{
  "best_metric": 0.8657823204994202,
  "best_model_checkpoint": "experts/mistralic-expert-25/checkpoint-200",
  "epoch": 0.26205641492265697,
  "eval_steps": 200,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6586,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7382,
      "step": 2
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9771,
      "step": 3
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.239,
      "step": 4
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.1367,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 6
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8723,
      "step": 7
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 8
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9516,
      "step": 9
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7457,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 11
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8585,
      "step": 12
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8002,
      "step": 13
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0357,
      "step": 14
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 15
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9481,
      "step": 16
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0791,
      "step": 17
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9394,
      "step": 18
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8856,
      "step": 19
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7797,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.1097,
      "step": 21
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8909,
      "step": 22
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0102,
      "step": 23
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.4585,
      "step": 24
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8925,
      "step": 25
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9697,
      "step": 26
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0254,
      "step": 27
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0075,
      "step": 28
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 29
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0522,
      "step": 30
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.973,
      "step": 31
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.559,
      "step": 32
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.4682,
      "step": 33
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0415,
      "step": 34
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 35
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8066,
      "step": 36
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5727,
      "step": 37
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0717,
      "step": 38
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 39
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8141,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 41
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0932,
      "step": 42
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.971,
      "step": 43
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9128,
      "step": 44
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8409,
      "step": 45
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0842,
      "step": 46
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8916,
      "step": 47
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7288,
      "step": 48
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.889,
      "step": 49
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9508,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.854,
      "step": 51
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 52
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8621,
      "step": 53
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.4238,
      "step": 54
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.2063,
      "step": 55
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7552,
      "step": 56
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8874,
      "step": 57
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8812,
      "step": 58
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 59
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 60
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1526,
      "step": 61
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5679,
      "step": 62
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.794,
      "step": 63
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9858,
      "step": 64
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 65
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0108,
      "step": 66
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1895,
      "step": 67
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5368,
      "step": 68
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 69
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 70
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9624,
      "step": 71
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.656,
      "step": 72
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7712,
      "step": 73
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8823,
      "step": 74
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6565,
      "step": 75
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0115,
      "step": 76
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.2936,
      "step": 77
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8947,
      "step": 78
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6353,
      "step": 79
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5608,
      "step": 80
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0677,
      "step": 81
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8707,
      "step": 82
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0677,
      "step": 83
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7594,
      "step": 84
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6097,
      "step": 85
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1323,
      "step": 86
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6166,
      "step": 87
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 88
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9475,
      "step": 89
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5055,
      "step": 90
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9368,
      "step": 91
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0028,
      "step": 92
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7944,
      "step": 93
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9518,
      "step": 94
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6992,
      "step": 95
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7948,
      "step": 96
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.5601,
      "step": 97
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 98
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9327,
      "step": 99
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0734,
      "step": 100
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7085,
      "step": 101
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8345,
      "step": 102
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 103
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0132,
      "step": 104
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6714,
      "step": 105
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1224,
      "step": 106
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9094,
      "step": 107
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7111,
      "step": 108
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8389,
      "step": 109
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.994,
      "step": 110
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7135,
      "step": 111
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 112
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8392,
      "step": 113
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 114
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5622,
      "step": 115
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0738,
      "step": 116
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9969,
      "step": 117
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1434,
      "step": 118
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7824,
      "step": 119
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6799,
      "step": 120
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6665,
      "step": 121
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.2051,
      "step": 122
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7486,
      "step": 123
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8522,
      "step": 124
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.07,
      "step": 125
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9415,
      "step": 126
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.622,
      "step": 127
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1949,
      "step": 128
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9893,
      "step": 129
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6507,
      "step": 130
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6626,
      "step": 131
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6973,
      "step": 132
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0162,
      "step": 133
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8232,
      "step": 134
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0537,
      "step": 135
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9964,
      "step": 136
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0375,
      "step": 137
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9369,
      "step": 138
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7409,
      "step": 139
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7771,
      "step": 140
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7179,
      "step": 141
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0834,
      "step": 142
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 143
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0296,
      "step": 144
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9016,
      "step": 145
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 146
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.739,
      "step": 147
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 148
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.3007,
      "step": 149
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.3336,
      "step": 150
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8723,
      "step": 151
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1111,
      "step": 152
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1674,
      "step": 153
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8059,
      "step": 154
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7778,
      "step": 155
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7531,
      "step": 156
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6723,
      "step": 157
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8483,
      "step": 158
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0657,
      "step": 159
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1322,
      "step": 160
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9167,
      "step": 161
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9756,
      "step": 162
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 163
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.135,
      "step": 164
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7567,
      "step": 165
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0317,
      "step": 166
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1242,
      "step": 167
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.977,
      "step": 168
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7498,
      "step": 169
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8805,
      "step": 170
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7917,
      "step": 171
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5351,
      "step": 172
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.869,
      "step": 173
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.2608,
      "step": 174
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8666,
      "step": 175
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 176
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8882,
      "step": 177
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0029,
      "step": 178
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 179
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8234,
      "step": 180
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 181
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8096,
      "step": 182
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8343,
      "step": 183
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 184
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0987,
      "step": 185
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8363,
      "step": 186
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0218,
      "step": 187
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.838,
      "step": 188
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9084,
      "step": 189
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5031,
      "step": 190
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 191
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8532,
      "step": 192
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9322,
      "step": 193
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0271,
      "step": 194
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 195
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8817,
      "step": 196
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8641,
      "step": 197
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6634,
      "step": 198
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8597,
      "step": 199
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 200
    },
    {
      "epoch": 0.03,
      "eval_loss": 0.8657823204994202,
      "eval_runtime": 150.7329,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 3.317,
      "step": 200
    },
    {
      "epoch": 0.03,
      "mmlu_eval_accuracy": 0.5909590703251556,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.3941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.5003863334188885,
      "step": 200
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8749,
      "step": 201
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8911,
      "step": 202
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8477,
      "step": 203
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 204
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 205
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1289,
      "step": 206
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6094,
      "step": 207
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7295,
      "step": 208
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9877,
      "step": 209
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4836,
      "step": 210
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.902,
      "step": 211
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0939,
      "step": 212
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8077,
      "step": 213
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1104,
      "step": 214
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0361,
      "step": 215
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 216
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 217
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.76,
      "step": 218
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9476,
      "step": 219
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 220
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.116,
      "step": 221
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8784,
      "step": 222
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0402,
      "step": 223
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0433,
      "step": 224
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 225
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 226
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8211,
      "step": 227
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9549,
      "step": 228
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1082,
      "step": 229
