{
  "best_metric": 0.9176025986671448,
  "best_model_checkpoint": "experts/mistralic-expert-31/checkpoint-1200",
  "epoch": 0.49367479173094725,
  "eval_steps": 200,
  "global_step": 1600,
  "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": 0.997,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0488,
      "step": 2
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9417,
      "step": 3
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0655,
      "step": 4
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9966,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.1248,
      "step": 6
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0554,
      "step": 7
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.1163,
      "step": 8
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 9
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.1657,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9985,
      "step": 11
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0346,
      "step": 12
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8882,
      "step": 13
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0175,
      "step": 14
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.877,
      "step": 15
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8384,
      "step": 16
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0233,
      "step": 17
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0147,
      "step": 18
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1521,
      "step": 19
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9418,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1608,
      "step": 21
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9631,
      "step": 22
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0209,
      "step": 23
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0619,
      "step": 24
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.2694,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9997,
      "step": 26
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9234,
      "step": 27
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9679,
      "step": 28
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9442,
      "step": 29
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9708,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0319,
      "step": 31
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0278,
      "step": 32
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0284,
      "step": 33
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8464,
      "step": 34
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9889,
      "step": 35
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.3341,
      "step": 36
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0814,
      "step": 37
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9036,
      "step": 38
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1541,
      "step": 39
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1549,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0072,
      "step": 41
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0638,
      "step": 42
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0451,
      "step": 43
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.1216,
      "step": 44
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0063,
      "step": 45
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8003,
      "step": 46
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0479,
      "step": 47
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7849,
      "step": 48
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0547,
      "step": 49
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9508,
      "step": 50
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.033,
      "step": 51
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7773,
      "step": 52
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.887,
      "step": 53
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8354,
      "step": 54
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9288,
      "step": 55
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1684,
      "step": 56
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0553,
      "step": 57
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8693,
      "step": 58
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8337,
      "step": 59
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9965,
      "step": 60
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9892,
      "step": 61
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9308,
      "step": 62
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 63
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0411,
      "step": 64
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7696,
      "step": 65
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7961,
      "step": 66
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8911,
      "step": 67
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0532,
      "step": 68
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0721,
      "step": 69
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 70
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 71
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8478,
      "step": 72
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.02,
      "step": 73
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.2677,
      "step": 74
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9836,
      "step": 75
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1115,
      "step": 76
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1138,
      "step": 77
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8705,
      "step": 78
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9836,
      "step": 79
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.068,
      "step": 80
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.043,
      "step": 81
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0843,
      "step": 82
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.974,
      "step": 83
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9724,
      "step": 84
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8375,
      "step": 85
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1525,
      "step": 86
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9822,
      "step": 87
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0685,
      "step": 88
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.2392,
      "step": 89
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 90
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8264,
      "step": 91
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1202,
      "step": 92
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0901,
      "step": 93
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1147,
      "step": 94
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 95
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1012,
      "step": 96
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9222,
      "step": 97
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0704,
      "step": 98
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.1067,
      "step": 99
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8698,
      "step": 100
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.2041,
      "step": 101
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8054,
      "step": 102
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0614,
      "step": 103
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9635,
      "step": 104
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0252,
      "step": 105
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.962,
      "step": 106
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9954,
      "step": 107
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9895,
      "step": 108
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9363,
      "step": 109
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9849,
      "step": 110
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0117,
      "step": 111
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0286,
      "step": 112
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9685,
      "step": 113
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9529,
      "step": 114
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0492,
      "step": 115
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7994,
      "step": 116
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.757,
      "step": 117
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9638,
      "step": 118
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0389,
      "step": 119
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9388,
      "step": 120
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9704,
      "step": 121
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 122
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1912,
      "step": 123
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0663,
      "step": 124
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0062,
      "step": 125
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0792,
      "step": 126
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 127
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 128
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0731,
      "step": 129
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0924,
      "step": 130
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9461,
      "step": 131
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0235,
      "step": 132
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 133
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 134
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0812,
      "step": 135
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9319,
      "step": 136
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 137
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9424,
      "step": 138
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0331,
      "step": 139
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0472,
      "step": 140
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.998,
      "step": 141
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9577,
      "step": 142
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.1829,
      "step": 143
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.049,
      "step": 144
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9152,
      "step": 145
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0662,
      "step": 146
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9938,
      "step": 147
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9318,
      "step": 148
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.3438,
      "step": 149
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9698,
      "step": 150
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9604,
      "step": 151
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8512,
      "step": 152
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8416,
      "step": 153
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 154
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.883,
      "step": 155
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.2187,
      "step": 156
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 157
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0157,
      "step": 158
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0114,
      "step": 159
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9243,
      "step": 160
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 161
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1389,
      "step": 162
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9449,
      "step": 163
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0677,
      "step": 164
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0569,
      "step": 165
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9678,
      "step": 166
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0145,
      "step": 167
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0152,
      "step": 168
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.016,
      "step": 169
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 170
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.073,
      "step": 171
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8966,
      "step": 172
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8771,
      "step": 173
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.075,
      "step": 174
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9856,
      "step": 175
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 176
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0132,
      "step": 177
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8739,
      "step": 178
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.965,
      "step": 179
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 180
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0677,
      "step": 181
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0182,
      "step": 182
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7834,
      "step": 183
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1045,
      "step": 184
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8176,
      "step": 185
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7513,
      "step": 186
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7464,
      "step": 187
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0779,
      "step": 188
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9206,
      "step": 189
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8617,
      "step": 190
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0357,
      "step": 191
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0274,
      "step": 192
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9239,
      "step": 193
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0465,
      "step": 194
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9282,
      "step": 195
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0537,
      "step": 196
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1715,
      "step": 197
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 198
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0934,
      "step": 199
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8832,
      "step": 200
    },
    {
      "epoch": 0.06,
      "eval_loss": 0.9499853253364563,
      "eval_runtime": 90.0334,
      "eval_samples_per_second": 11.107,
      "eval_steps_per_second": 5.553,
      "step": 200
    },
    {
      "epoch": 0.06,
      "mmlu_eval_accuracy": 0.583239780938983,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "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": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "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": 0.3448275862068966,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.3941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.5736223915227088,
      "step": 200
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.925,
