{
  "best_metric": 0.5674206614494324,
  "best_model_checkpoint": "experts/mistralic-expert-23/checkpoint-1200",
  "epoch": 0.28040658955485454,
  "eval_steps": 200,
  "global_step": 1200,
  "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": 0.6926,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 2
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.539,
      "step": 3
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5802,
      "step": 4
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5176,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6371,
      "step": 6
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5969,
      "step": 7
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 8
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7967,
      "step": 9
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5244,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9098,
      "step": 11
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 12
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5279,
      "step": 13
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 14
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9971,
      "step": 15
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 16
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6404,
      "step": 17
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8234,
      "step": 18
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.529,
      "step": 19
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5078,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.4981,
      "step": 21
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8841,
      "step": 22
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5563,
      "step": 23
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5759,
      "step": 24
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5739,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5891,
      "step": 26
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.454,
      "step": 27
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5685,
      "step": 28
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5723,
      "step": 29
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.516,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 31
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7529,
      "step": 32
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6331,
      "step": 33
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7149,
      "step": 34
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7413,
      "step": 35
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5488,
      "step": 36
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8207,
      "step": 37
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5589,
      "step": 38
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5907,
      "step": 39
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5317,
      "step": 41
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5762,
      "step": 42
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5986,
      "step": 43
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5758,
      "step": 44
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 45
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6161,
      "step": 46
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5554,
      "step": 47
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5997,
      "step": 48
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5616,
      "step": 49
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6303,
      "step": 51
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 52
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 53
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5734,
      "step": 54
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5353,
      "step": 55
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5713,
      "step": 56
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6006,
      "step": 57
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 58
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5286,
      "step": 59
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 60
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6916,
      "step": 61
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 62
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5933,
      "step": 63
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 64
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5764,
      "step": 65
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.4551,
      "step": 66
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6122,
      "step": 67
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6257,
      "step": 68
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6027,
      "step": 69
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 70
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6817,
      "step": 71
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5495,
      "step": 72
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5997,
      "step": 73
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.837,
      "step": 74
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5642,
      "step": 75
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 76
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5589,
      "step": 77
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6021,
      "step": 78
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6688,
      "step": 79
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5716,
      "step": 80
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.75,
      "step": 81
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6068,
      "step": 82
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 83
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 84
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5786,
      "step": 85
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6502,
      "step": 86
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 87
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6432,
      "step": 88
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5064,
      "step": 89
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5188,
      "step": 90
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9853,
      "step": 91
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5188,
      "step": 92
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8395,
      "step": 93
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5267,
      "step": 94
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5357,
      "step": 95
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5574,
      "step": 96
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5883,
      "step": 97
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 98
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5527,
      "step": 99
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6302,
      "step": 100
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8618,
      "step": 101
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6703,
      "step": 102
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5242,
      "step": 103
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.653,
      "step": 104
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 105
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 106
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5107,
      "step": 107
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6885,
      "step": 108
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5642,
      "step": 109
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6477,
      "step": 110
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 111
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5575,
      "step": 112
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 113
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5312,
      "step": 114
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 115
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4778,
      "step": 116
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 117
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4322,
      "step": 118
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 119
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6603,
      "step": 120
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4729,
      "step": 121
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8347,
      "step": 122
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6119,
      "step": 123
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "step": 124
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5018,
      "step": 125
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 126
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.606,
      "step": 127
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5727,
      "step": 128
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6316,
      "step": 129
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 130
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5735,
      "step": 131
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4528,
      "step": 132
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 133
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6042,
      "step": 134
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.76,
      "step": 135
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4522,
      "step": 136
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5826,
      "step": 137
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5408,
      "step": 138
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5659,
      "step": 139
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.551,
      "step": 140
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5766,
      "step": 141
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 142
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 143
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5066,
      "step": 144
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 145
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5727,
      "step": 146
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6109,
      "step": 147
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 148
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6917,
      "step": 149
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6373,
      "step": 150
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6928,
      "step": 151
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5866,
      "step": 152
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 153
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5483,
      "step": 154
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5335,
      "step": 155
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 156
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5694,
      "step": 157
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 158
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 159
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 160
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4149,