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 230
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.72,
      "step": 231
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.3027,
      "step": 232
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 233
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 234
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1021,
      "step": 235
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8727,
      "step": 236
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.14,
      "step": 237
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0598,
      "step": 238
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9005,
      "step": 239
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 240
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7344,
      "step": 241
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 242
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5014,
      "step": 243
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0667,
      "step": 244
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 245
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8805,
      "step": 246
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.943,
      "step": 247
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7799,
      "step": 248
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6903,
      "step": 249
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.078,
      "step": 250
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8822,
      "step": 251
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8223,
      "step": 252
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7609,
      "step": 253
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.66,
      "step": 254
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 255
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9438,
      "step": 256
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7693,
      "step": 257
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8977,
      "step": 258
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0147,
      "step": 259
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7311,
      "step": 260
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1428,
      "step": 261
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9863,
      "step": 262
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1905,
      "step": 263
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0178,
      "step": 264
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0601,
      "step": 265
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8216,
      "step": 266
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9855,
      "step": 267
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9618,
      "step": 268
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9469,
      "step": 269
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1452,
      "step": 270
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6944,
      "step": 271
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0098,
      "step": 272
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 273
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6379,
      "step": 274
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0303,
      "step": 275
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0032,
      "step": 276
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0306,
      "step": 277
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8565,
      "step": 278
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.264,
      "step": 279
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8809,
      "step": 280
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9519,
      "step": 281
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 282
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 283
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 284
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6762,
      "step": 285
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6035,
      "step": 286
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9632,
      "step": 287
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.78,
      "step": 288
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0164,
      "step": 289
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1476,
      "step": 290
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8958,
      "step": 291
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1306,
      "step": 292
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1924,
      "step": 293
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5887,
      "step": 294
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7278,
      "step": 295
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1009,
      "step": 296
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9431,
      "step": 297
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0868,
      "step": 298
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.962,
      "step": 299
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.944,
      "step": 300
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1359,
      "step": 301
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1752,
      "step": 302
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.2978,
      "step": 303
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1673,
      "step": 304
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6636,
      "step": 305
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6604,
      "step": 306
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9764,
      "step": 307
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9834,
      "step": 308
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6069,
      "step": 309
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8305,
      "step": 310
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8291,
      "step": 311
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4372,
      "step": 312
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 313
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4996,
      "step": 314
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9616,
      "step": 315
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1558,
      "step": 316
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5072,
      "step": 317
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8821,
      "step": 318
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9128,
      "step": 319
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.134,
      "step": 320
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.842,
      "step": 321
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0481,
      "step": 322
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0161,
      "step": 323
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0036,
      "step": 324
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6041,
      "step": 325
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1767,
      "step": 326
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6586,
      "step": 327
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 328
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 329
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 330
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8548,
      "step": 331
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9349,
      "step": 332
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7302,
      "step": 333
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.295,
      "step": 334
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 335
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5129,
      "step": 336
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.837,
      "step": 337
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0422,
      "step": 338
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9438,
      "step": 339
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 340
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5119,
      "step": 341
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6394,
      "step": 342
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1076,
      "step": 343
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0006,
      "step": 344
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7598,
      "step": 345
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8456,
      "step": 346
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4541,
      "step": 347
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7358,
      "step": 348
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8682,
      "step": 349
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9197,
      "step": 350
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7392,
      "step": 351
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.96,
      "step": 352
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8633,
      "step": 353
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 354
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7632,
      "step": 355
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8844,
      "step": 356
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1651,
      "step": 357
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 358
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9419,
      "step": 359
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9078,
      "step": 360
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5853,
      "step": 361
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.3986,
      "step": 362
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8585,
      "step": 363
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 364
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9836,
      "step": 365
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8018,
      "step": 366
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 367
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.261,
      "step": 368
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9301,
      "step": 369
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7423,
      "step": 370
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4561,
      "step": 371
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8535,
      "step": 372
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7937,
      "step": 373
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0679,
      "step": 374
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6891,
      "step": 375
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9629,
      "step": 376
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5769,
      "step": 377
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.2939,
      "step": 378
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4939,
      "step": 379
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4871,
      "step": 380
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9562,
      "step": 381
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0184,
      "step": 382
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 383
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.91,
      "step": 384
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8464,
      "step": 385
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4652,
      "step": 386
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9297,
      "step": 387
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1218,
      "step": 388
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 389
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 390
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0776,
      "step": 391
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9132,
      "step": 392
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.028,
      "step": 393
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1235,
      "step": 394
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8227,
      "step": 395
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0908,
      "step": 396
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5499,
      "step": 397
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9925,
      "step": 398
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9207,
      "step": 399
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8515,
      "step": 400
    },
    {
      "epoch": 0.06,
      "eval_loss": 0.876522958278656,
      "eval_runtime": 150.8131,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 3.315,
      "step": 400
    },
    {
      "epoch": 0.06,
      "mmlu_eval_accuracy": 0.5887418611751553,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 1.0,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1819022774851975,
      "step": 400
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.3854,
      "step": 401
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9,
      "step": 402
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 403
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7944,
      "step": 404
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 405
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.825,
      "step": 406
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.027,
      "step": 407
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8133,
      "step": 408
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8375,
      "step": 409
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.867,
      "step": 410
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 411
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0908,
      "step": 412
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 413
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0551,
      "step": 414
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7884,
      "step": 415
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.196,
      "step": 416
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9368,
      "step": 417
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7889,
      "step": 418
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.92,
      "step": 419
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.923,
      "step": 420
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.505,
      "step": 421
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6246,
      "step": 422
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5029,
      "step": 423
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6821,
      "step": 424
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0521,
      "step": 425
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.042,
      "step": 426
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1292,
      "step": 427
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6349,
      "step": 428
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1455,
      "step": 429
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1963,
      "step": 430
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.2138,
      "step": 431
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8752,
      "step": 432
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0621,
      "step": 433
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6651,
      "step": 434
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.859,
      "step": 435
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.716,
      "step": 436
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1293,
      "step": 437
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.839,
      "step": 438
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7725,
      "step": 439
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7605,
      "step": 440
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 441
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7387,
      "step": 442
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7862,
      "step": 443
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6836,
      "step": 444
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7421,
      "step": 445
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 446
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8383,
      "step": 447
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9174,
      "step": 448
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0507,
      "step": 449
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 450
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9912,
      "step": 451
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.3324,
      "step": 452
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9734,
      "step": 453
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8301,
      "step": 454
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8535,
      "step": 455
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0383,
      "step": 456
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 457
