 2930
    },
    {
      "epoch": 39.61,
      "learning_rate": 0.00010513636363636363,
      "loss": 0.0016,
      "step": 2931
    },
    {
      "epoch": 39.62,
      "learning_rate": 0.00010499999999999999,
      "loss": 0.0012,
      "step": 2932
    },
    {
      "epoch": 39.64,
      "learning_rate": 0.00010486363636363635,
      "loss": 0.0397,
      "step": 2933
    },
    {
      "epoch": 39.65,
      "learning_rate": 0.00010472727272727272,
      "loss": 0.0011,
      "step": 2934
    },
    {
      "epoch": 39.66,
      "learning_rate": 0.00010459090909090908,
      "loss": 0.0045,
      "step": 2935
    },
    {
      "epoch": 39.68,
      "learning_rate": 0.00010445454545454545,
      "loss": 0.0269,
      "step": 2936
    },
    {
      "epoch": 39.69,
      "learning_rate": 0.00010431818181818181,
      "loss": 0.0266,
      "step": 2937
    },
    {
      "epoch": 39.7,
      "learning_rate": 0.00010418181818181818,
      "loss": 0.0043,
      "step": 2938
    },
    {
      "epoch": 39.72,
      "learning_rate": 0.00010404545454545454,
      "loss": 0.0083,
      "step": 2939
    },
    {
      "epoch": 39.73,
      "learning_rate": 0.00010390909090909091,
      "loss": 0.0013,
      "step": 2940
    },
    {
      "epoch": 39.74,
      "learning_rate": 0.00010377272727272725,
      "loss": 0.0025,
      "step": 2941
    },
    {
      "epoch": 39.76,
      "learning_rate": 0.00010363636363636362,
      "loss": 0.001,
      "step": 2942
    },
    {
      "epoch": 39.77,
      "learning_rate": 0.00010349999999999998,
      "loss": 0.0014,
      "step": 2943
    },
    {
      "epoch": 39.78,
      "learning_rate": 0.00010336363636363635,
      "loss": 0.002,
      "step": 2944
    },
    {
      "epoch": 39.8,
      "learning_rate": 0.00010322727272727272,
      "loss": 0.0019,
      "step": 2945
    },
    {
      "epoch": 39.81,
      "learning_rate": 0.00010309090909090908,
      "loss": 0.0739,
      "step": 2946
    },
    {
      "epoch": 39.82,
      "learning_rate": 0.00010295454545454545,
      "loss": 0.0011,
      "step": 2947
    },
    {
      "epoch": 39.84,
      "learning_rate": 0.00010281818181818181,
      "loss": 0.001,
      "step": 2948
    },
    {
      "epoch": 39.85,
      "learning_rate": 0.00010268181818181818,
      "loss": 0.007,
      "step": 2949
    },
    {
      "epoch": 39.86,
      "learning_rate": 0.00010254545454545453,
      "loss": 0.0041,
      "step": 2950
    },
    {
      "epoch": 39.88,
      "learning_rate": 0.0001024090909090909,
      "loss": 0.0033,
      "step": 2951
    },
    {
      "epoch": 39.89,
      "learning_rate": 0.00010227272727272726,
      "loss": 0.0011,
      "step": 2952
    },
    {
      "epoch": 39.91,
      "learning_rate": 0.00010213636363636363,
      "loss": 0.0024,
      "step": 2953
    },
    {
      "epoch": 39.92,
      "learning_rate": 0.000102,
      "loss": 0.0075,
      "step": 2954
    },
    {
      "epoch": 39.93,
      "learning_rate": 0.00010186363636363636,
      "loss": 0.0026,
      "step": 2955
    },
    {
      "epoch": 39.95,
      "learning_rate": 0.00010172727272727273,
      "loss": 0.0045,
      "step": 2956
    },
    {
      "epoch": 39.96,
      "learning_rate": 0.00010159090909090907,
      "loss": 0.1503,
      "step": 2957
    },
    {
      "epoch": 39.97,
      "learning_rate": 0.00010145454545454544,
      "loss": 0.0018,
      "step": 2958
    },
    {
      "epoch": 39.99,
      "learning_rate": 0.0001013181818181818,
      "loss": 0.0042,
      "step": 2959
    },
    {
      "epoch": 40.0,
      "learning_rate": 0.00010118181818181817,
      "loss": 0.0009,
      "step": 2960
    },
    {
      "epoch": 40.01,
      "learning_rate": 0.00010104545454545453,
      "loss": 0.0754,
      "step": 2961
    },
    {
      "epoch": 40.03,
      "learning_rate": 0.0001009090909090909,
      "loss": 0.0108,
      "step": 2962
    },
    {
      "epoch": 40.04,
      "learning_rate": 0.00010077272727272727,
      "loss": 0.0027,
      "step": 2963
    },
    {
      "epoch": 40.05,
      "learning_rate": 0.00010063636363636363,
      "loss": 0.0434,
      "step": 2964
    },
    {
      "epoch": 40.07,
      "learning_rate": 0.0001005,
      "loss": 0.0579,
      "step": 2965
    },
    {
      "epoch": 40.08,
      "learning_rate": 0.00010036363636363635,
      "loss": 0.0711,
      "step": 2966
    },
    {
      "epoch": 40.09,
      "learning_rate": 0.00010022727272727272,
      "loss": 0.0019,
      "step": 2967
    },
    {
      "epoch": 40.11,
      "learning_rate": 0.00010009090909090908,
      "loss": 0.0007,
      "step": 2968
    },
    {
      "epoch": 40.12,
      "learning_rate": 9.995454545454545e-05,
      "loss": 0.0056,
      "step": 2969
    },
    {
      "epoch": 40.14,
      "learning_rate": 9.98181818181818e-05,
      "loss": 0.0597,
      "step": 2970
    },
    {
      "epoch": 40.15,
      "learning_rate": 9.968181818181818e-05,
      "loss": 0.003,
      "step": 2971
    },
    {
      "epoch": 40.16,
      "learning_rate": 9.954545454545455e-05,
      "loss": 0.11,
      "step": 2972
    },
    {
      "epoch": 40.18,
      "learning_rate": 9.94090909090909e-05,
      "loss": 0.0054,
      "step": 2973
    },
    {
      "epoch": 40.19,
      "learning_rate": 9.927272727272726e-05,
      "loss": 0.025,
      "step": 2974
    },
    {
      "epoch": 40.2,
      "learning_rate": 9.913636363636362e-05,
      "loss": 0.0008,
      "step": 2975
    },
    {
      "epoch": 40.22,
      "learning_rate": 9.9e-05,
      "loss": 0.001,
      "step": 2976
    },
    {
      "epoch": 40.23,
      "learning_rate": 9.886363636363635e-05,
      "loss": 0.0038,
      "step": 2977
    },
    {
      "epoch": 40.24,
      "learning_rate": 9.872727272727272e-05,
      "loss": 0.0009,
      "step": 2978
    },
    {
      "epoch": 40.26,
      "learning_rate": 9.859090909090908e-05,
      "loss": 0.033,
      "step": 2979
    },
    {
      "epoch": 40.27,
      "learning_rate": 9.845454545454545e-05,
      "loss": 0.0062,
      "step": 2980
    },
    {
      "epoch": 40.28,
      "learning_rate": 9.83181818181818e-05,
      "loss": 0.0785,
      "step": 2981
    },
    {
      "epoch": 40.3,
      "learning_rate": 9.818181818181817e-05,
      "loss": 0.0626,
      "step": 2982
    },
    {
      "epoch": 40.31,
      "learning_rate": 9.804545454545454e-05,
      "loss": 0.0091,
      "step": 2983
    },
    {
      "epoch": 40.32,
      "learning_rate": 9.79090909090909e-05,
      "loss": 0.0027,
      "step": 2984
    },
    {
      "epoch": 40.34,
      "learning_rate": 9.777272727272727e-05,
      "loss": 0.0863,
      "step": 2985
    },
    {
      "epoch": 40.35,
      "learning_rate": 9.763636363636363e-05,
      "loss": 0.0118,
      "step": 2986
    },
    {
      "epoch": 40.36,
      "learning_rate": 9.75e-05,
      "loss": 0.009,
      "step": 2987
    },
    {
      "epoch": 40.38,
      "learning_rate": 9.736363636363634e-05,
      "loss": 0.0007,
      "step": 2988
    },
    {
      "epoch": 40.39,
      "learning_rate": 9.722727272727271e-05,
      "loss": 0.0729,
      "step": 2989
    },
    {
      "epoch": 40.41,
      "learning_rate": 9.709090909090907e-05,
      "loss": 0.0008,
      "step": 2990
    },
    {
      "epoch": 40.42,
      "learning_rate": 9.695454545454544e-05,
      "loss": 0.087,
      "step": 2991
    },
    {
      "epoch": 40.43,
      "learning_rate": 9.681818181818181e-05,
      "loss": 0.0011,
      "step": 2992
    },
    {
      "epoch": 40.45,
      "learning_rate": 9.668181818181817e-05,
      "loss": 0.0983,
      "step": 2993
    },
    {
      "epoch": 40.46,
      "learning_rate": 9.654545454545454e-05,
      "loss": 0.006,
      "step": 2994
    },
    {
      "epoch": 40.47,
      "learning_rate": 9.64090909090909e-05,
      "loss": 0.0033,
      "step": 2995
    },
    {
      "epoch": 40.49,
      "learning_rate": 9.627272727272727e-05,
      "loss": 0.0059,
      "step": 2996
    },
    {
      "epoch": 40.5,
      "learning_rate": 9.613636363636362e-05,
      "loss": 0.0215,
      "step": 2997
    },
    {
      "epoch": 40.51,
      "learning_rate": 9.599999999999999e-05,
      "loss": 0.0047,
      "step": 2998
    },
    {
      "epoch": 40.53,
      "learning_rate": 9.586363636363635e-05,
      "loss": 0.0102,
      "step": 2999
    },
    {
      "epoch": 40.54,
      "learning_rate": 9.572727272727272e-05,
      "loss": 0.0436,
      "step": 3000
    },
    {
      "epoch": 40.54,
      "eval_accuracy": 0.9390243902439024,
      "eval_f1": 0.9396996103665722,
      "eval_loss": 0.3159126043319702,
      "eval_runtime": 126.6167,
      "eval_samples_per_second": 12.952,
      "eval_steps_per_second": 0.813,
      "step": 3000
    },
    {
      "epoch": 40.55,
      "learning_rate": 9.559090909090909e-05,
      "loss": 0.0025,
      "step": 3001
    },
    {
      "epoch": 40.57,
      "learning_rate": 9.545454545454545e-05,
      "loss": 0.0297,
      "step": 3002
    },
    {
      "epoch": 40.58,
      "learning_rate": 9.531818181818182e-05,
      "loss": 0.0059,
      "step": 3003
    },
    {
      "epoch": 40.59,
      "learning_rate": 9.518181818181818e-05,
      "loss": 0.0264,
      "step": 3004
    },
    {
      "epoch": 40.61,
      "learning_rate": 9.504545454545453e-05,
      "loss": 0.0034,
      "step": 3005
    },
    {
      "epoch": 40.62,
      "learning_rate": 9.490909090909089e-05,
      "loss": 0.0011,
      "step": 3006
    },
    {
      "epoch": 40.64,
      "learning_rate": 9.477272727272726e-05,
      "loss": 0.0008,
      "step": 3007
    },
    {
      "epoch": 40.65,
      "learning_rate": 9.463636363636362e-05,
      "loss": 0.0006,
      "step": 3008
    },
    {
      "epoch": 40.66,
      "learning_rate": 9.449999999999999e-05,
      "loss": 0.0075,
      "step": 3009
    },
    {
      "epoch": 40.68,
      "learning_rate": 9.436363636363636e-05,
      "loss": 0.0918,
      "step": 3010
    },
    {
      "epoch": 40.69,
      "learning_rate": 9.422727272727272e-05,
      "loss": 0.0982,
      "step": 3011
    },
    {
      "epoch": 40.7,
      "learning_rate": 9.40909090909091e-05,
      "loss": 0.0028,
      "step": 3012
    },
    {
      "epoch": 40.72,
      "learning_rate": 9.395454545454544e-05,
      "loss": 0.212,
      "step": 3013
    },
    {
      "epoch": 40.73,
      "learning_rate": 9.381818181818181e-05,
      "loss": 0.0545,
      "step": 3014
    },
    {
      "epoch": 40.74,
      "learning_rate": 9.368181818181817e-05,
      "loss": 0.0058,
      "step": 3015
    },
    {
      "epoch": 40.76,
      "learning_rate": 9.354545454545454e-05,
      "loss": 0.0258,
      "step": 3016
    },
    {
      "epoch": 40.77,
      "learning_rate": 9.34090909090909e-05,
      "loss": 0.0039,
      "step": 3017
    },
    {
      "epoch": 40.78,
      "learning_rate": 9.327272727272727e-05,
      "loss": 0.0008,
      "step": 3018
    },
    {
      "epoch": 40.8,
      "learning_rate": 9.313636363636363e-05,
      "loss": 0.0013,
      "step": 3019
    },
    {
      "epoch": 40.81,
      "learning_rate": 9.3e-05,
      "loss": 0.0061,
      "step": 3020
    },
    {
      "epoch": 40.82,
      "learning_rate": 9.286363636363637e-05,
      "loss": 0.001,
      "step": 3021
    },
    {
      "epoch": 40.84,
      "learning_rate": 9.272727272727271e-05,
      "loss": 0.0293,
      "step": 3022
    },
    {
      "epoch": 40.85,
      "learning_rate": 9.259090909090908e-05,
      "loss": 0.0746,
      "step": 3023
    },
    {
      "epoch": 40.86,
      "learning_rate": 9.245454545454544e-05,
      "loss": 0.0018,
      "step": 3024
    },
    {
      "epoch": 40.88,
      "learning_rate": 9.231818181818181e-05,
      "loss": 0.0808,
      "step": 3025
    },
    {
      "epoch": 40.89,
      "learning_rate": 9.218181818181817e-05,
