{
  "best_metric": 0.7507924437522888,
  "best_model_checkpoint": "experts/mistralic-expert-15/checkpoint-2200",
  "epoch": 0.36147052782912303,
  "eval_steps": 200,
  "global_step": 2200,
  "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": 0.8743,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0512,
      "step": 2
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 1.0398,
      "step": 3
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8428,
      "step": 4
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8539,
      "step": 5
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6822,
      "step": 6
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7817,
      "step": 7
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8251,
      "step": 8
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.641,
      "step": 9
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9213,
      "step": 10
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 11
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 12
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7953,
      "step": 13
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6879,
      "step": 14
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 15
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7557,
      "step": 16
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6147,
      "step": 17
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7999,
      "step": 18
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.9288,
      "step": 19
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 20
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6837,
      "step": 21
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8564,
      "step": 22
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6611,
      "step": 23
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6956,
      "step": 24
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.5079,
      "step": 25
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.6142,
      "step": 26
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7226,
      "step": 27
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 28
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8852,
      "step": 29
    },
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8991,
      "step": 31
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 32
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6147,
      "step": 33
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9075,
      "step": 34
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 35
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8367,
      "step": 36
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 37
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9249,
      "step": 38
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 39
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8997,
      "step": 40
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 41
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8314,
      "step": 42
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 43
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 44
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8334,
      "step": 45
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9775,
      "step": 46
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7298,
      "step": 47
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9334,
      "step": 48
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 49
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8088,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 51
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.4508,
      "step": 52
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.4574,
      "step": 53
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 54
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 55
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5989,
      "step": 56
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 1.0519,
      "step": 57
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9321,
      "step": 58
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 59
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 60
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9561,
      "step": 61
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5952,
      "step": 62
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 63
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 64
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.4376,
      "step": 65
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6991,
      "step": 66
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6598,
      "step": 67
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 68
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7757,
      "step": 69
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 70
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8523,
      "step": 71
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6573,
      "step": 72
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5618,
      "step": 73
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6906,
      "step": 74
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 75
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8629,
      "step": 76
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6166,
      "step": 77
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8685,
      "step": 78
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8057,
      "step": 79
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8424,
      "step": 80
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9356,
      "step": 81
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8011,
      "step": 82
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 83
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 84
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 85
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.6061,
      "step": 86
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5743,
      "step": 87
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.609,
      "step": 88
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 89
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9335,
      "step": 90
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.678,
      "step": 91
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7919,
      "step": 92
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6743,
      "step": 93
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9325,
      "step": 94
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 95
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5631,
      "step": 96
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8618,
      "step": 97
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.683,
      "step": 98
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 99
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 100
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8628,
      "step": 101
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 102
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8235,
      "step": 103
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.916,
      "step": 104
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0372,
      "step": 105
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9372,
      "step": 106
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6709,
      "step": 107
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1999,
      "step": 108
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7315,
      "step": 109
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7147,
      "step": 110
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6933,
      "step": 111
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.844,
      "step": 112
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7172,
      "step": 113
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7477,
      "step": 114
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.835,
      "step": 115
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.0017,
      "step": 116
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5665,
      "step": 117
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6992,
      "step": 118
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9693,
      "step": 119
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 120
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8181,
      "step": 121
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 122
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.85,
      "step": 123
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6986,
      "step": 124
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "step": 125
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7909,
      "step": 126
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7143,
      "step": 127
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 128
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7625,
      "step": 129
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 130
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.3807,
      "step": 131
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.322,
      "step": 132
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.582,
      "step": 133
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 134
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8918,
      "step": 135
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 1.1352,
      "step": 136
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9177,
      "step": 137
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 138
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5509,
      "step": 139
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9554,
      "step": 140
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5094,
      "step": 141
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5929,
      "step": 142
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9733,
      "step": 143
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.5117,
      "step": 144
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6024,
      "step": 145
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6401,
      "step": 146
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.924,
      "step": 147
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.9399,
      "step": 148
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.972,
      "step": 149
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6407,
      "step": 150
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.8665,
      "step": 151
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 152
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6445,
      "step": 153
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7775,
      "step": 154
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7517,
      "step": 155
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9464,
      "step": 156
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5912,
      "step": 157
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6161,
      "step": 158
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8201,
      "step": 159
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6348,
      "step": 160
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0841,
      "step": 161
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 162
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9317,
      "step": 163
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7158,
      "step": 164
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9286,
      "step": 165
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9918,
      "step": 166
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 167
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7513,
      "step": 168
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8818,
      "step": 169
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8644,
      "step": 170
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7351,
      "step": 171
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7574,
      "step": 172
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6465,
      "step": 173
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 174
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8849,
      "step": 175
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0791,
      "step": 176
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8741,
      "step": 177
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7822,
      "step": 178
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8417,
      "step": 179
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 180
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0245,
      "step": 181
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 182
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6059,
      "step": 183
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9795,
      "step": 184
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9011,
      "step": 185
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6384,
      "step": 186
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7801,
      "step": 187
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 188
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8018,
      "step": 189
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 190
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7468,
      "step": 191
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8598,
      "step": 192
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 193
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 194
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 195
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8057,
      "step": 196
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 197
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 198
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8127,
      "step": 199
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9115,
      "step": 200
    },
    {
      "epoch": 0.03,
      "eval_loss": 0.7577090859413147,
      "eval_runtime": 131.6982,
      "eval_samples_per_second": 7.593,
      "eval_steps_per_second": 3.797,
      "step": 200
    },
    {
      "epoch": 0.03,
      "mmlu_eval_accuracy": 0.5965101027344765,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "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": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 1.0,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "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": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.7647058823529411,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.4,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1431999449708132,
      "step": 200
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 201
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4792,
      "step": 202
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9803,
      "step": 203
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9486,
      "step": 204
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 205
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 206
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4869,
      "step": 207
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6713,
      "step": 208
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.6963,
      "step": 209
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 210
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 1.0783,
      "step": 211
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.9185,
      "step": 212
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.812,
      "step": 213
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9379,
      "step": 214
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9187,
      "step": 215
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6081,
      "step": 216
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7071,
      "step": 217
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7934,
      "step": 218
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7151,
      "step": 219
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8808,
      "step": 220
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6993,
      "step": 221
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8453,
      "step": 222
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 223
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6957,
      "step": 224
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9296,
      "step": 225
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 226
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5761,
      "step": 227
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 228
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6516,
      "step": 229
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 230
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.3882,
      "step": 231
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0478,
      "step": 232
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 233
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7188,
      "step": 234
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9145,
      "step": 235
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8007,
      "step": 236
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 237
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5326,
      "step": 238
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.4012,
      "step": 239
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9156,
      "step": 240
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6651,
      "step": 241
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.9193,
      "step": 242
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7638,
      "step": 243
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 244
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7544,
      "step": 245
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 246
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 247
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 248
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8149,
      "step": 249
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.534,
      "step": 250
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7246,
      "step": 251
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7361,
      "step": 252
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7693,
      "step": 253
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8553,
      "step": 254
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6042,
      "step": 255
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6617,
      "step": 256
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 257
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 258
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8238,
      "step": 259
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 260
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 261
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 262
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0992,
      "step": 263
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7046,
      "step": 264
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4757,
      "step": 265
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4978,
      "step": 266
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8656,
      "step": 267
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8534,
      "step": 268
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 1.0157,
      "step": 269
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6066,
      "step": 270
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8877,
      "step": 271
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8856,
      "step": 272
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8631,
      "step": 273
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0294,
      "step": 274
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 275
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8029,
      "step": 276
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6784,
      "step": 277
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 278
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9855,
      "step": 279
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7293,
      "step": 280
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8036,