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9024,
      "step": 458
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9093,
      "step": 459
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0652,
      "step": 460
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8565,
      "step": 461
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1394,
      "step": 462
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8269,
      "step": 463
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8476,
      "step": 464
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9835,
      "step": 465
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1196,
      "step": 466
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1287,
      "step": 467
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.2022,
      "step": 468
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.128,
      "step": 469
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 470
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 471
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 472
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8201,
      "step": 473
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7534,
      "step": 474
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5769,
      "step": 475
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8533,
      "step": 476
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.811,
      "step": 477
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9281,
      "step": 478
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 479
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 480
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9395,
      "step": 481
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1276,
      "step": 482
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7285,
      "step": 483
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.3029,
      "step": 484
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0498,
      "step": 485
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8145,
      "step": 486
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0636,
      "step": 487
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 488
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.3619,
      "step": 489
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0884,
      "step": 490
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4766,
      "step": 491
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1152,
      "step": 492
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 493
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0537,
      "step": 494
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8822,
      "step": 495
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7957,
      "step": 496
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 497
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8098,
      "step": 498
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9354,
      "step": 499
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8849,
      "step": 500
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9698,
      "step": 501
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9888,
      "step": 502
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5669,
      "step": 503
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.3472,
      "step": 504
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0173,
      "step": 505
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 506
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0662,
      "step": 507
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9613,
      "step": 508
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0865,
      "step": 509
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 510
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0889,
      "step": 511
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8027,
      "step": 512
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 513
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 514
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 515
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7503,
      "step": 516
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5418,
      "step": 517
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7422,
      "step": 518
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1012,
      "step": 519
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9694,
      "step": 520
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 521
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8709,
      "step": 522
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0756,
      "step": 523
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1526,
      "step": 524
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7537,
      "step": 525
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0647,
      "step": 526
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0072,
      "step": 527
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8717,
      "step": 528
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7804,
      "step": 529
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0055,
      "step": 530
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5341,
      "step": 531
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8038,
      "step": 532
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.654,
      "step": 533
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8137,
      "step": 534
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 535
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7722,
      "step": 536
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6854,
      "step": 537
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6428,
      "step": 538
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9156,
      "step": 539
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9922,
      "step": 540
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7666,
      "step": 541
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6726,
      "step": 542
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7685,
      "step": 543
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7058,
      "step": 544
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0806,
      "step": 545
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8948,
      "step": 546
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5322,
      "step": 547
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8362,
      "step": 548
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8556,
      "step": 549
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.777,
      "step": 550
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 551
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1689,
      "step": 552
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 553
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.515,
      "step": 554
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1138,
      "step": 555
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0415,
      "step": 556
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9963,
      "step": 557
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9907,
      "step": 558
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8903,
      "step": 559
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9432,
      "step": 560
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 561
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0427,
      "step": 562
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.819,
      "step": 563
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 564
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8071,
      "step": 565
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 566
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9555,
      "step": 567
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9625,
      "step": 568
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8593,
      "step": 569
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1361,
      "step": 570
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8922,
      "step": 571
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.481,
      "step": 572
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.937,
      "step": 573
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6731,
      "step": 574
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.863,
      "step": 575
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.4361,
      "step": 576
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1296,
      "step": 577
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7308,
      "step": 578
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 579
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7462,
      "step": 580
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9368,
      "step": 581
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6612,
      "step": 582
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.3348,
      "step": 583
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7409,
      "step": 584
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8789,
      "step": 585
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8893,
      "step": 586
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.952,
      "step": 587
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 588
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0583,
      "step": 589
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.2175,
      "step": 590
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6296,
      "step": 591
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6771,
      "step": 592
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0436,
      "step": 593
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9851,
      "step": 594
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 595
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1432,
      "step": 596
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 597
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 598
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6763,
      "step": 599
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7246,
      "step": 600
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.8749128580093384,
      "eval_runtime": 150.7177,
      "eval_samples_per_second": 6.635,
      "eval_steps_per_second": 3.317,
      "step": 600
    },
    {
      "epoch": 0.09,
      "mmlu_eval_accuracy": 0.5945356668386175,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.8125,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.300421111344046,
      "step": 600
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7933,
      "step": 601
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 602
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 603
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6045,
      "step": 604
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 605
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7786,
      "step": 606
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 607
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8977,
      "step": 608
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6011,
      "step": 609
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9039,
      "step": 610
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0054,
      "step": 611
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.884,
      "step": 612
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.103,
      "step": 613
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9803,
      "step": 614
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8419,
      "step": 615
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.2294,
      "step": 616
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9067,
      "step": 617
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 618
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7901,
      "step": 619
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9912,
      "step": 620
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1933,
      "step": 621
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0804,
      "step": 622
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1767,
      "step": 623
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1722,
      "step": 624
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0863,
      "step": 625
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5893,
      "step": 626
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1878,
      "step": 627
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 628
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0097,
      "step": 629
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9198,
      "step": 630
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9796,
      "step": 631
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7624,
      "step": 632
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8333,
      "step": 633
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.7832,
      "step": 634
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8904,
      "step": 635
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0267,
      "step": 636
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6716,
      "step": 637
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7147,
      "step": 638
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 639
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0915,
      "step": 640
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7105,
      "step": 641
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 642
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6649,
      "step": 643
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5645,
      "step": 644
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.938,
      "step": 645
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6768,
      "step": 646
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8096,
      "step": 647
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9646,
      "step": 648
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0146,
      "step": 649
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1331,
      "step": 650
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9202,
      "step": 651
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.671,
      "step": 652
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5388,
      "step": 653
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7091,
      "step": 654
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1427,
      "step": 655
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9963,
      "step": 656
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5947,
      "step": 657
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9741,
      "step": 658
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4016,
      "step": 659
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7136,
      "step": 660
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0186,
      "step": 661
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 662
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6063,
      "step": 663
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0557,
      "step": 664
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6535,
      "step": 665
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9654,
      "step": 666
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8981,
      "step": 667
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8701,
      "step": 668
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.888,
      "step": 669
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6904,
      "step": 670
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8483,
      "step": 671
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 672
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0177,
      "step": 673
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 674
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6163,
      "step": 675
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7424,
      "step": 676
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.175,
      "step": 677
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0164,
      "step": 678
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 679
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9453,
      "step": 680
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6732,
      "step": 681
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1588,
      "step": 682
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8268,
      "step": 683
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.2667,
      "step": 684
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0797,
      "step": 685