      "loss": 0.0012,
      "step": 3026
    },
    {
      "epoch": 40.91,
      "learning_rate": 9.204545454545454e-05,
      "loss": 0.0031,
      "step": 3027
    },
    {
      "epoch": 40.92,
      "learning_rate": 9.190909090909089e-05,
      "loss": 0.0427,
      "step": 3028
    },
    {
      "epoch": 40.93,
      "learning_rate": 9.177272727272726e-05,
      "loss": 0.0006,
      "step": 3029
    },
    {
      "epoch": 40.95,
      "learning_rate": 9.163636363636363e-05,
      "loss": 0.0377,
      "step": 3030
    },
    {
      "epoch": 40.96,
      "learning_rate": 9.149999999999999e-05,
      "loss": 0.0006,
      "step": 3031
    },
    {
      "epoch": 40.97,
      "learning_rate": 9.136363636363636e-05,
      "loss": 0.0008,
      "step": 3032
    },
    {
      "epoch": 40.99,
      "learning_rate": 9.122727272727272e-05,
      "loss": 0.0007,
      "step": 3033
    },
    {
      "epoch": 41.0,
      "learning_rate": 9.109090909090909e-05,
      "loss": 0.0013,
      "step": 3034
    },
    {
      "epoch": 41.01,
      "learning_rate": 9.095454545454545e-05,
      "loss": 0.0059,
      "step": 3035
    },
    {
      "epoch": 41.03,
      "learning_rate": 9.081818181818182e-05,
      "loss": 0.0283,
      "step": 3036
    },
    {
      "epoch": 41.04,
      "learning_rate": 9.068181818181816e-05,
      "loss": 0.001,
      "step": 3037
    },
    {
      "epoch": 41.05,
      "learning_rate": 9.054545454545453e-05,
      "loss": 0.1788,
      "step": 3038
    },
    {
      "epoch": 41.07,
      "learning_rate": 9.04090909090909e-05,
      "loss": 0.07,
      "step": 3039
    },
    {
      "epoch": 41.08,
      "learning_rate": 9.027272727272726e-05,
      "loss": 0.0047,
      "step": 3040
    },
    {
      "epoch": 41.09,
      "learning_rate": 9.013636363636363e-05,
      "loss": 0.1036,
      "step": 3041
    },
    {
      "epoch": 41.11,
      "learning_rate": 8.999999999999999e-05,
      "loss": 0.0486,
      "step": 3042
    },
    {
      "epoch": 41.12,
      "learning_rate": 8.986363636363636e-05,
      "loss": 0.0013,
      "step": 3043
    },
    {
      "epoch": 41.14,
      "learning_rate": 8.972727272727271e-05,
      "loss": 0.0012,
      "step": 3044
    },
    {
      "epoch": 41.15,
      "learning_rate": 8.959090909090908e-05,
      "loss": 0.0103,
      "step": 3045
    },
    {
      "epoch": 41.16,
      "learning_rate": 8.945454545454544e-05,
      "loss": 0.1589,
      "step": 3046
    },
    {
      "epoch": 41.18,
      "learning_rate": 8.931818181818181e-05,
      "loss": 0.0343,
      "step": 3047
    },
    {
      "epoch": 41.19,
      "learning_rate": 8.918181818181818e-05,
      "loss": 0.0658,
      "step": 3048
    },
    {
      "epoch": 41.2,
      "learning_rate": 8.904545454545454e-05,
      "loss": 0.1007,
      "step": 3049
    },
    {
      "epoch": 41.22,
      "learning_rate": 8.890909090909091e-05,
      "loss": 0.0007,
      "step": 3050
    },
    {
      "epoch": 41.23,
      "learning_rate": 8.877272727272727e-05,
      "loss": 0.0065,
      "step": 3051
    },
    {
      "epoch": 41.24,
      "learning_rate": 8.863636363636364e-05,
      "loss": 0.0028,
      "step": 3052
    },
    {
      "epoch": 41.26,
      "learning_rate": 8.849999999999998e-05,
      "loss": 0.0009,
      "step": 3053
    },
    {
      "epoch": 41.27,
      "learning_rate": 8.836363636363635e-05,
      "loss": 0.0008,
      "step": 3054
    },
    {
      "epoch": 41.28,
      "learning_rate": 8.822727272727271e-05,
      "loss": 0.0013,
      "step": 3055
    },
    {
      "epoch": 41.3,
      "learning_rate": 8.809090909090908e-05,
      "loss": 0.044,
      "step": 3056
    },
    {
      "epoch": 41.31,
      "learning_rate": 8.795454545454544e-05,
      "loss": 0.0294,
      "step": 3057
    },
    {
      "epoch": 41.32,
      "learning_rate": 8.781818181818181e-05,
      "loss": 0.001,
      "step": 3058
    },
    {
      "epoch": 41.34,
      "learning_rate": 8.768181818181818e-05,
      "loss": 0.0189,
      "step": 3059
    },
    {
      "epoch": 41.35,
      "learning_rate": 8.754545454545454e-05,
      "loss": 0.0113,
      "step": 3060
    },
    {
      "epoch": 41.36,
      "learning_rate": 8.74090909090909e-05,
      "loss": 0.0123,
      "step": 3061
    },
    {
      "epoch": 41.38,
      "learning_rate": 8.727272727272726e-05,
      "loss": 0.0015,
      "step": 3062
    },
    {
      "epoch": 41.39,
      "learning_rate": 8.713636363636363e-05,
      "loss": 0.0028,
      "step": 3063
    },
    {
      "epoch": 41.41,
      "learning_rate": 8.699999999999999e-05,
      "loss": 0.0174,
      "step": 3064
    },
    {
      "epoch": 41.42,
      "learning_rate": 8.686363636363636e-05,
      "loss": 0.0611,
      "step": 3065
    },
    {
      "epoch": 41.43,
      "learning_rate": 8.672727272727272e-05,
      "loss": 0.0034,
      "step": 3066
    },
    {
      "epoch": 41.45,
      "learning_rate": 8.659090909090909e-05,
      "loss": 0.0017,
      "step": 3067
    },
    {
      "epoch": 41.46,
      "learning_rate": 8.645454545454546e-05,
      "loss": 0.0019,
      "step": 3068
    },
    {
      "epoch": 41.47,
      "learning_rate": 8.63181818181818e-05,
      "loss": 0.0302,
      "step": 3069
    },
    {
      "epoch": 41.49,
      "learning_rate": 8.618181818181817e-05,
      "loss": 0.0009,
      "step": 3070
    },
    {
      "epoch": 41.5,
      "learning_rate": 8.604545454545453e-05,
      "loss": 0.0014,
      "step": 3071
    },
    {
      "epoch": 41.51,
      "learning_rate": 8.59090909090909e-05,
      "loss": 0.0206,
      "step": 3072
    },
    {
      "epoch": 41.53,
      "learning_rate": 8.577272727272726e-05,
      "loss": 0.0158,
      "step": 3073
    },
    {
      "epoch": 41.54,
      "learning_rate": 8.563636363636363e-05,
      "loss": 0.0012,
      "step": 3074
    },
    {
      "epoch": 41.55,
      "learning_rate": 8.549999999999999e-05,
      "loss": 0.0079,
      "step": 3075
    },
    {
      "epoch": 41.57,
      "learning_rate": 8.536363636363636e-05,
      "loss": 0.0035,
      "step": 3076
    },
    {
      "epoch": 41.58,
      "learning_rate": 8.522727272727273e-05,
      "loss": 0.0305,
      "step": 3077
    },
    {
      "epoch": 41.59,
      "learning_rate": 8.509090909090908e-05,
      "loss": 0.0204,
      "step": 3078
    },
    {
      "epoch": 41.61,
      "learning_rate": 8.495454545454545e-05,
      "loss": 0.0134,
      "step": 3079
    },
    {
      "epoch": 41.62,
      "learning_rate": 8.481818181818181e-05,
      "loss": 0.0018,
      "step": 3080
    },
    {
      "epoch": 41.64,
      "learning_rate": 8.468181818181818e-05,
      "loss": 0.0012,
      "step": 3081
    },
    {
      "epoch": 41.65,
      "learning_rate": 8.454545454545454e-05,
      "loss": 0.015,
      "step": 3082
    },
    {
      "epoch": 41.66,
      "learning_rate": 8.440909090909091e-05,
      "loss": 0.0011,
      "step": 3083
    },
    {
      "epoch": 41.68,
      "learning_rate": 8.427272727272725e-05,
      "loss": 0.0194,
      "step": 3084
    },
    {
      "epoch": 41.69,
      "learning_rate": 8.413636363636362e-05,
      "loss": 0.0026,
      "step": 3085
    },
    {
      "epoch": 41.7,
      "learning_rate": 8.4e-05,
      "loss": 0.0026,
      "step": 3086
    },
    {
      "epoch": 41.72,
      "learning_rate": 8.386363636363635e-05,
      "loss": 0.0203,
      "step": 3087
    },
    {
      "epoch": 41.73,
      "learning_rate": 8.372727272727272e-05,
      "loss": 0.001,
      "step": 3088
    },
    {
      "epoch": 41.74,
      "learning_rate": 8.359090909090908e-05,
      "loss": 0.0038,
      "step": 3089
    },
    {
      "epoch": 41.76,
      "learning_rate": 8.345454545454545e-05,
      "loss": 0.0071,
      "step": 3090
    },
    {
      "epoch": 41.77,
      "learning_rate": 8.331818181818181e-05,
      "loss": 0.0013,
      "step": 3091
    },
    {
      "epoch": 41.78,
      "learning_rate": 8.318181818181818e-05,
      "loss": 0.0025,
      "step": 3092
    },
    {
      "epoch": 41.8,
      "learning_rate": 8.304545454545453e-05,
      "loss": 0.0018,
      "step": 3093
    },
    {
      "epoch": 41.81,
      "learning_rate": 8.29090909090909e-05,
      "loss": 0.0208,
      "step": 3094
    },
    {
      "epoch": 41.82,
      "learning_rate": 8.277272727272726e-05,
      "loss": 0.0014,
      "step": 3095
    },
    {
      "epoch": 41.84,
      "learning_rate": 8.263636363636363e-05,
      "loss": 0.0028,
      "step": 3096
    },
    {
      "epoch": 41.85,
      "learning_rate": 8.25e-05,
      "loss": 0.02,
      "step": 3097
    },
    {
      "epoch": 41.86,
      "learning_rate": 8.236363636363636e-05,
      "loss": 0.0012,
      "step": 3098
    },
    {
      "epoch": 41.88,
      "learning_rate": 8.222727272727273e-05,
      "loss": 0.0026,
      "step": 3099
    },
    {
      "epoch": 41.89,
      "learning_rate": 8.209090909090907e-05,
      "loss": 0.0401,
      "step": 3100
    },
    {
      "epoch": 41.91,
      "learning_rate": 8.195454545454544e-05,
      "loss": 0.0012,
      "step": 3101
    },
    {
      "epoch": 41.92,
      "learning_rate": 8.18181818181818e-05,
      "loss": 0.0008,
      "step": 3102
    },
    {
      "epoch": 41.93,
      "learning_rate": 8.168181818181817e-05,
      "loss": 0.0046,
      "step": 3103
    },
    {
      "epoch": 41.95,
      "learning_rate": 8.154545454545453e-05,
      "loss": 0.0017,
      "step": 3104
    },
    {
      "epoch": 41.96,
      "learning_rate": 8.14090909090909e-05,
      "loss": 0.0008,
      "step": 3105
    },
    {
      "epoch": 41.97,
      "learning_rate": 8.127272727272727e-05,
      "loss": 0.0167,
      "step": 3106
    },
    {
      "epoch": 41.99,
      "learning_rate": 8.113636363636363e-05,
      "loss": 0.0043,
      "step": 3107
    },
    {
      "epoch": 42.0,
      "learning_rate": 8.1e-05,
      "loss": 0.0011,
      "step": 3108
    },
    {
      "epoch": 42.01,
      "learning_rate": 8.086363636363635e-05,
      "loss": 0.0528,
      "step": 3109
    },
    {
      "epoch": 42.03,
      "learning_rate": 8.072727272727272e-05,
      "loss": 0.0021,
      "step": 3110
    },
    {
      "epoch": 42.04,
      "learning_rate": 8.059090909090908e-05,
      "loss": 0.0029,
      "step": 3111
    },
    {
      "epoch": 42.05,
      "learning_rate": 8.045454545454545e-05,
      "loss": 0.0027,
      "step": 3112
    },
    {
      "epoch": 42.07,
      "learning_rate": 8.03181818181818e-05,
      "loss": 0.0013,
      "step": 3113
    },
    {
      "epoch": 42.08,
      "learning_rate": 8.018181818181818e-05,
      "loss": 0.004,
      "step": 3114
    },
    {
      "epoch": 42.09,
      "learning_rate": 8.004545454545455e-05,
      "loss": 0.0683,
      "step": 3115
    },
    {
      "epoch": 42.11,
      "learning_rate": 7.99090909090909e-05,
      "loss": 0.0006,
      "step": 3116
    },
    {
      "epoch": 42.12,
      "learning_rate": 7.977272727272727e-05,
      "loss": 0.0029,
      "step": 3117
    },
    {
      "epoch": 42.14,
      "learning_rate": 7.963636363636362e-05,
      "loss": 0.0013,
      "step": 3118
    },
    {
      "epoch": 42.15,
      "learning_rate": 7.95e-05,
      "loss": 0.0019,
      "step": 3119
    },
    {
      "epoch": 42.16,
      "learning_rate": 7.936363636363635e-05,
      "loss": 0.0235,
      "step": 3120
    },
    {
      "epoch": 42.18,
      "learning_rate": 7.922727272727272e-05,
      "loss": 0.0033,
      "step": 3121
    },
    {
      "epoch": 42.19,
      "learning_rate": 7.909090909090908e-05,
      "loss": 0.0214,
      "step": 3122
    },
    {
      "epoch": 42.2,
      "learning_rate": 7.895454545454545e-05,
      "loss": 0.1215,
      "step": 3123
    },
    {
      "epoch": 42.22,
      "learning_rate": 7.881818181818182e-05,