      "step": 161
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 162
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5949,
      "step": 163
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6294,
      "step": 164
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5303,
      "step": 165
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7256,
      "step": 166
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 167
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7129,
      "step": 168
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 169
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4735,
      "step": 170
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4963,
      "step": 171
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8293,
      "step": 172
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 173
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5835,
      "step": 174
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7986,
      "step": 175
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6569,
      "step": 176
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6295,
      "step": 177
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 178
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5401,
      "step": 179
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6203,
      "step": 180
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5045,
      "step": 181
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5779,
      "step": 182
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5386,
      "step": 183
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4483,
      "step": 184
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5543,
      "step": 185
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0266,
      "step": 186
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5621,
      "step": 187
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5884,
      "step": 188
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 189
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 190
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5938,
      "step": 191
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6306,
      "step": 192
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 193
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6613,
      "step": 194
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4926,
      "step": 195
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4874,
      "step": 196
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6966,
      "step": 197
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 198
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.62,
      "step": 199
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8005,
      "step": 200
    },
    {
      "epoch": 0.05,
      "eval_loss": 0.600193977355957,
      "eval_runtime": 156.5363,
      "eval_samples_per_second": 6.388,
      "eval_steps_per_second": 3.194,
      "step": 200
    },
    {
      "epoch": 0.05,
      "mmlu_eval_accuracy": 0.6110525160674652,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "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": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.37,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.7407407407407407,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
      "mmlu_loss": 1.4672958862532524,
      "step": 200
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 201
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5071,
      "step": 202
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7988,
      "step": 203
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.542,
      "step": 204
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.501,
      "step": 205
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6016,
      "step": 206
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7817,
      "step": 207
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5444,
      "step": 208
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7871,
      "step": 209
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6683,
      "step": 210
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4517,
      "step": 211
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8223,
      "step": 212
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.536,
      "step": 213
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5169,
      "step": 214
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 215
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 216
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5211,
      "step": 217
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5449,
      "step": 218
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5034,
      "step": 219
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8192,
      "step": 220
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5212,
      "step": 221
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 222
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 223
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5851,
      "step": 224
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 225
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 226
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8992,
      "step": 227
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5907,
      "step": 228
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.602,
      "step": 229
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5298,
      "step": 230
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4669,
      "step": 231
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8426,
      "step": 232
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8515,
      "step": 233
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5737,
      "step": 234
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5501,
      "step": 235
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4967,
      "step": 236
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6454,
      "step": 237
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4659,
      "step": 238
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 239
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 240
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5715,
      "step": 241
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4989,
      "step": 242
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6546,
      "step": 243
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7433,
      "step": 244
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5624,
      "step": 245
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5985,
      "step": 246
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5385,
      "step": 247
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 248
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5485,
      "step": 249
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6936,
      "step": 250
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.551,
      "step": 251
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 252
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5189,
      "step": 253
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7124,
      "step": 254
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 255
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 256
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4779,
      "step": 257
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5552,
      "step": 258
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 259
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5334,
      "step": 260
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6678,
      "step": 261
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7578,
      "step": 262
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 263
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5694,
      "step": 264
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 265
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4679,
      "step": 266
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6381,
      "step": 267
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5931,
      "step": 268
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5818,
      "step": 269
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 270
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6526,
      "step": 271
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5631,
      "step": 272
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7555,
      "step": 273
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 274
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5831,
      "step": 275
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5269,
      "step": 276
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4996,
      "step": 277
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8797,
      "step": 278
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4943,
      "step": 279
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8619,
      "step": 280
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6191,
      "step": 281
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4933,
      "step": 282
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.744,
      "step": 283
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4481,
      "step": 284
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5396,
      "step": 285
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5108,
      "step": 286
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 287
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5603,
      "step": 288
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5778,
      "step": 289
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 290
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.52,
      "step": 291
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6123,
      "step": 292
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5051,
      "step": 293
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5057,
      "step": 294
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.601,
      "step": 295
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9173,
      "step": 296
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5194,
      "step": 297
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6205,
      "step": 298
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 299
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8153,
      "step": 300
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5872,
      "step": 301
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 302
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4371,
      "step": 303
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 304
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5014,
      "step": 305
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4599,
      "step": 306
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5638,
      "step": 307
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4859,
      "step": 308
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5348,
      "step": 309
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4462,