      "step": 281
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8581,
      "step": 282
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6941,
      "step": 283
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9493,
      "step": 284
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9682,
      "step": 285
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7554,
      "step": 286
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8585,
      "step": 287
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 288
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.814,
      "step": 289
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7303,
      "step": 290
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7244,
      "step": 291
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7552,
      "step": 292
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0223,
      "step": 293
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7623,
      "step": 294
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7383,
      "step": 295
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5089,
      "step": 296
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4504,
      "step": 297
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.2205,
      "step": 298
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9076,
      "step": 299
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9371,
      "step": 300
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1473,
      "step": 301
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8726,
      "step": 302
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7765,
      "step": 303
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7015,
      "step": 304
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 305
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.1318,
      "step": 306
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 307
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 308
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9389,
      "step": 309
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0817,
      "step": 310
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6492,
      "step": 311
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8057,
      "step": 312
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8286,
      "step": 313
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7106,
      "step": 314
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9558,
      "step": 315
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7318,
      "step": 316
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 1.0106,
      "step": 317
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.923,
      "step": 318
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8891,
      "step": 319
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7447,
      "step": 320
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7532,
      "step": 321
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 322
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6687,
      "step": 323
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6815,
      "step": 324
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9777,
      "step": 325
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7946,
      "step": 326
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 327
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8379,
      "step": 328
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 329
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6723,
      "step": 330
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6847,
      "step": 331
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.9873,
      "step": 332
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6398,
      "step": 333
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.6382,
      "step": 334
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5199,
      "step": 335
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6147,
      "step": 336
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 337
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7803,
      "step": 338
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8861,
      "step": 339
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 340
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6291,
      "step": 341
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 342
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8301,
      "step": 343
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 344
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6998,
      "step": 345
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6413,
      "step": 346
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6048,
      "step": 347
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 348
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 349
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8123,
      "step": 350
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7396,
      "step": 351
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6787,
      "step": 352
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7128,
      "step": 353
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7682,
      "step": 354
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4681,
      "step": 355
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 356
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6903,
      "step": 357
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.054,
      "step": 358
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0388,
      "step": 359
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 360
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9394,
      "step": 361
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8511,
      "step": 362
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 363
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8453,
      "step": 364
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8057,
      "step": 365
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7747,
      "step": 366
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7674,
      "step": 367
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7861,
      "step": 368
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7925,
      "step": 369
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.951,
      "step": 370
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9278,
      "step": 371
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.781,
      "step": 372
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6763,
      "step": 373
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5281,
      "step": 374
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5816,
      "step": 375
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9666,
      "step": 376
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 377
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.0995,
      "step": 378
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 1.2138,
      "step": 379
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8365,
      "step": 380
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6347,
      "step": 381
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7131,
      "step": 382
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 383
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.5475,
      "step": 384
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6053,
      "step": 385
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9546,
      "step": 386
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9732,
      "step": 387
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 388
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 389
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6867,
      "step": 390
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8615,
      "step": 391
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 392
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.678,
      "step": 393
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.881,
      "step": 394
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.435,
      "step": 395
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 396
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7088,
      "step": 397
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8921,
      "step": 398
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 399
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7468,
      "step": 400
    },
    {
      "epoch": 0.07,
      "eval_loss": 0.7549907565116882,
      "eval_runtime": 131.6241,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.799,
      "step": 400
    },
    {
      "epoch": 0.07,
      "mmlu_eval_accuracy": 0.5905395038007227,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.8125,
      "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "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": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 1.0,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "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": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1786961650412637,
      "step": 400
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 401
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 402
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 403
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9732,
      "step": 404
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8672,
      "step": 405
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.3649,
      "step": 406
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 407
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9357,
      "step": 408
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8635,
      "step": 409
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 410
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.3983,
      "step": 411
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7475,
      "step": 412
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6109,
      "step": 413
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6539,
      "step": 414
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6245,
      "step": 415
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8199,
      "step": 416
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9231,
      "step": 417
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.739,
      "step": 418
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9367,
      "step": 419
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.71,
      "step": 420
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 421
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 422
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9306,
      "step": 423
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 424
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7127,
      "step": 425
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6587,
      "step": 426
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 427
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7736,
      "step": 428
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.016,
      "step": 429
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7706,
      "step": 430
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9474,
      "step": 431
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7483,
      "step": 432
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.811,
      "step": 433
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7163,
      "step": 434
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9745,
      "step": 435
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7279,
      "step": 436
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9218,
      "step": 437
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8183,
      "step": 438
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8229,
      "step": 439
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 440
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7858,
      "step": 441
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0076,
      "step": 442
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.9241,
      "step": 443
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8994,
      "step": 444
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.438,
      "step": 445
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7214,
      "step": 446
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5063,
      "step": 447
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 448
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6496,
      "step": 449
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7387,
      "step": 450
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.824,
      "step": 451
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.5313,
      "step": 452
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7908,
      "step": 453
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8242,
      "step": 454
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8876,
      "step": 455
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 456
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 457
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.4918,
      "step": 458
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.855,
      "step": 459
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 460
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7973,
      "step": 461
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9598,
      "step": 462
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.791,
      "step": 463
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5249,
      "step": 464
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6198,
      "step": 465
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 466
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 467
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8863,
      "step": 468
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6965,
      "step": 469
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7057,
      "step": 470
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8994,
      "step": 471
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6775,
      "step": 472
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5412,
      "step": 473
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9663,
      "step": 474
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9682,
      "step": 475
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9535,
      "step": 476
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 477
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8014,
      "step": 478
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8144,
      "step": 479
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.713,
      "step": 480
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6225,
      "step": 481
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 482
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5586,
      "step": 483
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8996,
      "step": 484
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7745,
      "step": 485
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9241,
      "step": 486
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7356,
      "step": 487
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6449,
      "step": 488
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7228,
      "step": 489
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6462,
      "step": 490
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8052,
      "step": 491
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9312,
      "step": 492
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7832,
      "step": 493
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 494
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 495
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7811,
      "step": 496
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8659,
      "step": 497
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7348,
      "step": 498
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7742,
      "step": 499
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8397,
      "step": 500
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 501
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8492,
      "step": 502
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0731,
      "step": 503
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.669,
      "step": 504
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6464,
      "step": 505
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9044,
      "step": 506
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.174,
      "step": 507
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9606,
      "step": 508
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0218,
      "step": 509
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5801,
      "step": 510
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0433,
      "step": 511
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8027,
      "step": 512
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0204,
      "step": 513
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.9669,
      "step": 514
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5325,
      "step": 515
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.0595,
      "step": 516
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 1.002,
      "step": 517
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8445,
      "step": 518
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8647,
      "step": 519
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7309,
      "step": 520
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5321,
      "step": 521
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7862,
      "step": 522
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8507,
      "step": 523
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9764,
      "step": 524
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 525
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6066,
      "step": 526
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7368,
      "step": 527
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0567,
      "step": 528
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8779,
      "step": 529
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0253,
      "step": 530
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 531
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 532
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.5222,
      "step": 533
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 534
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 535
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8593,
      "step": 536
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8867,
      "step": 537
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6727,
      "step": 538
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6709,
      "step": 539
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8352,
      "step": 540
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.0596,
      "step": 541
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 542
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.57,
      "step": 543
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7398,
      "step": 544
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9338,
      "step": 545
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7498,
      "step": 546
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6023,
      "step": 547
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8713,
      "step": 548
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8121,
      "step": 549
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 1.2485,
      "step": 550
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8921,
      "step": 551
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9471,
      "step": 552
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4399,
      "step": 553
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7338,
      "step": 554
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8261,
      "step": 555
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7494,
      "step": 556
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7305,
      "step": 557