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 686
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0952,
      "step": 687
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.686,
      "step": 688
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 689
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9482,
      "step": 690
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0309,
      "step": 691
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9224,
      "step": 692
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6838,
      "step": 693
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5536,
      "step": 694
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 695
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.567,
      "step": 696
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1088,
      "step": 697
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8476,
      "step": 698
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1508,
      "step": 699
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 700
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1719,
      "step": 701
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7328,
      "step": 702
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1598,
      "step": 703
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 704
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 705
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9277,
      "step": 706
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.3337,
      "step": 707
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8181,
      "step": 708
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 709
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.2111,
      "step": 710
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9376,
      "step": 711
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8854,
      "step": 712
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.999,
      "step": 713
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9933,
      "step": 714
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.3753,
      "step": 715
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.85,
      "step": 716
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9298,
      "step": 717
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.936,
      "step": 718
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7745,
      "step": 719
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5277,
      "step": 720
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1924,
      "step": 721
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5876,
      "step": 722
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8001,
      "step": 723
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7527,
      "step": 724
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6356,
      "step": 725
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 726
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.649,
      "step": 727
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.219,
      "step": 728
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1549,
      "step": 729
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8438,
      "step": 730
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 731
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8514,
      "step": 732
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0244,
      "step": 733
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6488,
      "step": 734
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9765,
      "step": 735
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0243,
      "step": 736
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1347,
      "step": 737
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4407,
      "step": 738
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8852,
      "step": 739
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6979,
      "step": 740
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0013,
      "step": 741
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7102,
      "step": 742
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5903,
      "step": 743
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 744
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 745
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0039,
      "step": 746
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1458,
      "step": 747
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9443,
      "step": 748
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 749
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1454,
      "step": 750
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0936,
      "step": 751
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8144,
      "step": 752
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 753
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1301,
      "step": 754
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8505,
      "step": 755
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 756
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.839,
      "step": 757
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0665,
      "step": 758
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5422,
      "step": 759
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1388,
      "step": 760
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6734,
      "step": 761
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.2507,
      "step": 762
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7401,
      "step": 763
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9891,
      "step": 764
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8424,
      "step": 765
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.749,
      "step": 766
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9126,
      "step": 767
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9094,
      "step": 768
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9753,
      "step": 769
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9845,
      "step": 770
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0428,
      "step": 771
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.2379,
      "step": 772
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0308,
      "step": 773
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.2342,
      "step": 774
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 775
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1706,
      "step": 776
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9846,
      "step": 777
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.624,
      "step": 778
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 779
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 780
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1015,
      "step": 781
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.67,
      "step": 782
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 783
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7011,
      "step": 784
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7409,
      "step": 785
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0441,
      "step": 786
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1641,
      "step": 787
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7083,
      "step": 788
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0735,
      "step": 789
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7574,
      "step": 790
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1546,
      "step": 791
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8884,
      "step": 792
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1335,
      "step": 793
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0672,
      "step": 794
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6803,
      "step": 795
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9414,
      "step": 796
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7552,
      "step": 797
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6815,
      "step": 798
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 799
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8239,
      "step": 800
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.8725178241729736,
      "eval_runtime": 150.6633,
      "eval_samples_per_second": 6.637,
      "eval_steps_per_second": 3.319,
      "step": 800
    },
    {
      "epoch": 0.12,
      "mmlu_eval_accuracy": 0.5608323161302006,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.7241379310344828,
      "mmlu_eval_accuracy_college_biology": 0.75,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.31,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.3734735648874825,
      "step": 800
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4724,
      "step": 801
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8192,
      "step": 802
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8416,
      "step": 803
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.169,
      "step": 804
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0249,
      "step": 805
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7157,
      "step": 806
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.3436,
      "step": 807
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 808
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1136,
      "step": 809
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6544,
      "step": 810
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7476,
      "step": 811
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8617,
      "step": 812
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9752,
      "step": 813
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.3313,
      "step": 814
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 815
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9164,
      "step": 816
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4844,
      "step": 817
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7642,
      "step": 818
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 819
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8774,
      "step": 820
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6074,
      "step": 821
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0901,
      "step": 822
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 823
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 824
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 825
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 826
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9657,
      "step": 827
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1283,
      "step": 828
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 829
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 830
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 831
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.606,
      "step": 832
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9659,
      "step": 833
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9891,
      "step": 834
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 835
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 836
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7557,
      "step": 837
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9977,
      "step": 838
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1685,
      "step": 839
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0044,
      "step": 840
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 841
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7377,
      "step": 842
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9704,
      "step": 843
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 844
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7844,
      "step": 845
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.038,
      "step": 846
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 847
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8121,
      "step": 848
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.497,
      "step": 849
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 850
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7443,
      "step": 851
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.2137,
      "step": 852
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0378,
      "step": 853
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8415,
      "step": 854
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9782,
      "step": 855
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1036,
      "step": 856
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9824,
      "step": 857
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9456,
      "step": 858
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9204,
      "step": 859
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1695,
      "step": 860
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.078,
      "step": 861
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0776,
      "step": 862
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 863
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4595,
      "step": 864
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0924,
      "step": 865
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8739,
      "step": 866
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9144,
      "step": 867
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.3307,
      "step": 868
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 869
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.2225,
      "step": 870
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 871
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8892,
      "step": 872
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8897,
      "step": 873
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0038,
      "step": 874
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 875
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9416,
      "step": 876
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8887,
      "step": 877
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9856,
      "step": 878
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9444,
      "step": 879
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 880
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8736,
      "step": 881
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 882
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6133,
      "step": 883
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4861,
      "step": 884
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 885
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1137,
      "step": 886
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 887
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0886,
      "step": 888
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8159,
      "step": 889
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 890
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7435,
      "step": 891
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0296,
      "step": 892
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7441,
      "step": 893
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 894
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7911,
      "step": 895
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 896
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6761,
      "step": 897
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 898
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.713,
      "step": 899
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 900
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5847,
      "step": 901
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8759,
      "step": 902
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9002,
      "step": 903
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1938,
      "step": 904
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9768,
      "step": 905
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 906
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0816,
      "step": 907
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.001,
      "step": 908
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.3129,
      "step": 909
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8504,
      "step": 910
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0414,
      "step": 911
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1041,
      "step": 912
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 913
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4841,
      "step": 914
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 915
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0808,