      "step": 201
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 202
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8345,
      "step": 203
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8449,
      "step": 204
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 205
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.1645,
      "step": 206
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0478,
      "step": 207
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0589,
      "step": 208
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9737,
      "step": 209
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0603,
      "step": 210
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0362,
      "step": 211
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9799,
      "step": 212
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8978,
      "step": 213
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9695,
      "step": 214
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9839,
      "step": 215
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0147,
      "step": 216
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9018,
      "step": 217
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1482,
      "step": 218
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9104,
      "step": 219
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0003,
      "step": 220
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8485,
      "step": 221
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8124,
      "step": 222
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9756,
      "step": 223
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.2614,
      "step": 224
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1372,
      "step": 225
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0422,
      "step": 226
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0229,
      "step": 227
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9819,
      "step": 228
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.917,
      "step": 229
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0221,
      "step": 230
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0757,
      "step": 231
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.1495,
      "step": 232
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.019,
      "step": 233
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9665,
      "step": 234
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0675,
      "step": 235
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8124,
      "step": 236
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0367,
      "step": 237
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.001,
      "step": 238
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7586,
      "step": 239
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0411,
      "step": 240
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8661,
      "step": 241
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9123,
      "step": 242
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.83,
      "step": 243
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8899,
      "step": 244
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0077,
      "step": 245
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9373,
      "step": 246
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1275,
      "step": 247
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7653,
      "step": 248
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 249
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1311,
      "step": 250
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8657,
      "step": 251
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1071,
      "step": 252
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 253
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8067,
      "step": 254
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1368,
      "step": 255
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8264,
      "step": 256
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9937,
      "step": 257
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0458,
      "step": 258
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9694,
      "step": 259
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0525,
      "step": 260
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8832,
      "step": 261
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8712,
      "step": 262
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0039,
      "step": 263
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1049,
      "step": 264
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7728,
      "step": 265
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8612,
      "step": 266
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9169,
      "step": 267
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9523,
      "step": 268
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8041,
      "step": 269
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8771,
      "step": 270
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.1153,
      "step": 271
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8168,
      "step": 272
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 273
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 274
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9758,
      "step": 275
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1362,
      "step": 276
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 277
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 278
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9961,
      "step": 279
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 280
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9617,
      "step": 281
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.3358,
      "step": 282
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.959,
      "step": 283
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9922,
      "step": 284
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0314,
      "step": 285
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9845,
      "step": 286
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.887,
      "step": 287
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0033,
      "step": 288
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0781,
      "step": 289
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1126,
      "step": 290
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9803,
      "step": 291
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0852,
      "step": 292
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 293
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9783,
      "step": 294
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9389,
      "step": 295
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0074,
      "step": 296
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7229,
      "step": 297
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0911,
      "step": 298
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.952,
      "step": 299
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0237,
      "step": 300
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.14,
      "step": 301
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7661,
      "step": 302
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9887,
      "step": 303
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.1378,
      "step": 304
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0622,
      "step": 305
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8306,
      "step": 306
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.71,
      "step": 307
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7812,
      "step": 308
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0251,
      "step": 309
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1483,
      "step": 310
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.041,
      "step": 311
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0637,
      "step": 312
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0051,
      "step": 313
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.021,
      "step": 314
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.03,
      "step": 315
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.2441,
      "step": 316
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8661,
      "step": 317
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9945,
      "step": 318
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1312,
      "step": 319
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.083,
      "step": 320
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9382,
      "step": 321
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9763,
      "step": 322
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9111,
      "step": 323
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0146,
      "step": 324
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.1616,
      "step": 325
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 326
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9968,
      "step": 327
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.022,
      "step": 328
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0467,
      "step": 329
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.188,
      "step": 330
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.028,
      "step": 331
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.982,
      "step": 332
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0669,
      "step": 333
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.875,
      "step": 334
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9985,
      "step": 335
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9164,
      "step": 336
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0613,
      "step": 337
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 338
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8803,
      "step": 339
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 340
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0023,
      "step": 341
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0508,
      "step": 342
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.83,
      "step": 343
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0438,
      "step": 344
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0424,
      "step": 345
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0642,
      "step": 346
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9585,
      "step": 347
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8884,
      "step": 348
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9328,
      "step": 349
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8091,
      "step": 350
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0103,
      "step": 351
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.2493,
      "step": 352
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9404,
      "step": 353
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9181,
      "step": 354
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8925,
      "step": 355
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 356
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9405,
      "step": 357
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9222,
      "step": 358
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0356,
      "step": 359
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0183,
      "step": 360
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 361
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9507,
      "step": 362
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9158,
      "step": 363
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 364
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8619,
      "step": 365
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1063,
      "step": 366
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 367
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0938,
      "step": 368
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 369
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8846,
      "step": 370
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.1139,
      "step": 371
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.871,
      "step": 372
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0847,
      "step": 373
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9088,
      "step": 374
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8681,
      "step": 375
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9996,
      "step": 376
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8648,
      "step": 377
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.85,
      "step": 378
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 379
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.2312,
      "step": 380
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9006,
      "step": 381
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8996,
      "step": 382
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1244,
      "step": 383
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9585,
      "step": 384
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 385
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0706,
      "step": 386
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1169,
      "step": 387
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8343,
      "step": 388
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9726,
      "step": 389
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9733,
      "step": 390
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9283,
      "step": 391
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0867,
      "step": 392
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8833,
      "step": 393
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1476,
      "step": 394
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9248,
      "step": 395
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9299,
      "step": 396
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7002,
      "step": 397
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9364,
      "step": 398
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9676,
      "step": 399
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0887,
      "step": 400
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.9337849020957947,
      "eval_runtime": 90.0647,
      "eval_samples_per_second": 11.103,
      "eval_steps_per_second": 5.552,
      "step": 400
    },
    {
      "epoch": 0.12,
      "mmlu_eval_accuracy": 0.5981609425223685,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.96,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.31,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.40588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.4234340800940213,
      "step": 400
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0558,
      "step": 401
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8911,
      "step": 402
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.1997,