      "step": 310
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6522,
      "step": 311
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7238,
      "step": 312
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4864,
      "step": 313
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 314
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4141,
      "step": 315
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 316
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4621,
      "step": 317
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5194,
      "step": 318
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7686,
      "step": 319
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7024,
      "step": 320
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6979,
      "step": 321
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 322
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5238,
      "step": 323
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5048,
      "step": 324
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5091,
      "step": 325
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7207,
      "step": 326
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5248,
      "step": 327
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8079,
      "step": 328
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 329
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6146,
      "step": 330
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5315,
      "step": 331
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.53,
      "step": 332
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.794,
      "step": 333
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.4968,
      "step": 334
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5732,
      "step": 335
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 336
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.469,
      "step": 337
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.638,
      "step": 338
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7457,
      "step": 339
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5851,
      "step": 340
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6549,
      "step": 341
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5412,
      "step": 342
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7014,
      "step": 343
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.727,
      "step": 344
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5415,
      "step": 345
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 346
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 347
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5198,
      "step": 348
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5274,
      "step": 349
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5288,
      "step": 350
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5649,
      "step": 351
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5385,
      "step": 352
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6064,
      "step": 353
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5652,
      "step": 354
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.4739,
      "step": 355
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.4917,
      "step": 356
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6023,
      "step": 357
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7409,
      "step": 358
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 359
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 360
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5896,
      "step": 361
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5608,
      "step": 362
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 363
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5431,
      "step": 364
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 365
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5106,
      "step": 366
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4998,
      "step": 367
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5058,
      "step": 368
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.536,
      "step": 369
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5248,
      "step": 370
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.728,
      "step": 371
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4739,
      "step": 372
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7373,
      "step": 373
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 374
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5658,
      "step": 375
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4556,
      "step": 376
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 377
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5215,
      "step": 378
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4547,
      "step": 379
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6626,
      "step": 380
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 381
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7555,
      "step": 382
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 383
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6711,
      "step": 384
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 385
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6033,
      "step": 386
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 387
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5267,
      "step": 388
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.664,
      "step": 389
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6835,
      "step": 390
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5874,
      "step": 391
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5413,
      "step": 392
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 393
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5294,
      "step": 394
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5736,
      "step": 395
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5794,
      "step": 396
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6177,
      "step": 397
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5843,
      "step": 398
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6333,
      "step": 399
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 400
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.5899355411529541,
      "eval_runtime": 156.5886,
      "eval_samples_per_second": 6.386,
      "eval_steps_per_second": 3.193,
      "step": 400
    },
    {
      "epoch": 0.09,
      "mmlu_eval_accuracy": 0.6068635133301271,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.8125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.37,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7058823529411765,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.4117647058823529,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.9545454545454546,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.496563289402048,
      "step": 400
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5013,
      "step": 401
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.509,
      "step": 402
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6059,
      "step": 403
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 404
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5252,
      "step": 405
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5063,
      "step": 406
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4901,
      "step": 407
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 408
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5539,
      "step": 409
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6965,
      "step": 410
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 411
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 412
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.488,
      "step": 413
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7407,
      "step": 414
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 415
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5734,
      "step": 416
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5481,
      "step": 417
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 418
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4902,
      "step": 419
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6816,
      "step": 420
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5906,
      "step": 421
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 422
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5178,
      "step": 423
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5605,
      "step": 424
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 425
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 426
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5083,
      "step": 427
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4889,
      "step": 428
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5004,
      "step": 429
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4759,
      "step": 430
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 431
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5438,
      "step": 432
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 433
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.839,
      "step": 434
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5581,
      "step": 435
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5813,
      "step": 436
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4878,
      "step": 437
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5984,
      "step": 438
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7623,
      "step": 439
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 440
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5112,
      "step": 441
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4414,
      "step": 442
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5551,
      "step": 443
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5107,
      "step": 444
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5318,
      "step": 445
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 446
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 447
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 448
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7416,
      "step": 449
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5064,
      "step": 450
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4845,
      "step": 451
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8342,
      "step": 452
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 453
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4511,
      "step": 454
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4945,
      "step": 455
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5943,
      "step": 456
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 457
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5331,
      "step": 458
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6232,