      "step": 916
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0586,
      "step": 917
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4607,
      "step": 918
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.71,
      "step": 919
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7603,
      "step": 920
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.283,
      "step": 921
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8549,
      "step": 922
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.2537,
      "step": 923
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9673,
      "step": 924
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5941,
      "step": 925
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5671,
      "step": 926
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1963,
      "step": 927
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 928
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9514,
      "step": 929
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7437,
      "step": 930
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9262,
      "step": 931
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 932
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 933
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7905,
      "step": 934
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9442,
      "step": 935
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 936
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9452,
      "step": 937
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0586,
      "step": 938
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6365,
      "step": 939
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9588,
      "step": 940
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7316,
      "step": 941
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8662,
      "step": 942
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 943
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8829,
      "step": 944
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0404,
      "step": 945
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6349,
      "step": 946
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0867,
      "step": 947
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.2428,
      "step": 948
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5699,
      "step": 949
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7636,
      "step": 950
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8744,
      "step": 951
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0051,
      "step": 952
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.31,
      "step": 953
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9168,
      "step": 954
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1918,
      "step": 955
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 956
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9934,
      "step": 957
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7472,
      "step": 958
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8533,
      "step": 959
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4424,
      "step": 960
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9007,
      "step": 961
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1104,
      "step": 962
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8961,
      "step": 963
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1111,
      "step": 964
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9645,
      "step": 965
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7034,
      "step": 966
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8797,
      "step": 967
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9625,
      "step": 968
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0512,
      "step": 969
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1515,
      "step": 970
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7157,
      "step": 971
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9643,
      "step": 972
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9941,
      "step": 973
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9834,
      "step": 974
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7637,
      "step": 975
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9422,
      "step": 976
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7724,
      "step": 977
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0439,
      "step": 978
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9078,
      "step": 979
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0477,
      "step": 980
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8008,
      "step": 981
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1732,
      "step": 982
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 983
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.173,
      "step": 984
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 985
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9224,
      "step": 986
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.599,
      "step": 987
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.056,
      "step": 988
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1333,
      "step": 989
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6025,
      "step": 990
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 991
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7187,
      "step": 992
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0451,
      "step": 993
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6228,
      "step": 994
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6876,
      "step": 995
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8525,
      "step": 996
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5607,
      "step": 997
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9217,
      "step": 998
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6697,
      "step": 999
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.993,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "eval_loss": 0.878848135471344,
      "eval_runtime": 150.6107,
      "eval_samples_per_second": 6.64,
      "eval_steps_per_second": 3.32,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "mmlu_eval_accuracy": 0.5734379096548112,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.3074099578294056,
      "step": 1000
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0667,
      "step": 1001
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1677,
      "step": 1002
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0216,
      "step": 1003
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7015,
      "step": 1004
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1563,
      "step": 1005
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 1006
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0494,
      "step": 1007
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9357,
      "step": 1008
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0718,
      "step": 1009
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.884,
      "step": 1010
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8736,
      "step": 1011
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.987,
      "step": 1012
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9063,
      "step": 1013
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7958,
      "step": 1014
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9232,
      "step": 1015
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.801,
      "step": 1016
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7013,
      "step": 1017
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 1018
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 1019
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1493,
      "step": 1020
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0065,
      "step": 1021
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.761,
      "step": 1022
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.656,
      "step": 1023
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8785,
      "step": 1024
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9281,
      "step": 1025
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.2239,
      "step": 1026
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0734,
      "step": 1027
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6647,
      "step": 1028
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0428,
      "step": 1029
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1885,
      "step": 1030
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9653,
      "step": 1031
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9913,
      "step": 1032
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8068,
      "step": 1033
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.2011,
      "step": 1034
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8751,
      "step": 1035
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 1036
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.694,
      "step": 1037
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9489,
      "step": 1038
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0471,
      "step": 1039
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4731,
      "step": 1040
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8806,
      "step": 1041
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7787,
      "step": 1042
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 1043
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4076,
      "step": 1044
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 1045
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0319,
      "step": 1046
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7897,
      "step": 1047
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0875,
      "step": 1048
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0006,
      "step": 1049
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.997,
      "step": 1050
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 1051
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6084,
      "step": 1052
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0805,
      "step": 1053
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.83,
      "step": 1054
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9635,
      "step": 1055
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 1056
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9552,
      "step": 1057
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1284,
      "step": 1058
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.237,
      "step": 1059
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9764,
      "step": 1060
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1781,
      "step": 1061
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6368,
      "step": 1062
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6834,
      "step": 1063
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8429,
      "step": 1064
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6826,
      "step": 1065
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.856,
      "step": 1066
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4964,
      "step": 1067
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7132,
      "step": 1068
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.872,
      "step": 1069
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0755,
      "step": 1070
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 1071
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1262,
      "step": 1072
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9963,
      "step": 1073
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.779,
      "step": 1074
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0618,
      "step": 1075
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9294,
      "step": 1076
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 1077
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0818,
      "step": 1078
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8857,
      "step": 1079
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 1080
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.3785,
      "step": 1081
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 1082
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7046,
      "step": 1083
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.928,
      "step": 1084
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4103,
      "step": 1085
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6407,
      "step": 1086
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0444,
      "step": 1087
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6457,
      "step": 1088
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9871,
      "step": 1089
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1242,
      "step": 1090
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7638,
      "step": 1091
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0511,
      "step": 1092
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5871,
      "step": 1093
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9146,
      "step": 1094
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5565,
      "step": 1095
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0031,
      "step": 1096
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0724,
      "step": 1097
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.94,
      "step": 1098
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6378,
      "step": 1099
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6874,
      "step": 1100
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5745,
      "step": 1101
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 1102
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0592,
      "step": 1103
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8215,
      "step": 1104
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 1105
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.4083,
      "step": 1106
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8962,
      "step": 1107
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5276,
      "step": 1108
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 1109
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0799,
      "step": 1110
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8959,
      "step": 1111
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.2023,
      "step": 1112
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8383,
      "step": 1113
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 1114
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 1115
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0278,
      "step": 1116
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8829,
      "step": 1117
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8967,
      "step": 1118
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8988,
      "step": 1119
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6602,
      "step": 1120
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6501,
      "step": 1121
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 1122
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 1123
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5433,
      "step": 1124
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 1125
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6557,
      "step": 1126
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.025,
      "step": 1127
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8426,
      "step": 1128
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 1129
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1424,
      "step": 1130
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4489,
      "step": 1131
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4552,
      "step": 1132
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6344,
      "step": 1133
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8552,
      "step": 1134
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 1135
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0369,
      "step": 1136
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1725,
      "step": 1137
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5666,
      "step": 1138
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.127,
      "step": 1139
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6571,
      "step": 1140
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0431,
      "step": 1141
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7279,