      "loss": 0.0194,
      "step": 3124
    },
    {
      "epoch": 42.23,
      "learning_rate": 7.868181818181817e-05,
      "loss": 0.0006,
      "step": 3125
    },
    {
      "epoch": 42.24,
      "learning_rate": 7.854545454545454e-05,
      "loss": 0.0017,
      "step": 3126
    },
    {
      "epoch": 42.26,
      "learning_rate": 7.84090909090909e-05,
      "loss": 0.0636,
      "step": 3127
    },
    {
      "epoch": 42.27,
      "learning_rate": 7.827272727272727e-05,
      "loss": 0.0008,
      "step": 3128
    },
    {
      "epoch": 42.28,
      "learning_rate": 7.813636363636363e-05,
      "loss": 0.0143,
      "step": 3129
    },
    {
      "epoch": 42.3,
      "learning_rate": 7.8e-05,
      "loss": 0.0123,
      "step": 3130
    },
    {
      "epoch": 42.31,
      "learning_rate": 7.786363636363636e-05,
      "loss": 0.0008,
      "step": 3131
    },
    {
      "epoch": 42.32,
      "learning_rate": 7.772727272727273e-05,
      "loss": 0.0019,
      "step": 3132
    },
    {
      "epoch": 42.34,
      "learning_rate": 7.759090909090907e-05,
      "loss": 0.001,
      "step": 3133
    },
    {
      "epoch": 42.35,
      "learning_rate": 7.745454545454544e-05,
      "loss": 0.0031,
      "step": 3134
    },
    {
      "epoch": 42.36,
      "learning_rate": 7.731818181818181e-05,
      "loss": 0.0059,
      "step": 3135
    },
    {
      "epoch": 42.38,
      "learning_rate": 7.718181818181817e-05,
      "loss": 0.0009,
      "step": 3136
    },
    {
      "epoch": 42.39,
      "learning_rate": 7.704545454545454e-05,
      "loss": 0.0213,
      "step": 3137
    },
    {
      "epoch": 42.41,
      "learning_rate": 7.69090909090909e-05,
      "loss": 0.0009,
      "step": 3138
    },
    {
      "epoch": 42.42,
      "learning_rate": 7.677272727272727e-05,
      "loss": 0.1451,
      "step": 3139
    },
    {
      "epoch": 42.43,
      "learning_rate": 7.663636363636362e-05,
      "loss": 0.0121,
      "step": 3140
    },
    {
      "epoch": 42.45,
      "learning_rate": 7.649999999999999e-05,
      "loss": 0.0046,
      "step": 3141
    },
    {
      "epoch": 42.46,
      "learning_rate": 7.636363636363635e-05,
      "loss": 0.0021,
      "step": 3142
    },
    {
      "epoch": 42.47,
      "learning_rate": 7.622727272727272e-05,
      "loss": 0.0363,
      "step": 3143
    },
    {
      "epoch": 42.49,
      "learning_rate": 7.609090909090909e-05,
      "loss": 0.0008,
      "step": 3144
    },
    {
      "epoch": 42.5,
      "learning_rate": 7.595454545454545e-05,
      "loss": 0.0092,
      "step": 3145
    },
    {
      "epoch": 42.51,
      "learning_rate": 7.581818181818182e-05,
      "loss": 0.0008,
      "step": 3146
    },
    {
      "epoch": 42.53,
      "learning_rate": 7.568181818181818e-05,
      "loss": 0.0827,
      "step": 3147
    },
    {
      "epoch": 42.54,
      "learning_rate": 7.554545454545455e-05,
      "loss": 0.011,
      "step": 3148
    },
    {
      "epoch": 42.55,
      "learning_rate": 7.540909090909089e-05,
      "loss": 0.0034,
      "step": 3149
    },
    {
      "epoch": 42.57,
      "learning_rate": 7.527272727272726e-05,
      "loss": 0.0006,
      "step": 3150
    },
    {
      "epoch": 42.58,
      "learning_rate": 7.513636363636362e-05,
      "loss": 0.0648,
      "step": 3151
    },
    {
      "epoch": 42.59,
      "learning_rate": 7.5e-05,
      "loss": 0.0013,
      "step": 3152
    },
    {
      "epoch": 42.61,
      "learning_rate": 7.486363636363635e-05,
      "loss": 0.0186,
      "step": 3153
    },
    {
      "epoch": 42.62,
      "learning_rate": 7.472727272727272e-05,
      "loss": 0.0091,
      "step": 3154
    },
    {
      "epoch": 42.64,
      "learning_rate": 7.459090909090908e-05,
      "loss": 0.0007,
      "step": 3155
    },
    {
      "epoch": 42.65,
      "learning_rate": 7.445454545454544e-05,
      "loss": 0.0519,
      "step": 3156
    },
    {
      "epoch": 42.66,
      "learning_rate": 7.431818181818181e-05,
      "loss": 0.0784,
      "step": 3157
    },
    {
      "epoch": 42.68,
      "learning_rate": 7.418181818181818e-05,
      "loss": 0.0029,
      "step": 3158
    },
    {
      "epoch": 42.69,
      "learning_rate": 7.404545454545454e-05,
      "loss": 0.007,
      "step": 3159
    },
    {
      "epoch": 42.7,
      "learning_rate": 7.390909090909091e-05,
      "loss": 0.0006,
      "step": 3160
    },
    {
      "epoch": 42.72,
      "learning_rate": 7.377272727272727e-05,
      "loss": 0.0027,
      "step": 3161
    },
    {
      "epoch": 42.73,
      "learning_rate": 7.363636363636363e-05,
      "loss": 0.0019,
      "step": 3162
    },
    {
      "epoch": 42.74,
      "learning_rate": 7.35e-05,
      "loss": 0.0024,
      "step": 3163
    },
    {
      "epoch": 42.76,
      "learning_rate": 7.336363636363636e-05,
      "loss": 0.0019,
      "step": 3164
    },
    {
      "epoch": 42.77,
      "learning_rate": 7.322727272727271e-05,
      "loss": 0.0059,
      "step": 3165
    },
    {
      "epoch": 42.78,
      "learning_rate": 7.309090909090908e-05,
      "loss": 0.0038,
      "step": 3166
    },
    {
      "epoch": 42.8,
      "learning_rate": 7.295454545454546e-05,
      "loss": 0.024,
      "step": 3167
    },
    {
      "epoch": 42.81,
      "learning_rate": 7.281818181818181e-05,
      "loss": 0.1597,
      "step": 3168
    },
    {
      "epoch": 42.82,
      "learning_rate": 7.268181818181817e-05,
      "loss": 0.005,
      "step": 3169
    },
    {
      "epoch": 42.84,
      "learning_rate": 7.254545454545454e-05,
      "loss": 0.0006,
      "step": 3170
    },
    {
      "epoch": 42.85,
      "learning_rate": 7.24090909090909e-05,
      "loss": 0.1035,
      "step": 3171
    },
    {
      "epoch": 42.86,
      "learning_rate": 7.227272727272726e-05,
      "loss": 0.0181,
      "step": 3172
    },
    {
      "epoch": 42.88,
      "learning_rate": 7.213636363636363e-05,
      "loss": 0.0009,
      "step": 3173
    },
    {
      "epoch": 42.89,
      "learning_rate": 7.199999999999999e-05,
      "loss": 0.1255,
      "step": 3174
    },
    {
      "epoch": 42.91,
      "learning_rate": 7.186363636363636e-05,
      "loss": 0.0094,
      "step": 3175
    },
    {
      "epoch": 42.92,
      "learning_rate": 7.172727272727273e-05,
      "loss": 0.0006,
      "step": 3176
    },
    {
      "epoch": 42.93,
      "learning_rate": 7.159090909090909e-05,
      "loss": 0.0137,
      "step": 3177
    },
    {
      "epoch": 42.95,
      "learning_rate": 7.145454545454545e-05,
      "loss": 0.0073,
      "step": 3178
    },
    {
      "epoch": 42.96,
      "learning_rate": 7.131818181818182e-05,
      "loss": 0.0008,
      "step": 3179
    },
    {
      "epoch": 42.97,
      "learning_rate": 7.118181818181818e-05,
      "loss": 0.0012,
      "step": 3180
    },
    {
      "epoch": 42.99,
      "learning_rate": 7.104545454545453e-05,
      "loss": 0.001,
      "step": 3181
    },
    {
      "epoch": 43.0,
      "learning_rate": 7.09090909090909e-05,
      "loss": 0.0017,
      "step": 3182
    },
    {
      "epoch": 43.01,
      "learning_rate": 7.077272727272726e-05,
      "loss": 0.0074,
      "step": 3183
    },
    {
      "epoch": 43.03,
      "learning_rate": 7.063636363636362e-05,
      "loss": 0.0007,
      "step": 3184
    },
    {
      "epoch": 43.04,
      "learning_rate": 7.049999999999999e-05,
      "loss": 0.0009,
      "step": 3185
    },
    {
      "epoch": 43.05,
      "learning_rate": 7.036363636363636e-05,
      "loss": 0.0008,
      "step": 3186
    },
    {
      "epoch": 43.07,
      "learning_rate": 7.022727272727272e-05,
      "loss": 0.002,
      "step": 3187
    },
    {
      "epoch": 43.08,
      "learning_rate": 7.009090909090909e-05,
      "loss": 0.1703,
      "step": 3188
    },
    {
      "epoch": 43.09,
      "learning_rate": 6.995454545454545e-05,
      "loss": 0.0416,
      "step": 3189
    },
    {
      "epoch": 43.11,
      "learning_rate": 6.981818181818181e-05,
      "loss": 0.0027,
      "step": 3190
    },
    {
      "epoch": 43.12,
      "learning_rate": 6.968181818181818e-05,
      "loss": 0.0015,
      "step": 3191
    },
    {
      "epoch": 43.14,
      "learning_rate": 6.954545454545454e-05,
      "loss": 0.0006,
      "step": 3192
    },
    {
      "epoch": 43.15,
      "learning_rate": 6.94090909090909e-05,
      "loss": 0.0009,
      "step": 3193
    },
    {
      "epoch": 43.16,
      "learning_rate": 6.927272727272727e-05,
      "loss": 0.0011,
      "step": 3194
    },
    {
      "epoch": 43.18,
      "learning_rate": 6.913636363636364e-05,
      "loss": 0.0007,
      "step": 3195
    },
    {
      "epoch": 43.19,
      "learning_rate": 6.9e-05,
      "loss": 0.0031,
      "step": 3196
    },
    {
      "epoch": 43.2,
      "learning_rate": 6.886363636363635e-05,
      "loss": 0.0289,
      "step": 3197
    },
    {
      "epoch": 43.22,
      "learning_rate": 6.872727272727273e-05,
      "loss": 0.0007,
      "step": 3198
    },
    {
      "epoch": 43.23,
      "learning_rate": 6.859090909090908e-05,
      "loss": 0.0008,
      "step": 3199
    },
    {
      "epoch": 43.24,
      "learning_rate": 6.845454545454544e-05,
      "loss": 0.0455,
      "step": 3200
    },
    {
      "epoch": 43.24,
      "eval_accuracy": 0.9390243902439024,
      "eval_f1": 0.9393269204505507,
      "eval_loss": 0.2962814271450043,
      "eval_runtime": 126.4543,
      "eval_samples_per_second": 12.969,
      "eval_steps_per_second": 0.815,
      "step": 3200
    },
    {
      "epoch": 43.26,
      "learning_rate": 6.831818181818181e-05,
      "loss": 0.0009,
      "step": 3201
    },
    {
      "epoch": 43.27,
      "learning_rate": 6.818181818181817e-05,
      "loss": 0.0437,
      "step": 3202
    },
    {
      "epoch": 43.28,
      "learning_rate": 6.804545454545454e-05,
      "loss": 0.0234,
      "step": 3203
    },
    {
      "epoch": 43.3,
      "learning_rate": 6.79090909090909e-05,
      "loss": 0.0007,
      "step": 3204
    },
    {
      "epoch": 43.31,
      "learning_rate": 6.777272727272727e-05,
      "loss": 0.0044,
      "step": 3205
    },
    {
      "epoch": 43.32,
      "learning_rate": 6.763636363636363e-05,
      "loss": 0.0008,
      "step": 3206
    },
    {
      "epoch": 43.34,
      "learning_rate": 6.75e-05,
      "loss": 0.0142,
      "step": 3207
    },
    {
      "epoch": 43.35,
      "learning_rate": 6.736363636363636e-05,
      "loss": 0.0921,
      "step": 3208
    },
    {
      "epoch": 43.36,
      "learning_rate": 6.722727272727272e-05,
      "loss": 0.035,
      "step": 3209
    },
    {
      "epoch": 43.38,
      "learning_rate": 6.709090909090909e-05,
      "loss": 0.0014,
      "step": 3210
    },
    {
      "epoch": 43.39,
      "learning_rate": 6.695454545454545e-05,
      "loss": 0.001,
      "step": 3211
    },
    {
      "epoch": 43.41,
      "learning_rate": 6.68181818181818e-05,
      "loss": 0.0005,
      "step": 3212
    },
    {
      "epoch": 43.42,
      "learning_rate": 6.668181818181818e-05,
      "loss": 0.0006,
      "step": 3213
    },
    {
      "epoch": 43.43,
      "learning_rate": 6.654545454545455e-05,
      "loss": 0.0005,
      "step": 3214
    },
    {
      "epoch": 43.45,
      "learning_rate": 6.64090909090909e-05,
      "loss": 0.001,
      "step": 3215
    },
    {
      "epoch": 43.46,
      "learning_rate": 6.627272727272726e-05,
      "loss": 0.0009,
      "step": 3216
    },
    {
      "epoch": 43.47,
      "learning_rate": 6.613636363636363e-05,
      "loss": 0.0359,
      "step": 3217
    },
    {
      "epoch": 43.49,
      "learning_rate": 6.599999999999999e-05,
      "loss": 0.0006,
      "step": 3218
    },
    {
      "epoch": 43.5,
      "learning_rate": 6.586363636363636e-05,
      "loss": 0.0172,
      "step": 3219
    },
    {
      "epoch": 43.51,
      "learning_rate": 6.572727272727272e-05,
