{
  "best_metric": 0.749654233455658,
  "best_model_checkpoint": "experts/mistralic-expert-15/checkpoint-3400",
  "epoch": 2.3002669952762376,
  "eval_steps": 200,
  "global_step": 14000,
  "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
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.5828,
      "step": 2201
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 2202
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7077,
      "step": 2203
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 2204
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6926,
      "step": 2205
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 2206
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8934,
      "step": 2207
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8543,
      "step": 2208
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 2209
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 2210
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8518,
      "step": 2211
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9768,
      "step": 2212
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.9592,
      "step": 2213
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.1021,
      "step": 2214
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0696,
      "step": 2215
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8984,
      "step": 2216
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8487,
      "step": 2217
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.6364,
      "step": 2218
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 2219
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 1.0047,
      "step": 2220
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7262,
      "step": 2221
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8957,
      "step": 2222
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.641,
      "step": 2223
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.5702,
      "step": 2224
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.764,
      "step": 2225
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7372,
      "step": 2226
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.754,
      "step": 2227
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8521,
      "step": 2228
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.1058,
      "step": 2229
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8239,
      "step": 2230
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0771,
      "step": 2231
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6587,
      "step": 2232
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7997,
      "step": 2233
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0849,
      "step": 2234
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.5827,
      "step": 2235
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 2236
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6897,
      "step": 2237
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7522,
      "step": 2238
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7877,
      "step": 2239
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.805,
      "step": 2240
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.5968,
      "step": 2241
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8151,
      "step": 2242
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 2243
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 2244
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8798,
      "step": 2245
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8186,
      "step": 2246
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 2247
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8388,
      "step": 2248
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8149,
      "step": 2249
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7716,
      "step": 2250
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 2251
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 2252
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7596,
      "step": 2253
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9931,
      "step": 2254
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.981,
      "step": 2255
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 2256
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 2257
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6767,
      "step": 2258
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.4725,
      "step": 2259
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 2260
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8727,
      "step": 2261
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.5604,
      "step": 2262
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 2263
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7527,
      "step": 2264
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.779,
      "step": 2265
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.943,
      "step": 2266
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7045,
      "step": 2267
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7053,
      "step": 2268
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9895,
      "step": 2269
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 2270
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 2271
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8361,
      "step": 2272
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 2273
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.646,
      "step": 2274
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 1.0571,
      "step": 2275
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.648,
      "step": 2276
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.7741,
      "step": 2277
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.863,
      "step": 2278
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.6895,
      "step": 2279
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8982,
      "step": 2280
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9069,
      "step": 2281
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.9023,
      "step": 2282
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.559,
      "step": 2283
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7736,
      "step": 2284
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6028,
      "step": 2285
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.3765,
      "step": 2286
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8312,
      "step": 2287
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9462,
      "step": 2288
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7134,
      "step": 2289
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.539,
      "step": 2290
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9708,
      "step": 2291
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6094,
      "step": 2292
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 2293
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8019,
      "step": 2294
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0196,
      "step": 2295
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.936,
      "step": 2296
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8383,
      "step": 2297
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0003,
      "step": 2298
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9363,
      "step": 2299
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9608,
      "step": 2300
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6128,
      "step": 2301
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 2302
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7403,
      "step": 2303
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 2304
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 2305
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 2306
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.4683,
      "step": 2307
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 2308
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.753,
      "step": 2309
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7064,
      "step": 2310
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8319,
      "step": 2311
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.5496,
      "step": 2312
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8146,
      "step": 2313
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0419,
      "step": 2314
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9729,
      "step": 2315
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 2316
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.4441,
      "step": 2317
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 2318
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7556,
      "step": 2319
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8206,
      "step": 2320
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8836,
      "step": 2321
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7174,
      "step": 2322
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 2323
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 2324
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7144,
      "step": 2325
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6857,
      "step": 2326
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8967,
      "step": 2327
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7278,
      "step": 2328
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 2329
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 2330
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7028,
      "step": 2331
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8076,
      "step": 2332
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 2333
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.888,
      "step": 2334
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6905,
      "step": 2335
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9466,
      "step": 2336
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.965,
      "step": 2337
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.9968,
      "step": 2338
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 1.0619,
      "step": 2339
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8506,
      "step": 2340
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 2341
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.7377,
      "step": 2342
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8507,
      "step": 2343
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7841,
      "step": 2344
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7395,
      "step": 2345
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.844,
      "step": 2346
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5406,
      "step": 2347
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7586,
      "step": 2348
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8497,
      "step": 2349
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7953,
      "step": 2350
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9833,
      "step": 2351
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6546,
      "step": 2352
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9985,
      "step": 2353
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6249,
      "step": 2354
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.933,
      "step": 2355
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8876,
      "step": 2356
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6898,
      "step": 2357
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0062,
      "step": 2358
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7294,
      "step": 2359
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9061,
      "step": 2360
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.866,
      "step": 2361
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 2362
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8759,
      "step": 2363
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 2364
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.714,
      "step": 2365
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.1328,
      "step": 2366
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7433,
      "step": 2367
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6027,
      "step": 2368
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7573,
      "step": 2369
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.094,
      "step": 2370
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9156,
      "step": 2371
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9604,
      "step": 2372
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 2373
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.1049,
      "step": 2374
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7491,
      "step": 2375
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 2376
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 2377
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 2378
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 2379
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8568,
      "step": 2380
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5992,
      "step": 2381
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.511,
      "step": 2382
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5195,
      "step": 2383
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8785,
      "step": 2384
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5884,
      "step": 2385
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.772,
      "step": 2386
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.1139,
      "step": 2387
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8165,
      "step": 2388
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 2389
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5957,
      "step": 2390
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9219,
      "step": 2391
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 2392
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5522,
      "step": 2393
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7682,
      "step": 2394
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7594,
      "step": 2395
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.732,
      "step": 2396
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7478,
      "step": 2397
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6377,
      "step": 2398
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.9237,
      "step": 2399
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0092,
      "step": 2400
    },
    {
      "epoch": 0.39,
      "eval_loss": 0.754124641418457,
      "eval_runtime": 131.5696,
      "eval_samples_per_second": 7.601,
      "eval_steps_per_second": 3.8,
      "step": 2400
    },
    {
      "epoch": 0.39,
      "mmlu_eval_accuracy": 0.5488713452213336,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.0,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "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": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
      "mmlu_eval_accuracy_high_school_psychology": 0.75,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.46153846153846156,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.7058823529411765,
      "mmlu_eval_accuracy_prehistory": 0.7428571428571429,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.223464566017878,
      "step": 2400
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.7608,
      "step": 2401
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.6459,
      "step": 2402
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 1.0475,
      "step": 2403
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.5478,
      "step": 2404
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8448,
      "step": 2405
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6834,
      "step": 2406
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 2407
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8986,
      "step": 2408
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8555,
      "step": 2409
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7626,
      "step": 2410
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 2411
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0853,
      "step": 2412
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8623,
      "step": 2413
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7921,
      "step": 2414
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.5459,
      "step": 2415
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6145,
      "step": 2416
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7401,
      "step": 2417
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8739,
      "step": 2418
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.4755,
      "step": 2419
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7535,
      "step": 2420
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.1141,
      "step": 2421
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8943,
      "step": 2422
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9819,
      "step": 2423
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 2424
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.4587,
      "step": 2425
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7576,
      "step": 2426
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8637,
      "step": 2427
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0658,
      "step": 2428
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 2429
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.01,
      "step": 2430
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7207,
      "step": 2431
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 2432
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6631,
      "step": 2433
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 2434
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8288,
      "step": 2435
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8355,
      "step": 2436
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8267,
      "step": 2437
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8579,
      "step": 2438
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8763,
      "step": 2439
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 2440
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.833,
      "step": 2441
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9564,
      "step": 2442
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9249,
      "step": 2443
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8565,
      "step": 2444
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.5787,
      "step": 2445
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.5709,
      "step": 2446
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8362,
      "step": 2447
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 2448
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8708,
      "step": 2449
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7568,
      "step": 2450
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7025,
      "step": 2451
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 2452
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6966,
      "step": 2453
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9525,
      "step": 2454
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8237,
      "step": 2455
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6726,
      "step": 2456
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6591,
      "step": 2457
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 2458
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.6239,
      "step": 2459
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 1.0067,
      "step": 2460
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.5825,
      "step": 2461
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 2462
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.9009,
      "step": 2463
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 2464
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 2465
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6286,
      "step": 2466
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.772,
      "step": 2467
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 2468
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8407,
      "step": 2469
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7503,
      "step": 2470
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6329,
      "step": 2471
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.712,
      "step": 2472
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9667,
      "step": 2473
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7194,
      "step": 2474
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 2475
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8703,
      "step": 2476
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6371,
      "step": 2477
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 2478
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8817,
      "step": 2479
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.601,
      "step": 2480
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5872,
      "step": 2481
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6102,
      "step": 2482
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9108,
      "step": 2483
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 2484
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5531,
      "step": 2485
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 2486
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9283,
      "step": 2487
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0349,
      "step": 2488
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 2489
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6669,
      "step": 2490
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8719,
      "step": 2491
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.4049,
      "step": 2492
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6762,
      "step": 2493
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7887,
      "step": 2494
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 2495
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 2496
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6862,
      "step": 2497
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7137,
      "step": 2498
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7956,
      "step": 2499
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7892,
      "step": 2500
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9866,
      "step": 2501
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6448,
      "step": 2502
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8001,
      "step": 2503
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8254,
      "step": 2504
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9118,
      "step": 2505
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5775,
      "step": 2506
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 2507
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.0153,
      "step": 2508
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9019,
      "step": 2509
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5547,
      "step": 2510
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8908,
      "step": 2511
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6691,
      "step": 2512
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5401,
      "step": 2513
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8547,
      "step": 2514
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9645,
      "step": 2515
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 1.102,
      "step": 2516
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8085,
      "step": 2517
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 2518
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.5556,
      "step": 2519
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.6201,
      "step": 2520
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7761,
      "step": 2521
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.628,
      "step": 2522
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9584,
      "step": 2523
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.9155,
      "step": 2524
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.7465,
      "step": 2525
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7352,
      "step": 2526
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7346,
      "step": 2527
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.6878,
      "step": 2528
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7621,
      "step": 2529
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.5621,
      "step": 2530
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 2531
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7636,
      "step": 2532
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 2533
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9402,
      "step": 2534
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8057,
      "step": 2535
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 2536
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9906,
      "step": 2537
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8246,
      "step": 2538
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.4754,
      "step": 2539
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7922,
      "step": 2540
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.818,
      "step": 2541
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 2542
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8984,
      "step": 2543
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 2544
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7439,
      "step": 2545
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 2546
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 2547
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 2548
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 2549
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.6229,
      "step": 2550
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 2551
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8911,
      "step": 2552
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.2983,
      "step": 2553
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.846,
      "step": 2554
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7868,
      "step": 2555
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7506,
      "step": 2556
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7047,
      "step": 2557
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7249,
      "step": 2558
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.6154,
      "step": 2559
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8087,
      "step": 2560
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9299,
      "step": 2561
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8161,
      "step": 2562
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0844,
      "step": 2563
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.863,
      "step": 2564
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.4191,
      "step": 2565
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7249,
      "step": 2566
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 2567
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.9462,
      "step": 2568
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8241,
      "step": 2569
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.5093,
      "step": 2570
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8147,
      "step": 2571
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7835,
      "step": 2572
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 1.0354,
      "step": 2573
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 2574
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8681,
      "step": 2575
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 2576
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8671,
      "step": 2577
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8738,
      "step": 2578
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8246,
      "step": 2579
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.6113,
      "step": 2580
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 2581
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7734,
      "step": 2582
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 2583
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.832,
      "step": 2584
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8859,
      "step": 2585
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 2586
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 2587
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.5618,
      "step": 2588
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6341,
      "step": 2589
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 2590
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 2591
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0108,
      "step": 2592
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7801,
      "step": 2593
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 2594
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.5425,
      "step": 2595
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7461,
      "step": 2596
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4163,
      "step": 2597
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 2598
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9198,
      "step": 2599
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.0873,
      "step": 2600
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.7570953369140625,
      "eval_runtime": 131.5868,
      "eval_samples_per_second": 7.6,
      "eval_steps_per_second": 3.8,
      "step": 2600
    },
    {
      "epoch": 0.43,
      "mmlu_eval_accuracy": 0.5581231690028787,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.6744186046511628,
      "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.47058823529411764,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.6666666666666666,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.9545454545454546,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0118672493310261,
      "step": 2600
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8037,
      "step": 2601
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7605,
      "step": 2602
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8871,
      "step": 2603
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.3938,
      "step": 2604
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 2605
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7479,
      "step": 2606
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8584,
      "step": 2607
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.5362,
      "step": 2608
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8184,
      "step": 2609
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9594,
      "step": 2610
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6906,
      "step": 2611
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6427,
      "step": 2612
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 1.1499,
      "step": 2613
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 2614
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 2615
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8639,
      "step": 2616
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6335,
      "step": 2617
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9219,
      "step": 2618
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8471,
      "step": 2619
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7907,
      "step": 2620
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6762,
      "step": 2621
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9186,
      "step": 2622
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6569,
      "step": 2623
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 2624
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7623,
      "step": 2625
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7855,
      "step": 2626
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8738,
      "step": 2627
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7203,
      "step": 2628
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4544,
      "step": 2629
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 2630
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4496,
      "step": 2631
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.5314,
      "step": 2632
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8078,
      "step": 2633
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8064,
      "step": 2634
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4411,
      "step": 2635
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9895,
      "step": 2636
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 2637
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.5229,
      "step": 2638
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7742,
      "step": 2639
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9026,
      "step": 2640
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4865,
      "step": 2641
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4471,
      "step": 2642
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8359,
      "step": 2643
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 2644
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.9158,
      "step": 2645
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 2646
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.6178,
      "step": 2647
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7435,
      "step": 2648
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 2649
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8002,
      "step": 2650
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.426,
      "step": 2651
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 2652
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 2653
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 2654
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7446,
      "step": 2655
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6941,
      "step": 2656
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7354,
      "step": 2657
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6183,
      "step": 2658
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8515,
      "step": 2659
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8874,
      "step": 2660
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6188,
      "step": 2661
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9337,
      "step": 2662
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7617,
      "step": 2663
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.5362,
      "step": 2664
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 2665
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7054,
      "step": 2666
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7383,
      "step": 2667
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 2668
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8373,
      "step": 2669
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 2670
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 2671
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 2672
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8043,
      "step": 2673
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6454,
      "step": 2674
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8217,
      "step": 2675
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 2676
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7064,
      "step": 2677
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0354,
      "step": 2678
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 2679
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6881,
      "step": 2680
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6349,
      "step": 2681
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.9402,
      "step": 2682
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0755,
      "step": 2683
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7402,
      "step": 2684
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.663,
      "step": 2685
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0114,
      "step": 2686
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 2687
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 2688
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8142,
      "step": 2689
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.5404,
      "step": 2690
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.914,
      "step": 2691
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 2692
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0168,
      "step": 2693
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.727,
      "step": 2694
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8568,
      "step": 2695
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7015,
      "step": 2696
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 2697