      "step": 403
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0156,
      "step": 404
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9449,
      "step": 405
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0698,
      "step": 406
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0274,
      "step": 407
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9211,
      "step": 408
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8839,
      "step": 409
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0304,
      "step": 410
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0496,
      "step": 411
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9538,
      "step": 412
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9134,
      "step": 413
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8762,
      "step": 414
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0562,
      "step": 415
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9841,
      "step": 416
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8964,
      "step": 417
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8863,
      "step": 418
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.854,
      "step": 419
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9016,
      "step": 420
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.938,
      "step": 421
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9199,
      "step": 422
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9212,
      "step": 423
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8638,
      "step": 424
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0294,
      "step": 425
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9055,
      "step": 426
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9948,
      "step": 427
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0142,
      "step": 428
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8889,
      "step": 429
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8868,
      "step": 430
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9281,
      "step": 431
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1905,
      "step": 432
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0471,
      "step": 433
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8069,
      "step": 434
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9336,
      "step": 435
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 436
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9063,
      "step": 437
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8862,
      "step": 438
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9648,
      "step": 439
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9775,
      "step": 440
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9181,
      "step": 441
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0634,
      "step": 442
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9453,
      "step": 443
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7123,
      "step": 444
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0168,
      "step": 445
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8962,
      "step": 446
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.038,
      "step": 447
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8036,
      "step": 448
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9838,
      "step": 449
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6722,
      "step": 450
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9898,
      "step": 451
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.2471,
      "step": 452
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9111,
      "step": 453
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1423,
      "step": 454
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6952,
      "step": 455
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0062,
      "step": 456
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9,
      "step": 457
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9786,
      "step": 458
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0072,
      "step": 459
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0386,
      "step": 460
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.013,
      "step": 461
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9622,
      "step": 462
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9577,
      "step": 463
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9358,
      "step": 464
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.853,
      "step": 465
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.1448,
      "step": 466
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.0459,
      "step": 467
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9386,
      "step": 468
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.98,
      "step": 469
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.995,
      "step": 470
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9831,
      "step": 471
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0843,
      "step": 472
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9832,
      "step": 473
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9159,
      "step": 474
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8165,
      "step": 475
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 476
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9109,
      "step": 477
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9167,
      "step": 478
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 479
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0069,
      "step": 480
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9082,
      "step": 481
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9815,
      "step": 482
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8454,
      "step": 483
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8938,
      "step": 484
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9268,
      "step": 485
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 486
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.141,
      "step": 487
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 488
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 489
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9348,
      "step": 490
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9177,
      "step": 491
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.007,
      "step": 492
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1407,
      "step": 493
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 494
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9559,
      "step": 495
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8707,
      "step": 496
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0767,
      "step": 497
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8656,
      "step": 498
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9875,
      "step": 499
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8553,
      "step": 500
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.2432,
      "step": 501
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9875,
      "step": 502
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0564,
      "step": 503
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0728,
      "step": 504
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8802,
      "step": 505
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0073,
      "step": 506
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0117,
      "step": 507
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9091,
      "step": 508
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0593,
      "step": 509
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7527,
      "step": 510
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.884,
      "step": 511
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6971,
      "step": 512
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 513
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 514
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0796,
      "step": 515
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9112,
      "step": 516
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0223,
      "step": 517
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1011,
      "step": 518
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9859,
      "step": 519
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9615,
      "step": 520
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9893,
      "step": 521
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 522
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 523
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8454,
      "step": 524
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7769,
      "step": 525
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0176,
      "step": 526
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9641,
      "step": 527
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 528
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.125,
      "step": 529
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1411,
      "step": 530
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8413,
      "step": 531
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.1159,
      "step": 532
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0399,
      "step": 533
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9691,
      "step": 534
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.038,
      "step": 535
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 536
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9709,
      "step": 537
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9577,
      "step": 538
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9499,
      "step": 539
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9988,
      "step": 540
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9525,
      "step": 541
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 542
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9578,
      "step": 543
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0587,
      "step": 544
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0067,
      "step": 545
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.955,
      "step": 546
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0419,
      "step": 547
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1421,
      "step": 548
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9016,
      "step": 549
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1797,
      "step": 550
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8918,
      "step": 551
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.978,
      "step": 552
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 553
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9897,
      "step": 554
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0817,
      "step": 555
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9277,
      "step": 556
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7438,
      "step": 557
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 558
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9218,
      "step": 559
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9851,
      "step": 560
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9524,
      "step": 561
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8747,
      "step": 562
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9278,
      "step": 563
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9328,
      "step": 564
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9843,
      "step": 565
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7991,
      "step": 566
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.976,
      "step": 567
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8677,
      "step": 568
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.965,
      "step": 569
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8707,
      "step": 570
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0468,
      "step": 571
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9707,
      "step": 572
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9433,
      "step": 573
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0574,
      "step": 574
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0994,
      "step": 575
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8638,
      "step": 576
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8006,
      "step": 577
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 578
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8249,
      "step": 579
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0455,
      "step": 580
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9342,
      "step": 581
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 582
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 583
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1784,
      "step": 584
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9906,
      "step": 585
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 586
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9585,
      "step": 587
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 588
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9779,
      "step": 589
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.3717,
      "step": 590
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9579,
      "step": 591
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 592
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9953,
      "step": 593
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 594
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8172,
      "step": 595
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0521,
      "step": 596
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9594,
      "step": 597
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.988,
      "step": 598
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1481,
      "step": 599
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9953,
      "step": 600
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.9254454374313354,
      "eval_runtime": 89.8948,
      "eval_samples_per_second": 11.124,
      "eval_steps_per_second": 5.562,
      "step": 600
    },
    {
      "epoch": 0.19,
      "mmlu_eval_accuracy": 0.5822425888854337,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "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": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.37,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6521739130434783,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "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": 0.8421052631578947,
      "mmlu_loss": 1.3702809761227266,
      "step": 600
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9535,
      "step": 601
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0501,
      "step": 602
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9769,
      "step": 603
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9042,