      "step": 1142
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6063,
      "step": 1143
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7581,
      "step": 1144
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 1145
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0069,
      "step": 1146
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6891,
      "step": 1147
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0225,
      "step": 1148
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.125,
      "step": 1149
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 1150
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8938,
      "step": 1151
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.3053,
      "step": 1152
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.2755,
      "step": 1153
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7661,
      "step": 1154
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0705,
      "step": 1155
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4421,
      "step": 1156
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0128,
      "step": 1157
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9538,
      "step": 1158
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7103,
      "step": 1159
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7642,
      "step": 1160
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8113,
      "step": 1161
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8415,
      "step": 1162
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8951,
      "step": 1163
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0651,
      "step": 1164
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 1165
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9646,
      "step": 1166
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7951,
      "step": 1167
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0334,
      "step": 1168
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.898,
      "step": 1169
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7447,
      "step": 1170
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8513,
      "step": 1171
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.694,
      "step": 1172
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1225,
      "step": 1173
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9221,
      "step": 1174
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9866,
      "step": 1175
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7634,
      "step": 1176
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8042,
      "step": 1177
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6925,
      "step": 1178
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0775,
      "step": 1179
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.2687,
      "step": 1180
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6053,
      "step": 1181
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1032,
      "step": 1182
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8699,
      "step": 1183
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9148,
      "step": 1184
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 1185
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9439,
      "step": 1186
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7358,
      "step": 1187
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 1188
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7556,
      "step": 1189
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9338,
      "step": 1190
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 1191
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8895,
      "step": 1192
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7087,
      "step": 1193
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9065,
      "step": 1194
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9182,
      "step": 1195
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9524,
      "step": 1196
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9308,
      "step": 1197
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0177,
      "step": 1198
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6573,
      "step": 1199
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0266,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "eval_loss": 0.8731595277786255,
      "eval_runtime": 150.81,
      "eval_samples_per_second": 6.631,
      "eval_steps_per_second": 3.315,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "mmlu_eval_accuracy": 0.5655083950728274,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.5365853658536586,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705,
      "mmlu_eval_accuracy_high_school_psychology": 0.9166666666666666,
      "mmlu_eval_accuracy_high_school_statistics": 0.21739130434782608,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.631578947368421,
      "mmlu_eval_accuracy_moral_scenarios": 0.38,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.227738902828712,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0574,
      "step": 1201
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0258,
      "step": 1202
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.82,
      "step": 1203
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7978,
      "step": 1204
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1839,
      "step": 1205
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9996,
      "step": 1206
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8803,
      "step": 1207
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.2029,
      "step": 1208
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 1209
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8134,
      "step": 1210
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.2752,
      "step": 1211
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6257,
      "step": 1212
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6906,
      "step": 1213
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6703,
      "step": 1214
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4906,
      "step": 1215
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 1216
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0812,
      "step": 1217
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0253,
      "step": 1218
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8225,
      "step": 1219
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4269,
      "step": 1220
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6597,
      "step": 1221
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5044,
      "step": 1222
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8305,
      "step": 1223
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 1224
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8687,
      "step": 1225
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8474,
      "step": 1226
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7057,
      "step": 1227
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5403,
      "step": 1228
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9693,
      "step": 1229
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9484,
      "step": 1230
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8662,
      "step": 1231
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1856,
      "step": 1232
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9941,
      "step": 1233
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.914,
      "step": 1234
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9935,
      "step": 1235
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6557,
      "step": 1236
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6215,
      "step": 1237
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7099,
      "step": 1238
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7738,
      "step": 1239
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 1240
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7753,
      "step": 1241
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8679,
      "step": 1242
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 1243
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9892,
      "step": 1244
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1887,
      "step": 1245
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.092,
      "step": 1246
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9123,
      "step": 1247
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0886,
      "step": 1248
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1109,
      "step": 1249
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 1250
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9847,
      "step": 1251
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5364,
      "step": 1252
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9933,
      "step": 1253
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9881,
      "step": 1254
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0078,
      "step": 1255
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6977,
      "step": 1256
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0643,
      "step": 1257
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9667,
      "step": 1258
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6888,
      "step": 1259
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7366,
      "step": 1260
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.721,
      "step": 1261
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 1262
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9149,
      "step": 1263
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 1264
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8448,
      "step": 1265
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4008,
      "step": 1266
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.837,
      "step": 1267
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7404,
      "step": 1268
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5514,
      "step": 1269
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 1270
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8624,
      "step": 1271
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 1272
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.199,
      "step": 1273
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1555,
      "step": 1274
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 1275
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.967,
      "step": 1276
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9065,
      "step": 1277
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4052,
      "step": 1278
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1346,
      "step": 1279
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7915,
      "step": 1280
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1923,
      "step": 1281
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0185,
      "step": 1282
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4596,
      "step": 1283
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7876,
      "step": 1284
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.832,
      "step": 1285
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4757,
      "step": 1286
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8247,
      "step": 1287
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9283,
      "step": 1288
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9071,
      "step": 1289
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.191,
      "step": 1290
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7962,
      "step": 1291
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8896,
      "step": 1292
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7145,
      "step": 1293
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9079,
      "step": 1294
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9214,
      "step": 1295
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.2769,
      "step": 1296
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8162,
      "step": 1297
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8889,
      "step": 1298
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9111,
      "step": 1299
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8728,
      "step": 1300
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8175,
      "step": 1301
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6789,
      "step": 1302
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9805,
      "step": 1303
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9906,
      "step": 1304
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.507,
      "step": 1305
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8996,
      "step": 1306
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0435,
      "step": 1307
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0742,
      "step": 1308
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5387,
      "step": 1309
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9834,
      "step": 1310
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0843,
      "step": 1311
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8038,
      "step": 1312
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1885,
      "step": 1313
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 1314
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9753,
      "step": 1315
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8098,
      "step": 1316
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5895,
      "step": 1317
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.707,
      "step": 1318
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9639,
      "step": 1319
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8746,
      "step": 1320
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.91,
      "step": 1321
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1887,
      "step": 1322
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8152,
      "step": 1323
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1066,
      "step": 1324
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0189,
      "step": 1325
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 1326
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9112,
      "step": 1327
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1207,
      "step": 1328
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.022,
      "step": 1329
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9741,
      "step": 1330
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7867,
      "step": 1331
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 1332
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6663,
      "step": 1333
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7408,
      "step": 1334
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.3928,
      "step": 1335
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6765,
      "step": 1336
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6276,
      "step": 1337
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9062,
      "step": 1338
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8599,
      "step": 1339
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0781,
      "step": 1340
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9473,
      "step": 1341
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0255,
      "step": 1342
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.086,
      "step": 1343
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.992,
      "step": 1344
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 1345
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 1346
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1587,
      "step": 1347
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 1348
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5941,
      "step": 1349
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0102,
      "step": 1350
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1519,
      "step": 1351
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0301,
      "step": 1352
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7078,
      "step": 1353
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8709,
      "step": 1354
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0033,
      "step": 1355
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8076,
      "step": 1356
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0318,
      "step": 1357
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9687,
      "step": 1358
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5623,
      "step": 1359
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0747,
      "step": 1360