      "step": 459
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7195,
      "step": 460
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7247,
      "step": 461
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7871,
      "step": 462
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 463
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5728,
      "step": 464
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5263,
      "step": 465
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 466
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 467
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5316,
      "step": 468
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4836,
      "step": 469
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4781,
      "step": 470
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.645,
      "step": 471
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6311,
      "step": 472
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5401,
      "step": 473
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 474
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7188,
      "step": 475
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.587,
      "step": 476
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.46,
      "step": 477
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5724,
      "step": 478
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5907,
      "step": 479
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.484,
      "step": 480
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 481
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5197,
      "step": 482
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 483
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4964,
      "step": 484
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4705,
      "step": 485
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4844,
      "step": 486
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6951,
      "step": 487
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5948,
      "step": 488
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5728,
      "step": 489
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 490
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 491
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5119,
      "step": 492
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6653,
      "step": 493
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 494
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5486,
      "step": 495
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7635,
      "step": 496
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 497
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5336,
      "step": 498
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5103,
      "step": 499
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4841,
      "step": 500
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5479,
      "step": 501
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6147,
      "step": 502
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4874,
      "step": 503
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7884,
      "step": 504
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6535,
      "step": 505
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6987,
      "step": 506
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6371,
      "step": 507
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 508
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7897,
      "step": 509
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4869,
      "step": 510
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6293,
      "step": 511
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7633,
      "step": 512
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 513
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5522,
      "step": 514
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5368,
      "step": 515
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5667,
      "step": 516
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5541,
      "step": 517
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5003,
      "step": 518
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5217,
      "step": 519
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.467,
      "step": 520
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4416,
      "step": 521
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4778,
      "step": 522
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5085,
      "step": 523
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 524
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4561,
      "step": 525
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.56,
      "step": 526
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5205,
      "step": 527
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4748,
      "step": 528
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5813,
      "step": 529
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 530
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 531
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4624,
      "step": 532
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.546,
      "step": 533
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9573,
      "step": 534
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5634,
      "step": 535
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.3786,
      "step": 536
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7356,
      "step": 537
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5748,
      "step": 538
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6774,
      "step": 539
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5361,
      "step": 540
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.465,
      "step": 541
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5905,
      "step": 542
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 543
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 544
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5201,
      "step": 545
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7455,
      "step": 546
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 547
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 548
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 549
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5438,
      "step": 550
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4865,
      "step": 551
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 552
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6324,
      "step": 553
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4357,
      "step": 554
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4484,
      "step": 555
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 556
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5309,
      "step": 557
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5603,
      "step": 558
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7126,
      "step": 559
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4893,
      "step": 560
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5384,
      "step": 561
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4642,
      "step": 562
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4378,
      "step": 563
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5831,
      "step": 564
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5491,
      "step": 565
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6361,
      "step": 566
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0347,
      "step": 567
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4843,
      "step": 568
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.762,
      "step": 569
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7908,
      "step": 570
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5221,
      "step": 571
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8129,
      "step": 572
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.542,
      "step": 573
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5179,
      "step": 574
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 575
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6367,
      "step": 576
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 577
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5287,
      "step": 578
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 579
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6124,
      "step": 580
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8515,
      "step": 581
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 582
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5048,
      "step": 583
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.559,
      "step": 584
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5914,
      "step": 585
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5469,
      "step": 586
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.719,
      "step": 587
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 588
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4804,
      "step": 589
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 590
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4929,
      "step": 591
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6596,
      "step": 592
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5239,
      "step": 593
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 594
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7026,
      "step": 595
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7988,
      "step": 596
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 597
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7739,
      "step": 598
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5263,
      "step": 599
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 600
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.5821242928504944,
      "eval_runtime": 156.5384,
      "eval_samples_per_second": 6.388,
      "eval_steps_per_second": 3.194,
      "step": 600
    },
    {
      "epoch": 0.14,
      "mmlu_eval_accuracy": 0.5980259299067578,
      "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": 0.6206896551724138,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "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": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "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": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.31,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.7058823529411765,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.5698812617178686,
      "step": 600
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.797,
      "step": 601
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5773,
      "step": 602
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4958,
      "step": 603
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5999,
      "step": 604
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5027,