      "step": 604
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8794,
      "step": 605
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9339,
      "step": 606
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9829,
      "step": 607
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8145,
      "step": 608
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9466,
      "step": 609
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8049,
      "step": 610
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1801,
      "step": 611
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9824,
      "step": 612
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9801,
      "step": 613
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9277,
      "step": 614
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0793,
      "step": 615
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9145,
      "step": 616
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9077,
      "step": 617
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9271,
      "step": 618
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8139,
      "step": 619
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1319,
      "step": 620
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0052,
      "step": 621
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.773,
      "step": 622
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0893,
      "step": 623
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9802,
      "step": 624
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8839,
      "step": 625
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 626
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.1832,
      "step": 627
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9879,
      "step": 628
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0714,
      "step": 629
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 630
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.999,
      "step": 631
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9246,
      "step": 632
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9175,
      "step": 633
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9249,
      "step": 634
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0007,
      "step": 635
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0112,
      "step": 636
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9751,
      "step": 637
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 638
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9026,
      "step": 639
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0026,
      "step": 640
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0469,
      "step": 641
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8268,
      "step": 642
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1288,
      "step": 643
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9288,
      "step": 644
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.995,
      "step": 645
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7639,
      "step": 646
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9523,
      "step": 647
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0845,
      "step": 648
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8971,
      "step": 649
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9431,
      "step": 650
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0578,
      "step": 651
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9411,
      "step": 652
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9653,
      "step": 653
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0232,
      "step": 654
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9372,
      "step": 655
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9865,
      "step": 656
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0774,
      "step": 657
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8414,
      "step": 658
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9289,
      "step": 659
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0405,
      "step": 660
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1074,
      "step": 661
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0345,
      "step": 662
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.013,
      "step": 663
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9731,
      "step": 664
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.197,
      "step": 665
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9602,
      "step": 666
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0854,
      "step": 667
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0112,
      "step": 668
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.058,
      "step": 669
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0063,
      "step": 670
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.852,
      "step": 671
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0859,
      "step": 672
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1411,
      "step": 673
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8524,
      "step": 674
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.809,
      "step": 675
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.014,
      "step": 676
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0507,
      "step": 677
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 678
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 679
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7784,
      "step": 680
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8716,
      "step": 681
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9756,
      "step": 682
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8185,
      "step": 683
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7187,
      "step": 684
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8743,
      "step": 685
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9534,
      "step": 686
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0375,
      "step": 687
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0418,
      "step": 688
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.1439,
      "step": 689
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0074,
      "step": 690
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9489,
      "step": 691
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0332,
      "step": 692
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7165,
      "step": 693
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9881,
      "step": 694
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.964,
      "step": 695
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0508,
      "step": 696
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9705,
      "step": 697
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0737,
      "step": 698
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9181,
      "step": 699
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9864,
      "step": 700
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8444,
      "step": 701
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8833,
      "step": 702
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9513,
      "step": 703
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0633,
      "step": 704
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9659,
      "step": 705
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9193,
      "step": 706
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9114,
      "step": 707
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7952,
      "step": 708
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 709
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 710
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0004,
      "step": 711
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9765,
      "step": 712
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0734,
      "step": 713
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9609,
      "step": 714
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0316,
      "step": 715
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8484,
      "step": 716
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9762,
      "step": 717
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0791,
      "step": 718
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.1466,
      "step": 719
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9606,
      "step": 720
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9582,
      "step": 721
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0114,
      "step": 722
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 723
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9429,
      "step": 724
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9376,
      "step": 725
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8704,
      "step": 726
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.985,
      "step": 727
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7996,
      "step": 728
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9308,
      "step": 729
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9079,
      "step": 730
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7765,
      "step": 731
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0763,
      "step": 732
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8579,
      "step": 733
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9353,
      "step": 734
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0666,
      "step": 735
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0547,
      "step": 736
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0175,
      "step": 737
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0049,
      "step": 738
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0471,
      "step": 739
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0488,
      "step": 740
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9471,
      "step": 741
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9221,
      "step": 742
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0767,
      "step": 743
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 744
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8834,
      "step": 745
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0069,
      "step": 746
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9915,
      "step": 747
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7124,
      "step": 748
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9043,
      "step": 749
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9733,
      "step": 750
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0334,
      "step": 751
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0691,
      "step": 752
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.827,
      "step": 753
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.2009,
      "step": 754
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9517,
      "step": 755
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0902,
      "step": 756
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 757
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9987,
      "step": 758
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9658,
      "step": 759
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9657,
      "step": 760
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7932,
      "step": 761
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8968,
      "step": 762
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9415,
      "step": 763
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9308,
      "step": 764
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7555,
      "step": 765
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9491,
      "step": 766
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.2491,
      "step": 767
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8267,
      "step": 768
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9367,
      "step": 769
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0347,
      "step": 770
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9795,
      "step": 771
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9948,
      "step": 772
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9551,
      "step": 773
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9297,
      "step": 774
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9208,
      "step": 775
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8997,
      "step": 776
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 777
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9908,
      "step": 778
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.2474,
      "step": 779
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8084,
      "step": 780
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0014,
      "step": 781
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.915,
      "step": 782
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0353,
      "step": 783
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9012,
      "step": 784
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9215,
      "step": 785
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9937,
      "step": 786
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9894,
      "step": 787
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9932,
      "step": 788
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1575,
      "step": 789
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9016,
      "step": 790
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0525,
      "step": 791
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.1081,
      "step": 792
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9134,
      "step": 793
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9649,
      "step": 794
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9309,
      "step": 795
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 796
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 797
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9005,
      "step": 798
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8663,
      "step": 799
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9574,
      "step": 800
    },
    {
      "epoch": 0.25,
      "eval_loss": 0.9275184273719788,
      "eval_runtime": 89.9113,
      "eval_samples_per_second": 11.122,
      "eval_steps_per_second": 5.561,
      "step": 800
    },
    {
      "epoch": 0.25,
      "mmlu_eval_accuracy": 0.5848031166258056,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "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": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.9,
      "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": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "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": 0.8421052631578947,
      "mmlu_loss": 1.4141516886242997,
      "step": 800
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9427,
      "step": 801
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0135,
      "step": 802