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8432,
      "step": 2698
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 2699
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8496,
      "step": 2700
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 2701
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.0566,
      "step": 2702
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7125,
      "step": 2703
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7536,
      "step": 2704
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 2705
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 1.11,
      "step": 2706
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 2707
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 2708
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8618,
      "step": 2709
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7046,
      "step": 2710
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6036,
      "step": 2711
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9455,
      "step": 2712
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8699,
      "step": 2713
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6548,
      "step": 2714
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.817,
      "step": 2715
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.799,
      "step": 2716
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 2717
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 2718
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.2124,
      "step": 2719
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.1396,
      "step": 2720
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 2721
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 2722
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7494,
      "step": 2723
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 2724
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7537,
      "step": 2725
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7779,
      "step": 2726
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.5883,
      "step": 2727
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0681,
      "step": 2728
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.5107,
      "step": 2729
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0429,
      "step": 2730
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7311,
      "step": 2731
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0758,
      "step": 2732
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.5067,
      "step": 2733
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 2734
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7282,
      "step": 2735
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.664,
      "step": 2736
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9432,
      "step": 2737
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6337,
      "step": 2738
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 2739
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6309,
      "step": 2740
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.792,
      "step": 2741
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9338,
      "step": 2742
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0635,
      "step": 2743
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8012,
      "step": 2744
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0958,
      "step": 2745
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8876,
      "step": 2746
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9664,
      "step": 2747
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7179,
      "step": 2748
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.609,
      "step": 2749
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6172,
      "step": 2750
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.1386,
      "step": 2751
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7308,
      "step": 2752
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 2753
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8759,
      "step": 2754
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8531,
      "step": 2755
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7749,
      "step": 2756
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8547,
      "step": 2757
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8355,
      "step": 2758
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8094,
      "step": 2759
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7815,
      "step": 2760
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 2761
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6334,
      "step": 2762
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 2763
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.6815,
      "step": 2764
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9356,
      "step": 2765
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.7899,
      "step": 2766
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 1.0199,
      "step": 2767
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.73,
      "step": 2768
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.9788,
      "step": 2769
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8299,
      "step": 2770
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7428,
      "step": 2771
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9578,
      "step": 2772
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6592,
      "step": 2773
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8675,
      "step": 2774
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 2775
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 2776
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.4909,
      "step": 2777
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6649,
      "step": 2778
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8044,
      "step": 2779
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7484,
      "step": 2780
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9253,
      "step": 2781
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 2782
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 2783
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9379,
      "step": 2784
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8074,
      "step": 2785
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 2786
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7403,
      "step": 2787
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 2788
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6428,
      "step": 2789
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.5931,
      "step": 2790
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 2791
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.5879,
      "step": 2792
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8604,
      "step": 2793
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.5792,
      "step": 2794
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 2795
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7078,
      "step": 2796
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7481,
      "step": 2797
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 2798
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 2799
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9685,
      "step": 2800
    },
    {
      "epoch": 0.46,
      "eval_loss": 0.7530816197395325,
      "eval_runtime": 131.6881,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 2800
    },
    {
      "epoch": 0.46,
      "mmlu_eval_accuracy": 0.5530778539637836,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.7272727272727273,
      "mmlu_eval_accuracy_college_physics": 0.8181818181818182,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "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": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.78125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
      "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "mmlu_eval_accuracy_human_sexuality": 0.6666666666666666,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.39,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.4294117647058823,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "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": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.0421120278712352,
      "step": 2800
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8098,
      "step": 2801
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9656,
      "step": 2802
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8368,
      "step": 2803
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6248,
      "step": 2804
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6017,
      "step": 2805
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 2806
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.1513,
      "step": 2807
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8444,
      "step": 2808
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8169,
      "step": 2809
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8228,
      "step": 2810
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0234,
      "step": 2811
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 2812
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.773,
      "step": 2813
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9613,
      "step": 2814
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.86,
      "step": 2815
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6061,
      "step": 2816
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8257,
      "step": 2817
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8071,
      "step": 2818
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 1.0167,
      "step": 2819
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 2820
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9094,
      "step": 2821
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6907,
      "step": 2822
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.6652,
      "step": 2823
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8172,
      "step": 2824
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.5711,
      "step": 2825
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.5757,
      "step": 2826
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 2827
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.9677,
      "step": 2828
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8585,
      "step": 2829
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.7479,
      "step": 2830
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 2831
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9481,
      "step": 2832
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.3661,
      "step": 2833
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8192,
      "step": 2834
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.4574,
      "step": 2835
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8958,
      "step": 2836
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7935,
      "step": 2837
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8905,
      "step": 2838
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 2839
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.978,
      "step": 2840
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 2841
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8611,
      "step": 2842
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7029,
      "step": 2843
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6195,
      "step": 2844
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 2845
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.5484,
      "step": 2846
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9817,
      "step": 2847
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9278,
      "step": 2848
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.671,
      "step": 2849
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 2850
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6611,
      "step": 2851
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 2852
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9665,
      "step": 2853
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.942,
      "step": 2854
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6561,
      "step": 2855
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9651,
      "step": 2856
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.5337,
      "step": 2857
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.792,
      "step": 2858
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.1503,
      "step": 2859
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8918,
      "step": 2860
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6808,
      "step": 2861
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.602,
      "step": 2862
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6973,
      "step": 2863
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8459,
      "step": 2864
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8665,
      "step": 2865
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 2866
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6571,
      "step": 2867
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 2868
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.5655,
      "step": 2869
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.9379,
      "step": 2870
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6435,
      "step": 2871
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7375,
      "step": 2872
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.0998,
      "step": 2873
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8005,
      "step": 2874
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 2875
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6895,
      "step": 2876
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 2877
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6207,
      "step": 2878
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.495,
      "step": 2879
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7,
      "step": 2880
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8074,
      "step": 2881
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7798,
      "step": 2882
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6458,
      "step": 2883
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 2884
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 2885
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7649,
      "step": 2886
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.7728,
      "step": 2887
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.5118,
      "step": 2888
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.826,
      "step": 2889
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 1.148,
      "step": 2890
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6991,
      "step": 2891
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.609,
      "step": 2892
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7536,
      "step": 2893
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7738,
      "step": 2894
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 2895
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5314,
      "step": 2896
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.4833,
      "step": 2897
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 2898
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6338,
      "step": 2899
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.4124,
      "step": 2900
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7652,
      "step": 2901
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 2902
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9406,
      "step": 2903
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5314,
      "step": 2904
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8301,
      "step": 2905
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8097,
      "step": 2906
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6065,
      "step": 2907
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.405,
      "step": 2908
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7592,
      "step": 2909
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6914,
      "step": 2910
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0258,
      "step": 2911
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.929,
      "step": 2912
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.733,
      "step": 2913
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6774,
      "step": 2914
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.803,
      "step": 2915
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5683,
      "step": 2916
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 2917
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8319,
      "step": 2918
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7761,
      "step": 2919
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9458,
      "step": 2920
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "step": 2921
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7603,
      "step": 2922
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7693,
      "step": 2923
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8695,
      "step": 2924
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.1133,
      "step": 2925
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5037,
      "step": 2926
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 2927
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6465,
      "step": 2928
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9632,
      "step": 2929
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6949,
      "step": 2930
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6652,
      "step": 2931
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.4068,
      "step": 2932
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7214,
      "step": 2933
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0307,
      "step": 2934
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6681,
      "step": 2935
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5881,
      "step": 2936
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7022,
      "step": 2937
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8272,
      "step": 2938
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5194,
      "step": 2939
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5115,
      "step": 2940
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8319,
      "step": 2941
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.7172,
      "step": 2942
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8612,
      "step": 2943
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9976,
      "step": 2944
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 2945
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.979,
      "step": 2946
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9404,
      "step": 2947
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.885,
      "step": 2948
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.9053,
      "step": 2949
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 2950
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 1.0637,
      "step": 2951
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 2952
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7503,
      "step": 2953
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 2954
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8625,
      "step": 2955
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8117,
      "step": 2956
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.66,
      "step": 2957
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0334,
      "step": 2958
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0704,
      "step": 2959
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8582,
      "step": 2960
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 2961
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.1001,
      "step": 2962
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7449,
      "step": 2963
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0588,
      "step": 2964
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7596,
      "step": 2965
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 2966
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7837,
      "step": 2967
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.056,
      "step": 2968
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.996,
      "step": 2969
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8761,
      "step": 2970
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8035,
      "step": 2971
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 2972
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7333,
      "step": 2973
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6356,
      "step": 2974
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7532,
      "step": 2975
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6252,
      "step": 2976
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7604,
      "step": 2977
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 2978
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.938,
      "step": 2979
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 2980
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9421,
      "step": 2981
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 2982
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 2983
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8903,
      "step": 2984
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.1069,
      "step": 2985
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.5866,
      "step": 2986
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 2987
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.715,
      "step": 2988
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0569,
      "step": 2989
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.5055,
      "step": 2990
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9878,
      "step": 2991
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8048,
      "step": 2992
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9757,
      "step": 2993
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.4064,
      "step": 2994
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8918,
      "step": 2995
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.703,
      "step": 2996
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0976,
      "step": 2997
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 2998
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.5379,
      "step": 2999
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.5319,
      "step": 3000
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.7562741041183472,
      "eval_runtime": 131.6412,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 3000
    },
    {
      "epoch": 0.49,
      "mmlu_eval_accuracy": 0.5657216197300882,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.6,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "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": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.8076923076923077,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.39,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.091038808109866,
      "step": 3000
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8781,
      "step": 3001
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6487,
      "step": 3002
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0339,
      "step": 3003
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.9526,
      "step": 3004
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.5935,
      "step": 3005
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6707,
      "step": 3006
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6641,
      "step": 3007
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 3008
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.7678,
      "step": 3009
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 1.0386,
      "step": 3010
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.772,
      "step": 3011
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.6929,
      "step": 3012
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8485,
      "step": 3013
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 3014
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 3015
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 3016
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.838,
      "step": 3017
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9496,
      "step": 3018
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 1.0716,
      "step": 3019
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 3020
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 3021
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7043,
      "step": 3022
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6221,
      "step": 3023
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 3024
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7658,
      "step": 3025
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9514,
      "step": 3026
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7065,
      "step": 3027
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5111,
      "step": 3028
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8038,
      "step": 3029
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.808,
      "step": 3030
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8133,
      "step": 3031
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7117,
      "step": 3032
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 3033
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8345,
      "step": 3034
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 3035
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 3036
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.618,
      "step": 3037
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7699,
      "step": 3038
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 3039
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7765,
      "step": 3040
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7144,
      "step": 3041
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5507,
      "step": 3042
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 1.0901,
      "step": 3043
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7746,
      "step": 3044
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9164,
      "step": 3045
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.614,
      "step": 3046
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6068,
      "step": 3047
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9984,
      "step": 3048
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9879,
      "step": 3049
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9075,
      "step": 3050
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6982,
      "step": 3051
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.927,
      "step": 3052
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 3053
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5627,
      "step": 3054
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7256,
      "step": 3055
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7678,
      "step": 3056
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5627,
      "step": 3057
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7685,
      "step": 3058
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8924,
      "step": 3059
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 3060
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6875,
      "step": 3061
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8683,
      "step": 3062
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.768,
      "step": 3063
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7801,
      "step": 3064
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.9501,
      "step": 3065
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 3066
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.6508,
      "step": 3067
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 3068
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 3069
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.5179,
      "step": 3070
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8495,
      "step": 3071
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.7408,
      "step": 3072
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8953,
      "step": 3073
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9477,
      "step": 3074
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9135,
      "step": 3075
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 3076
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 3077
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8012,
      "step": 3078
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8357,
      "step": 3079
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7435,
      "step": 3080
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.5473,
      "step": 3081
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 3082
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 3083
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6618,
      "step": 3084
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9276,
      "step": 3085
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 3086
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.654,
      "step": 3087
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8015,
      "step": 3088
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 1.0513,
      "step": 3089
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8357,
      "step": 3090
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 3091
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.695,
      "step": 3092
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7898,
      "step": 3093
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7437,
      "step": 3094
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 3095
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.729,
      "step": 3096
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7076,
      "step": 3097
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 3098
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 3099
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8445,
      "step": 3100
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.5858,
      "step": 3101
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.738,
      "step": 3102
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6476,
      "step": 3103
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.697,
      "step": 3104
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8828,
      "step": 3105
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9003,
      "step": 3106
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8344,
      "step": 3107
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7103,
      "step": 3108
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.637,
      "step": 3109
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 1.1437,
      "step": 3110
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7657,
      "step": 3111
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 3112
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 1.0407,
      "step": 3113
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 3114
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 3115
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 1.3087,
      "step": 3116
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9146,
      "step": 3117
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8138,
      "step": 3118
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 3119
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8417,
      "step": 3120
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.5309,
      "step": 3121
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6276,
      "step": 3122
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 1.0024,
      "step": 3123
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9719,
      "step": 3124
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.5478,
      "step": 3125
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8333,
      "step": 3126
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.5024,
      "step": 3127
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.88,
      "step": 3128
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 3129
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9953,
      "step": 3130
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8652,
      "step": 3131
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.6715,
      "step": 3132
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.9498,
      "step": 3133
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 3134
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 3135
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 3136
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.955,
      "step": 3137
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7728,
      "step": 3138
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6917,
      "step": 3139
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.929,
      "step": 3140
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8458,
      "step": 3141
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.915,
      "step": 3142
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 3143
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 3144
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6397,
      "step": 3145
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8762,
      "step": 3146
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7542,
      "step": 3147
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7003,
      "step": 3148
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9587,
      "step": 3149
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9843,
      "step": 3150
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 3151
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6476,
      "step": 3152
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 1.017,
      "step": 3153
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8428,
      "step": 3154
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7956,
      "step": 3155
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7353,
      "step": 3156
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6851,
      "step": 3157
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.5245,
      "step": 3158
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 3159
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8815,
      "step": 3160
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 3161
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9293,
      "step": 3162
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9643,
      "step": 3163
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6312,
      "step": 3164
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7596,
      "step": 3165
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 3166
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8905,
      "step": 3167
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6303,
      "step": 3168
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7482,
      "step": 3169
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.5716,
      "step": 3170
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8766,
      "step": 3171
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9254,
      "step": 3172
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7601,
      "step": 3173
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7328,
      "step": 3174
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 1.0365,
      "step": 3175
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 1.131,
      "step": 3176
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6634,
      "step": 3177
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 1.039,
      "step": 3178
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 3179
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 3180
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6879,
      "step": 3181
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 3182
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7998,
      "step": 3183
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6229,
      "step": 3184
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.711,
      "step": 3185
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 3186
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8632,
      "step": 3187
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 1.0824,
      "step": 3188
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9087,
      "step": 3189
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 3190
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.9201,
      "step": 3191
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6367,
      "step": 3192
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8713,
      "step": 3193
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.6782,
      "step": 3194