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9084,
      "step": 558
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 559
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 560
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 561
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8769,
      "step": 562
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7692,
      "step": 563
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9968,
      "step": 564
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6536,
      "step": 565
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7829,
      "step": 566
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7614,
      "step": 567
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8699,
      "step": 568
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8991,
      "step": 569
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 570
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7599,
      "step": 571
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6944,
      "step": 572
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.748,
      "step": 573
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6174,
      "step": 574
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 575
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8565,
      "step": 576
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.9175,
      "step": 577
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.578,
      "step": 578
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7151,
      "step": 579
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 580
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6315,
      "step": 581
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 582
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8364,
      "step": 583
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5496,
      "step": 584
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7928,
      "step": 585
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8136,
      "step": 586
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5404,
      "step": 587
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6769,
      "step": 588
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6331,
      "step": 589
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5495,
      "step": 590
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7055,
      "step": 591
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0395,
      "step": 592
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8604,
      "step": 593
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8628,
      "step": 594
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.4613,
      "step": 595
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 596
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9055,
      "step": 597
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5592,
      "step": 598
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0462,
      "step": 599
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7479,
      "step": 600
    },
    {
      "epoch": 0.1,
      "eval_loss": 0.7511235475540161,
      "eval_runtime": 131.6743,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 600
    },
    {
      "epoch": 0.1,
      "mmlu_eval_accuracy": 0.5854937013814044,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.8181818181818182,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "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": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.92,
      "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": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
      "mmlu_loss": 0.8422373449697351,
      "step": 600
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 601
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 602
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9892,
      "step": 603
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 604
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5368,
      "step": 605
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 606
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5438,
      "step": 607
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.5921,
      "step": 608
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8068,
      "step": 609
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8916,
      "step": 610
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 611
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9221,
      "step": 612
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8893,
      "step": 613
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 614
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 615
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7934,
      "step": 616
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9781,
      "step": 617
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8513,
      "step": 618
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7201,
      "step": 619
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9233,
      "step": 620
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6561,
      "step": 621
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 622
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8383,
      "step": 623
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8131,
      "step": 624
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7621,
      "step": 625
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.9067,
      "step": 626
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0717,
      "step": 627
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 628
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.0598,
      "step": 629
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8181,
      "step": 630
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8185,
      "step": 631
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 632
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6738,
      "step": 633
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 1.2803,
      "step": 634
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7685,
      "step": 635
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7091,
      "step": 636
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.779,
      "step": 637
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 638
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7769,
      "step": 639
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 640
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 641
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5228,
      "step": 642
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.3634,
      "step": 643
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8811,
      "step": 644
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9722,
      "step": 645
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 646
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7172,
      "step": 647
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0123,
      "step": 648
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.171,
      "step": 649
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8549,
      "step": 650
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9442,
      "step": 651
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8082,
      "step": 652
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 653
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7786,
      "step": 654
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5764,
      "step": 655
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5766,
      "step": 656
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.2254,
      "step": 657
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7426,
      "step": 658
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 659
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9842,
      "step": 660
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8422,
      "step": 661
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.785,
      "step": 662
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7621,
      "step": 663
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8803,
      "step": 664
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8239,
      "step": 665
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8327,
      "step": 666
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9772,
      "step": 667
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8029,
      "step": 668
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 669
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8794,
      "step": 670
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8821,
      "step": 671
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 672
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 673
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7127,
      "step": 674
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 675
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 676
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.9527,
      "step": 677
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8083,
      "step": 678
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4753,
      "step": 679
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 680
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5191,
      "step": 681
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 682
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6917,
      "step": 683
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6707,
      "step": 684
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6223,
      "step": 685
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7281,
      "step": 686
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8363,
      "step": 687
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8984,
      "step": 688
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 1.0086,
      "step": 689
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8176,
      "step": 690
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6743,
      "step": 691
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7653,
      "step": 692
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.908,
      "step": 693
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8402,
      "step": 694
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8425,
      "step": 695
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 696
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8191,
      "step": 697
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.5398,
      "step": 698
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6728,
      "step": 699
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8391,
      "step": 700
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7766,
      "step": 701
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6116,
      "step": 702
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9081,
      "step": 703
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6774,
      "step": 704
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8251,
      "step": 705
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9524,
      "step": 706
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.535,
      "step": 707
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6306,
      "step": 708
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 709
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9689,
      "step": 710
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9917,
      "step": 711
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 712
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.587,
      "step": 713
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5918,
      "step": 714
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 715
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 716
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 717
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 718
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7452,
      "step": 719
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 720
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8305,
      "step": 721
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 722
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 723
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5891,
      "step": 724
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.94,
      "step": 725
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7092,
      "step": 726
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7654,
      "step": 727
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 728
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 729
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5904,
      "step": 730
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8591,
      "step": 731
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9372,
      "step": 732
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8514,
      "step": 733
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8774,
      "step": 734
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 735
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 736
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8392,
      "step": 737
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 738
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 739
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8915,
      "step": 740
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7966,
      "step": 741
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 742
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 743
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 1.0216,
      "step": 744
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9432,
      "step": 745
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6548,
      "step": 746
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8015,
      "step": 747
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 748
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 749
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6494,
      "step": 750
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8614,
      "step": 751
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.948,
      "step": 752
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 753
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8588,
      "step": 754
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 755
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 756
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7218,
      "step": 757
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.9789,
      "step": 758
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6592,
      "step": 759
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.6305,
      "step": 760
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.866,
      "step": 761
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7438,
      "step": 762
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 763
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0191,
      "step": 764
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7692,
      "step": 765
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6723,
      "step": 766
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8188,
      "step": 767
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6793,
      "step": 768
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 769
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9137,
      "step": 770
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7096,
      "step": 771
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8164,
      "step": 772
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7262,
      "step": 773
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 774
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.2896,
      "step": 775
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8491,
      "step": 776
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 777
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8435,
      "step": 778
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7388,
      "step": 779
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.941,
      "step": 780
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8939,
      "step": 781
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 782
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 783
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 784
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 785
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5049,
      "step": 786
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7308,
      "step": 787
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7585,
      "step": 788
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8932,
      "step": 789
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0651,
      "step": 790
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 791
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 792
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6963,
      "step": 793
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9918,
      "step": 794
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9083,
      "step": 795
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9174,
      "step": 796
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8295,
      "step": 797
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.004,
      "step": 798
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 799
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.5399,
      "step": 800
    },
    {
      "epoch": 0.13,
      "eval_loss": 0.7548895478248596,
      "eval_runtime": 131.5977,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.799,
      "step": 800
    },
    {
      "epoch": 0.13,
      "mmlu_eval_accuracy": 0.5966805174146648,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6511627906976745,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "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": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.6774193548387096,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.7407407407407407,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
      "mmlu_loss": 0.9437568160868189,
      "step": 800
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 801
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7918,
      "step": 802
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 803
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8476,
      "step": 804
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7063,
      "step": 805
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8846,
      "step": 806
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 807
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.1473,
      "step": 808
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6126,
      "step": 809
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 810
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8065,
      "step": 811
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.9831,
      "step": 812
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7944,
      "step": 813
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6154,
      "step": 814
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7495,
      "step": 815
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 816
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6591,
      "step": 817
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.823,
      "step": 818
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7436,
      "step": 819
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.6991,
      "step": 820
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 1.0781,
      "step": 821
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 822
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9573,
      "step": 823
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 824
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9445,
      "step": 825
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8332,
      "step": 826
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 827
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.676,
      "step": 828
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 829
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5332,
      "step": 830
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6207,
      "step": 831