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 1361
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5379,
      "step": 1362
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1723,
      "step": 1363
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5623,
      "step": 1364
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1928,
      "step": 1365
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9919,
      "step": 1366
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 1367
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8815,
      "step": 1368
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.3812,
      "step": 1369
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7928,
      "step": 1370
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9124,
      "step": 1371
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 1372
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1718,
      "step": 1373
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6073,
      "step": 1374
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9901,
      "step": 1375
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 1376
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9299,
      "step": 1377
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9218,
      "step": 1378
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0773,
      "step": 1379
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 1380
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 1381
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9756,
      "step": 1382
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9728,
      "step": 1383
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1297,
      "step": 1384
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 1385
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7017,
      "step": 1386
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0745,
      "step": 1387
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8743,
      "step": 1388
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.48,
      "step": 1389
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 1390
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1489,
      "step": 1391
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7346,
      "step": 1392
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8834,
      "step": 1393
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9018,
      "step": 1394
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 1395
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7575,
      "step": 1396
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6201,
      "step": 1397
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1445,
      "step": 1398
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 1399
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.2336,
      "step": 1400
    },
    {
      "epoch": 0.2,
      "eval_loss": 0.8739627599716187,
      "eval_runtime": 150.6319,
      "eval_samples_per_second": 6.639,
      "eval_steps_per_second": 3.319,
      "step": 1400
    },
    {
      "epoch": 0.2,
      "mmlu_eval_accuracy": 0.5500602797223483,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.0,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 1.0,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.38,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.3384172773625769,
      "step": 1400
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.2293,
      "step": 1401
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6188,
      "step": 1402
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6192,
      "step": 1403
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5638,
      "step": 1404
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 1405
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6752,
      "step": 1406
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.805,
      "step": 1407
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8851,
      "step": 1408
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 1409
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8885,
      "step": 1410
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0146,
      "step": 1411
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0,
      "step": 1412
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.891,
      "step": 1413
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9117,
      "step": 1414
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7778,
      "step": 1415
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8771,
      "step": 1416
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8266,
      "step": 1417
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9444,
      "step": 1418
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9446,
      "step": 1419
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.3018,
      "step": 1420
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 1421
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5656,
      "step": 1422
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9266,
      "step": 1423
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9555,
      "step": 1424
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.054,
      "step": 1425
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6832,
      "step": 1426
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1318,
      "step": 1427
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9118,
      "step": 1428
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 1429
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.042,
      "step": 1430
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 1431
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6273,
      "step": 1432
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 1433
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1376,
      "step": 1434
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6583,
      "step": 1435
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0828,
      "step": 1436
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 1437
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 1438
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8013,
      "step": 1439
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 1440
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.2602,
      "step": 1441
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9369,
      "step": 1442
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.763,
      "step": 1443
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7517,
      "step": 1444
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7372,
      "step": 1445
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 1446
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 1447
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8075,
      "step": 1448
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.3385,
      "step": 1449
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7837,
      "step": 1450
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8372,
      "step": 1451
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8659,
      "step": 1452
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.668,
      "step": 1453
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 1454
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0566,
      "step": 1455
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.072,
      "step": 1456
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.471,
      "step": 1457
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.627,
      "step": 1458
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9232,
      "step": 1459
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0123,
      "step": 1460
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9645,
      "step": 1461
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7607,
      "step": 1462
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 1463
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5244,
      "step": 1464
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9529,
      "step": 1465
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8553,
      "step": 1466
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7937,
      "step": 1467
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8821,
      "step": 1468
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6856,
      "step": 1469
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8954,
      "step": 1470
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5814,
      "step": 1471
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6571,
      "step": 1472
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8668,
      "step": 1473
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1068,
      "step": 1474
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0992,
      "step": 1475
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1961,
      "step": 1476
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8368,
      "step": 1477
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6789,
      "step": 1478
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 1479
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0346,
      "step": 1480
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8459,
      "step": 1481
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 1482
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5587,
      "step": 1483
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0182,
      "step": 1484
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9832,
      "step": 1485
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 1486
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8287,
      "step": 1487
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4769,
      "step": 1488
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 1489
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8253,
      "step": 1490
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9484,
      "step": 1491
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 1492
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8256,
      "step": 1493
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9483,
      "step": 1494
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.1382,
      "step": 1495
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9424,
      "step": 1496
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8086,
      "step": 1497
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 1498
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6305,
      "step": 1499
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.1586,
      "step": 1500
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6502,
      "step": 1501
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9841,
      "step": 1502
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8351,
      "step": 1503
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8305,
      "step": 1504
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9804,
      "step": 1505
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9819,
      "step": 1506
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0511,
      "step": 1507
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6889,
      "step": 1508
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8127,
      "step": 1509
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.076,
      "step": 1510
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8027,
      "step": 1511
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7472,
      "step": 1512
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9433,
      "step": 1513
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 1514
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9434,
      "step": 1515
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7514,
      "step": 1516
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0196,
      "step": 1517
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7835,
      "step": 1518
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5263,
      "step": 1519
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8456,
      "step": 1520
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8293,
      "step": 1521
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.482,
      "step": 1522
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9047,
      "step": 1523
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7908,
      "step": 1524
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.812,
      "step": 1525
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 1526
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9137,
      "step": 1527
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8687,
      "step": 1528
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.929,
      "step": 1529
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 1530
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8347,
      "step": 1531
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9098,
      "step": 1532
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4818,
      "step": 1533
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8213,
      "step": 1534
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 1535
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9698,
      "step": 1536
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 1537
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.956,
      "step": 1538
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6987,
      "step": 1539
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9879,
      "step": 1540
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0679,
      "step": 1541
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7127,
      "step": 1542
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7991,
      "step": 1543
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 1544
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.925,
      "step": 1545
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0526,
      "step": 1546
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9478,
      "step": 1547
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8397,
      "step": 1548
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1229,
      "step": 1549
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8077,
      "step": 1550
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 1551
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7921,
      "step": 1552
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0494,
      "step": 1553
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5759,
      "step": 1554
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9064,
      "step": 1555
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6335,
      "step": 1556
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0769,
      "step": 1557
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9749,
      "step": 1558
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5599,
      "step": 1559
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.123,
      "step": 1560
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7494,
      "step": 1561
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7932,
      "step": 1562
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.3049,
      "step": 1563
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0138,
      "step": 1564
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8754,
      "step": 1565
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0756,
      "step": 1566
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 1567
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.805,
      "step": 1568
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7284,
      "step": 1569
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1413,
      "step": 1570
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7051,
      "step": 1571
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1664,
      "step": 1572
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4825,
      "step": 1573
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.412,
      "step": 1574
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 1575
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0357,
      "step": 1576
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7634,
      "step": 1577
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.962,
      "step": 1578
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9989,
      "step": 1579
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7018,
      "step": 1580
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9159,
      "step": 1581
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0805,
      "step": 1582
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0007,
      "step": 1583
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0586,
      "step": 1584
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 1585