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.657,
      "step": 3195
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9913,
      "step": 3196
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7744,
      "step": 3197
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 3198
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5735,
      "step": 3199
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7163,
      "step": 3200
    },
    {
      "epoch": 0.53,
      "eval_loss": 0.7516635656356812,
      "eval_runtime": 131.726,
      "eval_samples_per_second": 7.592,
      "eval_steps_per_second": 3.796,
      "step": 3200
    },
    {
      "epoch": 0.53,
      "mmlu_eval_accuracy": 0.5737180213767722,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "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": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.42857142857142855,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.71875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 0.8832502972636148,
      "step": 3200
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 3201
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5147,
      "step": 3202
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9274,
      "step": 3203
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6697,
      "step": 3204
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.727,
      "step": 3205
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6729,
      "step": 3206
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7744,
      "step": 3207
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9278,
      "step": 3208
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.676,
      "step": 3209
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 3210
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9212,
      "step": 3211
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 3212
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9432,
      "step": 3213
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 3214
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 3215
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8767,
      "step": 3216
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.668,
      "step": 3217
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8151,
      "step": 3218
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5956,
      "step": 3219
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8103,
      "step": 3220
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7316,
      "step": 3221
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.665,
      "step": 3222
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 3223
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9532,
      "step": 3224
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5412,
      "step": 3225
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7871,
      "step": 3226
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7102,
      "step": 3227
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 3228
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8141,
      "step": 3229
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 3230
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9752,
      "step": 3231
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9711,
      "step": 3232
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7897,
      "step": 3233
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 3234
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 3235
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 3236
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6455,
      "step": 3237
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5552,
      "step": 3238
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7528,
      "step": 3239
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.5483,
      "step": 3240
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 3241
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.3298,
      "step": 3242
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7539,
      "step": 3243
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9876,
      "step": 3244
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.4657,
      "step": 3245
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 3246
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6767,
      "step": 3247
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9049,
      "step": 3248
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8022,
      "step": 3249
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.7796,
      "step": 3250
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.9956,
      "step": 3251
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.6867,
      "step": 3252
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.4339,
      "step": 3253
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.775,
      "step": 3254
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.802,
      "step": 3255
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8255,
      "step": 3256
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7897,
      "step": 3257
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7778,
      "step": 3258
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.6077,
      "step": 3259
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 3260
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.6567,
      "step": 3261
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.6769,
      "step": 3262
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.712,
      "step": 3263
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.836,
      "step": 3264
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 1.0661,
      "step": 3265
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.6398,
      "step": 3266
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8791,
      "step": 3267
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.5367,
      "step": 3268
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8647,
      "step": 3269
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 3270
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.553,
      "step": 3271
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 3272
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7861,
      "step": 3273
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 3274
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.6838,
      "step": 3275
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.9176,
      "step": 3276
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7397,
      "step": 3277
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8605,
      "step": 3278
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 3279
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7457,
      "step": 3280
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8959,
      "step": 3281
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 3282
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7365,
      "step": 3283
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 3284
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7365,
      "step": 3285
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8435,
      "step": 3286
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 3287
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.528,
      "step": 3288
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8281,
      "step": 3289
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.4378,
      "step": 3290
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8019,
      "step": 3291
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 1.0369,
      "step": 3292
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.87,
      "step": 3293
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 3294
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 3295
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7638,
      "step": 3296
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.9255,
      "step": 3297
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 1.0927,
      "step": 3298
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 3299
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8979,
      "step": 3300
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8632,
      "step": 3301
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8038,
      "step": 3302
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7759,
      "step": 3303
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 3304
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.5661,
      "step": 3305
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8136,
      "step": 3306
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8669,
      "step": 3307
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7742,
      "step": 3308
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7886,
      "step": 3309
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.7035,
      "step": 3310
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 3311
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.9369,
      "step": 3312
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.9092,
      "step": 3313
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.775,
      "step": 3314
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.4413,
      "step": 3315
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.501,
      "step": 3316
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 3317
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.675,
      "step": 3318
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5488,
      "step": 3319
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7659,
      "step": 3320
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 3321
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8153,
      "step": 3322
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8325,
      "step": 3323
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8006,
      "step": 3324
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 3325
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8116,
      "step": 3326
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8767,
      "step": 3327
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5521,
      "step": 3328
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 3329
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5055,
      "step": 3330
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8363,
      "step": 3331
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.0262,
      "step": 3332
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6862,
      "step": 3333
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6646,
      "step": 3334
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.9674,
      "step": 3335
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 3336
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7075,
      "step": 3337
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 3338
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.914,
      "step": 3339
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 3340
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.0676,
      "step": 3341
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7941,
      "step": 3342
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 3343
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.002,
      "step": 3344
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.1118,
      "step": 3345
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 3346
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6462,
      "step": 3347
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5919,
      "step": 3348
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.1466,
      "step": 3349
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7977,
      "step": 3350
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5158,
      "step": 3351
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.904,
      "step": 3352
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8702,
      "step": 3353
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 3354
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.0569,
      "step": 3355
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 3356
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8772,
      "step": 3357
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5319,
      "step": 3358
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.0922,
      "step": 3359
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5511,
      "step": 3360
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8385,
      "step": 3361
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 3362
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7977,
      "step": 3363
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7771,
      "step": 3364
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.9893,
      "step": 3365
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 3366
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6704,
      "step": 3367
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 3368
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 1.049,
      "step": 3369
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 3370
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.5175,
      "step": 3371
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.632,
      "step": 3372
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 3373
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 3374
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8201,
      "step": 3375
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.7912,
      "step": 3376
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.8123,
      "step": 3377
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5863,
      "step": 3378
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7737,
      "step": 3379
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 3380
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 3381
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.923,
      "step": 3382
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7479,
      "step": 3383
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5268,
      "step": 3384
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 3385
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9283,
      "step": 3386
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7676,
      "step": 3387
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9106,
      "step": 3388
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9585,
      "step": 3389
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5391,
      "step": 3390
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.745,
      "step": 3391
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 1.3179,
      "step": 3392
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7739,
      "step": 3393
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5233,
      "step": 3394
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6186,
      "step": 3395
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 3396
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8342,
      "step": 3397
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 3398
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.587,
      "step": 3399
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 3400
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.749654233455658,
      "eval_runtime": 131.6498,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 3400
    },
    {
      "epoch": 0.56,
      "mmlu_eval_accuracy": 0.574503484588032,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "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": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.26,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "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": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.9545454545454546,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.0488827172326356,
      "step": 3400
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7714,
      "step": 3401
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 3402
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8402,
      "step": 3403
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7598,
      "step": 3404
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9918,
      "step": 3405
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7602,
      "step": 3406
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6071,
      "step": 3407
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 3408
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 3409
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6201,
      "step": 3410
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 3411
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6829,
      "step": 3412
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9137,
      "step": 3413
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5523,
      "step": 3414
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6845,
      "step": 3415
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 1.1485,
      "step": 3416
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8524,
      "step": 3417
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.934,
      "step": 3418
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5627,
      "step": 3419
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9024,
      "step": 3420
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7054,
      "step": 3421
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7798,
      "step": 3422
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8881,
      "step": 3423
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 3424
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6502,
      "step": 3425
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.9398,
      "step": 3426
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 3427
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 3428
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6628,
      "step": 3429
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5828,
      "step": 3430
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.813,
      "step": 3431
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8269,
      "step": 3432
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5469,
      "step": 3433
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7982,
      "step": 3434
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5248,
      "step": 3435
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8588,
      "step": 3436
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.7306,
      "step": 3437
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.6578,
      "step": 3438
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9038,
      "step": 3439
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 3440
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8225,
      "step": 3441
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8381,
      "step": 3442
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.6786,
      "step": 3443
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9323,
      "step": 3444
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.4972,
      "step": 3445
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9031,
      "step": 3446
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.0749,
      "step": 3447
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7974,
      "step": 3448
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 3449
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7745,
      "step": 3450
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8584,
      "step": 3451
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.4844,
      "step": 3452
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 3453
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7858,
      "step": 3454
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8104,
      "step": 3455
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 3456
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8718,
      "step": 3457
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.533,
      "step": 3458
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 3459
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9081,
      "step": 3460
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.6779,
      "step": 3461
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.0281,
      "step": 3462
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7136,
      "step": 3463
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7337,
      "step": 3464
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.1414,
      "step": 3465
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.615,
      "step": 3466
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8462,
      "step": 3467
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.6567,
      "step": 3468
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9807,
      "step": 3469
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.79,
      "step": 3470
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.1021,
      "step": 3471
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7497,
      "step": 3472
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9982,
      "step": 3473
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9731,
      "step": 3474
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.0387,
      "step": 3475
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.5922,
      "step": 3476
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8949,
      "step": 3477
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 3478
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.6747,
      "step": 3479
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.5461,
      "step": 3480
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.899,
      "step": 3481
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.0509,
      "step": 3482
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 1.0077,
      "step": 3483
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.5613,
      "step": 3484
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 3485
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9096,
      "step": 3486
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9492,
      "step": 3487
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.5355,
      "step": 3488
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8666,
      "step": 3489
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8679,
      "step": 3490
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.4945,
      "step": 3491
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 3492
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7361,
      "step": 3493
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 3494
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.6511,
      "step": 3495
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7011,
      "step": 3496
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.9052,
      "step": 3497
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.7592,
      "step": 3498
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8194,
      "step": 3499
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6785,
      "step": 3500
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7941,
      "step": 3501
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8527,
      "step": 3502
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9948,
      "step": 3503
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5943,
      "step": 3504
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6741,
      "step": 3505
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 1.2698,
      "step": 3506
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6925,
      "step": 3507
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 3508
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 1.0184,
      "step": 3509
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9142,
      "step": 3510
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8805,
      "step": 3511
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8892,
      "step": 3512
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5916,
      "step": 3513
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 3514
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6256,
      "step": 3515
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.4009,
      "step": 3516
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 1.0093,
      "step": 3517
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7941,
      "step": 3518
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 3519
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8771,
      "step": 3520
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 3521
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9429,
      "step": 3522
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 3523
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9307,
      "step": 3524
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9908,
      "step": 3525
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 3526
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 3527
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.4572,
      "step": 3528
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.4444,
      "step": 3529
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8649,
      "step": 3530
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.795,
      "step": 3531
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6956,
      "step": 3532
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9454,
      "step": 3533
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9834,
      "step": 3534
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8243,
      "step": 3535
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9817,
      "step": 3536
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5175,
      "step": 3537
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5396,
      "step": 3538
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9076,
      "step": 3539
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7235,
      "step": 3540
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7607,
      "step": 3541
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7125,
      "step": 3542
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9536,
      "step": 3543
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 3544
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9378,
      "step": 3545
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8266,
      "step": 3546
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8743,
      "step": 3547
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 1.0378,
      "step": 3548
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.969,
      "step": 3549
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 3550
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5435,
      "step": 3551
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.9508,
      "step": 3552
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.5541,
      "step": 3553
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 3554
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.6647,
      "step": 3555
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.902,
      "step": 3556
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 3557
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.954,
      "step": 3558
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7288,
      "step": 3559
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.7382,
      "step": 3560
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6404,
      "step": 3561
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9829,
      "step": 3562
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.827,
      "step": 3563
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6734,
      "step": 3564
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7555,
      "step": 3565
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8572,
      "step": 3566
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9734,
      "step": 3567
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9008,
      "step": 3568
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 1.0214,
      "step": 3569
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.614,
      "step": 3570
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6459,
      "step": 3571
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7835,
      "step": 3572
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8504,
      "step": 3573
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 3574
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 1.0251,
      "step": 3575
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.874,
      "step": 3576
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7773,
      "step": 3577
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.5916,
      "step": 3578
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7578,
      "step": 3579
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8065,
      "step": 3580
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8056,
      "step": 3581
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7554,
      "step": 3582
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 3583
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6385,
      "step": 3584
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7085,
      "step": 3585
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.895,
      "step": 3586
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9523,
      "step": 3587
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7818,
      "step": 3588
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6468,
      "step": 3589
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9299,
      "step": 3590
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.4926,
      "step": 3591
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8229,
      "step": 3592
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6897,
      "step": 3593
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8863,
      "step": 3594
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8045,
      "step": 3595
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.742,
      "step": 3596
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7566,
      "step": 3597
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 3598
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 3599
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8159,
      "step": 3600
    },
    {
      "epoch": 0.59,
      "eval_loss": 0.7506269812583923,
      "eval_runtime": 131.59,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.8,
      "step": 3600
    },
    {
      "epoch": 0.59,
      "mmlu_eval_accuracy": 0.5826522403805086,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.8125,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
      "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": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.71875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
      "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": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1334416739078477,
      "step": 3600
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6198,
      "step": 3601
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.5997,
      "step": 3602
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 3603
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 3604
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.661,
      "step": 3605
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.4949,
      "step": 3606
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7743,
      "step": 3607
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9543,
      "step": 3608
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 1.0559,
      "step": 3609
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 3610
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.9093,
      "step": 3611
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8173,
      "step": 3612
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8253,
      "step": 3613
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.5855,
      "step": 3614
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 3615
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 3616
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6732,
      "step": 3617
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6782,
      "step": 3618
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 3619
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.6727,
      "step": 3620
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.4883,
      "step": 3621
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 3622
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6871,
      "step": 3623
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8523,
      "step": 3624
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7614,
      "step": 3625
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.9922,
      "step": 3626
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8352,
      "step": 3627
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 3628
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7674,
      "step": 3629
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8229,
      "step": 3630
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7351,
      "step": 3631
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8084,
      "step": 3632
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 3633
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6576,
      "step": 3634
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.433,
      "step": 3635
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.9451,
      "step": 3636
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8166,
      "step": 3637
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8891,
      "step": 3638
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6921,
      "step": 3639
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.4516,
      "step": 3640
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6829,
      "step": 3641
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.9014,
      "step": 3642
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.5958,
      "step": 3643
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.3924,
      "step": 3644
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 3645
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7725,
      "step": 3646
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 3647
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7656,
      "step": 3648
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 1.0481,
      "step": 3649
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.9214,
      "step": 3650
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6203,
      "step": 3651
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 3652
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8814,
      "step": 3653
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.462,
      "step": 3654
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7891,
      "step": 3655
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6932,
      "step": 3656
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 3657
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 3658
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8812,
      "step": 3659
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8939,
      "step": 3660
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.989,
      "step": 3661
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7318,
      "step": 3662
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8295,
      "step": 3663
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.447,
      "step": 3664
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 3665
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 3666
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6583,
      "step": 3667
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.5963,
      "step": 3668
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 3669
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7756,
      "step": 3670
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7339,
      "step": 3671
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6942,
      "step": 3672
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8506,
      "step": 3673
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 3674
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7236,
      "step": 3675
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 3676
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.5004,
      "step": 3677
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.7134,
      "step": 3678
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 1.0034,
      "step": 3679
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 3680
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.5623,
      "step": 3681
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 3682
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7176,
      "step": 3683
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 3684
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9916,
      "step": 3685
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 3686
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6481,