      "loss": 0.0005,
      "step": 3220
    },
    {
      "epoch": 43.53,
      "learning_rate": 6.559090909090908e-05,
      "loss": 0.0163,
      "step": 3221
    },
    {
      "epoch": 43.54,
      "learning_rate": 6.545454545454545e-05,
      "loss": 0.0032,
      "step": 3222
    },
    {
      "epoch": 43.55,
      "learning_rate": 6.531818181818181e-05,
      "loss": 0.0012,
      "step": 3223
    },
    {
      "epoch": 43.57,
      "learning_rate": 6.518181818181818e-05,
      "loss": 0.0537,
      "step": 3224
    },
    {
      "epoch": 43.58,
      "learning_rate": 6.504545454545454e-05,
      "loss": 0.0608,
      "step": 3225
    },
    {
      "epoch": 43.59,
      "learning_rate": 6.490909090909091e-05,
      "loss": 0.0009,
      "step": 3226
    },
    {
      "epoch": 43.61,
      "learning_rate": 6.477272727272727e-05,
      "loss": 0.0007,
      "step": 3227
    },
    {
      "epoch": 43.62,
      "learning_rate": 6.463636363636362e-05,
      "loss": 0.1407,
      "step": 3228
    },
    {
      "epoch": 43.64,
      "learning_rate": 6.45e-05,
      "loss": 0.0009,
      "step": 3229
    },
    {
      "epoch": 43.65,
      "learning_rate": 6.436363636363635e-05,
      "loss": 0.0019,
      "step": 3230
    },
    {
      "epoch": 43.66,
      "learning_rate": 6.422727272727272e-05,
      "loss": 0.0972,
      "step": 3231
    },
    {
      "epoch": 43.68,
      "learning_rate": 6.409090909090908e-05,
      "loss": 0.0827,
      "step": 3232
    },
    {
      "epoch": 43.69,
      "learning_rate": 6.395454545454545e-05,
      "loss": 0.0011,
      "step": 3233
    },
    {
      "epoch": 43.7,
      "learning_rate": 6.381818181818181e-05,
      "loss": 0.0005,
      "step": 3234
    },
    {
      "epoch": 43.72,
      "learning_rate": 6.368181818181818e-05,
      "loss": 0.0006,
      "step": 3235
    },
    {
      "epoch": 43.73,
      "learning_rate": 6.354545454545454e-05,
      "loss": 0.0006,
      "step": 3236
    },
    {
      "epoch": 43.74,
      "learning_rate": 6.34090909090909e-05,
      "loss": 0.0008,
      "step": 3237
    },
    {
      "epoch": 43.76,
      "learning_rate": 6.327272727272727e-05,
      "loss": 0.0083,
      "step": 3238
    },
    {
      "epoch": 43.77,
      "learning_rate": 6.313636363636363e-05,
      "loss": 0.0007,
      "step": 3239
    },
    {
      "epoch": 43.78,
      "learning_rate": 6.299999999999999e-05,
      "loss": 0.0008,
      "step": 3240
    },
    {
      "epoch": 43.8,
      "learning_rate": 6.286363636363636e-05,
      "loss": 0.0008,
      "step": 3241
    },
    {
      "epoch": 43.81,
      "learning_rate": 6.272727272727272e-05,
      "loss": 0.0141,
      "step": 3242
    },
    {
      "epoch": 43.82,
      "learning_rate": 6.259090909090909e-05,
      "loss": 0.0019,
      "step": 3243
    },
    {
      "epoch": 43.84,
      "learning_rate": 6.245454545454544e-05,
      "loss": 0.0007,
      "step": 3244
    },
    {
      "epoch": 43.85,
      "learning_rate": 6.231818181818182e-05,
      "loss": 0.0019,
      "step": 3245
    },
    {
      "epoch": 43.86,
      "learning_rate": 6.218181818181817e-05,
      "loss": 0.0731,
      "step": 3246
    },
    {
      "epoch": 43.88,
      "learning_rate": 6.204545454545455e-05,
      "loss": 0.0012,
      "step": 3247
    },
    {
      "epoch": 43.89,
      "learning_rate": 6.19090909090909e-05,
      "loss": 0.0009,
      "step": 3248
    },
    {
      "epoch": 43.91,
      "learning_rate": 6.177272727272726e-05,
      "loss": 0.0192,
      "step": 3249
    },
    {
      "epoch": 43.92,
      "learning_rate": 6.163636363636363e-05,
      "loss": 0.0546,
      "step": 3250
    },
    {
      "epoch": 43.93,
      "learning_rate": 6.149999999999999e-05,
      "loss": 0.0094,
      "step": 3251
    },
    {
      "epoch": 43.95,
      "learning_rate": 6.136363636363636e-05,
      "loss": 0.0007,
      "step": 3252
    },
    {
      "epoch": 43.96,
      "learning_rate": 6.122727272727272e-05,
      "loss": 0.0216,
      "step": 3253
    },
    {
      "epoch": 43.97,
      "learning_rate": 6.109090909090909e-05,
      "loss": 0.0151,
      "step": 3254
    },
    {
      "epoch": 43.99,
      "learning_rate": 6.095454545454545e-05,
      "loss": 0.0106,
      "step": 3255
    },
    {
      "epoch": 44.0,
      "learning_rate": 6.0818181818181814e-05,
      "loss": 0.0005,
      "step": 3256
    },
    {
      "epoch": 44.01,
      "learning_rate": 6.068181818181818e-05,
      "loss": 0.0027,
      "step": 3257
    },
    {
      "epoch": 44.03,
      "learning_rate": 6.0545454545454536e-05,
      "loss": 0.0005,
      "step": 3258
    },
    {
      "epoch": 44.04,
      "learning_rate": 6.04090909090909e-05,
      "loss": 0.0019,
      "step": 3259
    },
    {
      "epoch": 44.05,
      "learning_rate": 6.0272727272727265e-05,
      "loss": 0.0009,
      "step": 3260
    },
    {
      "epoch": 44.07,
      "learning_rate": 6.013636363636363e-05,
      "loss": 0.0008,
      "step": 3261
    },
    {
      "epoch": 44.08,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 0.0022,
      "step": 3262
    },
    {
      "epoch": 44.09,
      "learning_rate": 5.986363636363636e-05,
      "loss": 0.001,
      "step": 3263
    },
    {
      "epoch": 44.11,
      "learning_rate": 5.9727272727272724e-05,
      "loss": 0.0737,
      "step": 3264
    },
    {
      "epoch": 44.12,
      "learning_rate": 5.959090909090909e-05,
      "loss": 0.0004,
      "step": 3265
    },
    {
      "epoch": 44.14,
      "learning_rate": 5.9454545454545447e-05,
      "loss": 0.0457,
      "step": 3266
    },
    {
      "epoch": 44.15,
      "learning_rate": 5.931818181818181e-05,
      "loss": 0.0283,
      "step": 3267
    },
    {
      "epoch": 44.16,
      "learning_rate": 5.9181818181818176e-05,
      "loss": 0.1305,
      "step": 3268
    },
    {
      "epoch": 44.18,
      "learning_rate": 5.904545454545454e-05,
      "loss": 0.0117,
      "step": 3269
    },
    {
      "epoch": 44.19,
      "learning_rate": 5.89090909090909e-05,
      "loss": 0.0301,
      "step": 3270
    },
    {
      "epoch": 44.2,
      "learning_rate": 5.877272727272727e-05,
      "loss": 0.0005,
      "step": 3271
    },
    {
      "epoch": 44.22,
      "learning_rate": 5.8636363636363634e-05,
      "loss": 0.0011,
      "step": 3272
    },
    {
      "epoch": 44.23,
      "learning_rate": 5.85e-05,
      "loss": 0.0022,
      "step": 3273
    },
    {
      "epoch": 44.24,
      "learning_rate": 5.836363636363636e-05,
      "loss": 0.0122,
      "step": 3274
    },
    {
      "epoch": 44.26,
      "learning_rate": 5.822727272727272e-05,
      "loss": 0.0006,
      "step": 3275
    },
    {
      "epoch": 44.27,
      "learning_rate": 5.8090909090909086e-05,
      "loss": 0.0004,
      "step": 3276
    },
    {
      "epoch": 44.28,
      "learning_rate": 5.795454545454545e-05,
      "loss": 0.0011,
      "step": 3277
    },
    {
      "epoch": 44.3,
      "learning_rate": 5.781818181818181e-05,
      "loss": 0.0016,
      "step": 3278
    },
    {
      "epoch": 44.31,
      "learning_rate": 5.768181818181817e-05,
      "loss": 0.0007,
      "step": 3279
    },
    {
      "epoch": 44.32,
      "learning_rate": 5.754545454545454e-05,
      "loss": 0.029,
      "step": 3280
    },
    {
      "epoch": 44.34,
      "learning_rate": 5.740909090909091e-05,
      "loss": 0.0038,
      "step": 3281
    },
    {
      "epoch": 44.35,
      "learning_rate": 5.727272727272727e-05,
      "loss": 0.0011,
      "step": 3282
    },
    {
      "epoch": 44.36,
      "learning_rate": 5.713636363636363e-05,
      "loss": 0.0005,
      "step": 3283
    },
    {
      "epoch": 44.38,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 0.0006,
      "step": 3284
    },
    {
      "epoch": 44.39,
      "learning_rate": 5.686363636363636e-05,
      "loss": 0.0008,
      "step": 3285
    },
    {
      "epoch": 44.41,
      "learning_rate": 5.672727272727272e-05,
      "loss": 0.0914,
      "step": 3286
    },
    {
      "epoch": 44.42,
      "learning_rate": 5.6590909090909083e-05,
      "loss": 0.0013,
      "step": 3287
    },
    {
      "epoch": 44.43,
      "learning_rate": 5.645454545454545e-05,
      "loss": 0.0005,
      "step": 3288
    },
    {
      "epoch": 44.45,
      "learning_rate": 5.631818181818181e-05,
      "loss": 0.0012,
      "step": 3289
    },
    {
      "epoch": 44.46,
      "learning_rate": 5.618181818181818e-05,
      "loss": 0.0006,
      "step": 3290
    },
    {
      "epoch": 44.47,
      "learning_rate": 5.604545454545454e-05,
      "loss": 0.0007,
      "step": 3291
    },
    {
      "epoch": 44.49,
      "learning_rate": 5.590909090909091e-05,
      "loss": 0.0028,
      "step": 3292
    },
    {
      "epoch": 44.5,
      "learning_rate": 5.577272727272727e-05,
      "loss": 0.0008,
      "step": 3293
    },
    {
      "epoch": 44.51,
      "learning_rate": 5.563636363636363e-05,
      "loss": 0.0009,
      "step": 3294
    },
    {
      "epoch": 44.53,
      "learning_rate": 5.5499999999999994e-05,
      "loss": 0.0089,
      "step": 3295
    },
    {
      "epoch": 44.54,
      "learning_rate": 5.536363636363636e-05,
      "loss": 0.0009,
      "step": 3296
    },
    {
      "epoch": 44.55,
      "learning_rate": 5.522727272727272e-05,
      "loss": 0.1454,
      "step": 3297
    },
    {
      "epoch": 44.57,
      "learning_rate": 5.509090909090908e-05,
      "loss": 0.0025,
      "step": 3298
    },
    {
      "epoch": 44.58,
      "learning_rate": 5.495454545454545e-05,
      "loss": 0.0006,
      "step": 3299
    },
    {
      "epoch": 44.59,
      "learning_rate": 5.481818181818182e-05,
      "loss": 0.0054,
      "step": 3300
    },
    {
      "epoch": 44.61,
      "learning_rate": 5.468181818181818e-05,
      "loss": 0.0007,
      "step": 3301
    },
    {
      "epoch": 44.62,
      "learning_rate": 5.454545454545454e-05,
      "loss": 0.0005,
      "step": 3302
    },
    {
      "epoch": 44.64,
      "learning_rate": 5.4409090909090904e-05,
      "loss": 0.0008,
      "step": 3303
    },
    {
      "epoch": 44.65,
      "learning_rate": 5.427272727272727e-05,
      "loss": 0.0067,
      "step": 3304
    },
    {
      "epoch": 44.66,
      "learning_rate": 5.413636363636363e-05,
      "loss": 0.0005,
      "step": 3305
    },
    {
      "epoch": 44.68,
      "learning_rate": 5.399999999999999e-05,
      "loss": 0.0022,
      "step": 3306
    },
    {
      "epoch": 44.69,
      "learning_rate": 5.3863636363636356e-05,
      "loss": 0.0006,
      "step": 3307
    },
    {
      "epoch": 44.7,
      "learning_rate": 5.372727272727272e-05,
      "loss": 0.0005,
      "step": 3308
    },
    {
      "epoch": 44.72,
      "learning_rate": 5.359090909090909e-05,
      "loss": 0.0048,
      "step": 3309
    },
    {
      "epoch": 44.73,
      "learning_rate": 5.345454545454545e-05,
      "loss": 0.0013,
      "step": 3310
    },
    {
      "epoch": 44.74,
      "learning_rate": 5.3318181818181814e-05,
      "loss": 0.0051,
      "step": 3311
    },
    {
      "epoch": 44.76,
      "learning_rate": 5.318181818181818e-05,
      "loss": 0.0006,
      "step": 3312
    },
    {
      "epoch": 44.77,
      "learning_rate": 5.3045454545454544e-05,
      "loss": 0.005,
      "step": 3313
    },
    {
      "epoch": 44.78,
      "learning_rate": 5.29090909090909e-05,
      "loss": 0.0341,
      "step": 3314
    },
    {
      "epoch": 44.8,
      "learning_rate": 5.2772727272727266e-05,
      "loss": 0.0004,
      "step": 3315
    },
    {
      "epoch": 44.81,
      "learning_rate": 5.263636363636363e-05,
      "loss": 0.0013,
      "step": 3316
    },
    {
      "epoch": 44.82,
      "learning_rate": 5.2499999999999995e-05,
      "loss": 0.0007,
      "step": 3317
    },
    {
      "epoch": 44.84,