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9614,
      "step": 832
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8742,
      "step": 833
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9492,
      "step": 834
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7094,
      "step": 835
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.889,
      "step": 836
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 837
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.2038,
      "step": 838
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9353,
      "step": 839
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.864,
      "step": 840
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8414,
      "step": 841
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 842
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8535,
      "step": 843
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7975,
      "step": 844
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8976,
      "step": 845
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9244,
      "step": 846
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.797,
      "step": 847
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4536,
      "step": 848
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8585,
      "step": 849
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.457,
      "step": 850
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5783,
      "step": 851
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 852
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6791,
      "step": 853
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7751,
      "step": 854
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.2038,
      "step": 855
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6359,
      "step": 856
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 1.2393,
      "step": 857
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 858
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6458,
      "step": 859
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 860
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7906,
      "step": 861
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6737,
      "step": 862
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8032,
      "step": 863
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5549,
      "step": 864
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6854,
      "step": 865
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.879,
      "step": 866
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5739,
      "step": 867
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9231,
      "step": 868
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8842,
      "step": 869
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 870
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 871
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.697,
      "step": 872
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7969,
      "step": 873
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9802,
      "step": 874
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.8818,
      "step": 875
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 876
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5309,
      "step": 877
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7392,
      "step": 878
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9443,
      "step": 879
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9115,
      "step": 880
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.5311,
      "step": 881
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9616,
      "step": 882
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 883
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9842,
      "step": 884
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8423,
      "step": 885
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9754,
      "step": 886
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6544,
      "step": 887
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 888
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6975,
      "step": 889
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 890
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7876,
      "step": 891
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0582,
      "step": 892
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 893
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6425,
      "step": 894
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.815,
      "step": 895
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4249,
      "step": 896
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6071,
      "step": 897
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6565,
      "step": 898
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 899
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8633,
      "step": 900
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.4776,
      "step": 901
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6176,
      "step": 902
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5023,
      "step": 903
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.683,
      "step": 904
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6307,
      "step": 905
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 906
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6745,
      "step": 907
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8103,
      "step": 908
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 909
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.814,
      "step": 910
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7424,
      "step": 911
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 912
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8703,
      "step": 913
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9498,
      "step": 914
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.954,
      "step": 915
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5388,
      "step": 916
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7777,
      "step": 917
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8466,
      "step": 918
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8867,
      "step": 919
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6023,
      "step": 920
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.0075,
      "step": 921
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5475,
      "step": 922
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.692,
      "step": 923
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8253,
      "step": 924
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8064,
      "step": 925
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.1761,
      "step": 926
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9338,
      "step": 927
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8504,
      "step": 928
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 929
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 930
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 931
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6123,
      "step": 932
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8887,
      "step": 933
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8548,
      "step": 934
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 1.3875,
      "step": 935
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6125,
      "step": 936
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 937
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7575,
      "step": 938
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 939
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 940
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.652,
      "step": 941
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8225,
      "step": 942
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 943
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6233,
      "step": 944
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6494,
      "step": 945
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6741,
      "step": 946
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6992,
      "step": 947
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0143,
      "step": 948
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.2206,
      "step": 949
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6951,
      "step": 950
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9564,
      "step": 951
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7238,
      "step": 952
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8592,
      "step": 953
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 954
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7797,
      "step": 955
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7822,
      "step": 956
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 957
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 958
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9142,
      "step": 959
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 960
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6945,
      "step": 961
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6274,
      "step": 962
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 963
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8954,
      "step": 964
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 965
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 966
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6027,
      "step": 967
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5063,
      "step": 968
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8569,
      "step": 969
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.0186,
      "step": 970
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 971
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8562,
      "step": 972
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9192,
      "step": 973
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9584,
      "step": 974
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7057,
      "step": 975
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7363,
      "step": 976
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7962,
      "step": 977
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8045,
      "step": 978
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 979
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 980
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6995,
      "step": 981
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5048,
      "step": 982
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9541,
      "step": 983
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 984
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.506,
      "step": 985
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8758,
      "step": 986
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6352,
      "step": 987
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 988
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8874,
      "step": 989
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.812,
      "step": 990
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.499,
      "step": 991
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 992
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9514,
      "step": 993
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9723,
      "step": 994
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.9755,
      "step": 995
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8997,
      "step": 996
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.82,
      "step": 997
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8012,
      "step": 998
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5464,
      "step": 999
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.7555892467498779,
      "eval_runtime": 131.6346,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "mmlu_eval_accuracy": 0.5856789112224339,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.6875,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 1.0,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.39,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "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": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8947368421052632,
      "mmlu_loss": 1.176194295909635,
      "step": 1000
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8221,
      "step": 1001
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 1.3423,
      "step": 1002
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.6437,
      "step": 1003
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8111,
      "step": 1004
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6787,
      "step": 1005
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5613,
      "step": 1006
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9071,
      "step": 1007
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 1008
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5898,
      "step": 1009
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7674,
      "step": 1010
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 1011
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 1012
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 1013
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6385,
      "step": 1014
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0248,
      "step": 1015
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7654,
      "step": 1016
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 1017
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 1018
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8954,
      "step": 1019
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9026,
      "step": 1020
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9989,
      "step": 1021
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7809,
      "step": 1022
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5978,
      "step": 1023
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 1024
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6645,
      "step": 1025
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8464,
      "step": 1026
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7771,
      "step": 1027
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9945,
      "step": 1028
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9833,
      "step": 1029
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6665,
      "step": 1030
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7047,
      "step": 1031
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7872,
      "step": 1032
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9812,
      "step": 1033
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6762,
      "step": 1034
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8424,
      "step": 1035
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9057,
      "step": 1036
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 1037
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 1038
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6887,
      "step": 1039
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8873,
      "step": 1040
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7545,
      "step": 1041
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 1042
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8972,
      "step": 1043
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8649,
      "step": 1044
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8409,
      "step": 1045
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.0243,
      "step": 1046
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 1047
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9103,
      "step": 1048
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.669,
      "step": 1049
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 1050
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8883,
      "step": 1051
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9743,
      "step": 1052
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.685,
      "step": 1053
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 1054
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 1055
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 1.1629,
      "step": 1056
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 1057
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5441,
      "step": 1058
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7121,
      "step": 1059
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5908,
      "step": 1060
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 1061
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 1062
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8256,
      "step": 1063
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.5057,
      "step": 1064
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 1065
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 1066
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9437,
      "step": 1067
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.1681,
      "step": 1068
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5524,
      "step": 1069
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8754,
      "step": 1070
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7274,
      "step": 1071
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6221,
      "step": 1072
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9309,
      "step": 1073
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8359,
      "step": 1074
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0487,
      "step": 1075
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 1076
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 1077
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4848,
      "step": 1078
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6176,
      "step": 1079
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7986,
      "step": 1080
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6757,
      "step": 1081
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8007,
      "step": 1082
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 1083
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.932,
      "step": 1084
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 1.0875,
      "step": 1085
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7853,
      "step": 1086
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 1087
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.944,
      "step": 1088
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6463,
      "step": 1089
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8139,
      "step": 1090
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9296,
      "step": 1091
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7712,
      "step": 1092
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9416,
      "step": 1093
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 1094
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 1095
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.711,
      "step": 1096
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7038,
      "step": 1097
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 1098
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5564,
      "step": 1099
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9269,
      "step": 1100
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9102,
      "step": 1101
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.628,
      "step": 1102
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7816,
      "step": 1103
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7092,
      "step": 1104
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7987,
      "step": 1105
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7363,
      "step": 1106
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 1107
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 1108
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 1109
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7048,
      "step": 1110
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 1111
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 1112