      "step": 3687
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6961,
      "step": 3688
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 3689
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 3690
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.605,
      "step": 3691
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5274,
      "step": 3692
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 3693
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9291,
      "step": 3694
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8049,
      "step": 3695
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5145,
      "step": 3696
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 1.003,
      "step": 3697
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9122,
      "step": 3698
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 1.0403,
      "step": 3699
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6969,
      "step": 3700
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9441,
      "step": 3701
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 1.0414,
      "step": 3702
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 3703
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8022,
      "step": 3704
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7118,
      "step": 3705
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 3706
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7644,
      "step": 3707
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 3708
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8302,
      "step": 3709
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.73,
      "step": 3710
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9252,
      "step": 3711
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8429,
      "step": 3712
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 3713
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9014,
      "step": 3714
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 3715
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 3716
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9826,
      "step": 3717
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6815,
      "step": 3718
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6904,
      "step": 3719
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 1.0041,
      "step": 3720
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5924,
      "step": 3721
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.6953,
      "step": 3722
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7888,
      "step": 3723
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8817,
      "step": 3724
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9264,
      "step": 3725
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 3726
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7993,
      "step": 3727
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 3728
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 3729
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5969,
      "step": 3730
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.85,
      "step": 3731
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 3732
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7647,
      "step": 3733
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8237,
      "step": 3734
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7258,
      "step": 3735
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9647,
      "step": 3736
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5279,
      "step": 3737
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.5559,
      "step": 3738
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 3739
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 3740
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.9443,
      "step": 3741
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.7924,
      "step": 3742
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8119,
      "step": 3743
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5553,
      "step": 3744
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 1.0719,
      "step": 3745
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6299,
      "step": 3746
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7441,
      "step": 3747
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 3748
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9058,
      "step": 3749
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7353,
      "step": 3750
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7452,
      "step": 3751
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9035,
      "step": 3752
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6045,
      "step": 3753
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 3754
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.648,
      "step": 3755
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7814,
      "step": 3756
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 1.024,
      "step": 3757
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 3758
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 1.0115,
      "step": 3759
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7928,
      "step": 3760
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 3761
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9024,
      "step": 3762
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7677,
      "step": 3763
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5106,
      "step": 3764
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6804,
      "step": 3765
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7812,
      "step": 3766
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7305,
      "step": 3767
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6532,
      "step": 3768
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9125,
      "step": 3769
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 3770
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7715,
      "step": 3771
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 1.0193,
      "step": 3772
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5958,
      "step": 3773
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 3774
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5232,
      "step": 3775
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5943,
      "step": 3776
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.97,
      "step": 3777
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.8649,
      "step": 3778
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7044,
      "step": 3779
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6559,
      "step": 3780
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 1.0746,
      "step": 3781
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.836,
      "step": 3782
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 3783
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7576,
      "step": 3784
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9034,
      "step": 3785
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6595,
      "step": 3786
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9853,
      "step": 3787
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9735,
      "step": 3788
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 3789
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 3790
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7765,
      "step": 3791
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9774,
      "step": 3792
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5509,
      "step": 3793
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9114,
      "step": 3794
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.4861,
      "step": 3795
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 3796
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.4111,
      "step": 3797
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9114,
      "step": 3798
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 3799
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9816,
      "step": 3800
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.7537699341773987,
      "eval_runtime": 131.6884,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 3800
    },
    {
      "epoch": 0.62,
      "mmlu_eval_accuracy": 0.5747248918142907,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6551724137931034,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "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": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
      "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": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.38235294117647056,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.0576872202730676,
      "step": 3800
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 3801
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.9108,
      "step": 3802
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 3803
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7245,
      "step": 3804
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 3805
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 3806
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9332,
      "step": 3807
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 1.2136,
      "step": 3808
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8751,
      "step": 3809
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 3810
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9675,
      "step": 3811
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7018,
      "step": 3812
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7989,
      "step": 3813
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7021,
      "step": 3814
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6984,
      "step": 3815
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 1.1223,
      "step": 3816
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 3817
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 3818
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8791,
      "step": 3819
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 3820
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.864,
      "step": 3821
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.4597,
      "step": 3822
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8548,
      "step": 3823
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.742,
      "step": 3824
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.5929,
      "step": 3825
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6599,
      "step": 3826
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9248,
      "step": 3827
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7489,
      "step": 3828
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7957,
      "step": 3829
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8486,
      "step": 3830
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8626,
      "step": 3831
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9905,
      "step": 3832
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7992,
      "step": 3833
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8073,
      "step": 3834
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.5318,
      "step": 3835
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 3836
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8805,
      "step": 3837
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 3838
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7076,
      "step": 3839
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6394,
      "step": 3840
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7299,
      "step": 3841
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9415,
      "step": 3842
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7803,
      "step": 3843
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9901,
      "step": 3844
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6428,
      "step": 3845
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7608,
      "step": 3846
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6747,
      "step": 3847
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9323,
      "step": 3848
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6931,
      "step": 3849
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 3850
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9666,
      "step": 3851
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 1.0202,
      "step": 3852
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 3853
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6166,
      "step": 3854
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 1.08,
      "step": 3855
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 1.164,
      "step": 3856
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9347,
      "step": 3857
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 3858
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 3859
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 3860
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.3595,
      "step": 3861
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.9754,
      "step": 3862
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.7815,
      "step": 3863
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.4558,
      "step": 3864
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9595,
      "step": 3865
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 3866
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.581,
      "step": 3867
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7417,
      "step": 3868
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7924,
      "step": 3869
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.5623,
      "step": 3870
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.488,
      "step": 3871
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.864,
      "step": 3872
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 3873
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7831,
      "step": 3874
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 3875
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 3876
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 3877
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 3878
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7008,
      "step": 3879
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 1.1345,
      "step": 3880
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 3881
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6671,
      "step": 3882
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6767,
      "step": 3883
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8261,
      "step": 3884
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7584,
      "step": 3885
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 3886
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8094,
      "step": 3887
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 1.03,
      "step": 3888
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 3889
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 3890
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.4543,
      "step": 3891
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.658,
      "step": 3892
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6426,
      "step": 3893
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8137,
      "step": 3894
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8807,
      "step": 3895
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9436,
      "step": 3896
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.892,
      "step": 3897
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8417,
      "step": 3898
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6452,
      "step": 3899
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 3900
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9157,
      "step": 3901
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6588,
      "step": 3902
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6912,
      "step": 3903
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7504,
      "step": 3904
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8242,
      "step": 3905
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9903,
      "step": 3906
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6505,
      "step": 3907
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 3908
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7766,
      "step": 3909
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 3910
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6193,
      "step": 3911
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7262,
      "step": 3912
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6989,
      "step": 3913
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8782,
      "step": 3914
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 3915
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 1.1232,
      "step": 3916
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 3917
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6064,
      "step": 3918
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.7359,
      "step": 3919
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.796,
      "step": 3920
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6751,
      "step": 3921
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 3922
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.5512,
      "step": 3923
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.9652,
      "step": 3924
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.5687,
      "step": 3925
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 3926
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.5293,
      "step": 3927
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.883,
      "step": 3928
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6263,
      "step": 3929
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 3930
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6,
      "step": 3931
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 3932
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8341,
      "step": 3933
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7435,
      "step": 3934
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 1.1017,
      "step": 3935
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 3936
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6318,
      "step": 3937
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 3938
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8693,
      "step": 3939
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9442,
      "step": 3940
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 3941
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 3942
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9228,
      "step": 3943
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.4507,
      "step": 3944
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6448,
      "step": 3945
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9261,
      "step": 3946
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8647,
      "step": 3947
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 3948
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6935,
      "step": 3949
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 1.1047,
      "step": 3950
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 3951
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9353,
      "step": 3952
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7937,
      "step": 3953
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9495,
      "step": 3954
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8424,
      "step": 3955
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6798,
      "step": 3956
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 1.117,
      "step": 3957
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.935,
      "step": 3958
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7301,
      "step": 3959
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6926,
      "step": 3960
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9344,
      "step": 3961
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 3962
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 3963
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8308,
      "step": 3964
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 3965
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 3966
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7405,
      "step": 3967
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8253,
      "step": 3968
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8218,
      "step": 3969
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6495,
      "step": 3970
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 3971
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8966,
      "step": 3972
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7902,
      "step": 3973
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6365,
      "step": 3974
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8687,
      "step": 3975
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 3976
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.6106,
      "step": 3977
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.5016,
      "step": 3978
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.906,
      "step": 3979
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8617,
      "step": 3980
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 3981
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 3982
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 1.191,
      "step": 3983
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8554,
      "step": 3984
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.4396,
      "step": 3985
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.9333,
      "step": 3986
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8592,
      "step": 3987
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6116,
      "step": 3988
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7387,
      "step": 3989
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8976,
      "step": 3990
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6268,
      "step": 3991
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8299,
      "step": 3992
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 3993
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8059,
      "step": 3994
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7929,
      "step": 3995
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.9452,
      "step": 3996
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 1.008,
      "step": 3997
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8184,
      "step": 3998
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 3999
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.9196,
      "step": 4000
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.7537485957145691,
      "eval_runtime": 131.7422,
      "eval_samples_per_second": 7.591,
      "eval_steps_per_second": 3.795,
      "step": 4000
    },
    {
      "epoch": 0.66,
      "mmlu_eval_accuracy": 0.5449337961098338,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "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": 0.4166666666666667,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 0.9393934273299598,
      "step": 4000
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8735,
      "step": 4001
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8306,
      "step": 4002
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6419,
      "step": 4003
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.4954,
      "step": 4004
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.606,
      "step": 4005
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 4006
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 1.1135,
      "step": 4007
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.4582,
      "step": 4008
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 4009
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 1.1929,
      "step": 4010
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6899,
      "step": 4011
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7652,
      "step": 4012
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.4438,
      "step": 4013
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6595,
      "step": 4014
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8755,
      "step": 4015
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.997,
      "step": 4016
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 4017
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 4018
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 4019
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8957,
      "step": 4020
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7544,
      "step": 4021
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.9787,
      "step": 4022
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.4722,
      "step": 4023
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8886,
      "step": 4024
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6733,
      "step": 4025
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6,
      "step": 4026
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8649,
      "step": 4027
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7694,
      "step": 4028
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7215,
      "step": 4029
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.9129,
      "step": 4030
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 4031
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 4032
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 4033
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 1.0315,
      "step": 4034
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 1.0094,
      "step": 4035
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.481,
      "step": 4036
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.7125,
      "step": 4037
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.5605,
      "step": 4038
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6978,
      "step": 4039
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 4040
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 4041
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8544,
      "step": 4042
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.544,
      "step": 4043
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.5517,
      "step": 4044
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.9996,
      "step": 4045
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 4046
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.892,
      "step": 4047
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 4048
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.045,
      "step": 4049
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.0,
      "step": 4050
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.803,
      "step": 4051
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 4052
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.951,
      "step": 4053
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6267,
      "step": 4054
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.3254,
      "step": 4055
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 4056
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6778,
      "step": 4057
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.1308,
      "step": 4058
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6028,
      "step": 4059
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7107,
      "step": 4060
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 4061
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7788,
      "step": 4062
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5525,
      "step": 4063
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7175,
      "step": 4064
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7034,
      "step": 4065
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.9934,
      "step": 4066
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.0798,
      "step": 4067
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6854,
      "step": 4068
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.3544,
      "step": 4069
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.35,
      "step": 4070
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6968,
      "step": 4071
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5759,
      "step": 4072
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.9113,
      "step": 4073
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.0737,
      "step": 4074
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8199,
      "step": 4075
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6423,
      "step": 4076
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8007,
      "step": 4077
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5192,
      "step": 4078
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5552,
      "step": 4079
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 4080
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 4081
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7497,
      "step": 4082
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 4083
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7074,
      "step": 4084
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.0309,
      "step": 4085
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8056,
      "step": 4086
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8165,
      "step": 4087
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.872,
      "step": 4088
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6163,
      "step": 4089
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8407,
      "step": 4090
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 4091
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.7988,
      "step": 4092
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6226,
      "step": 4093
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 1.0535,
      "step": 4094
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8436,
      "step": 4095
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5542,
      "step": 4096
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8667,
      "step": 4097
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.9178,
      "step": 4098
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8791,
      "step": 4099
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.5041,
      "step": 4100
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.6645,
      "step": 4101
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.968,
      "step": 4102
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.4261,
      "step": 4103
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8508,
      "step": 4104
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 4105
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.4507,
      "step": 4106
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.982,
      "step": 4107
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.8964,
      "step": 4108
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9159,
      "step": 4109
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7705,
      "step": 4110
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7375,
      "step": 4111
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9107,
      "step": 4112
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 4113
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8554,
      "step": 4114
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 4115
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6024,
      "step": 4116
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.5505,
      "step": 4117
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 4118
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.3902,
      "step": 4119
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6852,
      "step": 4120
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8564,
      "step": 4121
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6535,
      "step": 4122
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8605,
      "step": 4123
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 4124
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6501,
      "step": 4125
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 4126
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8045,
      "step": 4127
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6482,
      "step": 4128
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.5674,
      "step": 4129
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 4130
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9681,
      "step": 4131
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 4132
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9622,
      "step": 4133
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 4134
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6511,
      "step": 4135
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 4136
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8652,
      "step": 4137
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 4138
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.834,
      "step": 4139
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7341,
      "step": 4140
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 4141
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8801,
      "step": 4142
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7538,
      "step": 4143
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 4144
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 4145
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.2172,
      "step": 4146
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 4147
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8717,
      "step": 4148
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.003,
      "step": 4149
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9082,
      "step": 4150
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9477,
      "step": 4151
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.9291,
      "step": 4152
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.0087,
      "step": 4153
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.063,
      "step": 4154
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.849,
      "step": 4155
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.0245,
      "step": 4156
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.4341,
      "step": 4157
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7575,
      "step": 4158
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8221,
      "step": 4159
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 4160
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.4986,
      "step": 4161
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 1.0502,
      "step": 4162
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.921,
      "step": 4163
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 4164
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7228,
      "step": 4165
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7494,
      "step": 4166
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 4167
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 4168
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.4793,
      "step": 4169
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8861,
      "step": 4170
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8571,
      "step": 4171
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.1305,
      "step": 4172
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 4173
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.5141,
      "step": 4174
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.918,
      "step": 4175
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9921,
      "step": 4176
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7473,
      "step": 4177
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 4178
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7158,
      "step": 4179
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8161,
      "step": 4180
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8173,
      "step": 4181
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.768,
      "step": 4182
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.0386,
      "step": 4183
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7681,
      "step": 4184
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7326,
      "step": 4185
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9131,
      "step": 4186
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6131,
      "step": 4187
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.1967,
      "step": 4188
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7489,
      "step": 4189
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 4190
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8456,
      "step": 4191
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 4192
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8159,