      "learning_rate": 5.236363636363636e-05,
      "loss": 0.0004,
      "step": 3318
    },
    {
      "epoch": 44.85,
      "learning_rate": 5.2227272727272725e-05,
      "loss": 0.0743,
      "step": 3319
    },
    {
      "epoch": 44.86,
      "learning_rate": 5.209090909090909e-05,
      "loss": 0.0015,
      "step": 3320
    },
    {
      "epoch": 44.88,
      "learning_rate": 5.1954545454545454e-05,
      "loss": 0.0794,
      "step": 3321
    },
    {
      "epoch": 44.89,
      "learning_rate": 5.181818181818181e-05,
      "loss": 0.0005,
      "step": 3322
    },
    {
      "epoch": 44.91,
      "learning_rate": 5.1681818181818176e-05,
      "loss": 0.002,
      "step": 3323
    },
    {
      "epoch": 44.92,
      "learning_rate": 5.154545454545454e-05,
      "loss": 0.0005,
      "step": 3324
    },
    {
      "epoch": 44.93,
      "learning_rate": 5.1409090909090906e-05,
      "loss": 0.003,
      "step": 3325
    },
    {
      "epoch": 44.95,
      "learning_rate": 5.1272727272727264e-05,
      "loss": 0.0005,
      "step": 3326
    },
    {
      "epoch": 44.96,
      "learning_rate": 5.113636363636363e-05,
      "loss": 0.0019,
      "step": 3327
    },
    {
      "epoch": 44.97,
      "learning_rate": 5.1e-05,
      "loss": 0.0023,
      "step": 3328
    },
    {
      "epoch": 44.99,
      "learning_rate": 5.0863636363636364e-05,
      "loss": 0.0008,
      "step": 3329
    },
    {
      "epoch": 45.0,
      "learning_rate": 5.072727272727272e-05,
      "loss": 0.0009,
      "step": 3330
    },
    {
      "epoch": 45.01,
      "learning_rate": 5.059090909090909e-05,
      "loss": 0.0006,
      "step": 3331
    },
    {
      "epoch": 45.03,
      "learning_rate": 5.045454545454545e-05,
      "loss": 0.0006,
      "step": 3332
    },
    {
      "epoch": 45.04,
      "learning_rate": 5.0318181818181816e-05,
      "loss": 0.0005,
      "step": 3333
    },
    {
      "epoch": 45.05,
      "learning_rate": 5.0181818181818174e-05,
      "loss": 0.0004,
      "step": 3334
    },
    {
      "epoch": 45.07,
      "learning_rate": 5.004545454545454e-05,
      "loss": 0.0007,
      "step": 3335
    },
    {
      "epoch": 45.08,
      "learning_rate": 4.99090909090909e-05,
      "loss": 0.0005,
      "step": 3336
    },
    {
      "epoch": 45.09,
      "learning_rate": 4.9772727272727275e-05,
      "loss": 0.0009,
      "step": 3337
    },
    {
      "epoch": 45.11,
      "learning_rate": 4.963636363636363e-05,
      "loss": 0.001,
      "step": 3338
    },
    {
      "epoch": 45.12,
      "learning_rate": 4.95e-05,
      "loss": 0.0004,
      "step": 3339
    },
    {
      "epoch": 45.14,
      "learning_rate": 4.936363636363636e-05,
      "loss": 0.0143,
      "step": 3340
    },
    {
      "epoch": 45.15,
      "learning_rate": 4.9227272727272726e-05,
      "loss": 0.0005,
      "step": 3341
    },
    {
      "epoch": 45.16,
      "learning_rate": 4.9090909090909084e-05,
      "loss": 0.0008,
      "step": 3342
    },
    {
      "epoch": 45.18,
      "learning_rate": 4.895454545454545e-05,
      "loss": 0.0009,
      "step": 3343
    },
    {
      "epoch": 45.19,
      "learning_rate": 4.8818181818181813e-05,
      "loss": 0.0004,
      "step": 3344
    },
    {
      "epoch": 45.2,
      "learning_rate": 4.868181818181817e-05,
      "loss": 0.0009,
      "step": 3345
    },
    {
      "epoch": 45.22,
      "learning_rate": 4.8545454545454536e-05,
      "loss": 0.0004,
      "step": 3346
    },
    {
      "epoch": 45.23,
      "learning_rate": 4.840909090909091e-05,
      "loss": 0.0026,
      "step": 3347
    },
    {
      "epoch": 45.24,
      "learning_rate": 4.827272727272727e-05,
      "loss": 0.0017,
      "step": 3348
    },
    {
      "epoch": 45.26,
      "learning_rate": 4.8136363636363637e-05,
      "loss": 0.0005,
      "step": 3349
    },
    {
      "epoch": 45.27,
      "learning_rate": 4.7999999999999994e-05,
      "loss": 0.0007,
      "step": 3350
    },
    {
      "epoch": 45.28,
      "learning_rate": 4.786363636363636e-05,
      "loss": 0.0005,
      "step": 3351
    },
    {
      "epoch": 45.3,
      "learning_rate": 4.7727272727272724e-05,
      "loss": 0.0015,
      "step": 3352
    },
    {
      "epoch": 45.31,
      "learning_rate": 4.759090909090909e-05,
      "loss": 0.0004,
      "step": 3353
    },
    {
      "epoch": 45.32,
      "learning_rate": 4.7454545454545446e-05,
      "loss": 0.0005,
      "step": 3354
    },
    {
      "epoch": 45.34,
      "learning_rate": 4.731818181818181e-05,
      "loss": 0.0009,
      "step": 3355
    },
    {
      "epoch": 45.35,
      "learning_rate": 4.718181818181818e-05,
      "loss": 0.0007,
      "step": 3356
    },
    {
      "epoch": 45.36,
      "learning_rate": 4.704545454545455e-05,
      "loss": 0.0005,
      "step": 3357
    },
    {
      "epoch": 45.38,
      "learning_rate": 4.6909090909090905e-05,
      "loss": 0.0095,
      "step": 3358
    },
    {
      "epoch": 45.39,
      "learning_rate": 4.677272727272727e-05,
      "loss": 0.0007,
      "step": 3359
    },
    {
      "epoch": 45.41,
      "learning_rate": 4.6636363636363634e-05,
      "loss": 0.0391,
      "step": 3360
    },
    {
      "epoch": 45.42,
      "learning_rate": 4.65e-05,
      "loss": 0.1265,
      "step": 3361
    },
    {
      "epoch": 45.43,
      "learning_rate": 4.6363636363636356e-05,
      "loss": 0.0005,
      "step": 3362
    },
    {
      "epoch": 45.45,
      "learning_rate": 4.622727272727272e-05,
      "loss": 0.0005,
      "step": 3363
    },
    {
      "epoch": 45.46,
      "learning_rate": 4.6090909090909086e-05,
      "loss": 0.0008,
      "step": 3364
    },
    {
      "epoch": 45.47,
      "learning_rate": 4.5954545454545444e-05,
      "loss": 0.0541,
      "step": 3365
    },
    {
      "epoch": 45.49,
      "learning_rate": 4.5818181818181815e-05,
      "loss": 0.0004,
      "step": 3366
    },
    {
      "epoch": 45.5,
      "learning_rate": 4.568181818181818e-05,
      "loss": 0.0045,
      "step": 3367
    },
    {
      "epoch": 45.51,
      "learning_rate": 4.5545454545454544e-05,
      "loss": 0.001,
      "step": 3368
    },
    {
      "epoch": 45.53,
      "learning_rate": 4.540909090909091e-05,
      "loss": 0.011,
      "step": 3369
    },
    {
      "epoch": 45.54,
      "learning_rate": 4.527272727272727e-05,
      "loss": 0.0006,
      "step": 3370
    },
    {
      "epoch": 45.55,
      "learning_rate": 4.513636363636363e-05,
      "loss": 0.1165,
      "step": 3371
    },
    {
      "epoch": 45.57,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 0.0007,
      "step": 3372
    },
    {
      "epoch": 45.58,
      "learning_rate": 4.4863636363636354e-05,
      "loss": 0.0056,
      "step": 3373
    },
    {
      "epoch": 45.59,
      "learning_rate": 4.472727272727272e-05,
      "loss": 0.0033,
      "step": 3374
    },
    {
      "epoch": 45.61,
      "learning_rate": 4.459090909090909e-05,
      "loss": 0.0006,
      "step": 3375
    },
    {
      "epoch": 45.62,
      "learning_rate": 4.4454545454545455e-05,
      "loss": 0.0437,
      "step": 3376
    },
    {
      "epoch": 45.64,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.047,
      "step": 3377
    },
    {
      "epoch": 45.65,
      "learning_rate": 4.418181818181818e-05,
      "loss": 0.0606,
      "step": 3378
    },
    {
      "epoch": 45.66,
      "learning_rate": 4.404545454545454e-05,
      "loss": 0.0005,
      "step": 3379
    },
    {
      "epoch": 45.68,
      "learning_rate": 4.3909090909090906e-05,
      "loss": 0.0005,
      "step": 3380
    },
    {
      "epoch": 45.69,
      "learning_rate": 4.377272727272727e-05,
      "loss": 0.0005,
      "step": 3381
    },
    {
      "epoch": 45.7,
      "learning_rate": 4.363636363636363e-05,
      "loss": 0.0013,
      "step": 3382
    },
    {
      "epoch": 45.72,
      "learning_rate": 4.3499999999999993e-05,
      "loss": 0.0315,
      "step": 3383
    },
    {
      "epoch": 45.73,
      "learning_rate": 4.336363636363636e-05,
      "loss": 0.0006,
      "step": 3384
    },
    {
      "epoch": 45.74,
      "learning_rate": 4.322727272727273e-05,
      "loss": 0.0006,
      "step": 3385
    },
    {
      "epoch": 45.76,
      "learning_rate": 4.309090909090909e-05,
      "loss": 0.0018,
      "step": 3386
    },
    {
      "epoch": 45.77,
      "learning_rate": 4.295454545454545e-05,
      "loss": 0.0005,
      "step": 3387
    },
    {
      "epoch": 45.78,
      "learning_rate": 4.281818181818182e-05,
      "loss": 0.0008,
      "step": 3388
    },
    {
      "epoch": 45.8,
      "learning_rate": 4.268181818181818e-05,
      "loss": 0.0408,
      "step": 3389
    },
    {
      "epoch": 45.81,
      "learning_rate": 4.254545454545454e-05,
      "loss": 0.0129,
      "step": 3390
    },
    {
      "epoch": 45.82,
      "learning_rate": 4.2409090909090904e-05,
      "loss": 0.0831,
      "step": 3391
    },
    {
      "epoch": 45.84,
      "learning_rate": 4.227272727272727e-05,
      "loss": 0.0015,
      "step": 3392
    },
    {
      "epoch": 45.85,
      "learning_rate": 4.2136363636363626e-05,
      "loss": 0.0626,
      "step": 3393
    },
    {
      "epoch": 45.86,
      "learning_rate": 4.2e-05,
      "loss": 0.001,
      "step": 3394
    },
    {
      "epoch": 45.88,
      "learning_rate": 4.186363636363636e-05,
      "loss": 0.0005,
      "step": 3395
    },
    {
      "epoch": 45.89,
      "learning_rate": 4.172727272727273e-05,
      "loss": 0.0009,
      "step": 3396
    },
    {
      "epoch": 45.91,
      "learning_rate": 4.159090909090909e-05,
      "loss": 0.0022,
      "step": 3397
    },
    {
      "epoch": 45.92,
      "learning_rate": 4.145454545454545e-05,
      "loss": 0.015,
      "step": 3398
    },
    {
      "epoch": 45.93,
      "learning_rate": 4.1318181818181814e-05,
      "loss": 0.0007,
      "step": 3399
    },
    {
      "epoch": 45.95,
      "learning_rate": 4.118181818181818e-05,
      "loss": 0.046,
      "step": 3400
    },
    {
      "epoch": 45.95,
      "eval_accuracy": 0.9451219512195121,
      "eval_f1": 0.9456602760788949,
      "eval_loss": 0.29142892360687256,
      "eval_runtime": 126.8362,
      "eval_samples_per_second": 12.93,
      "eval_steps_per_second": 0.812,
      "step": 3400
    },
    {
      "epoch": 45.96,
      "learning_rate": 4.1045454545454537e-05,
      "loss": 0.0007,
      "step": 3401
    },
    {
      "epoch": 45.97,
      "learning_rate": 4.09090909090909e-05,
      "loss": 0.0329,
      "step": 3402
    },
    {
      "epoch": 45.99,
      "learning_rate": 4.0772727272727266e-05,
      "loss": 0.0958,
      "step": 3403
    },
    {
      "epoch": 46.0,
      "learning_rate": 4.063636363636364e-05,
      "loss": 0.0296,
      "step": 3404
    },
    {
      "epoch": 46.01,
      "learning_rate": 4.05e-05,
      "loss": 0.001,
      "step": 3405
    },
    {
      "epoch": 46.03,
      "learning_rate": 4.036363636363636e-05,
      "loss": 0.0023,
      "step": 3406
    },
    {
      "epoch": 46.04,
      "learning_rate": 4.0227272727272724e-05,
      "loss": 0.0089,
      "step": 3407
    },
    {
      "epoch": 46.05,
      "learning_rate": 4.009090909090909e-05,
      "loss": 0.0005,
      "step": 3408
    },
    {
      "epoch": 46.07,
      "learning_rate": 3.995454545454545e-05,
      "loss": 0.001,
      "step": 3409
    },
    {
      "epoch": 46.08,
      "learning_rate": 3.981818181818181e-05,
      "loss": 0.0068,
      "step": 3410
    },
    {
      "epoch": 46.09,
      "learning_rate": 3.9681818181818176e-05,
      "loss": 0.0394,
      "step": 3411
    },
    {
      "epoch": 46.11,
      "learning_rate": 3.954545454545454e-05,
      "loss": 0.0201,
      "step": 3412
    },
    {
      "epoch": 46.12,
      "learning_rate": 3.940909090909091e-05,
      "loss": 0.0015,
      "step": 3413
    },