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8713,
      "step": 803
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0833,
      "step": 804
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0042,
      "step": 805
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.959,
      "step": 806
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.013,
      "step": 807
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9253,
      "step": 808
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0935,
      "step": 809
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9374,
      "step": 810
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8282,
      "step": 811
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8206,
      "step": 812
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.1563,
      "step": 813
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9319,
      "step": 814
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9049,
      "step": 815
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8766,
      "step": 816
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8062,
      "step": 817
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 818
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0272,
      "step": 819
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.934,
      "step": 820
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9123,
      "step": 821
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5495,
      "step": 822
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0141,
      "step": 823
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9341,
      "step": 824
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0332,
      "step": 825
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0712,
      "step": 826
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0292,
      "step": 827
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8024,
      "step": 828
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0508,
      "step": 829
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.963,
      "step": 830
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0002,
      "step": 831
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9335,
      "step": 832
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 833
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8892,
      "step": 834
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9723,
      "step": 835
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0575,
      "step": 836
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.05,
      "step": 837
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8846,
      "step": 838
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 839
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0433,
      "step": 840
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9837,
      "step": 841
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9802,
      "step": 842
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0836,
      "step": 843
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1496,
      "step": 844
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9185,
      "step": 845
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0289,
      "step": 846
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0887,
      "step": 847
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0574,
      "step": 848
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0082,
      "step": 849
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.987,
      "step": 850
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.1394,
      "step": 851
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9435,
      "step": 852
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0265,
      "step": 853
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.262,
      "step": 854
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7882,
      "step": 855
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9496,
      "step": 856
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9872,
      "step": 857
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9108,
      "step": 858
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0647,
      "step": 859
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9214,
      "step": 860
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0681,
      "step": 861
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.973,
      "step": 862
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0409,
      "step": 863
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0346,
      "step": 864
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0088,
      "step": 865
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0624,
      "step": 866
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7368,
      "step": 867
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9073,
      "step": 868
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8833,
      "step": 869
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9638,
      "step": 870
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0025,
      "step": 871
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9592,
      "step": 872
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.859,
      "step": 873
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8741,
      "step": 874
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8943,
      "step": 875
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0382,
      "step": 876
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8844,
      "step": 877
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 878
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9284,
      "step": 879
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9539,
      "step": 880
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 881
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.1924,
      "step": 882
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.963,
      "step": 883
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8643,
      "step": 884
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9516,
      "step": 885
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 886
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0312,
      "step": 887
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0812,
      "step": 888
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0821,
      "step": 889
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0811,
      "step": 890
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0041,
      "step": 891
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0108,
      "step": 892
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9336,
      "step": 893
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0712,
      "step": 894
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9484,
      "step": 895
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0285,
      "step": 896
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.041,
      "step": 897
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 898
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0873,
      "step": 899
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9262,
      "step": 900
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 901
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.171,
      "step": 902
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0179,
      "step": 903
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0213,
      "step": 904
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9468,
      "step": 905
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0403,
      "step": 906
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9136,
      "step": 907
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9405,
      "step": 908
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9723,
      "step": 909
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 910
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 911
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0357,
      "step": 912
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8705,
      "step": 913
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9708,
      "step": 914
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0517,
      "step": 915
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.1077,
      "step": 916
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.1631,
      "step": 917
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.159,
      "step": 918
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0905,
      "step": 919
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.1051,
      "step": 920
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9867,
      "step": 921
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8435,
      "step": 922
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8128,
      "step": 923
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6869,
      "step": 924
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9651,
      "step": 925
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 926
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7858,
      "step": 927
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8522,
      "step": 928
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9778,
      "step": 929
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0316,
      "step": 930
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0194,
      "step": 931
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 932
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0097,
      "step": 933
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9328,
      "step": 934
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9256,
      "step": 935
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 936
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0078,
      "step": 937
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.2338,
      "step": 938
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9385,
      "step": 939
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.1942,
      "step": 940
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 941
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.935,
      "step": 942
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 943
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 944
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.112,
      "step": 945
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9498,
      "step": 946
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 947
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.783,
      "step": 948
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0608,
      "step": 949
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0487,
      "step": 950
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8814,
      "step": 951
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.1653,
      "step": 952
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0693,
      "step": 953
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9636,
      "step": 954
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.875,
      "step": 955
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9218,
      "step": 956
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8526,
      "step": 957
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9431,
      "step": 958
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.061,
      "step": 959
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9771,
      "step": 960
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9243,
      "step": 961
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.1043,
      "step": 962
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9182,
      "step": 963
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.1294,
      "step": 964
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8865,
      "step": 965
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0268,
      "step": 966
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8041,
      "step": 967
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9429,
      "step": 968
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7829,
      "step": 969
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8987,
      "step": 970
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 971
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9317,
      "step": 972
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0334,
      "step": 973
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8731,
      "step": 974
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8646,
      "step": 975
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.896,
      "step": 976
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 977
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.026,
      "step": 978
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0152,
      "step": 979
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0456,
      "step": 980
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.1195,
      "step": 981
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9388,
      "step": 982
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9499,
      "step": 983
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9198,
      "step": 984
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8698,
      "step": 985
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9208,
      "step": 986
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0626,
      "step": 987
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9059,
      "step": 988
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8268,
      "step": 989
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8676,
      "step": 990
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 991
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9051,
      "step": 992
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8135,
      "step": 993
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9547,
      "step": 994
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9808,
      "step": 995
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1119,
      "step": 996
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8935,
      "step": 997
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1607,
      "step": 998
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0976,
      "step": 999
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1611,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.9281554818153381,
      "eval_runtime": 90.092,
      "eval_samples_per_second": 11.1,
      "eval_steps_per_second": 5.55,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "mmlu_eval_accuracy": 0.5852498134904298,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.96,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
      "mmlu_eval_accuracy_professional_psychology": 0.6666666666666666,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.1828467012930166,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8763,
      "step": 1001
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0074,
      "step": 1002
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9437,
      "step": 1003
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9731,
      "step": 1004
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0313,