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 1113
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 1114
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7026,
      "step": 1115
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5051,
      "step": 1116
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.9068,
      "step": 1117
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7855,
      "step": 1118
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 1119
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 1120
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8529,
      "step": 1121
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 1122
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.5733,
      "step": 1123
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 1124
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.4714,
      "step": 1125
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.911,
      "step": 1126
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8629,
      "step": 1127
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6223,
      "step": 1128
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 1129
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9176,
      "step": 1130
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9618,
      "step": 1131
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5051,
      "step": 1132
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0598,
      "step": 1133
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9089,
      "step": 1134
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.657,
      "step": 1135
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8222,
      "step": 1136
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8729,
      "step": 1137
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8369,
      "step": 1138
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 1139
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8162,
      "step": 1140
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 1141
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 1142
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7976,
      "step": 1143
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6839,
      "step": 1144
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7471,
      "step": 1145
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6854,
      "step": 1146
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5067,
      "step": 1147
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6942,
      "step": 1148
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6953,
      "step": 1149
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 1150
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9129,
      "step": 1151
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5026,
      "step": 1152
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8202,
      "step": 1153
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0068,
      "step": 1154
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8909,
      "step": 1155
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7483,
      "step": 1156
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8111,
      "step": 1157
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 1158
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5368,
      "step": 1159
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.943,
      "step": 1160
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7594,
      "step": 1161
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8482,
      "step": 1162
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0019,
      "step": 1163
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 1164
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8401,
      "step": 1165
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0795,
      "step": 1166
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5373,
      "step": 1167
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4947,
      "step": 1168
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6115,
      "step": 1169
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5907,
      "step": 1170
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 1171
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6941,
      "step": 1172
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8686,
      "step": 1173
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8994,
      "step": 1174
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6574,
      "step": 1175
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 1.0021,
      "step": 1176
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.751,
      "step": 1177
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8691,
      "step": 1178
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5775,
      "step": 1179
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7803,
      "step": 1180
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 1181
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8704,
      "step": 1182
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 1183
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 1184
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7895,
      "step": 1185
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 1186
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6107,
      "step": 1187
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8162,
      "step": 1188
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 1189
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5487,
      "step": 1190
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.897,
      "step": 1191
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9297,
      "step": 1192
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7531,
      "step": 1193
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7013,
      "step": 1194
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8751,
      "step": 1195
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 1196
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 1197
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6135,
      "step": 1198
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0381,
      "step": 1199
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0329,
      "step": 1200
    },
    {
      "epoch": 0.2,
      "eval_loss": 0.7524652481079102,
      "eval_runtime": 131.7121,
      "eval_samples_per_second": 7.592,
      "eval_steps_per_second": 3.796,
      "step": 1200
    },
    {
      "epoch": 0.2,
      "mmlu_eval_accuracy": 0.5742141326455935,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "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": 0.45454545454545453,
      "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": 0.4146341463414634,
      "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": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.88,
      "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": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.9090909090909091,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.2399911641996773,
      "step": 1200
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5216,
      "step": 1201
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.618,
      "step": 1202
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 1203
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 1204
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.4266,
      "step": 1205
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8449,
      "step": 1206
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7134,
      "step": 1207
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 1208
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9778,
      "step": 1209
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8887,
      "step": 1210
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5954,
      "step": 1211
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7968,
      "step": 1212
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.719,
      "step": 1213
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.1052,
      "step": 1214
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8811,
      "step": 1215
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8834,
      "step": 1216
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 1217
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6735,
      "step": 1218
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6044,
      "step": 1219
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 1220
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9047,
      "step": 1221
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0172,
      "step": 1222
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6626,
      "step": 1223
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 1224
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 1225
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8591,
      "step": 1226
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 1227
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 1228
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7013,
      "step": 1229
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 1230
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7474,
      "step": 1231
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 1232
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.931,
      "step": 1233
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 1234
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9315,
      "step": 1235
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5211,
      "step": 1236
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6216,
      "step": 1237
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7627,
      "step": 1238
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8407,
      "step": 1239
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 1240
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 1.0086,
      "step": 1241
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 1242
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8143,
      "step": 1243
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8281,
      "step": 1244
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 1245
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.5096,
      "step": 1246
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.671,
      "step": 1247
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4925,
      "step": 1248
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 1249
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9885,
      "step": 1250
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7908,
      "step": 1251
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8172,
      "step": 1252
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 1253
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6758,
      "step": 1254
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 1255
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8398,
      "step": 1256
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7218,
      "step": 1257
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6242,
      "step": 1258
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6917,
      "step": 1259
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9483,
      "step": 1260
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7747,
      "step": 1261
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6527,
      "step": 1262
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8877,
      "step": 1263
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9047,
      "step": 1264
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9139,
      "step": 1265
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7574,
      "step": 1266
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8861,
      "step": 1267
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 1268
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7459,
      "step": 1269
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7822,
      "step": 1270
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7384,
      "step": 1271
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7143,
      "step": 1272
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5266,
      "step": 1273
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9644,
      "step": 1274
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7135,
      "step": 1275
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8658,
      "step": 1276
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6821,
      "step": 1277
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9459,
      "step": 1278
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7442,
      "step": 1279
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.781,
      "step": 1280
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.64,
      "step": 1281
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6969,
      "step": 1282
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4252,
      "step": 1283
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 1284
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9488,
      "step": 1285
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0964,
      "step": 1286
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7516,
      "step": 1287
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6294,
      "step": 1288
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.931,
      "step": 1289
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 1290
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 1291
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8139,
      "step": 1292
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 1293
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 1294
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7511,
      "step": 1295
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6742,
      "step": 1296
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 1297
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7265,
      "step": 1298
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9514,
      "step": 1299
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.9468,
      "step": 1300
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7451,
      "step": 1301
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7945,
      "step": 1302
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.825,
      "step": 1303
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 1304
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 1.0714,
      "step": 1305
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7463,
      "step": 1306
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 1307
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7203,
      "step": 1308
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 1309
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7341,
      "step": 1310
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7748,
      "step": 1311
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7815,
      "step": 1312
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8286,
      "step": 1313
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9297,
      "step": 1314
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.029,
      "step": 1315
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 1316
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 1317
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7323,
      "step": 1318
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 1319
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9488,
      "step": 1320
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6787,
      "step": 1321
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 1322
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8983,
      "step": 1323
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 1324
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0014,
      "step": 1325
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7568,
      "step": 1326
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0171,
      "step": 1327
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.3873,
      "step": 1328
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7452,
      "step": 1329
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7642,
      "step": 1330
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.753,
      "step": 1331
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 1332
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 1333
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0902,
      "step": 1334
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8505,
      "step": 1335
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7669,
      "step": 1336
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8207,
      "step": 1337
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6588,
      "step": 1338
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8637,
      "step": 1339
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6975,
      "step": 1340
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8329,
      "step": 1341
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9193,
      "step": 1342
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.926,
      "step": 1343
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 1344
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9468,
      "step": 1345
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 1346
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8462,
      "step": 1347
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 1.0115,
      "step": 1348
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8523,
      "step": 1349
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6957,
      "step": 1350
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9172,
      "step": 1351
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5134,
      "step": 1352
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 1353
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7291,
      "step": 1354
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.873,
      "step": 1355
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 1356
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8432,
      "step": 1357
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.9966,
      "step": 1358
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.99,
      "step": 1359
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 1360
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8726,
      "step": 1361
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.5941,
      "step": 1362
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6808,
      "step": 1363
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6686,
      "step": 1364
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8719,
      "step": 1365
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6443,
      "step": 1366
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.6231,
      "step": 1367
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7072,
      "step": 1368
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 1369
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.897,
      "step": 1370
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7346,
      "step": 1371
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.776,
      "step": 1372
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 1373
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8826,
      "step": 1374
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9369,
      "step": 1375
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8684,
      "step": 1376
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6926,
      "step": 1377
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8416,
      "step": 1378
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 1379
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 1380
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.924,
      "step": 1381
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.2479,
      "step": 1382
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.685,
      "step": 1383
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 1384
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.484,
      "step": 1385
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7056,
      "step": 1386
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0267,
      "step": 1387