    {
      "epoch": 46.14,
      "learning_rate": 3.927272727272727e-05,
      "loss": 0.0005,
      "step": 3414
    },
    {
      "epoch": 46.15,
      "learning_rate": 3.9136363636363635e-05,
      "loss": 0.0006,
      "step": 3415
    },
    {
      "epoch": 46.16,
      "learning_rate": 3.9e-05,
      "loss": 0.0018,
      "step": 3416
    },
    {
      "epoch": 46.18,
      "learning_rate": 3.8863636363636364e-05,
      "loss": 0.0016,
      "step": 3417
    },
    {
      "epoch": 46.19,
      "learning_rate": 3.872727272727272e-05,
      "loss": 0.0239,
      "step": 3418
    },
    {
      "epoch": 46.2,
      "learning_rate": 3.8590909090909086e-05,
      "loss": 0.0013,
      "step": 3419
    },
    {
      "epoch": 46.22,
      "learning_rate": 3.845454545454545e-05,
      "loss": 0.0009,
      "step": 3420
    },
    {
      "epoch": 46.23,
      "learning_rate": 3.831818181818181e-05,
      "loss": 0.0005,
      "step": 3421
    },
    {
      "epoch": 46.24,
      "learning_rate": 3.8181818181818174e-05,
      "loss": 0.0026,
      "step": 3422
    },
    {
      "epoch": 46.26,
      "learning_rate": 3.8045454545454545e-05,
      "loss": 0.0012,
      "step": 3423
    },
    {
      "epoch": 46.27,
      "learning_rate": 3.790909090909091e-05,
      "loss": 0.0006,
      "step": 3424
    },
    {
      "epoch": 46.28,
      "learning_rate": 3.7772727272727274e-05,
      "loss": 0.0006,
      "step": 3425
    },
    {
      "epoch": 46.3,
      "learning_rate": 3.763636363636363e-05,
      "loss": 0.0006,
      "step": 3426
    },
    {
      "epoch": 46.31,
      "learning_rate": 3.75e-05,
      "loss": 0.0005,
      "step": 3427
    },
    {
      "epoch": 46.32,
      "learning_rate": 3.736363636363636e-05,
      "loss": 0.0009,
      "step": 3428
    },
    {
      "epoch": 46.34,
      "learning_rate": 3.722727272727272e-05,
      "loss": 0.0028,
      "step": 3429
    },
    {
      "epoch": 46.35,
      "learning_rate": 3.709090909090909e-05,
      "loss": 0.0199,
      "step": 3430
    },
    {
      "epoch": 46.36,
      "learning_rate": 3.6954545454545455e-05,
      "loss": 0.0914,
      "step": 3431
    },
    {
      "epoch": 46.38,
      "learning_rate": 3.681818181818181e-05,
      "loss": 0.0005,
      "step": 3432
    },
    {
      "epoch": 46.39,
      "learning_rate": 3.668181818181818e-05,
      "loss": 0.0006,
      "step": 3433
    },
    {
      "epoch": 46.41,
      "learning_rate": 3.654545454545454e-05,
      "loss": 0.0018,
      "step": 3434
    },
    {
      "epoch": 46.42,
      "learning_rate": 3.640909090909091e-05,
      "loss": 0.0417,
      "step": 3435
    },
    {
      "epoch": 46.43,
      "learning_rate": 3.627272727272727e-05,
      "loss": 0.0135,
      "step": 3436
    },
    {
      "epoch": 46.45,
      "learning_rate": 3.613636363636363e-05,
      "loss": 0.0006,
      "step": 3437
    },
    {
      "epoch": 46.46,
      "learning_rate": 3.5999999999999994e-05,
      "loss": 0.0206,
      "step": 3438
    },
    {
      "epoch": 46.47,
      "learning_rate": 3.5863636363636366e-05,
      "loss": 0.002,
      "step": 3439
    },
    {
      "epoch": 46.49,
      "learning_rate": 3.5727272727272723e-05,
      "loss": 0.0551,
      "step": 3440
    },
    {
      "epoch": 46.5,
      "learning_rate": 3.559090909090909e-05,
      "loss": 0.0008,
      "step": 3441
    },
    {
      "epoch": 46.51,
      "learning_rate": 3.545454545454545e-05,
      "loss": 0.0033,
      "step": 3442
    },
    {
      "epoch": 46.53,
      "learning_rate": 3.531818181818181e-05,
      "loss": 0.0016,
      "step": 3443
    },
    {
      "epoch": 46.54,
      "learning_rate": 3.518181818181818e-05,
      "loss": 0.0006,
      "step": 3444
    },
    {
      "epoch": 46.55,
      "learning_rate": 3.5045454545454547e-05,
      "loss": 0.0005,
      "step": 3445
    },
    {
      "epoch": 46.57,
      "learning_rate": 3.4909090909090904e-05,
      "loss": 0.0005,
      "step": 3446
    },
    {
      "epoch": 46.58,
      "learning_rate": 3.477272727272727e-05,
      "loss": 0.0004,
      "step": 3447
    },
    {
      "epoch": 46.59,
      "learning_rate": 3.4636363636363634e-05,
      "loss": 0.008,
      "step": 3448
    },
    {
      "epoch": 46.61,
      "learning_rate": 3.45e-05,
      "loss": 0.0011,
      "step": 3449
    },
    {
      "epoch": 46.62,
      "learning_rate": 3.436363636363636e-05,
      "loss": 0.0441,
      "step": 3450
    },
    {
      "epoch": 46.64,
      "learning_rate": 3.422727272727272e-05,
      "loss": 0.0006,
      "step": 3451
    },
    {
      "epoch": 46.65,
      "learning_rate": 3.4090909090909085e-05,
      "loss": 0.0004,
      "step": 3452
    },
    {
      "epoch": 46.66,
      "learning_rate": 3.395454545454545e-05,
      "loss": 0.0046,
      "step": 3453
    },
    {
      "epoch": 46.68,
      "learning_rate": 3.3818181818181815e-05,
      "loss": 0.0007,
      "step": 3454
    },
    {
      "epoch": 46.69,
      "learning_rate": 3.368181818181818e-05,
      "loss": 0.0004,
      "step": 3455
    },
    {
      "epoch": 46.7,
      "learning_rate": 3.3545454545454544e-05,
      "loss": 0.0027,
      "step": 3456
    },
    {
      "epoch": 46.72,
      "learning_rate": 3.34090909090909e-05,
      "loss": 0.0054,
      "step": 3457
    },
    {
      "epoch": 46.73,
      "learning_rate": 3.327272727272727e-05,
      "loss": 0.0008,
      "step": 3458
    },
    {
      "epoch": 46.74,
      "learning_rate": 3.313636363636363e-05,
      "loss": 0.0004,
      "step": 3459
    },
    {
      "epoch": 46.76,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 0.0008,
      "step": 3460
    },
    {
      "epoch": 46.77,
      "learning_rate": 3.286363636363636e-05,
      "loss": 0.0005,
      "step": 3461
    },
    {
      "epoch": 46.78,
      "learning_rate": 3.2727272727272725e-05,
      "loss": 0.001,
      "step": 3462
    },
    {
      "epoch": 46.8,
      "learning_rate": 3.259090909090909e-05,
      "loss": 0.0011,
      "step": 3463
    },
    {
      "epoch": 46.81,
      "learning_rate": 3.2454545454545454e-05,
      "loss": 0.0019,
      "step": 3464
    },
    {
      "epoch": 46.82,
      "learning_rate": 3.231818181818181e-05,
      "loss": 0.0843,
      "step": 3465
    },
    {
      "epoch": 46.84,
      "learning_rate": 3.218181818181818e-05,
      "loss": 0.0007,
      "step": 3466
    },
    {
      "epoch": 46.85,
      "learning_rate": 3.204545454545454e-05,
      "loss": 0.005,
      "step": 3467
    },
    {
      "epoch": 46.86,
      "learning_rate": 3.1909090909090906e-05,
      "loss": 0.0006,
      "step": 3468
    },
    {
      "epoch": 46.88,
      "learning_rate": 3.177272727272727e-05,
      "loss": 0.0013,
      "step": 3469
    },
    {
      "epoch": 46.89,
      "learning_rate": 3.1636363636363635e-05,
      "loss": 0.0108,
      "step": 3470
    },
    {
      "epoch": 46.91,
      "learning_rate": 3.149999999999999e-05,
      "loss": 0.0072,
      "step": 3471
    },
    {
      "epoch": 46.92,
      "learning_rate": 3.136363636363636e-05,
      "loss": 0.0005,
      "step": 3472
    },
    {
      "epoch": 46.93,
      "learning_rate": 3.122727272727272e-05,
      "loss": 0.0016,
      "step": 3473
    },
    {
      "epoch": 46.95,
      "learning_rate": 3.109090909090909e-05,
      "loss": 0.0005,
      "step": 3474
    },
    {
      "epoch": 46.96,
      "learning_rate": 3.095454545454545e-05,
      "loss": 0.0005,
      "step": 3475
    },
    {
      "epoch": 46.97,
      "learning_rate": 3.0818181818181816e-05,
      "loss": 0.0611,
      "step": 3476
    },
    {
      "epoch": 46.99,
      "learning_rate": 3.068181818181818e-05,
      "loss": 0.0007,
      "step": 3477
    },
    {
      "epoch": 47.0,
      "learning_rate": 3.0545454545454546e-05,
      "loss": 0.0003,
      "step": 3478
    },
    {
      "epoch": 47.01,
      "learning_rate": 3.0409090909090907e-05,
      "loss": 0.0008,
      "step": 3479
    },
    {
      "epoch": 47.03,
      "learning_rate": 3.0272727272727268e-05,
      "loss": 0.0005,
      "step": 3480
    },
    {
      "epoch": 47.04,
      "learning_rate": 3.0136363636363633e-05,
      "loss": 0.002,
      "step": 3481
    },
    {
      "epoch": 47.05,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 0.0009,
      "step": 3482
    },
    {
      "epoch": 47.07,
      "learning_rate": 2.9863636363636362e-05,
      "loss": 0.0026,
      "step": 3483
    },
    {
      "epoch": 47.08,
      "learning_rate": 2.9727272727272723e-05,
      "loss": 0.0009,
      "step": 3484
    },
    {
      "epoch": 47.09,
      "learning_rate": 2.9590909090909088e-05,
      "loss": 0.0005,
      "step": 3485
    },
    {
      "epoch": 47.11,
      "learning_rate": 2.945454545454545e-05,
      "loss": 0.0057,
      "step": 3486
    },
    {
      "epoch": 47.12,
      "learning_rate": 2.9318181818181817e-05,
      "loss": 0.0005,
      "step": 3487
    },
    {
      "epoch": 47.14,
      "learning_rate": 2.918181818181818e-05,
      "loss": 0.0139,
      "step": 3488
    },
    {
      "epoch": 47.15,
      "learning_rate": 2.9045454545454543e-05,
      "loss": 0.0004,
      "step": 3489
    },
    {
      "epoch": 47.16,
      "learning_rate": 2.8909090909090904e-05,
      "loss": 0.0013,
      "step": 3490
    },
    {
      "epoch": 47.18,
      "learning_rate": 2.877272727272727e-05,
      "loss": 0.0457,
      "step": 3491
    },
    {
      "epoch": 47.19,
      "learning_rate": 2.8636363636363634e-05,
      "loss": 0.0005,
      "step": 3492
    },
    {
      "epoch": 47.2,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.0004,
      "step": 3493
    },
    {
      "epoch": 47.22,
      "learning_rate": 2.836363636363636e-05,
      "loss": 0.0927,
      "step": 3494
    },
    {
      "epoch": 47.23,
      "learning_rate": 2.8227272727272724e-05,
      "loss": 0.0006,
      "step": 3495
    },
    {
      "epoch": 47.24,
      "learning_rate": 2.809090909090909e-05,
      "loss": 0.0004,
      "step": 3496
    },
    {
      "epoch": 47.26,
      "learning_rate": 2.7954545454545453e-05,
      "loss": 0.0006,
      "step": 3497
    },
    {
      "epoch": 47.27,
      "learning_rate": 2.7818181818181815e-05,
      "loss": 0.0031,
      "step": 3498
    },
    {
      "epoch": 47.28,
      "learning_rate": 2.768181818181818e-05,
      "loss": 0.0008,
      "step": 3499
    },
    {
      "epoch": 47.3,
      "learning_rate": 2.754545454545454e-05,
      "loss": 0.0007,
      "step": 3500
    },
    {
      "epoch": 47.31,
      "learning_rate": 2.740909090909091e-05,
      "loss": 0.0005,
      "step": 3501
    },
    {
      "epoch": 47.32,
      "learning_rate": 2.727272727272727e-05,
      "loss": 0.0011,
      "step": 3502
    },
    {
      "epoch": 47.34,
      "learning_rate": 2.7136363636363634e-05,
      "loss": 0.0041,
      "step": 3503
    },
    {
      "epoch": 47.35,
      "learning_rate": 2.6999999999999996e-05,
      "loss": 0.0006,
      "step": 3504
    },
    {
      "epoch": 47.36,
      "learning_rate": 2.686363636363636e-05,
      "loss": 0.0004,
      "step": 3505
    },
    {
      "epoch": 47.38,
      "learning_rate": 2.6727272727272725e-05,
      "loss": 0.0166,
      "step": 3506
    },
    {
      "epoch": 47.39,
      "learning_rate": 2.659090909090909e-05,
      "loss": 0.0049,
      "step": 3507
    },
    {
      "epoch": 47.41,
      "learning_rate": 2.645454545454545e-05,
      "loss": 0.0021,
      "step": 3508
    },
    {
      "epoch": 47.42,
      "learning_rate": 2.6318181818181815e-05,
      "loss": 0.0005,
      "step": 3509
    },
    {
      "epoch": 47.43,
      "learning_rate": 2.618181818181818e-05,
      "loss": 0.0005,
      "step": 3510
    },
    {
      "epoch": 47.45,
      "learning_rate": 2.6045454545454545e-05,
      "loss": 0.0007,