      "step": 605
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 606
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.484,
      "step": 607
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4481,
      "step": 608
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9088,
      "step": 609
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4749,
      "step": 610
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8021,
      "step": 611
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 612
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4418,
      "step": 613
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 614
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 615
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7919,
      "step": 616
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 617
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 618
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5026,
      "step": 619
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 620
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5307,
      "step": 621
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7302,
      "step": 622
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 623
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5515,
      "step": 624
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5555,
      "step": 625
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 626
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5085,
      "step": 627
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 628
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 629
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5179,
      "step": 630
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.502,
      "step": 631
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.555,
      "step": 632
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5634,
      "step": 633
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 634
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5024,
      "step": 635
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6944,
      "step": 636
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 637
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5491,
      "step": 638
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5276,
      "step": 639
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 640
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5254,
      "step": 641
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5183,
      "step": 642
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5534,
      "step": 643
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 644
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6834,
      "step": 645
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5231,
      "step": 646
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7475,
      "step": 647
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5971,
      "step": 648
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5364,
      "step": 649
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 650
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5151,
      "step": 651
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5848,
      "step": 652
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 653
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7276,
      "step": 654
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4609,
      "step": 655
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6721,
      "step": 656
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6221,
      "step": 657
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6601,
      "step": 658
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5458,
      "step": 659
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5848,
      "step": 660
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5555,
      "step": 661
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5629,
      "step": 662
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5773,
      "step": 663
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5895,
      "step": 664
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.584,
      "step": 665
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4354,
      "step": 666
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6236,
      "step": 667
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 668
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 669
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5186,
      "step": 670
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6771,
      "step": 671
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5456,
      "step": 672
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5139,
      "step": 673
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 674
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6333,
      "step": 675
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5399,
      "step": 676
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.485,
      "step": 677
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5551,
      "step": 678
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4292,
      "step": 679
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5707,
      "step": 680
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5536,
      "step": 681
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 682
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.561,
      "step": 683
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5815,
      "step": 684
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 685
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 686
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4585,
      "step": 687
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5481,
      "step": 688
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5959,
      "step": 689
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5654,
      "step": 690
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7278,
      "step": 691
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8314,
      "step": 692
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5705,
      "step": 693
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 694
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 695
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5345,
      "step": 696
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5068,
      "step": 697
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5463,
      "step": 698
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4945,
      "step": 699
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8122,
      "step": 700
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7609,
      "step": 701
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4743,
      "step": 702
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 703
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 704
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 705
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5838,
      "step": 706
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 707
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.451,
      "step": 708
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6256,
      "step": 709
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 710
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4892,
      "step": 711
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.508,
      "step": 712
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5751,
      "step": 713
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5195,
      "step": 714
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5914,
      "step": 715
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 716
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6462,
      "step": 717
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5456,
      "step": 718
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6971,
      "step": 719
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 720
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5519,
      "step": 721
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4363,
      "step": 722
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5198,
      "step": 723
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 724
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5382,
      "step": 725
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5231,
      "step": 726
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9078,
      "step": 727
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 728
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5581,
      "step": 729
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5676,
      "step": 730
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5519,
      "step": 731
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7338,
      "step": 732
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6373,
      "step": 733
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5116,
      "step": 734
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 735
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5385,
      "step": 736
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5202,
      "step": 737
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6285,
      "step": 738
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 739
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5757,
      "step": 740
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4787,
      "step": 741
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4862,
      "step": 742
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 743
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5101,
      "step": 744
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5368,
      "step": 745
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 746
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6534,
      "step": 747
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4966,
      "step": 748
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5995,
      "step": 749
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5371,
      "step": 750
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 751
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4673,
      "step": 752
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 753
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6554,
      "step": 754
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8246,
      "step": 755
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 756
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5867,
      "step": 757
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5766,
      "step": 758
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4687,
      "step": 759
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5795,
      "step": 760
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4207,
      "step": 761
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.524,
      "step": 762
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 763
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4486,
      "step": 764