      "step": 4193
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.0059,
      "step": 4194
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.019,
      "step": 4195
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6476,
      "step": 4196
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9292,
      "step": 4197
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6789,
      "step": 4198
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 4199
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 4200
    },
    {
      "epoch": 0.69,
      "eval_loss": 0.7533943057060242,
      "eval_runtime": 131.6714,
      "eval_samples_per_second": 7.595,
      "eval_steps_per_second": 3.797,
      "step": 4200
    },
    {
      "epoch": 0.69,
      "mmlu_eval_accuracy": 0.5644995419094495,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "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": 0.8636363636363636,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.0887044936262908,
      "step": 4200
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7088,
      "step": 4201
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.826,
      "step": 4202
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9575,
      "step": 4203
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 4204
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6873,
      "step": 4205
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8586,
      "step": 4206
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.183,
      "step": 4207
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9128,
      "step": 4208
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.879,
      "step": 4209
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 1.0587,
      "step": 4210
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6885,
      "step": 4211
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8904,
      "step": 4212
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8077,
      "step": 4213
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 4214
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8644,
      "step": 4215
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7677,
      "step": 4216
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9061,
      "step": 4217
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.5827,
      "step": 4218
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7322,
      "step": 4219
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.9292,
      "step": 4220
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6685,
      "step": 4221
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 4222
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7416,
      "step": 4223
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.8206,
      "step": 4224
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.947,
      "step": 4225
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7442,
      "step": 4226
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.6038,
      "step": 4227
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7065,
      "step": 4228
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 4229
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9411,
      "step": 4230
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 4231
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 1.0305,
      "step": 4232
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6893,
      "step": 4233
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.728,
      "step": 4234
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9497,
      "step": 4235
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 4236
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.5091,
      "step": 4237
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 4238
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7998,
      "step": 4239
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9126,
      "step": 4240
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8319,
      "step": 4241
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 4242
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7115,
      "step": 4243
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 4244
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6007,
      "step": 4245
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6406,
      "step": 4246
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7189,
      "step": 4247
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6078,
      "step": 4248
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7395,
      "step": 4249
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 4250
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6547,
      "step": 4251
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 4252
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7467,
      "step": 4253
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 4254
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7927,
      "step": 4255
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 4256
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8082,
      "step": 4257
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7945,
      "step": 4258
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.5583,
      "step": 4259
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7802,
      "step": 4260
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7948,
      "step": 4261
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7878,
      "step": 4262
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.5858,
      "step": 4263
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 4264
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8607,
      "step": 4265
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 4266
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8997,
      "step": 4267
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9369,
      "step": 4268
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8915,
      "step": 4269
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7083,
      "step": 4270
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 1.1058,
      "step": 4271
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6546,
      "step": 4272
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7363,
      "step": 4273
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 4274
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9282,
      "step": 4275
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 4276
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 4277
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 4278
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8363,
      "step": 4279
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 4280
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 4281
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7878,
      "step": 4282
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9379,
      "step": 4283
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.5127,
      "step": 4284
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6401,
      "step": 4285
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.7128,
      "step": 4286
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8791,
      "step": 4287
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.9435,
      "step": 4288
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 4289
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 4290
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 4291
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8362,
      "step": 4292
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7557,
      "step": 4293
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8104,
      "step": 4294
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 4295
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6191,
      "step": 4296
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7229,
      "step": 4297
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 4298
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9561,
      "step": 4299
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7094,
      "step": 4300
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7046,
      "step": 4301
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7452,
      "step": 4302
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7965,
      "step": 4303
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9377,
      "step": 4304
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8339,
      "step": 4305
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.5201,
      "step": 4306
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.693,
      "step": 4307
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.4841,
      "step": 4308
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8746,
      "step": 4309
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7905,
      "step": 4310
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8549,
      "step": 4311
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 4312
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7492,
      "step": 4313
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8333,
      "step": 4314
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 4315
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.5097,
      "step": 4316
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.5333,
      "step": 4317
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8874,
      "step": 4318
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7506,
      "step": 4319
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.899,
      "step": 4320
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 4321
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 4322
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8668,
      "step": 4323
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 4324
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8459,
      "step": 4325
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6318,
      "step": 4326
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9774,
      "step": 4327
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7431,
      "step": 4328
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7021,
      "step": 4329
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 4330
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9133,
      "step": 4331
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9535,
      "step": 4332
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 4333
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.5714,
      "step": 4334
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8052,
      "step": 4335
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 1.0769,
      "step": 4336
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8726,
      "step": 4337
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8605,
      "step": 4338
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.862,
      "step": 4339
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8431,
      "step": 4340
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7999,
      "step": 4341
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8344,
      "step": 4342
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8283,
      "step": 4343
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.9465,
      "step": 4344
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 4345
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 4346
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.6611,
      "step": 4347
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7626,
      "step": 4348
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 4349
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 4350
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 4351
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 1.0379,
      "step": 4352
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 4353
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.865,
      "step": 4354
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8033,
      "step": 4355
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7056,
      "step": 4356
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8245,
      "step": 4357
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6756,
      "step": 4358
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 4359
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8485,
      "step": 4360
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 4361
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 4362
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 4363
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8121,
      "step": 4364
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6454,
      "step": 4365
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.5119,
      "step": 4366
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7997,
      "step": 4367
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7839,
      "step": 4368
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9646,
      "step": 4369
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7841,
      "step": 4370
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 1.0305,
      "step": 4371
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 4372
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8418,
      "step": 4373
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.781,
      "step": 4374
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9208,
      "step": 4375
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 1.0694,
      "step": 4376
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6828,
      "step": 4377
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7383,
      "step": 4378
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8482,
      "step": 4379
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8074,
      "step": 4380
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.74,
      "step": 4381
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 4382
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6981,
      "step": 4383
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7824,
      "step": 4384
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6468,
      "step": 4385
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 1.186,
      "step": 4386
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7951,
      "step": 4387
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9237,
      "step": 4388
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9582,
      "step": 4389
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 4390
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.3732,
      "step": 4391
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6837,
      "step": 4392
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 4393
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8395,
      "step": 4394
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.632,
      "step": 4395
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8551,
      "step": 4396
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8718,
      "step": 4397
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 4398
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6758,
      "step": 4399
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7194,
      "step": 4400
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.7532736659049988,
      "eval_runtime": 131.6439,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 4400
    },
    {
      "epoch": 0.72,
      "mmlu_eval_accuracy": 0.5489116992002624,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3764705882352941,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1256680759764215,
      "step": 4400
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 4401
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7805,
      "step": 4402
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9821,
      "step": 4403
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.9217,
      "step": 4404
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.4598,
      "step": 4405
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 4406
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7677,
      "step": 4407
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 1.0347,
      "step": 4408
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 4409
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6557,
      "step": 4410
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 4411
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.7251,
      "step": 4412
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 4413
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 4414
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8984,
      "step": 4415
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6837,
      "step": 4416
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7705,
      "step": 4417
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8128,
      "step": 4418
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.5027,
      "step": 4419
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9676,
      "step": 4420
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9938,
      "step": 4421
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 1.0139,
      "step": 4422
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8246,
      "step": 4423
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.4546,
      "step": 4424
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 4425
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9061,
      "step": 4426
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6387,
      "step": 4427
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.5553,
      "step": 4428
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8795,
      "step": 4429
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6621,
      "step": 4430
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7296,
      "step": 4431
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8448,
      "step": 4432
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.918,
      "step": 4433
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8295,
      "step": 4434
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.794,
      "step": 4435
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 4436
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 4437
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 4438
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7756,
      "step": 4439
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8451,
      "step": 4440
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.5788,
      "step": 4441
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9984,
      "step": 4442
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7915,
      "step": 4443
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 4444
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 4445
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7619,
      "step": 4446
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.5191,
      "step": 4447
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 4448
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7875,
      "step": 4449
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 4450
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6226,
      "step": 4451
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 4452
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.821,
      "step": 4453
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 4454
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 4455
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.61,
      "step": 4456
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 4457
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9688,
      "step": 4458
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8398,
      "step": 4459
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9748,
      "step": 4460
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6699,
      "step": 4461
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 4462
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.8977,
      "step": 4463
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7809,
      "step": 4464
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 4465
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6417,
      "step": 4466
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 1.0545,
      "step": 4467
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.9741,
      "step": 4468
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7331,
      "step": 4469
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 4470
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.811,
      "step": 4471
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7116,
      "step": 4472
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.6309,
      "step": 4473
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 4474
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.5559,
      "step": 4475
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8749,
      "step": 4476
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8352,
      "step": 4477
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.724,
      "step": 4478
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.692,
      "step": 4479
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9015,
      "step": 4480
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.5229,
      "step": 4481
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7055,
      "step": 4482
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7741,
      "step": 4483
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8178,
      "step": 4484
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 4485
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 4486
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 4487
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9261,
      "step": 4488
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7099,
      "step": 4489
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 4490
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 4491
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 4492
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9413,
      "step": 4493
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9017,
      "step": 4494
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8861,
      "step": 4495
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8836,
      "step": 4496
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7972,
      "step": 4497
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8568,
      "step": 4498
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7918,
      "step": 4499
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 4500
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7749,
      "step": 4501
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 1.0638,
      "step": 4502
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8119,
      "step": 4503
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8776,
      "step": 4504
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7677,
      "step": 4505
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7324,
      "step": 4506
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7398,
      "step": 4507
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.732,
      "step": 4508
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 4509
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.772,
      "step": 4510
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.886,
      "step": 4511
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.6252,
      "step": 4512
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 1.0659,
      "step": 4513
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 4514
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.6815,
      "step": 4515
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9449,
      "step": 4516
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7396,
      "step": 4517
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 1.0366,
      "step": 4518
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.6383,
      "step": 4519
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 4520
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.916,
      "step": 4521
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 4522
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.4668,
      "step": 4523
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7961,
      "step": 4524
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.6233,
      "step": 4525
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.5247,
      "step": 4526
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9137,
      "step": 4527
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 4528
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 1.1018,
      "step": 4529
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 4530
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 4531
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 4532
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.9354,
      "step": 4533
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.6268,
      "step": 4534
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.864,
      "step": 4535
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.576,
      "step": 4536
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7936,
      "step": 4537
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 1.0962,
      "step": 4538
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 4539
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7947,
      "step": 4540
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6836,
      "step": 4541
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6667,
      "step": 4542
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9211,
      "step": 4543
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 4544
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 4545
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9592,
      "step": 4546
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.906,
      "step": 4547
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6151,
      "step": 4548
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 1.1113,
      "step": 4549
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.723,
      "step": 4550
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8942,
      "step": 4551
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8589,
      "step": 4552
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6644,
      "step": 4553
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6037,
      "step": 4554
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 4555
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 4556
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9568,
      "step": 4557
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9693,
      "step": 4558
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 1.0103,
      "step": 4559
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8603,
      "step": 4560
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7311,
      "step": 4561
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8597,
      "step": 4562
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.684,
      "step": 4563
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.825,
      "step": 4564
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 4565
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6409,
      "step": 4566
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8368,
      "step": 4567
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7797,
      "step": 4568
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.4905,
      "step": 4569
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8342,
      "step": 4570
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9027,
      "step": 4571
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.754,
      "step": 4572
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 4573
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7125,
      "step": 4574
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6785,
      "step": 4575
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7762,
      "step": 4576
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6681,
      "step": 4577
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9039,
      "step": 4578
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 4579
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9517,
      "step": 4580
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9827,
      "step": 4581
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 1.1242,
      "step": 4582
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7676,
      "step": 4583
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 4584
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8196,
      "step": 4585
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 4586
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 4587
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.9508,
      "step": 4588
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6084,
      "step": 4589
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 4590
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6949,
      "step": 4591
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.7422,
      "step": 4592
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8267,
      "step": 4593
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6379,
      "step": 4594
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 4595
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.5569,
      "step": 4596
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7502,
      "step": 4597
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7659,
      "step": 4598
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8215,
      "step": 4599
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8463,
      "step": 4600
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.7545199990272522,
      "eval_runtime": 131.6197,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 4600
    },
    {
      "epoch": 0.76,
      "mmlu_eval_accuracy": 0.5579345039363026,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.47058823529411764,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "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": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6764705882352942,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5942028985507246,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1306793517372316,
      "step": 4600
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7911,
      "step": 4601
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 4602
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 4603
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 4604
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.5599,
      "step": 4605
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8866,
      "step": 4606
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7925,
      "step": 4607
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.4836,
      "step": 4608
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.629,
      "step": 4609
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8442,
      "step": 4610
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8111,
      "step": 4611
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.723,
      "step": 4612
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7955,
      "step": 4613
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 4614
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 4615
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.5537,
      "step": 4616
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7414,
      "step": 4617
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9449,
      "step": 4618
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.959,
      "step": 4619
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8925,
      "step": 4620
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.5446,
      "step": 4621
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 4622
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9294,
      "step": 4623
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 1.0393,
      "step": 4624
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8364,
      "step": 4625
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8019,
      "step": 4626
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8926,
      "step": 4627
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7648,
      "step": 4628
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9909,
      "step": 4629
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.4993,
      "step": 4630
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8648,
      "step": 4631
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7263,
      "step": 4632
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7969,
      "step": 4633
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 4634
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7987,
      "step": 4635
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8769,
      "step": 4636
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6689,
      "step": 4637
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6925,
      "step": 4638
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.5434,
      "step": 4639
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9533,
      "step": 4640
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9459,
      "step": 4641
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6551,
      "step": 4642
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8431,
      "step": 4643
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 4644
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8077,
      "step": 4645
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8904,
      "step": 4646
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.6417,
      "step": 4647
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.9134,
      "step": 4648
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7827,
      "step": 4649
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.901,
      "step": 4650
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.4452,
      "step": 4651
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 1.049,
      "step": 4652
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 1.0046,
      "step": 4653
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 4654
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.7016,
      "step": 4655
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7885,
      "step": 4656
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8188,
      "step": 4657
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7878,
      "step": 4658
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5348,
      "step": 4659
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 4660
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9717,
      "step": 4661
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9968,
      "step": 4662
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 4663
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9584,
      "step": 4664
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 4665
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8323,
      "step": 4666
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5385,
      "step": 4667
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8671,
      "step": 4668
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 1.2198,
      "step": 4669
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8682,
      "step": 4670
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.693,
      "step": 4671
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 4672
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5764,
      "step": 4673
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6759,
      "step": 4674
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 4675
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 4676
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8198,
      "step": 4677
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9619,
      "step": 4678
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9287,
      "step": 4679
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5996,
      "step": 4680
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 4681
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8839,
      "step": 4682
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 4683
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6976,
      "step": 4684
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9371,
      "step": 4685
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6739,
      "step": 4686
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9652,
      "step": 4687
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 4688
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5818,
      "step": 4689
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5674,
      "step": 4690
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.806,
      "step": 4691
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7965,
      "step": 4692
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6628,
      "step": 4693
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9155,
      "step": 4694