      "step": 1005
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1213,
      "step": 1006
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8985,
      "step": 1007
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9706,
      "step": 1008
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9968,
      "step": 1009
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0232,
      "step": 1010
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0898,
      "step": 1011
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0339,
      "step": 1012
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 1013
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8873,
      "step": 1014
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9396,
      "step": 1015
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.026,
      "step": 1016
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0399,
      "step": 1017
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9343,
      "step": 1018
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9436,
      "step": 1019
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 1020
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0722,
      "step": 1021
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0077,
      "step": 1022
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8781,
      "step": 1023
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9901,
      "step": 1024
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0017,
      "step": 1025
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7056,
      "step": 1026
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9322,
      "step": 1027
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 1028
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8537,
      "step": 1029
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.1457,
      "step": 1030
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8933,
      "step": 1031
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9028,
      "step": 1032
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0536,
      "step": 1033
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.1442,
      "step": 1034
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9895,
      "step": 1035
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0998,
      "step": 1036
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8625,
      "step": 1037
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0222,
      "step": 1038
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9688,
      "step": 1039
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8813,
      "step": 1040
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 1041
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9258,
      "step": 1042
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8294,
      "step": 1043
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8176,
      "step": 1044
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9517,
      "step": 1045
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 1046
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0241,
      "step": 1047
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8452,
      "step": 1048
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 1049
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9219,
      "step": 1050
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8537,
      "step": 1051
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 1052
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9874,
      "step": 1053
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9991,
      "step": 1054
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9132,
      "step": 1055
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9159,
      "step": 1056
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0384,
      "step": 1057
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0183,
      "step": 1058
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 1059
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0092,
      "step": 1060
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9433,
      "step": 1061
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8205,
      "step": 1062
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9983,
      "step": 1063
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8982,
      "step": 1064
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 1065
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9036,
      "step": 1066
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0081,
      "step": 1067
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.949,
      "step": 1068
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0288,
      "step": 1069
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8627,
      "step": 1070
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8879,
      "step": 1071
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9761,
      "step": 1072
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0231,
      "step": 1073
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0934,
      "step": 1074
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.1262,
      "step": 1075
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8889,
      "step": 1076
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 1077
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 1078
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 1079
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8197,
      "step": 1080
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.95,
      "step": 1081
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8528,
      "step": 1082
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8824,
      "step": 1083
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.0092,
      "step": 1084
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8331,
      "step": 1085
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9745,
      "step": 1086
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9619,
      "step": 1087
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0569,
      "step": 1088
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0464,
      "step": 1089
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0884,
      "step": 1090
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.1204,
      "step": 1091
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0618,
      "step": 1092
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.1045,
      "step": 1093
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9596,
      "step": 1094
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0873,
      "step": 1095
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9049,
      "step": 1096
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9232,
      "step": 1097
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8941,
      "step": 1098
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.1156,
      "step": 1099
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9576,
      "step": 1100
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 1101
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0408,
      "step": 1102
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 1103
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9923,
      "step": 1104
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0737,
      "step": 1105
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0137,
      "step": 1106
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9673,
      "step": 1107
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8285,
      "step": 1108
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 1109
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9518,
      "step": 1110
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9824,
      "step": 1111
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0127,
      "step": 1112
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0113,
      "step": 1113
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 1114
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9248,
      "step": 1115
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8637,
      "step": 1116
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9103,
      "step": 1117
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.105,
      "step": 1118
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0067,
      "step": 1119
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.973,
      "step": 1120
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0713,
      "step": 1121
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0587,
      "step": 1122
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8998,
      "step": 1123
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9329,
      "step": 1124
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.1487,
      "step": 1125
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 1126
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8467,
      "step": 1127
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0595,
      "step": 1128
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.1683,
      "step": 1129
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9936,
      "step": 1130
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.986,
      "step": 1131
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0121,
      "step": 1132
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 1133
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.904,
      "step": 1134
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0766,
      "step": 1135
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8939,
      "step": 1136
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0652,
      "step": 1137
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9253,
      "step": 1138
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0411,
      "step": 1139
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 1140
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0199,
      "step": 1141
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6923,
      "step": 1142
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8144,
      "step": 1143
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8892,
      "step": 1144
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 1145
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8501,
      "step": 1146
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.035,
      "step": 1147
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 1148
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9245,
      "step": 1149
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0358,
      "step": 1150
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.1745,
      "step": 1151
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8288,
      "step": 1152
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.1436,
      "step": 1153
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8763,
      "step": 1154
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9678,
      "step": 1155
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9621,
      "step": 1156
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8683,
      "step": 1157
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0735,
      "step": 1158
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 1159
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0767,
      "step": 1160
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9052,
      "step": 1161
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0587,
      "step": 1162
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.1583,
      "step": 1163
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 1164
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8088,
      "step": 1165
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0476,
      "step": 1166
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9264,
      "step": 1167
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 1168
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0706,
      "step": 1169
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0106,
      "step": 1170
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 1171
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9861,
      "step": 1172
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.1965,
      "step": 1173
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 1174
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9704,
      "step": 1175
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9458,
      "step": 1176
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.2611,
      "step": 1177
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 1178
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9203,
      "step": 1179
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0566,
      "step": 1180
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0108,
      "step": 1181
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9439,
      "step": 1182
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9775,
      "step": 1183
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.1282,
      "step": 1184
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0194,
      "step": 1185
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9753,
      "step": 1186
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.1286,
      "step": 1187
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6684,
      "step": 1188
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9649,
      "step": 1189
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9801,
      "step": 1190
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0684,
      "step": 1191
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0935,
      "step": 1192
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8894,
      "step": 1193
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9705,
      "step": 1194
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9077,
      "step": 1195
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8026,
      "step": 1196
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0893,
      "step": 1197
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9805,
      "step": 1198
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 1199
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0744,
      "step": 1200
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.9176025986671448,
      "eval_runtime": 89.901,
      "eval_samples_per_second": 11.123,
      "eval_steps_per_second": 5.562,
      "step": 1200
    },
    {
      "epoch": 0.37,
      "mmlu_eval_accuracy": 0.596570042543892,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.5121951219512195,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.27,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
      "mmlu_loss": 1.3288292476309187,
      "step": 1200
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.1542,
      "step": 1201
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7307,
      "step": 1202
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8367,
      "step": 1203
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8915,
      "step": 1204
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0608,
      "step": 1205
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0443,
      "step": 1206