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 765
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7085,
      "step": 766
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5446,
      "step": 767
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4995,
      "step": 768
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5232,
      "step": 769
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5164,
      "step": 770
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 771
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7406,
      "step": 772
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5989,
      "step": 773
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5384,
      "step": 774
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5186,
      "step": 775
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5206,
      "step": 776
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7905,
      "step": 777
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5032,
      "step": 778
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4573,
      "step": 779
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8265,
      "step": 780
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5497,
      "step": 781
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5017,
      "step": 782
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 783
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 784
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5182,
      "step": 785
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6738,
      "step": 786
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5116,
      "step": 787
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4761,
      "step": 788
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 789
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6349,
      "step": 790
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5733,
      "step": 791
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4627,
      "step": 792
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4988,
      "step": 793
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.691,
      "step": 794
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5198,
      "step": 795
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5535,
      "step": 796
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5272,
      "step": 797
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 798
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7339,
      "step": 799
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5,
      "step": 800
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.574880838394165,
      "eval_runtime": 156.4152,
      "eval_samples_per_second": 6.393,
      "eval_steps_per_second": 3.197,
      "step": 800
    },
    {
      "epoch": 0.19,
      "mmlu_eval_accuracy": 0.5932018544060925,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "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": 0.7272727272727273,
      "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": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.7272727272727273,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.3,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7058823529411765,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
      "mmlu_loss": 1.3265360119526441,
      "step": 800
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5165,
      "step": 801
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 802
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4657,
      "step": 803
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5278,
      "step": 804
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5308,
      "step": 805
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5479,
      "step": 806
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 807
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 808
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4815,
      "step": 809
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4946,
      "step": 810
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8382,
      "step": 811
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 812
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.517,
      "step": 813
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 814
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 815
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 816
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5489,
      "step": 817
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5801,
      "step": 818
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 819
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 820
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 821
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5404,
      "step": 822
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 823
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5629,
      "step": 824
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 825
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 826
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5907,
      "step": 827
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7593,
      "step": 828
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5988,
      "step": 829
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7543,
      "step": 830
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6409,
      "step": 831
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5226,
      "step": 832
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5747,
      "step": 833
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4743,
      "step": 834
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 835
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5089,
      "step": 836
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5515,
      "step": 837
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4745,
      "step": 838
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6076,
      "step": 839
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4837,
      "step": 840
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 841
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6733,
      "step": 842
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 843
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 844
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 845
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 846
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5534,
      "step": 847
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5574,
      "step": 848
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5164,
      "step": 849
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4638,
      "step": 850
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 851
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6846,
      "step": 852
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5436,
      "step": 853
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4961,
      "step": 854
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5504,
      "step": 855
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 856
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4132,
      "step": 857
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 858
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5085,
      "step": 859
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5122,
      "step": 860
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6464,
      "step": 861
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6277,
      "step": 862
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5759,
      "step": 863
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4963,
      "step": 864
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.531,
      "step": 865
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5125,
      "step": 866
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.496,
      "step": 867
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4053,
      "step": 868
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5383,
      "step": 869
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5676,
      "step": 870
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 871
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6998,
      "step": 872
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6198,
      "step": 873
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5652,
      "step": 874
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 875
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4955,
      "step": 876
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4391,
      "step": 877
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8064,
      "step": 878
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.726,
      "step": 879
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5287,
      "step": 880
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4968,
      "step": 881
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5153,
      "step": 882
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5129,
      "step": 883
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5616,
      "step": 884
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5374,
      "step": 885
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6629,
      "step": 886
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.544,
      "step": 887
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4661,
      "step": 888
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5176,
      "step": 889
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5545,
      "step": 890
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 891
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5179,
      "step": 892
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 893
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6187,
      "step": 894
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6309,
      "step": 895
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5833,
      "step": 896
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4936,
      "step": 897
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4672,
      "step": 898
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.448,
      "step": 899
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5412,
      "step": 900
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 901
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 902
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6641,
      "step": 903
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5018,
      "step": 904
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 905
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5706,
      "step": 906
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5344,
      "step": 907
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4701,
      "step": 908
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 909
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5901,
      "step": 910
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6398,
      "step": 911
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8337,
      "step": 912