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8472,
      "step": 4695
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7586,
      "step": 4696
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5,
      "step": 4697
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6968,
      "step": 4698
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.5077,
      "step": 4699
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.4783,
      "step": 4700
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8218,
      "step": 4701
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7666,
      "step": 4702
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 1.222,
      "step": 4703
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 1.1747,
      "step": 4704
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9013,
      "step": 4705
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8178,
      "step": 4706
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7968,
      "step": 4707
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6571,
      "step": 4708
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 4709
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.9057,
      "step": 4710
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7577,
      "step": 4711
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.6935,
      "step": 4712
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 1.1719,
      "step": 4713
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7239,
      "step": 4714
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8716,
      "step": 4715
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.7261,
      "step": 4716
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 4717
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8988,
      "step": 4718
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 4719
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5656,
      "step": 4720
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5795,
      "step": 4721
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8601,
      "step": 4722
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9158,
      "step": 4723
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 4724
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.908,
      "step": 4725
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8905,
      "step": 4726
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.831,
      "step": 4727
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 4728
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8823,
      "step": 4729
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6111,
      "step": 4730
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 4731
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.855,
      "step": 4732
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.489,
      "step": 4733
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 4734
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7705,
      "step": 4735
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.649,
      "step": 4736
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6189,
      "step": 4737
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 4738
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7817,
      "step": 4739
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6566,
      "step": 4740
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9041,
      "step": 4741
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.4254,
      "step": 4742
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.0887,
      "step": 4743
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.4048,
      "step": 4744
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7499,
      "step": 4745
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 4746
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 4747
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9799,
      "step": 4748
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7086,
      "step": 4749
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5226,
      "step": 4750
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7229,
      "step": 4751
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 4752
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.123,
      "step": 4753
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9267,
      "step": 4754
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 4755
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.0838,
      "step": 4756
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9511,
      "step": 4757
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 4758
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8848,
      "step": 4759
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5661,
      "step": 4760
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.634,
      "step": 4761
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9702,
      "step": 4762
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7348,
      "step": 4763
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7856,
      "step": 4764
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.0825,
      "step": 4765
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6216,
      "step": 4766
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5991,
      "step": 4767
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 4768
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 4769
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.5324,
      "step": 4770
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.908,
      "step": 4771
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 1.0117,
      "step": 4772
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 4773
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.6918,
      "step": 4774
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8191,
      "step": 4775
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.9084,
      "step": 4776
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.711,
      "step": 4777
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 4778
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5487,
      "step": 4779
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5357,
      "step": 4780
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6629,
      "step": 4781
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6267,
      "step": 4782
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7728,
      "step": 4783
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 4784
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.696,
      "step": 4785
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 4786
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9569,
      "step": 4787
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6884,
      "step": 4788
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6987,
      "step": 4789
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 4790
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8554,
      "step": 4791
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8516,
      "step": 4792
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6886,
      "step": 4793
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5816,
      "step": 4794
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 4795
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.822,
      "step": 4796
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5578,
      "step": 4797
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7181,
      "step": 4798
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9475,
      "step": 4799
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8014,
      "step": 4800
    },
    {
      "epoch": 0.79,
      "eval_loss": 0.7526820302009583,
      "eval_runtime": 131.6004,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.799,
      "step": 4800
    },
    {
      "epoch": 0.79,
      "mmlu_eval_accuracy": 0.5392650576015245,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.5454545454545454,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "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": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
      "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": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.26,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "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": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.2467338922129598,
      "step": 4800
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.4966,
      "step": 4801
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6529,
      "step": 4802
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8714,
      "step": 4803
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5789,
      "step": 4804
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.903,
      "step": 4805
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 4806
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6975,
      "step": 4807
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9127,
      "step": 4808
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7049,
      "step": 4809
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 1.0193,
      "step": 4810
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 4811
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.428,
      "step": 4812
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8916,
      "step": 4813
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.5671,
      "step": 4814
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7441,
      "step": 4815
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 4816
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7915,
      "step": 4817
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7546,
      "step": 4818
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6821,
      "step": 4819
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 4820
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.3551,
      "step": 4821
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7464,
      "step": 4822
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9629,
      "step": 4823
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 1.0849,
      "step": 4824
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.4646,
      "step": 4825
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 4826
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6032,
      "step": 4827
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 4828
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.725,
      "step": 4829
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.9246,
      "step": 4830
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6965,
      "step": 4831
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6399,
      "step": 4832
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 4833
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8358,
      "step": 4834
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7491,
      "step": 4835
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.6326,
      "step": 4836
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 1.1249,
      "step": 4837
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.7432,
      "step": 4838
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 1.0147,
      "step": 4839
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7459,
      "step": 4840
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 4841
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 4842
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9504,
      "step": 4843
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8959,
      "step": 4844
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.461,
      "step": 4845
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6699,
      "step": 4846
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4154,
      "step": 4847
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9485,
      "step": 4848
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 4849
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8109,
      "step": 4850
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8148,
      "step": 4851
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4654,
      "step": 4852
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8444,
      "step": 4853
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 4854
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6692,
      "step": 4855
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7235,
      "step": 4856
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7402,
      "step": 4857
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7794,
      "step": 4858
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9156,
      "step": 4859
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8724,
      "step": 4860
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6646,
      "step": 4861
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6552,
      "step": 4862
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8261,
      "step": 4863
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7549,
      "step": 4864
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4587,
      "step": 4865
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9461,
      "step": 4866
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9087,
      "step": 4867
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 4868
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8056,
      "step": 4869
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9365,
      "step": 4870
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6373,
      "step": 4871
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9882,
      "step": 4872
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 4873
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7489,
      "step": 4874
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 1.1003,
      "step": 4875
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6933,
      "step": 4876
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7414,
      "step": 4877
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4158,
      "step": 4878
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.648,
      "step": 4879
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7928,
      "step": 4880
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 4881
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7973,
      "step": 4882
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4256,
      "step": 4883
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9049,
      "step": 4884
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.907,
      "step": 4885
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6737,
      "step": 4886
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 1.1773,
      "step": 4887
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 4888
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7401,
      "step": 4889
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7418,
      "step": 4890
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 4891
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 4892
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 4893
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7898,
      "step": 4894
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 4895
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.7071,
      "step": 4896
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8681,
      "step": 4897
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.6333,
      "step": 4898
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 4899
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7741,
      "step": 4900
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 1.0184,
      "step": 4901
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9164,
      "step": 4902
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 4903
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6396,
      "step": 4904
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 1.2245,
      "step": 4905
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8713,
      "step": 4906
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.996,
      "step": 4907
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8079,
      "step": 4908
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9805,
      "step": 4909
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 4910
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9672,
      "step": 4911
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 4912
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7811,
      "step": 4913
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 1.0604,
      "step": 4914
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 4915
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9942,
      "step": 4916
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.948,
      "step": 4917
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7708,
      "step": 4918
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.644,
      "step": 4919
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8183,
      "step": 4920
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9998,
      "step": 4921
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6516,
      "step": 4922
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8257,
      "step": 4923
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8655,
      "step": 4924
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8193,
      "step": 4925
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6359,
      "step": 4926
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6756,
      "step": 4927
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.865,
      "step": 4928
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6023,
      "step": 4929
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8107,
      "step": 4930
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 4931
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5452,
      "step": 4932
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8593,
      "step": 4933
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 4934
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 4935
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7644,
      "step": 4936
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9559,
      "step": 4937
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 4938
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 4939
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7025,
      "step": 4940
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8732,
      "step": 4941
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8613,
      "step": 4942
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7975,
      "step": 4943
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 1.0406,
      "step": 4944
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 4945
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7925,
      "step": 4946
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 4947
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 4948
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.892,
      "step": 4949
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5657,
      "step": 4950
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 4951
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.6019,
      "step": 4952
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.4724,
      "step": 4953
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9311,
      "step": 4954
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7948,
      "step": 4955
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.7516,
      "step": 4956
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.4601,
      "step": 4957
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 4958
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.5278,
      "step": 4959
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.55,
      "step": 4960
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8358,
      "step": 4961
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.904,
      "step": 4962
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8669,
      "step": 4963
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6948,
      "step": 4964
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9742,
      "step": 4965
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8455,
      "step": 4966
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9986,
      "step": 4967
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 1.1176,
      "step": 4968
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7747,
      "step": 4969
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7824,
      "step": 4970
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 1.1169,
      "step": 4971
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.44,
      "step": 4972
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.4928,
      "step": 4973
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 4974
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 4975
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8774,
      "step": 4976
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8368,
      "step": 4977
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.447,
      "step": 4978
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.795,
      "step": 4979
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7965,
      "step": 4980
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7807,
      "step": 4981
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8919,
      "step": 4982
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9949,
      "step": 4983
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8772,
      "step": 4984
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 4985
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7467,
      "step": 4986
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7547,
      "step": 4987
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 4988
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6139,
      "step": 4989
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8463,
      "step": 4990
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6306,
      "step": 4991
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 4992
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.5543,
      "step": 4993
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 1.1244,
      "step": 4994
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9058,
      "step": 4995
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6997,
      "step": 4996
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9399,
      "step": 4997
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7832,
      "step": 4998
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6344,
      "step": 4999
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.691,
      "step": 5000
    },
    {
      "epoch": 0.82,
      "eval_loss": 0.7546924948692322,
      "eval_runtime": 131.6828,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 5000
    },
    {
      "epoch": 0.82,
      "mmlu_eval_accuracy": 0.5418852977811036,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
      "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": 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": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.37058823529411766,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.0400860130047673,
      "step": 5000
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 5001
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7547,
      "step": 5002
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 5003
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9166,
      "step": 5004
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8079,
      "step": 5005
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.9695,
      "step": 5006
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8322,
      "step": 5007
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.919,
      "step": 5008
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6851,
      "step": 5009
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6177,
      "step": 5010
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.869,
      "step": 5011
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.612,
      "step": 5012
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.7186,
      "step": 5013
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8265,
      "step": 5014
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8337,
      "step": 5015
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 5016
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8563,
      "step": 5017
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 5018
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8611,
      "step": 5019
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.6829,
      "step": 5020
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8569,
      "step": 5021
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6617,
      "step": 5022
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6274,
      "step": 5023
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.534,
      "step": 5024
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 1.1059,
      "step": 5025
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8075,
      "step": 5026
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9163,
      "step": 5027
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9912,
      "step": 5028
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6242,
      "step": 5029
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9923,
      "step": 5030
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 5031
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8626,
      "step": 5032
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.3738,
      "step": 5033
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9007,
      "step": 5034
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9533,
      "step": 5035
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7739,
      "step": 5036
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9124,
      "step": 5037
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5033,
      "step": 5038
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 5039
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8114,
      "step": 5040
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 1.0448,
      "step": 5041
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6149,
      "step": 5042
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7747,
      "step": 5043
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7328,
      "step": 5044
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7047,
      "step": 5045
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9814,
      "step": 5046
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 1.0747,
      "step": 5047
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7094,
      "step": 5048
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9872,
      "step": 5049
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9597,
      "step": 5050
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7743,
      "step": 5051
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6882,
      "step": 5052
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5816,
      "step": 5053
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 5054
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8066,
      "step": 5055
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8358,
      "step": 5056
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 5057
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9675,
      "step": 5058
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7536,
      "step": 5059
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9672,
      "step": 5060
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8778,
      "step": 5061
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.531,
      "step": 5062
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7825,
      "step": 5063
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8806,
      "step": 5064
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9126,
      "step": 5065
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 5066
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 5067
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.7333,
      "step": 5068
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.4986,
      "step": 5069
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5319,
      "step": 5070
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8008,
      "step": 5071
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.9164,
      "step": 5072
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8599,
      "step": 5073
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5823,
      "step": 5074
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5219,
      "step": 5075
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.5349,
      "step": 5076
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 5077
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8132,
      "step": 5078
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8611,
      "step": 5079
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 5080
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8415,
      "step": 5081
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.899,
      "step": 5082
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.71,
      "step": 5083
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.908,
      "step": 5084
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8432,
      "step": 5085
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6448,
      "step": 5086
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 5087
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7631,
      "step": 5088
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 5089
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6807,
      "step": 5090
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 5091
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7977,
      "step": 5092
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8451,
      "step": 5093
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9068,
      "step": 5094
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9989,
      "step": 5095
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8002,
      "step": 5096
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 5097
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 5098
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.4009,
      "step": 5099
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8492,
      "step": 5100
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.0456,
      "step": 5101
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.0063,
      "step": 5102
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7048,
      "step": 5103
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.93,
      "step": 5104
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.1099,
      "step": 5105
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.0477,
      "step": 5106
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9419,
      "step": 5107
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 5108
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.782,
      "step": 5109
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.0007,
      "step": 5110
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 5111
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8626,
      "step": 5112
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8566,
      "step": 5113
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.821,
      "step": 5114
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6982,
      "step": 5115
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8734,
      "step": 5116
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6952,
      "step": 5117
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6359,
      "step": 5118
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8674,
      "step": 5119
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9601,
      "step": 5120
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8296,
      "step": 5121
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 5122
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6152,
      "step": 5123
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9122,
      "step": 5124
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.3275,
      "step": 5125
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.1467,
      "step": 5126
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8109,
      "step": 5127
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8532,
      "step": 5128
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8827,
      "step": 5129
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6189,
      "step": 5130
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8095,
      "step": 5131
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 5132
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7072,
      "step": 5133
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8098,
      "step": 5134
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6731,
      "step": 5135
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8144,
      "step": 5136
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.0108,
      "step": 5137
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.6936,
      "step": 5138
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.7974,
      "step": 5139
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 5140
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.9056,
      "step": 5141
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 1.1622,
      "step": 5142
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.4363,
      "step": 5143
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8767,
      "step": 5144
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 5145
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.591,
      "step": 5146
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 5147
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8133,
      "step": 5148
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.869,
      "step": 5149
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.5649,
      "step": 5150
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6343,
      "step": 5151
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.5792,
      "step": 5152
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 5153
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7543,
      "step": 5154
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7557,
      "step": 5155
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6972,
      "step": 5156
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 5157
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 5158
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8267,
      "step": 5159
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.978,
      "step": 5160
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.5671,
      "step": 5161
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6536,
      "step": 5162
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 5163
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7226,
      "step": 5164
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.4204,
      "step": 5165
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6504,
      "step": 5166
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8797,
      "step": 5167
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6669,
      "step": 5168
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 5169
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6961,
      "step": 5170
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7178,
      "step": 5171
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.9813,
      "step": 5172
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 5173
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 5174
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7244,
      "step": 5175
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 1.0433,
      "step": 5176
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 5177
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7653,
      "step": 5178
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 5179
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 5180
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6232,
      "step": 5181
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8279,
      "step": 5182
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 5183
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.613,
      "step": 5184
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7917,
      "step": 5185
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7079,
      "step": 5186
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 1.2307,
      "step": 5187
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8599,
      "step": 5188
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.65,
      "step": 5189
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7754,
      "step": 5190
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.6245,
      "step": 5191
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.72,
      "step": 5192