      "step": 3511
    },
    {
      "epoch": 47.46,
      "learning_rate": 2.5909090909090906e-05,
      "loss": 0.0005,
      "step": 3512
    },
    {
      "epoch": 47.47,
      "learning_rate": 2.577272727272727e-05,
      "loss": 0.0003,
      "step": 3513
    },
    {
      "epoch": 47.49,
      "learning_rate": 2.5636363636363632e-05,
      "loss": 0.0218,
      "step": 3514
    },
    {
      "epoch": 47.5,
      "learning_rate": 2.55e-05,
      "loss": 0.0009,
      "step": 3515
    },
    {
      "epoch": 47.51,
      "learning_rate": 2.536363636363636e-05,
      "loss": 0.0005,
      "step": 3516
    },
    {
      "epoch": 47.53,
      "learning_rate": 2.5227272727272726e-05,
      "loss": 0.0006,
      "step": 3517
    },
    {
      "epoch": 47.54,
      "learning_rate": 2.5090909090909087e-05,
      "loss": 0.0009,
      "step": 3518
    },
    {
      "epoch": 47.55,
      "learning_rate": 2.495454545454545e-05,
      "loss": 0.0319,
      "step": 3519
    },
    {
      "epoch": 47.57,
      "learning_rate": 2.4818181818181816e-05,
      "loss": 0.0004,
      "step": 3520
    },
    {
      "epoch": 47.58,
      "learning_rate": 2.468181818181818e-05,
      "loss": 0.0008,
      "step": 3521
    },
    {
      "epoch": 47.59,
      "learning_rate": 2.4545454545454542e-05,
      "loss": 0.0032,
      "step": 3522
    },
    {
      "epoch": 47.61,
      "learning_rate": 2.4409090909090907e-05,
      "loss": 0.0107,
      "step": 3523
    },
    {
      "epoch": 47.62,
      "learning_rate": 2.4272727272727268e-05,
      "loss": 0.0016,
      "step": 3524
    },
    {
      "epoch": 47.64,
      "learning_rate": 2.4136363636363636e-05,
      "loss": 0.0004,
      "step": 3525
    },
    {
      "epoch": 47.65,
      "learning_rate": 2.3999999999999997e-05,
      "loss": 0.0004,
      "step": 3526
    },
    {
      "epoch": 47.66,
      "learning_rate": 2.3863636363636362e-05,
      "loss": 0.0006,
      "step": 3527
    },
    {
      "epoch": 47.68,
      "learning_rate": 2.3727272727272723e-05,
      "loss": 0.0004,
      "step": 3528
    },
    {
      "epoch": 47.69,
      "learning_rate": 2.359090909090909e-05,
      "loss": 0.0061,
      "step": 3529
    },
    {
      "epoch": 47.7,
      "learning_rate": 2.3454545454545452e-05,
      "loss": 0.0281,
      "step": 3530
    },
    {
      "epoch": 47.72,
      "learning_rate": 2.3318181818181817e-05,
      "loss": 0.0004,
      "step": 3531
    },
    {
      "epoch": 47.73,
      "learning_rate": 2.3181818181818178e-05,
      "loss": 0.0004,
      "step": 3532
    },
    {
      "epoch": 47.74,
      "learning_rate": 2.3045454545454543e-05,
      "loss": 0.0011,
      "step": 3533
    },
    {
      "epoch": 47.76,
      "learning_rate": 2.2909090909090908e-05,
      "loss": 0.0004,
      "step": 3534
    },
    {
      "epoch": 47.77,
      "learning_rate": 2.2772727272727272e-05,
      "loss": 0.0005,
      "step": 3535
    },
    {
      "epoch": 47.78,
      "learning_rate": 2.2636363636363633e-05,
      "loss": 0.0008,
      "step": 3536
    },
    {
      "epoch": 47.8,
      "learning_rate": 2.2499999999999998e-05,
      "loss": 0.0019,
      "step": 3537
    },
    {
      "epoch": 47.81,
      "learning_rate": 2.236363636363636e-05,
      "loss": 0.0004,
      "step": 3538
    },
    {
      "epoch": 47.82,
      "learning_rate": 2.2227272727272727e-05,
      "loss": 0.0017,
      "step": 3539
    },
    {
      "epoch": 47.84,
      "learning_rate": 2.209090909090909e-05,
      "loss": 0.0376,
      "step": 3540
    },
    {
      "epoch": 47.85,
      "learning_rate": 2.1954545454545453e-05,
      "loss": 0.0004,
      "step": 3541
    },
    {
      "epoch": 47.86,
      "learning_rate": 2.1818181818181814e-05,
      "loss": 0.0155,
      "step": 3542
    },
    {
      "epoch": 47.88,
      "learning_rate": 2.168181818181818e-05,
      "loss": 0.0005,
      "step": 3543
    },
    {
      "epoch": 47.89,
      "learning_rate": 2.1545454545454544e-05,
      "loss": 0.0004,
      "step": 3544
    },
    {
      "epoch": 47.91,
      "learning_rate": 2.140909090909091e-05,
      "loss": 0.0005,
      "step": 3545
    },
    {
      "epoch": 47.92,
      "learning_rate": 2.127272727272727e-05,
      "loss": 0.0019,
      "step": 3546
    },
    {
      "epoch": 47.93,
      "learning_rate": 2.1136363636363634e-05,
      "loss": 0.0005,
      "step": 3547
    },
    {
      "epoch": 47.95,
      "learning_rate": 2.1e-05,
      "loss": 0.0004,
      "step": 3548
    },
    {
      "epoch": 47.96,
      "learning_rate": 2.0863636363636363e-05,
      "loss": 0.0004,
      "step": 3549
    },
    {
      "epoch": 47.97,
      "learning_rate": 2.0727272727272725e-05,
      "loss": 0.0069,
      "step": 3550
    },
    {
      "epoch": 47.99,
      "learning_rate": 2.059090909090909e-05,
      "loss": 0.0007,
      "step": 3551
    },
    {
      "epoch": 48.0,
      "learning_rate": 2.045454545454545e-05,
      "loss": 0.0007,
      "step": 3552
    },
    {
      "epoch": 48.01,
      "learning_rate": 2.031818181818182e-05,
      "loss": 0.0004,
      "step": 3553
    },
    {
      "epoch": 48.03,
      "learning_rate": 2.018181818181818e-05,
      "loss": 0.0007,
      "step": 3554
    },
    {
      "epoch": 48.04,
      "learning_rate": 2.0045454545454544e-05,
      "loss": 0.0011,
      "step": 3555
    },
    {
      "epoch": 48.05,
      "learning_rate": 1.9909090909090906e-05,
      "loss": 0.0006,
      "step": 3556
    },
    {
      "epoch": 48.07,
      "learning_rate": 1.977272727272727e-05,
      "loss": 0.0004,
      "step": 3557
    },
    {
      "epoch": 48.08,
      "learning_rate": 1.9636363636363635e-05,
      "loss": 0.0004,
      "step": 3558
    },
    {
      "epoch": 48.09,
      "learning_rate": 1.95e-05,
      "loss": 0.0235,
      "step": 3559
    },
    {
      "epoch": 48.11,
      "learning_rate": 1.936363636363636e-05,
      "loss": 0.0008,
      "step": 3560
    },
    {
      "epoch": 48.12,
      "learning_rate": 1.9227272727272726e-05,
      "loss": 0.032,
      "step": 3561
    },
    {
      "epoch": 48.14,
      "learning_rate": 1.9090909090909087e-05,
      "loss": 0.0006,
      "step": 3562
    },
    {
      "epoch": 48.15,
      "learning_rate": 1.8954545454545455e-05,
      "loss": 0.0663,
      "step": 3563
    },
    {
      "epoch": 48.16,
      "learning_rate": 1.8818181818181816e-05,
      "loss": 0.0031,
      "step": 3564
    },
    {
      "epoch": 48.18,
      "learning_rate": 1.868181818181818e-05,
      "loss": 0.0108,
      "step": 3565
    },
    {
      "epoch": 48.19,
      "learning_rate": 1.8545454545454545e-05,
      "loss": 0.0005,
      "step": 3566
    },
    {
      "epoch": 48.2,
      "learning_rate": 1.8409090909090907e-05,
      "loss": 0.0378,
      "step": 3567
    },
    {
      "epoch": 48.22,
      "learning_rate": 1.827272727272727e-05,
      "loss": 0.0141,
      "step": 3568
    },
    {
      "epoch": 48.23,
      "learning_rate": 1.8136363636363636e-05,
      "loss": 0.001,
      "step": 3569
    },
    {
      "epoch": 48.24,
      "learning_rate": 1.7999999999999997e-05,
      "loss": 0.0004,
      "step": 3570
    },
    {
      "epoch": 48.26,
      "learning_rate": 1.7863636363636362e-05,
      "loss": 0.0005,
      "step": 3571
    },
    {
      "epoch": 48.27,
      "learning_rate": 1.7727272727272726e-05,
      "loss": 0.0008,
      "step": 3572
    },
    {
      "epoch": 48.28,
      "learning_rate": 1.759090909090909e-05,
      "loss": 0.0004,
      "step": 3573
    },
    {
      "epoch": 48.3,
      "learning_rate": 1.7454545454545452e-05,
      "loss": 0.0004,
      "step": 3574
    },
    {
      "epoch": 48.31,
      "learning_rate": 1.7318181818181817e-05,
      "loss": 0.0004,
      "step": 3575
    },
    {
      "epoch": 48.32,
      "learning_rate": 1.718181818181818e-05,
      "loss": 0.0008,
      "step": 3576
    },
    {
      "epoch": 48.34,
      "learning_rate": 1.7045454545454543e-05,
      "loss": 0.0007,
      "step": 3577
    },
    {
      "epoch": 48.35,
      "learning_rate": 1.6909090909090907e-05,
      "loss": 0.0004,
      "step": 3578
    },
    {
      "epoch": 48.36,
      "learning_rate": 1.6772727272727272e-05,
      "loss": 0.0004,
      "step": 3579
    },
    {
      "epoch": 48.38,
      "learning_rate": 1.6636363636363637e-05,
      "loss": 0.0013,
      "step": 3580
    },
    {
      "epoch": 48.39,
      "learning_rate": 1.6499999999999998e-05,
      "loss": 0.0354,
      "step": 3581
    },
    {
      "epoch": 48.41,
      "learning_rate": 1.6363636363636363e-05,
      "loss": 0.0004,
      "step": 3582
    },
    {
      "epoch": 48.42,
      "learning_rate": 1.6227272727272727e-05,
      "loss": 0.0005,
      "step": 3583
    },
    {
      "epoch": 48.43,
      "learning_rate": 1.609090909090909e-05,
      "loss": 0.0023,
      "step": 3584
    },
    {
      "epoch": 48.45,
      "learning_rate": 1.5954545454545453e-05,
      "loss": 0.0004,
      "step": 3585
    },
    {
      "epoch": 48.46,
      "learning_rate": 1.5818181818181818e-05,
      "loss": 0.0005,
      "step": 3586
    },
    {
      "epoch": 48.47,
      "learning_rate": 1.568181818181818e-05,
      "loss": 0.0004,
      "step": 3587
    },
    {
      "epoch": 48.49,
      "learning_rate": 1.5545454545454544e-05,
      "loss": 0.0027,
      "step": 3588
    },
    {
      "epoch": 48.5,
      "learning_rate": 1.5409090909090908e-05,
      "loss": 0.0016,
      "step": 3589
    },
    {
      "epoch": 48.51,
      "learning_rate": 1.5272727272727273e-05,
      "loss": 0.0005,
      "step": 3590
    },
    {
      "epoch": 48.53,
      "learning_rate": 1.5136363636363634e-05,
      "loss": 0.0004,
      "step": 3591
    },
    {
      "epoch": 48.54,
      "learning_rate": 1.4999999999999999e-05,
      "loss": 0.0017,
      "step": 3592
    },
    {
      "epoch": 48.55,
      "learning_rate": 1.4863636363636362e-05,
      "loss": 0.0004,
      "step": 3593
    },
    {
      "epoch": 48.57,
      "learning_rate": 1.4727272727272725e-05,
      "loss": 0.0005,
      "step": 3594
    },
    {
      "epoch": 48.58,
      "learning_rate": 1.459090909090909e-05,
      "loss": 0.0014,
      "step": 3595
    },
    {
      "epoch": 48.59,
      "learning_rate": 1.4454545454545452e-05,
      "loss": 0.0004,
      "step": 3596
    },
    {
      "epoch": 48.61,
      "learning_rate": 1.4318181818181817e-05,
      "loss": 0.0005,
      "step": 3597
    },
    {
      "epoch": 48.62,
      "learning_rate": 1.418181818181818e-05,
      "loss": 0.0254,
      "step": 3598
    },
    {
      "epoch": 48.64,
      "learning_rate": 1.4045454545454544e-05,
      "loss": 0.0011,
      "step": 3599
    },
    {
      "epoch": 48.65,
      "learning_rate": 1.3909090909090907e-05,
      "loss": 0.0042,
      "step": 3600
    },
    {
      "epoch": 48.65,
      "eval_accuracy": 0.9469512195121951,
      "eval_f1": 0.9474140798771133,
      "eval_loss": 0.28895431756973267,
      "eval_runtime": 126.3882,
      "eval_samples_per_second": 12.976,
      "eval_steps_per_second": 0.815,
      "step": 3600
    },
    {
      "epoch": 48.66,
      "learning_rate": 1.377272727272727e-05,
      "loss": 0.0726,
      "step": 3601
    },
    {
      "epoch": 48.68,
      "learning_rate": 1.3636363636363635e-05,
      "loss": 0.0004,
      "step": 3602
    },
    {
      "epoch": 48.69,
      "learning_rate": 1.3499999999999998e-05,
      "loss": 0.0003,
      "step": 3603
    },
    {
      "epoch": 48.7,
      "learning_rate": 1.3363636363636362e-05,
      "loss": 0.0242,
      "step": 3604
    },
    {
      "epoch": 48.72,
      "learning_rate": 1.3227272727272725e-05,
      "loss": 0.0004,
      "step": 3605
    },
    {
      "epoch": 48.73,
      "learning_rate": 1.309090909090909e-05,
      "loss": 0.0004,