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8326,
      "step": 1388
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7695,
      "step": 1389
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8156,
      "step": 1390
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 1391
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0382,
      "step": 1392
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9356,
      "step": 1393
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7705,
      "step": 1394
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9324,
      "step": 1395
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1027,
      "step": 1396
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 1397
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7028,
      "step": 1398
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8509,
      "step": 1399
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7476,
      "step": 1400
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.7565351724624634,
      "eval_runtime": 131.6911,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 1400
    },
    {
      "epoch": 0.23,
      "mmlu_eval_accuracy": 0.5850023557597949,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7790697674418605,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.39,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.295089167574367,
      "step": 1400
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7247,
      "step": 1401
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7038,
      "step": 1402
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8196,
      "step": 1403
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7339,
      "step": 1404
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.625,
      "step": 1405
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 1406
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9877,
      "step": 1407
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8991,
      "step": 1408
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 1409
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 1410
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.0762,
      "step": 1411
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7993,
      "step": 1412
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 1413
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6561,
      "step": 1414
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7605,
      "step": 1415
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 1.1527,
      "step": 1416
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 1417
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6008,
      "step": 1418
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 1419
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.597,
      "step": 1420
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6108,
      "step": 1421
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9615,
      "step": 1422
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.9308,
      "step": 1423
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6633,
      "step": 1424
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 1425
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8532,
      "step": 1426
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 1427
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6888,
      "step": 1428
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.6228,
      "step": 1429
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7561,
      "step": 1430
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6594,
      "step": 1431
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7793,
      "step": 1432
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 1433
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0135,
      "step": 1434
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6814,
      "step": 1435
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9646,
      "step": 1436
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 1437
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 1438
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 1439
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9395,
      "step": 1440
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9995,
      "step": 1441
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6857,
      "step": 1442
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8066,
      "step": 1443
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7895,
      "step": 1444
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8392,
      "step": 1445
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5989,
      "step": 1446
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0545,
      "step": 1447
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9732,
      "step": 1448
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6197,
      "step": 1449
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9419,
      "step": 1450
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7909,
      "step": 1451
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8606,
      "step": 1452
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6496,
      "step": 1453
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8656,
      "step": 1454
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.823,
      "step": 1455
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7034,
      "step": 1456
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8816,
      "step": 1457
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 1458
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8636,
      "step": 1459
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8487,
      "step": 1460
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8579,
      "step": 1461
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6756,
      "step": 1462
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 1463
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7554,
      "step": 1464
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8105,
      "step": 1465
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7906,
      "step": 1466
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9217,
      "step": 1467
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 1468
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.69,
      "step": 1469
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6041,
      "step": 1470
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7895,
      "step": 1471
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.6045,
      "step": 1472
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.835,
      "step": 1473
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9231,
      "step": 1474
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 1475
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 1476
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9674,
      "step": 1477
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9376,
      "step": 1478
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9131,
      "step": 1479
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9775,
      "step": 1480
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.3288,
      "step": 1481
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.593,
      "step": 1482
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9766,
      "step": 1483
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7524,
      "step": 1484
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9152,
      "step": 1485
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7092,
      "step": 1486
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7326,
      "step": 1487
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7106,
      "step": 1488
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0405,
      "step": 1489
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 1.0875,
      "step": 1490
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.9197,
      "step": 1491
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4796,
      "step": 1492
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9769,
      "step": 1493
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5788,
      "step": 1494
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 1495
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.674,
      "step": 1496
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 1497
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4968,
      "step": 1498
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 1499
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 1500
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0425,
      "step": 1501
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4321,
      "step": 1502
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 1503
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8082,
      "step": 1504
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5295,
      "step": 1505
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 1506
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5363,
      "step": 1507
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7775,
      "step": 1508
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 1509
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.432,
      "step": 1510
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6316,
      "step": 1511
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8131,
      "step": 1512
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.97,
      "step": 1513
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6833,
      "step": 1514
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4482,
      "step": 1515
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6134,
      "step": 1516
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 1517
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 1518
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7455,
      "step": 1519
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8642,
      "step": 1520
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.809,
      "step": 1521
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8495,
      "step": 1522
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 1523
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9561,
      "step": 1524
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7723,
      "step": 1525
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7609,
      "step": 1526
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 1527
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 1528
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6784,
      "step": 1529
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 1530
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8004,
      "step": 1531
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.1055,
      "step": 1532
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.726,
      "step": 1533
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 1534
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7927,
      "step": 1535
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 1536
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4738,
      "step": 1537
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0319,
      "step": 1538
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9731,
      "step": 1539
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8639,
      "step": 1540
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.1349,
      "step": 1541
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.9366,
      "step": 1542
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 1543
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7024,
      "step": 1544
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0042,
      "step": 1545
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0105,
      "step": 1546
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8245,
      "step": 1547
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.5256,
      "step": 1548
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 1.0531,
      "step": 1549
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7971,
      "step": 1550
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.6856,
      "step": 1551
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8739,
      "step": 1552
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0163,
      "step": 1553
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6874,
      "step": 1554
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 1555
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8165,
      "step": 1556
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7431,
      "step": 1557
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7756,
      "step": 1558
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 1559
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5777,
      "step": 1560
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7756,
      "step": 1561
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7613,
      "step": 1562
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6612,
      "step": 1563
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7807,
      "step": 1564
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8527,
      "step": 1565
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5728,
      "step": 1566
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 1567
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 1568
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 1569
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7762,
      "step": 1570
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8781,
      "step": 1571
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9252,
      "step": 1572
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6258,
      "step": 1573
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5836,
      "step": 1574
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9178,
      "step": 1575
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5666,
      "step": 1576
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4898,
      "step": 1577
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9138,
      "step": 1578
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5762,
      "step": 1579
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8416,
      "step": 1580
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4505,
      "step": 1581
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 1582
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8154,
      "step": 1583
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7504,
      "step": 1584
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7091,
      "step": 1585
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 1586
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9147,
      "step": 1587
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6601,
      "step": 1588
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8604,
      "step": 1589
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6772,
      "step": 1590
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7868,
      "step": 1591
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8426,
      "step": 1592
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.9511,
      "step": 1593
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7094,
      "step": 1594
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0559,
      "step": 1595
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7736,
      "step": 1596
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5268,
      "step": 1597
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7912,
      "step": 1598
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.764,
      "step": 1599
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5238,
      "step": 1600
    },
    {
      "epoch": 0.26,
      "eval_loss": 0.7527549862861633,
      "eval_runtime": 131.6921,
      "eval_samples_per_second": 7.593,
      "eval_steps_per_second": 3.797,
      "step": 1600
    },
    {
      "epoch": 0.26,
      "mmlu_eval_accuracy": 0.5841085819573677,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.41379310344827586,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.8181818181818182,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.031481397175291,
      "step": 1600
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8259,
      "step": 1601
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7597,
      "step": 1602
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 1603
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.739,
      "step": 1604
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6859,
      "step": 1605
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 1606
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8233,
      "step": 1607
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0097,
      "step": 1608
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 1.0705,
      "step": 1609
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.8505,
      "step": 1610
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4597,
      "step": 1611
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6459,
      "step": 1612
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 1613
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6851,
      "step": 1614
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 1615
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6669,
      "step": 1616
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0592,
      "step": 1617
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6426,
      "step": 1618
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8768,
      "step": 1619
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 1620
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7126,
      "step": 1621
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6475,
      "step": 1622
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 1623
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9468,
      "step": 1624
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9996,
      "step": 1625
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6853,
      "step": 1626
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7665,
      "step": 1627
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6701,
      "step": 1628
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.679,
      "step": 1629
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4117,
      "step": 1630
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0549,
      "step": 1631
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.942,
      "step": 1632
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 1633
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 1634
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9103,
      "step": 1635
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5898,
      "step": 1636
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4918,
      "step": 1637
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6478,
      "step": 1638
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 1639
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8687,
      "step": 1640
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9022,
      "step": 1641
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 1642
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9609,
      "step": 1643
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7365,
      "step": 1644
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6662,
      "step": 1645
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 1646
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5832,
      "step": 1647
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 1648
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8406,
      "step": 1649
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 1650
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 1651
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0056,
      "step": 1652
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5359,
      "step": 1653
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 1654
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9781,
      "step": 1655
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6859,
      "step": 1656
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 1657
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5682,
      "step": 1658
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7168,