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.5067,
      "step": 5193
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8264,
      "step": 5194
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8361,
      "step": 5195
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8509,
      "step": 5196
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7017,
      "step": 5197
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7595,
      "step": 5198
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8285,
      "step": 5199
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 1.0161,
      "step": 5200
    },
    {
      "epoch": 0.85,
      "eval_loss": 0.7570607662200928,
      "eval_runtime": 131.5936,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.8,
      "step": 5200
    },
    {
      "epoch": 0.85,
      "mmlu_eval_accuracy": 0.5391922125688261,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.75,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
      "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": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.2944577410622615,
      "step": 5200
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.7148,
      "step": 5201
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8881,
      "step": 5202
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 5203
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 5204
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5627,
      "step": 5205
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7415,
      "step": 5206
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8756,
      "step": 5207
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7867,
      "step": 5208
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7876,
      "step": 5209
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.9018,
      "step": 5210
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.662,
      "step": 5211
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8326,
      "step": 5212
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5151,
      "step": 5213
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8341,
      "step": 5214
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6266,
      "step": 5215
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8856,
      "step": 5216
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5794,
      "step": 5217
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 5218
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.3348,
      "step": 5219
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6973,
      "step": 5220
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7916,
      "step": 5221
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 5222
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 1.1066,
      "step": 5223
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 5224
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.9062,
      "step": 5225
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 5226
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 5227
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5829,
      "step": 5228
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8905,
      "step": 5229
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 5230
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8705,
      "step": 5231
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7131,
      "step": 5232
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8611,
      "step": 5233
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 5234
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6996,
      "step": 5235
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.91,
      "step": 5236
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 5237
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 5238
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7117,
      "step": 5239
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8512,
      "step": 5240
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 5241
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7058,
      "step": 5242
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7749,
      "step": 5243
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7576,
      "step": 5244
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 5245
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 5246
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7251,
      "step": 5247
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6967,
      "step": 5248
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 5249
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5726,
      "step": 5250
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6307,
      "step": 5251
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7522,
      "step": 5252
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5172,
      "step": 5253
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.46,
      "step": 5254
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 5255
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 5256
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 5257
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6034,
      "step": 5258
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 5259
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 5260
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.5615,
      "step": 5261
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 5262
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7543,
      "step": 5263
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 5264
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7863,
      "step": 5265
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 1.0483,
      "step": 5266
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 1.0626,
      "step": 5267
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7134,
      "step": 5268
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 1.2534,
      "step": 5269
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 5270
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7735,
      "step": 5271
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7985,
      "step": 5272
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6574,
      "step": 5273
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.5534,
      "step": 5274
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9228,
      "step": 5275
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.845,
      "step": 5276
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6409,
      "step": 5277
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7251,
      "step": 5278
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6542,
      "step": 5279
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8726,
      "step": 5280
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 5281
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.5973,
      "step": 5282
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9301,
      "step": 5283
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.4312,
      "step": 5284
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 5285
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8837,
      "step": 5286
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6669,
      "step": 5287
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 5288
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8299,
      "step": 5289
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6955,
      "step": 5290
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6232,
      "step": 5291
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6172,
      "step": 5292
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7974,
      "step": 5293
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 5294
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 5295
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8934,
      "step": 5296
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 1.0895,
      "step": 5297
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9629,
      "step": 5298
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8588,
      "step": 5299
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.94,
      "step": 5300
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 5301
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9395,
      "step": 5302
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.5857,
      "step": 5303
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6201,
      "step": 5304
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 5305
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6865,
      "step": 5306
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.693,
      "step": 5307
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 5308
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 1.1614,
      "step": 5309
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8339,
      "step": 5310
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6874,
      "step": 5311
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7138,
      "step": 5312
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 5313
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 5314
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7985,
      "step": 5315
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6018,
      "step": 5316
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9183,
      "step": 5317
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.5919,
      "step": 5318
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8873,
      "step": 5319
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 5320
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9526,
      "step": 5321
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 5322
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.7442,
      "step": 5323
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.9682,
      "step": 5324
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.6986,
      "step": 5325
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 1.0101,
      "step": 5326
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 5327
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 5328
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7464,
      "step": 5329
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6972,
      "step": 5330
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8185,
      "step": 5331
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9128,
      "step": 5332
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8034,
      "step": 5333
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9641,
      "step": 5334
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7911,
      "step": 5335
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8467,
      "step": 5336
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 5337
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7157,
      "step": 5338
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8093,
      "step": 5339
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 5340
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.878,
      "step": 5341
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.5611,
      "step": 5342
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9083,
      "step": 5343
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.743,
      "step": 5344
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6457,
      "step": 5345
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7839,
      "step": 5346
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6778,
      "step": 5347
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9833,
      "step": 5348
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.4485,
      "step": 5349
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 5350
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6182,
      "step": 5351
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.801,
      "step": 5352
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.5387,
      "step": 5353
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8203,
      "step": 5354
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.778,
      "step": 5355
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 5356
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6091,
      "step": 5357
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 5358
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7605,
      "step": 5359
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 5360
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9418,
      "step": 5361
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 5362
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.761,
      "step": 5363
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9497,
      "step": 5364
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8418,
      "step": 5365
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6609,
      "step": 5366
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8712,
      "step": 5367
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.5725,
      "step": 5368
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6984,
      "step": 5369
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6018,
      "step": 5370
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6502,
      "step": 5371
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 1.0874,
      "step": 5372
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.5237,
      "step": 5373
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6227,
      "step": 5374
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.4831,
      "step": 5375
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9457,
      "step": 5376
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.704,
      "step": 5377
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.712,
      "step": 5378
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 5379
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9111,
      "step": 5380
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 5381
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6228,
      "step": 5382
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.9846,
      "step": 5383
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8681,
      "step": 5384
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.6221,
      "step": 5385
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8166,
      "step": 5386
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.0206,
      "step": 5387
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8194,
      "step": 5388
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9925,
      "step": 5389
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9388,
      "step": 5390
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6106,
      "step": 5391
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.4923,
      "step": 5392
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 5393
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8685,
      "step": 5394
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 5395
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.894,
      "step": 5396
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 5397
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 5398
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 5399
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8313,
      "step": 5400
    },
    {
      "epoch": 0.89,
      "eval_loss": 0.7573027610778809,
      "eval_runtime": 131.61,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 5400
    },
    {
      "epoch": 0.89,
      "mmlu_eval_accuracy": 0.5535363827788407,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "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": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.7037037037037037,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.0044040840226116,
      "step": 5400
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.0322,
      "step": 5401
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7576,
      "step": 5402
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9805,
      "step": 5403
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8773,
      "step": 5404
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9102,
      "step": 5405
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9518,
      "step": 5406
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 5407
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 5408
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.1114,
      "step": 5409
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 5410
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8599,
      "step": 5411
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9889,
      "step": 5412
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7419,
      "step": 5413
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7471,
      "step": 5414
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 5415
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 5416
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7799,
      "step": 5417
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6111,
      "step": 5418
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7181,
      "step": 5419
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.0414,
      "step": 5420
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7894,
      "step": 5421
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7075,
      "step": 5422
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.0978,
      "step": 5423
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8059,
      "step": 5424
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 5425
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 5426
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9878,
      "step": 5427
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 5428
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.977,
      "step": 5429
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8139,
      "step": 5430
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.4639,
      "step": 5431
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8779,
      "step": 5432
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 5433
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.919,
      "step": 5434
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7097,
      "step": 5435
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 5436
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7733,
      "step": 5437
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 5438
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 5439
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7418,
      "step": 5440
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.6483,
      "step": 5441
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 1.0091,
      "step": 5442
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.5953,
      "step": 5443
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.9923,
      "step": 5444
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8842,
      "step": 5445
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.7261,
      "step": 5446
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8973,
      "step": 5447
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8355,
      "step": 5448
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9119,
      "step": 5449
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6546,
      "step": 5450
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 5451
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7652,
      "step": 5452
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9246,
      "step": 5453
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8217,
      "step": 5454
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6573,
      "step": 5455
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8254,
      "step": 5456
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9003,
      "step": 5457
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8375,
      "step": 5458
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 5459
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6549,
      "step": 5460
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 5461
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 5462
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 5463
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.709,
      "step": 5464
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.909,
      "step": 5465
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 5466
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6139,
      "step": 5467
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 5468
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.5793,
      "step": 5469
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9662,
      "step": 5470
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 5471
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7888,
      "step": 5472
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8934,
      "step": 5473
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8559,
      "step": 5474
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9941,
      "step": 5475
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9985,
      "step": 5476
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9379,
      "step": 5477
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9548,
      "step": 5478
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9312,
      "step": 5479
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8416,
      "step": 5480
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.823,
      "step": 5481
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6526,
      "step": 5482
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7048,
      "step": 5483
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8494,
      "step": 5484
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 5485
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.808,
      "step": 5486
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6146,
      "step": 5487
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8975,
      "step": 5488
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8299,
      "step": 5489
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.508,
      "step": 5490
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8859,
      "step": 5491
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.4986,
      "step": 5492
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 5493
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.5677,
      "step": 5494
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9249,
      "step": 5495
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.9316,
      "step": 5496
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7841,
      "step": 5497
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7884,
      "step": 5498
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.3761,
      "step": 5499
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8021,
      "step": 5500
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 5501
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8181,
      "step": 5502
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8774,
      "step": 5503
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.7942,
      "step": 5504
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8235,
      "step": 5505
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.6374,
      "step": 5506
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.4308,
      "step": 5507
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.4129,
      "step": 5508
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9604,
      "step": 5509
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8313,
      "step": 5510
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7495,
      "step": 5511
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9446,
      "step": 5512
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 5513
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9904,
      "step": 5514
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.989,
      "step": 5515
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6253,
      "step": 5516
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.5869,
      "step": 5517
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8086,
      "step": 5518
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 5519
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 1.1378,
      "step": 5520
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.5666,
      "step": 5521
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 5522
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8578,
      "step": 5523
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6893,
      "step": 5524
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.738,
      "step": 5525
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7215,
      "step": 5526
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6664,
      "step": 5527
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 1.0088,
      "step": 5528
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8572,
      "step": 5529
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9753,
      "step": 5530
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9261,
      "step": 5531
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9845,
      "step": 5532
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7809,
      "step": 5533
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.5954,
      "step": 5534
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7221,
      "step": 5535
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 5536
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8694,
      "step": 5537
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9574,
      "step": 5538
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 5539
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9122,
      "step": 5540
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8863,
      "step": 5541
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.4845,
      "step": 5542
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 5543
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 5544
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.925,
      "step": 5545
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7193,
      "step": 5546
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9306,
      "step": 5547
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 5548
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8898,
      "step": 5549
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7773,
      "step": 5550
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 5551
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8281,
      "step": 5552
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9012,
      "step": 5553
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 5554
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 5555
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9052,
      "step": 5556
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 5557
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.6887,
      "step": 5558
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7351,
      "step": 5559
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9455,
      "step": 5560
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7989,
      "step": 5561
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.733,
      "step": 5562
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8826,
      "step": 5563
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7459,
      "step": 5564
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7546,
      "step": 5565
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 5566
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.7619,
      "step": 5567
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 5568
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 5569
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.1281,
      "step": 5570
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 5571
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.0303,
      "step": 5572
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 5573
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 5574
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9057,
      "step": 5575
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9976,
      "step": 5576
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.0461,
      "step": 5577
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7773,
      "step": 5578
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.942,
      "step": 5579
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8754,
      "step": 5580
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 5581
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8193,
      "step": 5582
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6067,
      "step": 5583
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8549,
      "step": 5584
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 5585
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 5586
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7282,
      "step": 5587
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9881,
      "step": 5588
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.0898,
      "step": 5589
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8647,
      "step": 5590
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 5591
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5586,
      "step": 5592
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.589,
      "step": 5593
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8831,
      "step": 5594
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6811,
      "step": 5595
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 5596
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 5597
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6318,
      "step": 5598
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7892,
      "step": 5599
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 5600
    },
    {
      "epoch": 0.92,
      "eval_loss": 0.7568509578704834,
      "eval_runtime": 131.6466,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 5600
    },
    {
      "epoch": 0.92,
      "mmlu_eval_accuracy": 0.5302614353168894,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.25,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "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": 0.631578947368421,
      "mmlu_loss": 1.2112150888102173,
      "step": 5600
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 5601
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 5602
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6554,
      "step": 5603
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7234,
      "step": 5604
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.4037,
      "step": 5605
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9236,
      "step": 5606
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7781,
      "step": 5607
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.917,
      "step": 5608
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 5609
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5141,
      "step": 5610
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.0792,
      "step": 5611
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9425,
      "step": 5612
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 5613
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.824,
      "step": 5614
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.4964,
      "step": 5615
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 5616
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.5439,
      "step": 5617
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6681,
      "step": 5618
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.987,
      "step": 5619
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9177,
      "step": 5620
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 5621
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.9006,
      "step": 5622
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 5623
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 5624
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 1.3664,
      "step": 5625
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6688,
      "step": 5626
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7017,
      "step": 5627
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 5628
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.6384,
      "step": 5629
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.915,
      "step": 5630
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8656,
      "step": 5631
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 5632
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 5633
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.756,
      "step": 5634
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.532,
      "step": 5635
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7359,
      "step": 5636
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.691,
      "step": 5637
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8208,
      "step": 5638
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8302,
      "step": 5639
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6807,
      "step": 5640
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7715,
      "step": 5641
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 1.1177,
      "step": 5642
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8522,
      "step": 5643
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7695,
      "step": 5644
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8631,
      "step": 5645
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7542,
      "step": 5646
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8105,
      "step": 5647
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6049,
      "step": 5648
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 5649
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6373,
      "step": 5650
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 5651
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 1.0081,
      "step": 5652
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 5653
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 5654
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.5167,
      "step": 5655
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 5656
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7793,
      "step": 5657
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6875,
      "step": 5658
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8738,
      "step": 5659
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.5131,
      "step": 5660
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.4931,
      "step": 5661
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7163,
      "step": 5662
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7372,
      "step": 5663
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6434,
      "step": 5664
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8497,
      "step": 5665
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7592,
      "step": 5666
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.5747,
      "step": 5667
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 5668
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7951,
      "step": 5669
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7659,
      "step": 5670
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6826,
      "step": 5671
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 5672
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 1.0523,
      "step": 5673
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.76,
      "step": 5674
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.973,
      "step": 5675
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 5676
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.9101,
      "step": 5677
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7137,
      "step": 5678
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8016,
      "step": 5679
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8851,
      "step": 5680
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.77,
      "step": 5681
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7092,
      "step": 5682
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 1.1899,
      "step": 5683
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.5347,
      "step": 5684
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 5685
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 5686
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7049,
      "step": 5687
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 5688
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.9036,
      "step": 5689
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.7228,
      "step": 5690
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9574,
      "step": 5691
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.973,
      "step": 5692
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.754,
      "step": 5693
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7229,
      "step": 5694
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6008,
      "step": 5695