      "step": 3606
    },
    {
      "epoch": 48.74,
      "learning_rate": 1.2954545454545453e-05,
      "loss": 0.0052,
      "step": 3607
    },
    {
      "epoch": 48.76,
      "learning_rate": 1.2818181818181816e-05,
      "loss": 0.0114,
      "step": 3608
    },
    {
      "epoch": 48.77,
      "learning_rate": 1.268181818181818e-05,
      "loss": 0.0004,
      "step": 3609
    },
    {
      "epoch": 48.78,
      "learning_rate": 1.2545454545454543e-05,
      "loss": 0.001,
      "step": 3610
    },
    {
      "epoch": 48.8,
      "learning_rate": 1.2409090909090908e-05,
      "loss": 0.0008,
      "step": 3611
    },
    {
      "epoch": 48.81,
      "learning_rate": 1.2272727272727271e-05,
      "loss": 0.0005,
      "step": 3612
    },
    {
      "epoch": 48.82,
      "learning_rate": 1.2136363636363634e-05,
      "loss": 0.0013,
      "step": 3613
    },
    {
      "epoch": 48.84,
      "learning_rate": 1.1999999999999999e-05,
      "loss": 0.0008,
      "step": 3614
    },
    {
      "epoch": 48.85,
      "learning_rate": 1.1863636363636362e-05,
      "loss": 0.0005,
      "step": 3615
    },
    {
      "epoch": 48.86,
      "learning_rate": 1.1727272727272726e-05,
      "loss": 0.0007,
      "step": 3616
    },
    {
      "epoch": 48.88,
      "learning_rate": 1.1590909090909089e-05,
      "loss": 0.0008,
      "step": 3617
    },
    {
      "epoch": 48.89,
      "learning_rate": 1.1454545454545454e-05,
      "loss": 0.0004,
      "step": 3618
    },
    {
      "epoch": 48.91,
      "learning_rate": 1.1318181818181817e-05,
      "loss": 0.0004,
      "step": 3619
    },
    {
      "epoch": 48.92,
      "learning_rate": 1.118181818181818e-05,
      "loss": 0.0006,
      "step": 3620
    },
    {
      "epoch": 48.93,
      "learning_rate": 1.1045454545454544e-05,
      "loss": 0.0004,
      "step": 3621
    },
    {
      "epoch": 48.95,
      "learning_rate": 1.0909090909090907e-05,
      "loss": 0.0014,
      "step": 3622
    },
    {
      "epoch": 48.96,
      "learning_rate": 1.0772727272727272e-05,
      "loss": 0.0133,
      "step": 3623
    },
    {
      "epoch": 48.97,
      "learning_rate": 1.0636363636363635e-05,
      "loss": 0.0899,
      "step": 3624
    },
    {
      "epoch": 48.99,
      "learning_rate": 1.05e-05,
      "loss": 0.0004,
      "step": 3625
    },
    {
      "epoch": 49.0,
      "learning_rate": 1.0363636363636362e-05,
      "loss": 0.0002,
      "step": 3626
    },
    {
      "epoch": 49.01,
      "learning_rate": 1.0227272727272725e-05,
      "loss": 0.054,
      "step": 3627
    },
    {
      "epoch": 49.03,
      "learning_rate": 1.009090909090909e-05,
      "loss": 0.0005,
      "step": 3628
    },
    {
      "epoch": 49.04,
      "learning_rate": 9.954545454545453e-06,
      "loss": 0.0927,
      "step": 3629
    },
    {
      "epoch": 49.05,
      "learning_rate": 9.818181818181818e-06,
      "loss": 0.0031,
      "step": 3630
    },
    {
      "epoch": 49.07,
      "learning_rate": 9.68181818181818e-06,
      "loss": 0.0005,
      "step": 3631
    },
    {
      "epoch": 49.08,
      "learning_rate": 9.545454545454543e-06,
      "loss": 0.0006,
      "step": 3632
    },
    {
      "epoch": 49.09,
      "learning_rate": 9.409090909090908e-06,
      "loss": 0.0004,
      "step": 3633
    },
    {
      "epoch": 49.11,
      "learning_rate": 9.272727272727273e-06,
      "loss": 0.0474,
      "step": 3634
    },
    {
      "epoch": 49.12,
      "learning_rate": 9.136363636363636e-06,
      "loss": 0.0076,
      "step": 3635
    },
    {
      "epoch": 49.14,
      "learning_rate": 8.999999999999999e-06,
      "loss": 0.0003,
      "step": 3636
    },
    {
      "epoch": 49.15,
      "learning_rate": 8.863636363636363e-06,
      "loss": 0.0011,
      "step": 3637
    },
    {
      "epoch": 49.16,
      "learning_rate": 8.727272727272726e-06,
      "loss": 0.0025,
      "step": 3638
    },
    {
      "epoch": 49.18,
      "learning_rate": 8.59090909090909e-06,
      "loss": 0.1313,
      "step": 3639
    },
    {
      "epoch": 49.19,
      "learning_rate": 8.454545454545454e-06,
      "loss": 0.0907,
      "step": 3640
    },
    {
      "epoch": 49.2,
      "learning_rate": 8.318181818181818e-06,
      "loss": 0.0095,
      "step": 3641
    },
    {
      "epoch": 49.22,
      "learning_rate": 8.181818181818181e-06,
      "loss": 0.0007,
      "step": 3642
    },
    {
      "epoch": 49.23,
      "learning_rate": 8.045454545454544e-06,
      "loss": 0.054,
      "step": 3643
    },
    {
      "epoch": 49.24,
      "learning_rate": 7.909090909090909e-06,
      "loss": 0.0375,
      "step": 3644
    },
    {
      "epoch": 49.26,
      "learning_rate": 7.772727272727272e-06,
      "loss": 0.0003,
      "step": 3645
    },
    {
      "epoch": 49.27,
      "learning_rate": 7.636363636363636e-06,
      "loss": 0.0004,
      "step": 3646
    },
    {
      "epoch": 49.28,
      "learning_rate": 7.499999999999999e-06,
      "loss": 0.0004,
      "step": 3647
    },
    {
      "epoch": 49.3,
      "learning_rate": 7.363636363636362e-06,
      "loss": 0.008,
      "step": 3648
    },
    {
      "epoch": 49.31,
      "learning_rate": 7.227272727272726e-06,
      "loss": 0.0012,
      "step": 3649
    },
    {
      "epoch": 49.32,
      "learning_rate": 7.09090909090909e-06,
      "loss": 0.0141,
      "step": 3650
    },
    {
      "epoch": 49.34,
      "learning_rate": 6.954545454545454e-06,
      "loss": 0.0009,
      "step": 3651
    },
    {
      "epoch": 49.35,
      "learning_rate": 6.8181818181818174e-06,
      "loss": 0.0004,
      "step": 3652
    },
    {
      "epoch": 49.36,
      "learning_rate": 6.681818181818181e-06,
      "loss": 0.0027,
      "step": 3653
    },
    {
      "epoch": 49.38,
      "learning_rate": 6.545454545454545e-06,
      "loss": 0.0006,
      "step": 3654
    },
    {
      "epoch": 49.39,
      "learning_rate": 6.409090909090908e-06,
      "loss": 0.0014,
      "step": 3655
    },
    {
      "epoch": 49.41,
      "learning_rate": 6.272727272727272e-06,
      "loss": 0.0008,
      "step": 3656
    },
    {
      "epoch": 49.42,
      "learning_rate": 6.1363636363636355e-06,
      "loss": 0.0005,
      "step": 3657
    },
    {
      "epoch": 49.43,
      "learning_rate": 5.999999999999999e-06,
      "loss": 0.0008,
      "step": 3658
    },
    {
      "epoch": 49.45,
      "learning_rate": 5.863636363636363e-06,
      "loss": 0.0163,
      "step": 3659
    },
    {
      "epoch": 49.46,
      "learning_rate": 5.727272727272727e-06,
      "loss": 0.0004,
      "step": 3660
    },
    {
      "epoch": 49.47,
      "learning_rate": 5.59090909090909e-06,
      "loss": 0.0178,
      "step": 3661
    },
    {
      "epoch": 49.49,
      "learning_rate": 5.454545454545454e-06,
      "loss": 0.0003,
      "step": 3662
    },
    {
      "epoch": 49.5,
      "learning_rate": 5.318181818181817e-06,
      "loss": 0.0004,
      "step": 3663
    },
    {
      "epoch": 49.51,
      "learning_rate": 5.181818181818181e-06,
      "loss": 0.0005,
      "step": 3664
    },
    {
      "epoch": 49.53,
      "learning_rate": 5.045454545454545e-06,
      "loss": 0.0743,
      "step": 3665
    },
    {
      "epoch": 49.54,
      "learning_rate": 4.909090909090909e-06,
      "loss": 0.0009,
      "step": 3666
    },
    {
      "epoch": 49.55,
      "learning_rate": 4.772727272727272e-06,
      "loss": 0.0005,
      "step": 3667
    },
    {
      "epoch": 49.57,
      "learning_rate": 4.636363636363636e-06,
      "loss": 0.0004,
      "step": 3668
    },
    {
      "epoch": 49.58,
      "learning_rate": 4.499999999999999e-06,
      "loss": 0.0004,
      "step": 3669
    },
    {
      "epoch": 49.59,
      "learning_rate": 4.363636363636363e-06,
      "loss": 0.0737,
      "step": 3670
    },
    {
      "epoch": 49.61,
      "learning_rate": 4.227272727272727e-06,
      "loss": 0.0004,
      "step": 3671
    },
    {
      "epoch": 49.62,
      "learning_rate": 4.090909090909091e-06,
      "loss": 0.0023,
      "step": 3672
    },
    {
      "epoch": 49.64,
      "learning_rate": 3.954545454545454e-06,
      "loss": 0.0047,
      "step": 3673
    },
    {
      "epoch": 49.65,
      "learning_rate": 3.818181818181818e-06,
      "loss": 0.0007,
      "step": 3674
    },
    {
      "epoch": 49.66,
      "learning_rate": 3.681818181818181e-06,
      "loss": 0.0004,
      "step": 3675
    },
    {
      "epoch": 49.68,
      "learning_rate": 3.545454545454545e-06,
      "loss": 0.0012,
      "step": 3676
    },
    {
      "epoch": 49.69,
      "learning_rate": 3.4090909090909087e-06,
      "loss": 0.001,
      "step": 3677
    },
    {
      "epoch": 49.7,
      "learning_rate": 3.2727272727272725e-06,
      "loss": 0.0077,
      "step": 3678
    },
    {
      "epoch": 49.72,
      "learning_rate": 3.136363636363636e-06,
      "loss": 0.0004,
      "step": 3679
    },
    {
      "epoch": 49.73,
      "learning_rate": 2.9999999999999997e-06,
      "loss": 0.0003,
      "step": 3680
    },
    {
      "epoch": 49.74,
      "learning_rate": 2.8636363636363634e-06,
      "loss": 0.0006,
      "step": 3681
    },
    {
      "epoch": 49.76,
      "learning_rate": 2.727272727272727e-06,
      "loss": 0.0007,
      "step": 3682
    },
    {
      "epoch": 49.77,
      "learning_rate": 2.5909090909090906e-06,
      "loss": 0.0004,
      "step": 3683
    },
    {
      "epoch": 49.78,
      "learning_rate": 2.4545454545454544e-06,
      "loss": 0.0008,
      "step": 3684
    },
    {
      "epoch": 49.8,
      "learning_rate": 2.318181818181818e-06,
      "loss": 0.0004,
      "step": 3685
    },
    {
      "epoch": 49.81,
      "learning_rate": 2.1818181818181815e-06,
      "loss": 0.0011,
      "step": 3686
    },
    {
      "epoch": 49.82,
      "learning_rate": 2.0454545454545453e-06,
      "loss": 0.0012,
      "step": 3687
    },
    {
      "epoch": 49.84,
      "learning_rate": 1.909090909090909e-06,
      "loss": 0.0054,
      "step": 3688
    },
    {
      "epoch": 49.85,
      "learning_rate": 1.7727272727272725e-06,
      "loss": 0.0036,
      "step": 3689
    },
    {
      "epoch": 49.86,
      "learning_rate": 1.6363636363636363e-06,
      "loss": 0.0317,
      "step": 3690
    },
    {
      "epoch": 49.88,
      "learning_rate": 1.4999999999999998e-06,
      "loss": 0.0532,
      "step": 3691
    },
    {
      "epoch": 49.89,
      "learning_rate": 1.3636363636363634e-06,
      "loss": 0.0031,
      "step": 3692
    },
    {
      "epoch": 49.91,
      "learning_rate": 1.2272727272727272e-06,
      "loss": 0.0027,
      "step": 3693
    },
    {
      "epoch": 49.92,
      "learning_rate": 1.0909090909090908e-06,
      "loss": 0.0013,
      "step": 3694
    },
    {
      "epoch": 49.93,
      "learning_rate": 9.545454545454546e-07,
      "loss": 0.0027,
      "step": 3695
    },
    {
      "epoch": 49.95,
      "learning_rate": 8.181818181818181e-07,
      "loss": 0.0141,
      "step": 3696
    },
    {
      "epoch": 49.96,
      "learning_rate": 6.818181818181817e-07,
      "loss": 0.0007,
      "step": 3697
    },
    {
      "epoch": 49.97,
      "learning_rate": 5.454545454545454e-07,
      "loss": 0.0006,
      "step": 3698
    },
    {
      "epoch": 49.99,
      "learning_rate": 4.0909090909090906e-07,
      "loss": 0.0011,
      "step": 3699
    },
    {
      "epoch": 50.0,
      "learning_rate": 2.727272727272727e-07,
      "loss": 0.0004,
      "step": 3700
    },
    {
      "epoch": 50.0,
      "step": 3700,
      "total_flos": 7.505624772241051e+19,
      "train_loss": 0.4516502962981326,
      "train_runtime": 17200.5131,
      "train_samples_per_second": 13.604,
      "train_steps_per_second": 0.215
    }
  ],
  "max_steps": 3700,
  "num_train_epochs": 50,
  "total_flos": 7.505624772241051e+19,
  "trial_name": null,
  "trial_params": null
}