      "step": 1659
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 1.0028,
      "step": 1660
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 1661
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9537,
      "step": 1662
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.5807,
      "step": 1663
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.716,
      "step": 1664
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 1665
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 1666
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6855,
      "step": 1667
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6207,
      "step": 1668
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9699,
      "step": 1669
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.9061,
      "step": 1670
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6173,
      "step": 1671
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.6367,
      "step": 1672
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8444,
      "step": 1673
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8603,
      "step": 1674
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8036,
      "step": 1675
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9018,
      "step": 1676
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5705,
      "step": 1677
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6724,
      "step": 1678
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9261,
      "step": 1679
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 1680
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7732,
      "step": 1681
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.545,
      "step": 1682
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 1683
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 1684
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8507,
      "step": 1685
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9669,
      "step": 1686
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.0176,
      "step": 1687
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7758,
      "step": 1688
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9836,
      "step": 1689
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 1690
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 1691
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9225,
      "step": 1692
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 1693
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5515,
      "step": 1694
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 1695
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6856,
      "step": 1696
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5237,
      "step": 1697
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 1698
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8218,
      "step": 1699
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5861,
      "step": 1700
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 1701
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6571,
      "step": 1702
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8541,
      "step": 1703
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9976,
      "step": 1704
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9278,
      "step": 1705
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 1706
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 1707
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 1708
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6752,
      "step": 1709
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 1710
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.1475,
      "step": 1711
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.705,
      "step": 1712
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7031,
      "step": 1713
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6636,
      "step": 1714
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7738,
      "step": 1715
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5624,
      "step": 1716
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9593,
      "step": 1717
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7804,
      "step": 1718
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9136,
      "step": 1719
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 1.094,
      "step": 1720
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 1721
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6492,
      "step": 1722
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6876,
      "step": 1723
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7212,
      "step": 1724
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.5282,
      "step": 1725
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7868,
      "step": 1726
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9316,
      "step": 1727
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9305,
      "step": 1728
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.9418,
      "step": 1729
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7478,
      "step": 1730
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 1731
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7648,
      "step": 1732
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7998,
      "step": 1733
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 1734
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 1735
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8026,
      "step": 1736
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 1737
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8301,
      "step": 1738
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5024,
      "step": 1739
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8239,
      "step": 1740
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 1741
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0579,
      "step": 1742
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 1743
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5013,
      "step": 1744
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 1745
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 1746
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0144,
      "step": 1747
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.947,
      "step": 1748
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5192,
      "step": 1749
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9724,
      "step": 1750
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6316,
      "step": 1751
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0797,
      "step": 1752
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7627,
      "step": 1753
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8389,
      "step": 1754
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8919,
      "step": 1755
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6576,
      "step": 1756
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 1757
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8281,
      "step": 1758
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.831,
      "step": 1759
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7995,
      "step": 1760
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7144,
      "step": 1761
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0493,
      "step": 1762
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 1763
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7824,
      "step": 1764
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.762,
      "step": 1765
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8588,
      "step": 1766
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5117,
      "step": 1767
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7963,
      "step": 1768
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.797,
      "step": 1769
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 1770
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6041,
      "step": 1771
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9522,
      "step": 1772
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6672,
      "step": 1773
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 1774
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 1775
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 1776
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.6884,
      "step": 1777
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7675,
      "step": 1778
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5017,
      "step": 1779
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7173,
      "step": 1780
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9622,
      "step": 1781
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.5934,
      "step": 1782
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9902,
      "step": 1783
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 1784
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 1785
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8256,
      "step": 1786
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0602,
      "step": 1787
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9092,
      "step": 1788
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0037,
      "step": 1789
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 1.0462,
      "step": 1790
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8461,
      "step": 1791
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8195,
      "step": 1792
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.885,
      "step": 1793
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.4779,
      "step": 1794
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.9366,
      "step": 1795
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8306,
      "step": 1796
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7913,
      "step": 1797
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8551,
      "step": 1798
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7799,
      "step": 1799
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9489,
      "step": 1800
    },
    {
      "epoch": 0.3,
      "eval_loss": 0.7564910650253296,
      "eval_runtime": 131.6705,
      "eval_samples_per_second": 7.595,
      "eval_steps_per_second": 3.797,
      "step": 1800
    },
    {
      "epoch": 0.3,
      "mmlu_eval_accuracy": 0.5870383762445152,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.6363636363636364,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163,
      "mmlu_eval_accuracy_moral_disputes": 0.631578947368421,
      "mmlu_eval_accuracy_moral_scenarios": 0.42,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.9473684210526315,
      "mmlu_loss": 1.0342699467707863,
      "step": 1800
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8336,
      "step": 1801
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9158,
      "step": 1802
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6537,
      "step": 1803
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 1804
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.4424,
      "step": 1805
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8845,
      "step": 1806
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0182,
      "step": 1807
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7668,
      "step": 1808
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8308,
      "step": 1809
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6782,
      "step": 1810
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8811,
      "step": 1811
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 1812
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6867,
      "step": 1813
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 1814
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5932,
      "step": 1815
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.763,
      "step": 1816
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.3666,
      "step": 1817
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 1818
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.1761,
      "step": 1819
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 1820
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9538,
      "step": 1821
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6732,
      "step": 1822
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7191,
      "step": 1823
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7516,
      "step": 1824
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 1825
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8245,
      "step": 1826
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7211,
      "step": 1827
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8983,
      "step": 1828
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9165,
      "step": 1829
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8816,
      "step": 1830
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6977,
      "step": 1831
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5939,
      "step": 1832
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 1833
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5445,
      "step": 1834
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.814,
      "step": 1835
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.83,
      "step": 1836
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0397,
      "step": 1837
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5553,
      "step": 1838
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 1839
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8618,
      "step": 1840
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8449,
      "step": 1841
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8293,
      "step": 1842
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.7413,
      "step": 1843
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8758,
      "step": 1844
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8831,
      "step": 1845
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.578,
      "step": 1846
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8054,
      "step": 1847
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9829,
      "step": 1848
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8205,
      "step": 1849
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 1850
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 1.0171,
      "step": 1851
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8148,
      "step": 1852
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.5911,
      "step": 1853
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.9356,
      "step": 1854
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.6723,
      "step": 1855
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 1856
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9787,
      "step": 1857
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.0051,
      "step": 1858
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.4867,
      "step": 1859
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1325,
      "step": 1860
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.1187,
      "step": 1861
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8395,
      "step": 1862
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9261,
      "step": 1863
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8604,
      "step": 1864
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 1865
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 1866
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 1.2542,
      "step": 1867
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 1868
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6661,
      "step": 1869
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 1870
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7274,
      "step": 1871
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7226,
      "step": 1872
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6299,
      "step": 1873
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8366,
      "step": 1874
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 1875
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 1876
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.5952,
      "step": 1877
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8474,
      "step": 1878
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 1879
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 1880
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.856,
      "step": 1881
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.985,
      "step": 1882
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8828,
      "step": 1883
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7644,
      "step": 1884
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7251,
      "step": 1885
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 1886
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7388,
      "step": 1887
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6078,
      "step": 1888
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8609,
      "step": 1889
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.739,
      "step": 1890
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 1891
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.3554,
      "step": 1892
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9325,
      "step": 1893
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.608,
      "step": 1894
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7599,
      "step": 1895
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7637,
      "step": 1896
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 1897
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9611,
      "step": 1898
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.792,
      "step": 1899
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8939,
      "step": 1900
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 1901
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8245,
      "step": 1902
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9334,
      "step": 1903
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 1904
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7117,
      "step": 1905
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 1906
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 1907
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9798,
      "step": 1908
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 1909
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9791,
      "step": 1910
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7955,
      "step": 1911
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7784,
      "step": 1912
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8823,
      "step": 1913
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6896,
      "step": 1914
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.9004,
      "step": 1915
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.6769,
      "step": 1916
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 1917
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8832,
      "step": 1918
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.3216,
      "step": 1919
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8418,
      "step": 1920
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 1921
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9854,
      "step": 1922
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6857,
      "step": 1923
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 1924
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7502,
      "step": 1925
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8959,
      "step": 1926
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8317,
      "step": 1927
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.761,
      "step": 1928
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7882,
      "step": 1929
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8495,
      "step": 1930
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7368,
      "step": 1931
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.657,
      "step": 1932
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8864,
      "step": 1933
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 1934
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.5665,
      "step": 1935