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5989,
      "step": 913
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8263,
      "step": 914
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5399,
      "step": 915
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4453,
      "step": 916
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 917
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4743,
      "step": 918
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 919
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6518,
      "step": 920
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6899,
      "step": 921
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4602,
      "step": 922
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4718,
      "step": 923
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.743,
      "step": 924
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7759,
      "step": 925
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.53,
      "step": 926
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4967,
      "step": 927
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 928
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.3803,
      "step": 929
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.564,
      "step": 930
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6286,
      "step": 931
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.519,
      "step": 932
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 933
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6336,
      "step": 934
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 935
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6905,
      "step": 936
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 937
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5102,
      "step": 938
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4539,
      "step": 939
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5792,
      "step": 940
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5131,
      "step": 941
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4312,
      "step": 942
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6203,
      "step": 943
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 944
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8426,
      "step": 945
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.558,
      "step": 946
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4399,
      "step": 947
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6979,
      "step": 948
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 949
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 950
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5208,
      "step": 951
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7029,
      "step": 952
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5348,
      "step": 953
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5016,
      "step": 954
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.561,
      "step": 955
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6065,
      "step": 956
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.509,
      "step": 957
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5928,
      "step": 958
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5336,
      "step": 959
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4983,
      "step": 960
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5083,
      "step": 961
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5219,
      "step": 962
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6288,
      "step": 963
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5795,
      "step": 964
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5825,
      "step": 965
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4772,
      "step": 966
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6507,
      "step": 967
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7522,
      "step": 968
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 969
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4939,
      "step": 970
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4675,
      "step": 971
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5173,
      "step": 972
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 973
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5046,
      "step": 974
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6346,
      "step": 975
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4336,
      "step": 976
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.895,
      "step": 977
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5283,
      "step": 978
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5317,
      "step": 979
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4931,
      "step": 980
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5238,
      "step": 981
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 982
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.64,
      "step": 983
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6442,
      "step": 984
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5357,
      "step": 985
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5308,
      "step": 986
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4645,
      "step": 987
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5952,
      "step": 988
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7809,
      "step": 989
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.733,
      "step": 990
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 991
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 992
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6119,
      "step": 993
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4745,
      "step": 994
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6025,
      "step": 995
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6024,
      "step": 996
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4761,
      "step": 997
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8131,
      "step": 998
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5084,
      "step": 999
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5849,
      "step": 1000
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.5721054673194885,
      "eval_runtime": 156.5412,
      "eval_samples_per_second": 6.388,
      "eval_steps_per_second": 3.194,
      "step": 1000
    },
    {
      "epoch": 0.23,
      "mmlu_eval_accuracy": 0.5929489471474008,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.75,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "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": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.4235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "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": 0.9545454545454546,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
      "mmlu_loss": 1.4194403641684559,
      "step": 1000
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4694,
      "step": 1001
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4452,
      "step": 1002
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6288,
      "step": 1003
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 1004
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5607,
      "step": 1005
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4829,
      "step": 1006
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4815,
      "step": 1007
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 1008
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.523,
      "step": 1009
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5669,
      "step": 1010
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5128,
      "step": 1011
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 1012
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5051,
      "step": 1013
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6593,
      "step": 1014
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8449,
      "step": 1015
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 1016
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 1017
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.654,
      "step": 1018
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.555,
      "step": 1019
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 1020
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.539,
      "step": 1021
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5369,
      "step": 1022
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7462,
      "step": 1023
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4919,
      "step": 1024
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 1025
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6417,
      "step": 1026
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 1027
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 1028
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 1029
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8065,
      "step": 1030
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6074,
      "step": 1031
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5452,
      "step": 1032
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5269,
      "step": 1033
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4618,
      "step": 1034
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5843,
      "step": 1035
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 1036
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 1037
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4987,
      "step": 1038
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 1039
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4573,
      "step": 1040
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4609,
      "step": 1041
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5074,
      "step": 1042
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7367,
      "step": 1043
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5011,
      "step": 1044
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.362,
      "step": 1045
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5725,
      "step": 1046
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8629,
      "step": 1047
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5021,
      "step": 1048
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 1049
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 1050
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4932,
      "step": 1051
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5024,
      "step": 1052
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4504,
      "step": 1053
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4706,
      "step": 1054
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4296,
      "step": 1055
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5156,
      "step": 1056
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.755,
      "step": 1057
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 1058