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9116,
      "step": 1207
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8792,
      "step": 1208
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8015,
      "step": 1209
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9604,
      "step": 1210
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 1211
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7791,
      "step": 1212
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0618,
      "step": 1213
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8964,
      "step": 1214
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9088,
      "step": 1215
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0457,
      "step": 1216
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.117,
      "step": 1217
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9835,
      "step": 1218
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9824,
      "step": 1219
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9666,
      "step": 1220
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9044,
      "step": 1221
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.94,
      "step": 1222
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7952,
      "step": 1223
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.095,
      "step": 1224
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0297,
      "step": 1225
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9501,
      "step": 1226
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9679,
      "step": 1227
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0423,
      "step": 1228
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.662,
      "step": 1229
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0075,
      "step": 1230
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8543,
      "step": 1231
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 1232
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9565,
      "step": 1233
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8641,
      "step": 1234
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.209,
      "step": 1235
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9951,
      "step": 1236
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0175,
      "step": 1237
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.1686,
      "step": 1238
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 1239
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8106,
      "step": 1240
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 1241
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 1242
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0552,
      "step": 1243
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8743,
      "step": 1244
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9729,
      "step": 1245
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0373,
      "step": 1246
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9799,
      "step": 1247
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 1248
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8489,
      "step": 1249
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0098,
      "step": 1250
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0982,
      "step": 1251
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8332,
      "step": 1252
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8909,
      "step": 1253
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8685,
      "step": 1254
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 1255
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8793,
      "step": 1256
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.1403,
      "step": 1257
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9152,
      "step": 1258
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8975,
      "step": 1259
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9908,
      "step": 1260
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9897,
      "step": 1261
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9151,
      "step": 1262
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9559,
      "step": 1263
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9446,
      "step": 1264
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.1279,
      "step": 1265
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9222,
      "step": 1266
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.041,
      "step": 1267
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7995,
      "step": 1268
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8859,
      "step": 1269
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 1270
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8929,
      "step": 1271
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9573,
      "step": 1272
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.936,
      "step": 1273
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.046,
      "step": 1274
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9503,
      "step": 1275
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0669,
      "step": 1276
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9959,
      "step": 1277
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 1278
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9905,
      "step": 1279
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0656,
      "step": 1280
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9665,
      "step": 1281
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8301,
      "step": 1282
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9603,
      "step": 1283
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0745,
      "step": 1284
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.148,
      "step": 1285
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7732,
      "step": 1286
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0181,
      "step": 1287
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1092,
      "step": 1288
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0049,
      "step": 1289
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9572,
      "step": 1290
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9615,
      "step": 1291
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8475,
      "step": 1292
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8134,
      "step": 1293
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9617,
      "step": 1294
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7477,
      "step": 1295
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8937,
      "step": 1296
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1268,
      "step": 1297
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.368,
      "step": 1298
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 1299
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.853,
      "step": 1300
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.989,
      "step": 1301
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1176,
      "step": 1302
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0556,
      "step": 1303
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1562,
      "step": 1304
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.143,
      "step": 1305
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1175,
      "step": 1306
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.2488,
      "step": 1307
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9911,
      "step": 1308
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9566,
      "step": 1309
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9312,
      "step": 1310
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 1311
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1687,
      "step": 1312
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9811,
      "step": 1313
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8658,
      "step": 1314
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0036,
      "step": 1315
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9236,
      "step": 1316
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.144,
      "step": 1317
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9423,
      "step": 1318
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8951,
      "step": 1319
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9624,
      "step": 1320
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9367,
      "step": 1321
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9051,
      "step": 1322
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.057,
      "step": 1323
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8954,
      "step": 1324
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.909,
      "step": 1325
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8227,
      "step": 1326
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.847,
      "step": 1327
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0988,
      "step": 1328
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0409,
      "step": 1329
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.988,
      "step": 1330
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8617,
      "step": 1331
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9526,
      "step": 1332
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8341,
      "step": 1333
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.1904,
      "step": 1334
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9688,
      "step": 1335
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8901,
      "step": 1336
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8503,
      "step": 1337
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 1338
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0139,
      "step": 1339
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9395,
      "step": 1340
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 1341
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0484,
      "step": 1342
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8469,
      "step": 1343
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.1205,
      "step": 1344
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.1018,
      "step": 1345
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0853,
      "step": 1346
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9743,
      "step": 1347
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9451,
      "step": 1348
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7954,
      "step": 1349
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8221,
      "step": 1350
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8608,
      "step": 1351
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.1369,
      "step": 1352
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8639,
      "step": 1353
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8781,
      "step": 1354
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0263,
      "step": 1355
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0397,
      "step": 1356
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8503,
      "step": 1357
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.109,
      "step": 1358
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8384,
      "step": 1359
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0194,
      "step": 1360
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0931,
      "step": 1361
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7706,
      "step": 1362
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.842,
      "step": 1363
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0501,
      "step": 1364
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 1365
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8942,
      "step": 1366
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0004,
      "step": 1367
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0165,
      "step": 1368
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9966,
      "step": 1369
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9131,
      "step": 1370
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8004,
      "step": 1371
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0551,
      "step": 1372
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9318,
      "step": 1373
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0716,
      "step": 1374
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8975,
      "step": 1375
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9131,
      "step": 1376
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8181,
      "step": 1377
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8736,
      "step": 1378
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0262,
      "step": 1379
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9019,
      "step": 1380
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7965,
      "step": 1381
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0962,
      "step": 1382
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7301,
      "step": 1383
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0309,
      "step": 1384
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 1385
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0708,
      "step": 1386
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0523,
      "step": 1387
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8399,
      "step": 1388
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9749,
      "step": 1389
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9429,
      "step": 1390
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8521,
      "step": 1391
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.1477,
      "step": 1392
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8762,
      "step": 1393
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9633,
      "step": 1394
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.802,
      "step": 1395
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 1396
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9616,
      "step": 1397
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9362,
      "step": 1398
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9515,
      "step": 1399
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9866,
      "step": 1400
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.9190293550491333,
      "eval_runtime": 90.015,
      "eval_samples_per_second": 11.109,
      "eval_steps_per_second": 5.555,
      "step": 1400
    },
    {
      "epoch": 0.43,
      "mmlu_eval_accuracy": 0.5960061249588275,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.6363636363636364,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "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": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
      "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": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.7941176470588235,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6376811594202898,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
      "mmlu_loss": 1.0498463314794997,
      "step": 1400
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8492,
      "step": 1401