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8674,
      "step": 1586
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9625,
      "step": 1587
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9374,
      "step": 1588
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.86,
      "step": 1589
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6195,
      "step": 1590
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0528,
      "step": 1591
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1585,
      "step": 1592
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7111,
      "step": 1593
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.046,
      "step": 1594
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9216,
      "step": 1595
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 1596
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9915,
      "step": 1597
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.861,
      "step": 1598
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9136,
      "step": 1599
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8502,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.881350576877594,
      "eval_runtime": 150.7277,
      "eval_samples_per_second": 6.634,
      "eval_steps_per_second": 3.317,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "mmlu_eval_accuracy": 0.5581891903781571,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.4375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.42,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.7096774193548387,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1852335237725284,
      "step": 1600
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1021,
      "step": 1601
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9366,
      "step": 1602
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9467,
      "step": 1603
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6091,
      "step": 1604
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0294,
      "step": 1605
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7689,
      "step": 1606
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.89,
      "step": 1607
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 1608
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8634,
      "step": 1609
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7717,
      "step": 1610
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0527,
      "step": 1611
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.713,
      "step": 1612
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7274,
      "step": 1613
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 1614
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 1615
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9472,
      "step": 1616
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8614,
      "step": 1617
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8309,
      "step": 1618
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 1619
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 1620
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 1621
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6191,
      "step": 1622
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9833,
      "step": 1623
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 1624
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 1625
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0974,
      "step": 1626
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5714,
      "step": 1627
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7563,
      "step": 1628
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0776,
      "step": 1629
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 1630
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7059,
      "step": 1631
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1213,
      "step": 1632
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9185,
      "step": 1633
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7874,
      "step": 1634
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 1635
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7931,
      "step": 1636
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7402,
      "step": 1637
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 1638
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0603,
      "step": 1639
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6336,
      "step": 1640
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.2513,
      "step": 1641
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 1642
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0262,
      "step": 1643
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0361,
      "step": 1644
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8104,
      "step": 1645
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 1646
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1309,
      "step": 1647
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 1648
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 1649
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9672,
      "step": 1650
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1506,
      "step": 1651
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 1652
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4049,
      "step": 1653
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0255,
      "step": 1654
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.2515,
      "step": 1655
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 1656
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0357,
      "step": 1657
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5498,
      "step": 1658
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6804,
      "step": 1659
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 1660
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 1661
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.671,
      "step": 1662
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6949,
      "step": 1663
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.3788,
      "step": 1664
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9495,
      "step": 1665
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 1666
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8979,
      "step": 1667
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 1668
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8644,
      "step": 1669
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0155,
      "step": 1670
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8751,
      "step": 1671
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1636,
      "step": 1672
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8598,
      "step": 1673
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5069,
      "step": 1674
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9482,
      "step": 1675
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.133,
      "step": 1676
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1351,
      "step": 1677
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 1678
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1285,
      "step": 1679
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 1680
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6329,
      "step": 1681
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 1682
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8504,
      "step": 1683
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 1684
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7733,
      "step": 1685
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.818,
      "step": 1686
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9067,
      "step": 1687
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.2218,
      "step": 1688
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6305,
      "step": 1689
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 1690
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9255,
      "step": 1691
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8922,
      "step": 1692
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9592,
      "step": 1693
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.2001,
      "step": 1694
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7491,
      "step": 1695
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5827,
      "step": 1696
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7294,
      "step": 1697
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9185,
      "step": 1698
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9132,
      "step": 1699
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5498,
      "step": 1700
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8386,
      "step": 1701
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9912,
      "step": 1702
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9487,
      "step": 1703
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 1704
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9624,
      "step": 1705
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9179,
      "step": 1706
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.922,
      "step": 1707
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 1708
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6873,
      "step": 1709
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8312,
      "step": 1710
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4543,
      "step": 1711
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8561,
      "step": 1712
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5476,
      "step": 1713
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0949,
      "step": 1714
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8883,
      "step": 1715
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.1032,
      "step": 1716
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7074,
      "step": 1717
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0305,
      "step": 1718
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6343,
      "step": 1719
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8422,
      "step": 1720
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.879,
      "step": 1721
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9655,
      "step": 1722
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6115,
      "step": 1723
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8749,
      "step": 1724
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6542,
      "step": 1725
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.2118,
      "step": 1726
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7937,
      "step": 1727
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4274,
      "step": 1728
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 1729
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 1730
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.034,
      "step": 1731
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7667,
      "step": 1732
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 1733
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9594,
      "step": 1734
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0448,
      "step": 1735
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 1736
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7751,
      "step": 1737
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9858,
      "step": 1738
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 1739
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 1740
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9634,
      "step": 1741
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.831,
      "step": 1742
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9495,
      "step": 1743
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8425,
      "step": 1744
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8138,
      "step": 1745
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9572,
      "step": 1746
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9466,
      "step": 1747
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9175,
      "step": 1748
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6684,
      "step": 1749
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7323,
      "step": 1750
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8096,
      "step": 1751
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9258,
      "step": 1752
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 1753
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0351,
      "step": 1754
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.054,
      "step": 1755
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.2227,
      "step": 1756
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9301,
      "step": 1757
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0017,
      "step": 1758
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.558,
      "step": 1759
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7367,
      "step": 1760
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9185,
      "step": 1761
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.2052,
      "step": 1762
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 1763
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9693,
      "step": 1764
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8496,
      "step": 1765
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7695,
      "step": 1766
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1844,
      "step": 1767
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8114,
      "step": 1768
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8955,
      "step": 1769
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 1770
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7877,
      "step": 1771
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.979,
      "step": 1772
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8724,
      "step": 1773
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1764,
      "step": 1774
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0432,
      "step": 1775
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 1776
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9143,
      "step": 1777
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1882,
      "step": 1778
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 1779
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9327,
      "step": 1780
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6219,
      "step": 1781
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.884,
      "step": 1782
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0027,
      "step": 1783
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6589,
      "step": 1784
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 1785
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.808,
      "step": 1786
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.2178,
      "step": 1787
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 1788
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1546,
      "step": 1789
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0111,
      "step": 1790
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9663,
      "step": 1791
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8282,
      "step": 1792
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8979,
      "step": 1793
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9908,
      "step": 1794
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8076,
      "step": 1795
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5658,
      "step": 1796
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9425,
      "step": 1797
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8265,
      "step": 1798
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9913,
      "step": 1799
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7498,
      "step": 1800
    },
    {
      "epoch": 0.26,
      "eval_loss": 0.880131721496582,
      "eval_runtime": 150.694,
      "eval_samples_per_second": 6.636,
      "eval_steps_per_second": 3.318,
      "step": 1800
    },
    {
      "epoch": 0.26,
      "mmlu_eval_accuracy": 0.5745485859086488,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.6,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7777777777777778,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.27,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.40588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.2574062925172538,
      "step": 1800
    }
  ],
  "logging_steps": 1,
  "max_steps": 20604,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 4.2340572282873446e+17,
  "trial_name": null,
  "trial_params": null
}