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5421,
      "step": 1059
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5279,
      "step": 1060
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4957,
      "step": 1061
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5078,
      "step": 1062
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4285,
      "step": 1063
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 1064
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5213,
      "step": 1065
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5419,
      "step": 1066
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5335,
      "step": 1067
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6135,
      "step": 1068
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4099,
      "step": 1069
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4687,
      "step": 1070
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7218,
      "step": 1071
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6365,
      "step": 1072
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8037,
      "step": 1073
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6133,
      "step": 1074
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5327,
      "step": 1075
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 1076
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6197,
      "step": 1077
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 1078
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 1079
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5642,
      "step": 1080
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7326,
      "step": 1081
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 1082
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4875,
      "step": 1083
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6873,
      "step": 1084
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4772,
      "step": 1085
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4758,
      "step": 1086
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 1087
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7724,
      "step": 1088
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 1089
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5736,
      "step": 1090
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4825,
      "step": 1091
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 1092
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 1093
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 1094
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4171,
      "step": 1095
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4914,
      "step": 1096
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 1097
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4906,
      "step": 1098
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5104,
      "step": 1099
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7659,
      "step": 1100
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4673,
      "step": 1101
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4495,
      "step": 1102
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5212,
      "step": 1103
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8471,
      "step": 1104
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5485,
      "step": 1105
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5625,
      "step": 1106
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 1107
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 1108
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5568,
      "step": 1109
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4996,
      "step": 1110
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.41,
      "step": 1111
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7022,
      "step": 1112
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7582,
      "step": 1113
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.555,
      "step": 1114
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5697,
      "step": 1115
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5304,
      "step": 1116
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4333,
      "step": 1117
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 1118
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5237,
      "step": 1119
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6009,
      "step": 1120
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.476,
      "step": 1121
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5012,
      "step": 1122
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5294,
      "step": 1123
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4581,
      "step": 1124
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.3987,
      "step": 1125
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6608,
      "step": 1126
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9719,
      "step": 1127
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 1128
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 1129
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0379,
      "step": 1130
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5355,
      "step": 1131
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9247,
      "step": 1132
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7992,
      "step": 1133
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4367,
      "step": 1134
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5313,
      "step": 1135
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4346,
      "step": 1136
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5513,
      "step": 1137
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6981,
      "step": 1138
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.53,
      "step": 1139
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5565,
      "step": 1140
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8145,
      "step": 1141
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5173,
      "step": 1142
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7706,
      "step": 1143
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5209,
      "step": 1144
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5082,
      "step": 1145
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8339,
      "step": 1146
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 1147
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4769,
      "step": 1148
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 1149
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4841,
      "step": 1150
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5175,
      "step": 1151
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 1152
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4722,
      "step": 1153
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 1154
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5219,
      "step": 1155
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6622,
      "step": 1156
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7403,
      "step": 1157
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5171,
      "step": 1158
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 1159
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4686,
      "step": 1160
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4958,
      "step": 1161
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5461,
      "step": 1162
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4739,
      "step": 1163
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 1164
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 1165
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 1166
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6872,
      "step": 1167
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5438,
      "step": 1168
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 1169
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5235,
      "step": 1170
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 1171
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7884,
      "step": 1172
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.645,
      "step": 1173
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4354,
      "step": 1174
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4801,
      "step": 1175
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.608,
      "step": 1176
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 1177
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5196,
      "step": 1178
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6569,
      "step": 1179
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4905,
      "step": 1180
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 1181
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6434,
      "step": 1182
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 1183
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 1184
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4896,
      "step": 1185
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6711,
      "step": 1186
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7248,
      "step": 1187
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4794,
      "step": 1188
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.537,
      "step": 1189
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 1190
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4828,
      "step": 1191
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7808,
      "step": 1192
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6188,
      "step": 1193
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5699,
      "step": 1194
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7116,
      "step": 1195
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.503,
      "step": 1196
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.537,
      "step": 1197
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5936,
      "step": 1198
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 1199
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 1200
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.5674206614494324,
      "eval_runtime": 156.5334,
      "eval_samples_per_second": 6.388,
      "eval_steps_per_second": 3.194,
      "step": 1200
    },
    {
      "epoch": 0.28,
      "mmlu_eval_accuracy": 0.5949461253320095,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "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": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.058823529411764705,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "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": 0.6956521739130435,
      "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": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.31,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.3941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.4938532913166922,
      "step": 1200
    }
  ],
  "logging_steps": 1,
  "max_steps": 12837,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 2.941487578127401e+17,
  "trial_name": null,
  "trial_params": null
}