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9779,
      "step": 1402
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 1403
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8128,
      "step": 1404
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.927,
      "step": 1405
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0042,
      "step": 1406
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8674,
      "step": 1407
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 1408
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8691,
      "step": 1409
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8636,
      "step": 1410
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8793,
      "step": 1411
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9418,
      "step": 1412
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.051,
      "step": 1413
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0027,
      "step": 1414
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8758,
      "step": 1415
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 1416
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.1128,
      "step": 1417
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8966,
      "step": 1418
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.2294,
      "step": 1419
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9881,
      "step": 1420
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9916,
      "step": 1421
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7884,
      "step": 1422
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0125,
      "step": 1423
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9087,
      "step": 1424
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.1595,
      "step": 1425
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.951,
      "step": 1426
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9853,
      "step": 1427
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9789,
      "step": 1428
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.873,
      "step": 1429
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 1430
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 1431
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8371,
      "step": 1432
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.1315,
      "step": 1433
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9347,
      "step": 1434
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7906,
      "step": 1435
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 1436
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8991,
      "step": 1437
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9691,
      "step": 1438
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.043,
      "step": 1439
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8923,
      "step": 1440
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.047,
      "step": 1441
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8913,
      "step": 1442
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9225,
      "step": 1443
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9654,
      "step": 1444
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 1445
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8179,
      "step": 1446
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8242,
      "step": 1447
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0486,
      "step": 1448
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7388,
      "step": 1449
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8256,
      "step": 1450
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.1483,
      "step": 1451
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 1452
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.982,
      "step": 1453
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8125,
      "step": 1454
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.988,
      "step": 1455
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8206,
      "step": 1456
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0767,
      "step": 1457
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9239,
      "step": 1458
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0935,
      "step": 1459
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0311,
      "step": 1460
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9147,
      "step": 1461
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9811,
      "step": 1462
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.915,
      "step": 1463
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0595,
      "step": 1464
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9444,
      "step": 1465
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 1466
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8666,
      "step": 1467
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8093,
      "step": 1468
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 1469
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9899,
      "step": 1470
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.1735,
      "step": 1471
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.028,
      "step": 1472
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8825,
      "step": 1473
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 1474
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.055,
      "step": 1475
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 1476
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0391,
      "step": 1477
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8742,
      "step": 1478
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0402,
      "step": 1479
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9174,
      "step": 1480
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 1481
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 1482
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0083,
      "step": 1483
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.931,
      "step": 1484
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9527,
      "step": 1485
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 1486
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.981,
      "step": 1487
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7719,
      "step": 1488
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.1041,
      "step": 1489
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9527,
      "step": 1490
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9694,
      "step": 1491
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6862,
      "step": 1492
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9522,
      "step": 1493
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0251,
      "step": 1494
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9824,
      "step": 1495
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 1496
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.1202,
      "step": 1497
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.826,
      "step": 1498
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9202,
      "step": 1499
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 1500
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9665,
      "step": 1501
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.729,
      "step": 1502
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9902,
      "step": 1503
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 1504
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.977,
      "step": 1505
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9534,
      "step": 1506
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0151,
      "step": 1507
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.05,
      "step": 1508
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8793,
      "step": 1509
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7912,
      "step": 1510
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9835,
      "step": 1511
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9542,
      "step": 1512
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 1513
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9876,
      "step": 1514
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 1515
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9205,
      "step": 1516
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 1517
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.1473,
      "step": 1518
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.0178,
      "step": 1519
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 1520
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9166,
      "step": 1521
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9647,
      "step": 1522
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8987,
      "step": 1523
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9982,
      "step": 1524
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8679,
      "step": 1525
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9512,
      "step": 1526
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8242,
      "step": 1527
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9852,
      "step": 1528
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9217,
      "step": 1529
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.968,
      "step": 1530
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.0709,
      "step": 1531
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9476,
      "step": 1532
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9636,
      "step": 1533
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 1534
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9178,
      "step": 1535
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8641,
      "step": 1536
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.0786,
      "step": 1537
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.0666,
      "step": 1538
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9642,
      "step": 1539
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.923,
      "step": 1540
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9326,
      "step": 1541
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.2009,
      "step": 1542
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9392,
      "step": 1543
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0525,
      "step": 1544
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8799,
      "step": 1545
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.2003,
      "step": 1546
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.1773,
      "step": 1547
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0113,
      "step": 1548
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9682,
      "step": 1549
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.023,
      "step": 1550
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 1551
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.815,
      "step": 1552
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8883,
      "step": 1553
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 1554
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8967,
      "step": 1555
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9608,
      "step": 1556
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 1557
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.2526,
      "step": 1558
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0192,
      "step": 1559
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.1287,
      "step": 1560
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0127,
      "step": 1561
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9929,
      "step": 1562
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8429,
      "step": 1563
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0176,
      "step": 1564
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.914,
      "step": 1565
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9642,
      "step": 1566
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9213,
      "step": 1567
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8745,
      "step": 1568
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7676,
      "step": 1569
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.1113,
      "step": 1570
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9522,
      "step": 1571
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0345,
      "step": 1572
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0055,
      "step": 1573
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9273,
      "step": 1574
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9929,
      "step": 1575
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6779,
      "step": 1576
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 1577
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9575,
      "step": 1578
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8854,
      "step": 1579
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 1580
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8388,
      "step": 1581
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.1907,
      "step": 1582
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0232,
      "step": 1583
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 1584
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7673,
      "step": 1585
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.019,
      "step": 1586
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 1587
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9602,
      "step": 1588
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 1589
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 1590
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9493,
      "step": 1591
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8281,
      "step": 1592
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0627,
      "step": 1593
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0024,
      "step": 1594
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.903,
      "step": 1595
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9033,
      "step": 1596
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0493,
      "step": 1597
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.1023,
      "step": 1598
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 1599
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8118,
      "step": 1600
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.9203125834465027,
      "eval_runtime": 89.9199,
      "eval_samples_per_second": 11.121,
      "eval_steps_per_second": 5.561,
      "step": 1600
    },
    {
      "epoch": 0.49,
      "mmlu_eval_accuracy": 0.5875064924534396,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4878048780487805,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.627906976744186,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.96,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6231884057971014,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1526734829727103,
      "step": 1600
    }
  ],
  "logging_steps": 1,
  "max_steps": 9723,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 1.7401756479356928e+17,
  "trial_name": null,
  "trial_params": null
}