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6528,
      "step": 1936
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.4354,
      "step": 1937
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 1938
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 1939
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 1940
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.4796,
      "step": 1941
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7293,
      "step": 1942
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8563,
      "step": 1943
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0438,
      "step": 1944
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 1945
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 1946
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 1947
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6799,
      "step": 1948
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.2362,
      "step": 1949
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9167,
      "step": 1950
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7315,
      "step": 1951
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6957,
      "step": 1952
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6972,
      "step": 1953
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 1.0928,
      "step": 1954
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8077,
      "step": 1955
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9741,
      "step": 1956
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6326,
      "step": 1957
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6168,
      "step": 1958
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8294,
      "step": 1959
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8041,
      "step": 1960
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7237,
      "step": 1961
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7619,
      "step": 1962
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6153,
      "step": 1963
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 1964
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8367,
      "step": 1965
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.9133,
      "step": 1966
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8438,
      "step": 1967
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.6413,
      "step": 1968
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8013,
      "step": 1969
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.761,
      "step": 1970
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7426,
      "step": 1971
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8,
      "step": 1972
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.7121,
      "step": 1973
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8288,
      "step": 1974
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 1975
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8884,
      "step": 1976
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 1977
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8072,
      "step": 1978
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.4938,
      "step": 1979
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6872,
      "step": 1980
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9711,
      "step": 1981
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.974,
      "step": 1982
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8756,
      "step": 1983
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8436,
      "step": 1984
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6283,
      "step": 1985
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 1986
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 1987
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 1988
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9639,
      "step": 1989
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7354,
      "step": 1990
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 1991
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9131,
      "step": 1992
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.4003,
      "step": 1993
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.914,
      "step": 1994
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.4672,
      "step": 1995
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7574,
      "step": 1996
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7151,
      "step": 1997
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 1998
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 1999
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7852,
      "step": 2000
    },
    {
      "epoch": 0.33,
      "eval_loss": 0.752848744392395,
      "eval_runtime": 131.602,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.799,
      "step": 2000
    },
    {
      "epoch": 0.33,
      "mmlu_eval_accuracy": 0.5691849760946832,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6046511627906976,
      "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": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "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": 1.0,
      "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": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.43,
      "mmlu_eval_accuracy_nutrition": 0.7878787878787878,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1752762202727889,
      "step": 2000
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 2001
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.2648,
      "step": 2002
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 2003
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8376,
      "step": 2004
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.702,
      "step": 2005
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8086,
      "step": 2006
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8528,
      "step": 2007
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6656,
      "step": 2008
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8634,
      "step": 2009
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8844,
      "step": 2010
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 2011
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9794,
      "step": 2012
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.839,
      "step": 2013
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 2014
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7027,
      "step": 2015
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7342,
      "step": 2016
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6352,
      "step": 2017
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 2018
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.4453,
      "step": 2019
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 2020
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8115,
      "step": 2021
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.893,
      "step": 2022
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.789,
      "step": 2023
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 2024
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6548,
      "step": 2025
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 1.006,
      "step": 2026
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7132,
      "step": 2027
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 2028
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6992,
      "step": 2029
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.5689,
      "step": 2030
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6236,
      "step": 2031
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8304,
      "step": 2032
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.9352,
      "step": 2033
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8443,
      "step": 2034
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6599,
      "step": 2035
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8409,
      "step": 2036
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.6997,
      "step": 2037
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 2038
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9681,
      "step": 2039
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8544,
      "step": 2040
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9446,
      "step": 2041
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 2042
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0545,
      "step": 2043
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 2044
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7428,
      "step": 2045
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8388,
      "step": 2046
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7171,
      "step": 2047
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.646,
      "step": 2048
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8184,
      "step": 2049
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5731,
      "step": 2050
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8029,
      "step": 2051
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.98,
      "step": 2052
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5321,
      "step": 2053
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7811,
      "step": 2054
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0171,
      "step": 2055
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.753,
      "step": 2056
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9288,
      "step": 2057
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7526,
      "step": 2058
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 2059
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8702,
      "step": 2060
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8068,
      "step": 2061
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 2062
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 2063
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8021,
      "step": 2064
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7745,
      "step": 2065
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5975,
      "step": 2066
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5209,
      "step": 2067
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.9128,
      "step": 2068
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5747,
      "step": 2069
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5494,
      "step": 2070
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8932,
      "step": 2071
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5441,
      "step": 2072
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5504,
      "step": 2073
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5092,
      "step": 2074
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.1967,
      "step": 2075
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.694,
      "step": 2076
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 2077
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6906,
      "step": 2078
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 2079
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8694,
      "step": 2080
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6817,
      "step": 2081
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5004,
      "step": 2082
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7916,
      "step": 2083
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8507,
      "step": 2084
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5931,
      "step": 2085
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 2086
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 1.0013,
      "step": 2087
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 2088
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7849,
      "step": 2089
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6948,
      "step": 2090
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.488,
      "step": 2091
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 2092
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8818,
      "step": 2093
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 2094
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.889,
      "step": 2095
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.5147,
      "step": 2096
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.6197,
      "step": 2097
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7885,
      "step": 2098
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 2099
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7704,
      "step": 2100
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 2101
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.4656,
      "step": 2102
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 2103
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7891,
      "step": 2104
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 2105
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7552,
      "step": 2106
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0138,
      "step": 2107
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 2108
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0103,
      "step": 2109
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.851,
      "step": 2110
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 2111
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9706,
      "step": 2112
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5157,
      "step": 2113
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7643,
      "step": 2114
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7947,
      "step": 2115
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8266,
      "step": 2116
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5799,
      "step": 2117
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8642,
      "step": 2118
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8729,
      "step": 2119
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 2120
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7454,
      "step": 2121
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 2122
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 2123
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.524,
      "step": 2124
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7295,
      "step": 2125
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 2126
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 2127
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.623,
      "step": 2128
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 2129
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 2130
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8399,
      "step": 2131
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 2132
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8489,
      "step": 2133
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8093,
      "step": 2134
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7857,
      "step": 2135
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 2136
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.837,
      "step": 2137
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.64,
      "step": 2138
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5885,
      "step": 2139
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 2140
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8894,
      "step": 2141
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9435,
      "step": 2142
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8138,
      "step": 2143
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9148,
      "step": 2144
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8558,
      "step": 2145
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.4714,
      "step": 2146
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 2147
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6831,
      "step": 2148
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 2149
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.6415,
      "step": 2150
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5021,
      "step": 2151
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 2152
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8857,
      "step": 2153
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 2154
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9997,
      "step": 2155
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.5424,
      "step": 2156
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.0604,
      "step": 2157
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.9917,
      "step": 2158
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 1.1453,
      "step": 2159
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 2160
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 2161
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7995,
      "step": 2162
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.2655,
      "step": 2163
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7892,
      "step": 2164
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6177,
      "step": 2165
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 2166
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.4969,
      "step": 2167
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.5629,
      "step": 2168
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 2169
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8582,
      "step": 2170
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9504,
      "step": 2171
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6852,
      "step": 2172
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7765,
      "step": 2173
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 2174
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9861,
      "step": 2175
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6297,
      "step": 2176
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0347,
      "step": 2177
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.778,
      "step": 2178
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7029,
      "step": 2179
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6162,
      "step": 2180
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8231,
      "step": 2181
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.763,
      "step": 2182
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6419,
      "step": 2183
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.607,
      "step": 2184
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6823,
      "step": 2185
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7794,
      "step": 2186
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 2187
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9136,
      "step": 2188
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.4843,
      "step": 2189
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 2190
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0224,
      "step": 2191
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9855,
      "step": 2192
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.5692,
      "step": 2193
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6979,
      "step": 2194
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.5271,
      "step": 2195
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9303,
      "step": 2196
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.929,
      "step": 2197
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 2198
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 2199
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8963,
      "step": 2200
    },
    {
      "epoch": 0.36,
      "eval_loss": 0.7507924437522888,
      "eval_runtime": 131.5866,
      "eval_samples_per_second": 7.6,
      "eval_steps_per_second": 3.8,
      "step": 2200
    },
    {
      "epoch": 0.36,
      "mmlu_eval_accuracy": 0.5536695878813044,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.71875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.7714285714285715,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "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": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1485521083323202,
      "step": 2200
    }
  ],
  "logging_steps": 1,
  "max_steps": 18258,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 4.551301043799982e+17,
  "trial_name": null,
  "trial_params": null
}