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8347,
      "step": 5696
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 5697
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7682,
      "step": 5698
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 5699
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9489,
      "step": 5700
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 5701
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7936,
      "step": 5702
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6357,
      "step": 5703
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.4616,
      "step": 5704
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8291,
      "step": 5705
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5767,
      "step": 5706
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8355,
      "step": 5707
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 5708
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8414,
      "step": 5709
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9,
      "step": 5710
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8502,
      "step": 5711
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8572,
      "step": 5712
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.73,
      "step": 5713
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7497,
      "step": 5714
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9662,
      "step": 5715
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.713,
      "step": 5716
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7176,
      "step": 5717
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 5718
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8369,
      "step": 5719
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9989,
      "step": 5720
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6209,
      "step": 5721
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 5722
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.871,
      "step": 5723
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 5724
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5495,
      "step": 5725
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6294,
      "step": 5726
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6597,
      "step": 5727
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7899,
      "step": 5728
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5511,
      "step": 5729
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8029,
      "step": 5730
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9552,
      "step": 5731
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.689,
      "step": 5732
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8029,
      "step": 5733
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9937,
      "step": 5734
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7684,
      "step": 5735
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7458,
      "step": 5736
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8201,
      "step": 5737
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 1.2752,
      "step": 5738
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.884,
      "step": 5739
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 5740
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 1.0416,
      "step": 5741
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7158,
      "step": 5742
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8879,
      "step": 5743
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8423,
      "step": 5744
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 5745
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.5072,
      "step": 5746
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 1.1542,
      "step": 5747
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 5748
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8678,
      "step": 5749
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.6684,
      "step": 5750
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 5751
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 5752
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 5753
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.705,
      "step": 5754
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.853,
      "step": 5755
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8636,
      "step": 5756
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 5757
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 5758
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7943,
      "step": 5759
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 5760
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6332,
      "step": 5761
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 1.4062,
      "step": 5762
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8999,
      "step": 5763
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.3884,
      "step": 5764
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7841,
      "step": 5765
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8468,
      "step": 5766
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 5767
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6359,
      "step": 5768
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.42,
      "step": 5769
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 5770
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 5771
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6663,
      "step": 5772
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9856,
      "step": 5773
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9537,
      "step": 5774
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.794,
      "step": 5775
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8019,
      "step": 5776
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6929,
      "step": 5777
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7452,
      "step": 5778
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6483,
      "step": 5779
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7669,
      "step": 5780
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.5762,
      "step": 5781
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9958,
      "step": 5782
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 1.0292,
      "step": 5783
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.747,
      "step": 5784
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6999,
      "step": 5785
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 5786
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.696,
      "step": 5787
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7454,
      "step": 5788
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7263,
      "step": 5789
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8639,
      "step": 5790
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7553,
      "step": 5791
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.5242,
      "step": 5792
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8186,
      "step": 5793
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 1.2811,
      "step": 5794
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8176,
      "step": 5795
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7739,
      "step": 5796
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8576,
      "step": 5797
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7849,
      "step": 5798
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7429,
      "step": 5799
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6609,
      "step": 5800
    },
    {
      "epoch": 0.95,
      "eval_loss": 0.7572384476661682,
      "eval_runtime": 131.6327,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 5800
    },
    {
      "epoch": 0.95,
      "mmlu_eval_accuracy": 0.5230331495492276,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.1,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.6363636363636364,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.6086956521739131,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1871224955385722,
      "step": 5800
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9059,
      "step": 5801
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9874,
      "step": 5802
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 5803
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9475,
      "step": 5804
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8847,
      "step": 5805
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7602,
      "step": 5806
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.9615,
      "step": 5807
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8496,
      "step": 5808
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 5809
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 5810
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.6387,
      "step": 5811
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8825,
      "step": 5812
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.5843,
      "step": 5813
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7989,
      "step": 5814
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8118,
      "step": 5815
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6368,
      "step": 5816
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.0179,
      "step": 5817
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7691,
      "step": 5818
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.2382,
      "step": 5819
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 5820
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.5415,
      "step": 5821
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8365,
      "step": 5822
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6377,
      "step": 5823
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 5824
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.0803,
      "step": 5825
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7633,
      "step": 5826
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6959,
      "step": 5827
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.829,
      "step": 5828
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.002,
      "step": 5829
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.011,
      "step": 5830
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6827,
      "step": 5831
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.0655,
      "step": 5832
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8833,
      "step": 5833
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9107,
      "step": 5834
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6728,
      "step": 5835
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 5836
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8589,
      "step": 5837
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7529,
      "step": 5838
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6623,
      "step": 5839
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9483,
      "step": 5840
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8995,
      "step": 5841
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9031,
      "step": 5842
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.825,
      "step": 5843
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6552,
      "step": 5844
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8065,
      "step": 5845
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8009,
      "step": 5846
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7398,
      "step": 5847
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9893,
      "step": 5848
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7149,
      "step": 5849
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7349,
      "step": 5850
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.4984,
      "step": 5851
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 5852
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.799,
      "step": 5853
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7188,
      "step": 5854
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 5855
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.2405,
      "step": 5856
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.5321,
      "step": 5857
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.0541,
      "step": 5858
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7558,
      "step": 5859
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7913,
      "step": 5860
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8548,
      "step": 5861
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9363,
      "step": 5862
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7766,
      "step": 5863
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.9109,
      "step": 5864
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7699,
      "step": 5865
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.2606,
      "step": 5866
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.6255,
      "step": 5867
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.768,
      "step": 5868
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7604,
      "step": 5869
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 5870
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8899,
      "step": 5871
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 1.1141,
      "step": 5872
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.7358,
      "step": 5873
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5494,
      "step": 5874
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 5875
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9928,
      "step": 5876
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9858,
      "step": 5877
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 5878
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.683,
      "step": 5879
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 5880
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6677,
      "step": 5881
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8965,
      "step": 5882
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.873,
      "step": 5883
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.2479,
      "step": 5884
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5186,
      "step": 5885
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8986,
      "step": 5886
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 5887
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6354,
      "step": 5888
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6855,
      "step": 5889
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6733,
      "step": 5890
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.703,
      "step": 5891
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5373,
      "step": 5892
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 5893
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6242,
      "step": 5894
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 5895
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9275,
      "step": 5896
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8822,
      "step": 5897
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.0804,
      "step": 5898
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8364,
      "step": 5899
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 5900
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8575,
      "step": 5901
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 5902
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6238,
      "step": 5903
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.1837,
      "step": 5904
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 5905
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5047,
      "step": 5906
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9894,
      "step": 5907
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 5908
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 5909
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6154,
      "step": 5910
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9496,
      "step": 5911
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 5912
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.0598,
      "step": 5913
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9148,
      "step": 5914
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8715,
      "step": 5915
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8047,
      "step": 5916
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9877,
      "step": 5917
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 5918
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 5919
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5439,
      "step": 5920
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.808,
      "step": 5921
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.0274,
      "step": 5922
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7757,
      "step": 5923
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7065,
      "step": 5924
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9578,
      "step": 5925
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 5926
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.6899,
      "step": 5927
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7149,
      "step": 5928
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.5282,
      "step": 5929
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.9435,
      "step": 5930
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8707,
      "step": 5931
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8243,
      "step": 5932
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 1.0413,
      "step": 5933
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.7702,
      "step": 5934
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.4833,
      "step": 5935
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8733,
      "step": 5936
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 5937
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7631,
      "step": 5938
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.989,
      "step": 5939
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7686,
      "step": 5940
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 5941
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7277,
      "step": 5942
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 5943
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 5944
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6214,
      "step": 5945
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7048,
      "step": 5946
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6142,
      "step": 5947
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9274,
      "step": 5948
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9764,
      "step": 5949
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9148,
      "step": 5950
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.5132,
      "step": 5951
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9114,
      "step": 5952
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8297,
      "step": 5953
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8272,
      "step": 5954
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 5955
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 5956
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9263,
      "step": 5957
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9341,
      "step": 5958
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 1.0222,
      "step": 5959
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 5960
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7541,
      "step": 5961
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6096,
      "step": 5962
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 5963
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 5964
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8359,
      "step": 5965
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7218,
      "step": 5966
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8403,
      "step": 5967
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 5968
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9046,
      "step": 5969
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.5978,
      "step": 5970
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9524,
      "step": 5971
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 5972
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8263,
      "step": 5973
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 1.0198,
      "step": 5974
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7039,
      "step": 5975
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9132,
      "step": 5976
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6496,
      "step": 5977
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7322,
      "step": 5978
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8371,
      "step": 5979
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.9671,
      "step": 5980
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 5981
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8938,
      "step": 5982
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.661,
      "step": 5983
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8196,
      "step": 5984
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 1.0196,
      "step": 5985
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 1.1033,
      "step": 5986
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 5987
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8286,
      "step": 5988
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.586,
      "step": 5989
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 1.1123,
      "step": 5990
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8704,
      "step": 5991
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8366,
      "step": 5992
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8541,
      "step": 5993
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.8266,
      "step": 5994
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9463,
      "step": 5995
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7052,
      "step": 5996
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 1.1841,
      "step": 5997
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6531,
      "step": 5998
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.806,
      "step": 5999
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7816,
      "step": 6000
    },
    {
      "epoch": 0.99,
      "eval_loss": 0.7560673952102661,
      "eval_runtime": 131.7253,
      "eval_samples_per_second": 7.592,
      "eval_steps_per_second": 3.796,
      "step": 6000
    },
    {
      "epoch": 0.99,
      "mmlu_eval_accuracy": 0.5541866617056794,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
      "mmlu_eval_accuracy_high_school_mathematics": 0.4827586206896552,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.7272727272727273,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "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": 0.6842105263157895,
      "mmlu_loss": 0.9176231137921854,
      "step": 6000
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.92,
      "step": 6001
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6709,
      "step": 6002
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 6003
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 6004
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5218,
      "step": 6005
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5127,
      "step": 6006
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5466,
      "step": 6007
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6999,
      "step": 6008
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8014,
      "step": 6009
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5503,
      "step": 6010
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8462,
      "step": 6011
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.91,
      "step": 6012
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9856,
      "step": 6013
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8973,
      "step": 6014
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6229,
      "step": 6015
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 1.0128,
      "step": 6016
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8,
      "step": 6017
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8851,
      "step": 6018
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7596,
      "step": 6019
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5101,
      "step": 6020
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.647,
      "step": 6021
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6608,
      "step": 6022
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 6023
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8862,
      "step": 6024
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5658,
      "step": 6025
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6894,
      "step": 6026
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 6027
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9762,
      "step": 6028
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8543,
      "step": 6029
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8214,
      "step": 6030
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8155,
      "step": 6031
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6961,
      "step": 6032
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 6033
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8938,
      "step": 6034
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.964,
      "step": 6035
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7863,
      "step": 6036
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9647,
      "step": 6037
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6533,
      "step": 6038
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6301,
      "step": 6039
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5623,
      "step": 6040
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9503,
      "step": 6041
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9377,
      "step": 6042
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8947,
      "step": 6043
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.5348,
      "step": 6044
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9553,
      "step": 6045
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 1.0936,
      "step": 6046
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8397,
      "step": 6047
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6345,
      "step": 6048
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.4864,
      "step": 6049
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.9717,
      "step": 6050
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8439,
      "step": 6051
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 1.1819,
      "step": 6052
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.6844,
      "step": 6053
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.7877,
      "step": 6054
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.4692,
      "step": 6055
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8353,
      "step": 6056
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 6057
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7034,
      "step": 6058
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 6059
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7265,
      "step": 6060
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 1.0102,
      "step": 6061
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6727,
      "step": 6062
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.686,
      "step": 6063
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.452,
      "step": 6064
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 1.0397,
      "step": 6065
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7481,
      "step": 6066
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8179,
      "step": 6067
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 1.1399,
      "step": 6068
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 6069
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8474,
      "step": 6070
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7565,
      "step": 6071
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7641,
      "step": 6072
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 6073
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7553,
      "step": 6074
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.714,
      "step": 6075
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6999,
      "step": 6076
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6577,
      "step": 6077
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 1.0083,
      "step": 6078
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7476,
      "step": 6079
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7481,
      "step": 6080
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7459,
      "step": 6081
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8133,
      "step": 6082
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 6083
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 6084
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8323,
      "step": 6085
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6757,
      "step": 6086
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 6087
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5235,
      "step": 6088
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4337,
      "step": 6089
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 6090
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6134,
      "step": 6091
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5729,
      "step": 6092
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5863,
      "step": 6093
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4666,
      "step": 6094
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 6095
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4437,
      "step": 6096
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4508,
      "step": 6097
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4723,
      "step": 6098
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5512,
      "step": 6099
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.3863,
      "step": 6100
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6035,
      "step": 6101
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4541,
      "step": 6102
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4495,
      "step": 6103
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7307,
      "step": 6104
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7699,
      "step": 6105
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5502,
      "step": 6106
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 6107
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.3051,
      "step": 6108
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 6109
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 6110
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 6111
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5602,
      "step": 6112
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 6113
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.5315,
      "step": 6114
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 6115
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.9281,
      "step": 6116
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.2253,
      "step": 6117
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4907,
      "step": 6118
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.605,
      "step": 6119
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 6120
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7316,
      "step": 6121
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4795,
      "step": 6122
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7582,
      "step": 6123
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6188,
      "step": 6124
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5787,
      "step": 6125
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.3778,
      "step": 6126
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.456,
      "step": 6127
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5953,
      "step": 6128
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.9266,
      "step": 6129
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6317,
      "step": 6130
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5758,
      "step": 6131
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6978,
      "step": 6132
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4863,
      "step": 6133
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.538,
      "step": 6134
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6262,
      "step": 6135
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 6136
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6534,
      "step": 6137
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.3126,
      "step": 6138
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.62,
      "step": 6139
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5577,
      "step": 6140
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6544,
      "step": 6141
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.3914,
      "step": 6142
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5244,
      "step": 6143
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5589,
      "step": 6144
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6551,
      "step": 6145
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.8224,
      "step": 6146
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6967,
      "step": 6147
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6565,
      "step": 6148
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4591,
      "step": 6149
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.454,
      "step": 6150
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5841,
      "step": 6151
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6853,
      "step": 6152
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 6153
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4419,
      "step": 6154
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.3326,
      "step": 6155
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5264,
      "step": 6156
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6229,
      "step": 6157
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 6158
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.8161,
      "step": 6159
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5327,
      "step": 6160
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 6161
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.663,
      "step": 6162
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 6163
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5661,
      "step": 6164
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 6165
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6532,
      "step": 6166
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6723,
      "step": 6167
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6245,
      "step": 6168
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 6169
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4213,
      "step": 6170
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5742,
      "step": 6171
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.6358,
      "step": 6172
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4959,
      "step": 6173
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4134,
      "step": 6174
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.374,
      "step": 6175
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 6176
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.5818,
      "step": 6177
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.8435,
      "step": 6178
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 6179
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7051,
      "step": 6180
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6058,
      "step": 6181
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 6182
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5256,
      "step": 6183
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 6184
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7521,
      "step": 6185
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5981,
      "step": 6186
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.4882,
      "step": 6187
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 6188
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5135,
      "step": 6189
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5801,
      "step": 6190
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 6191
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 6192
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5081,
      "step": 6193
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.4595,
      "step": 6194
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.8701,
      "step": 6195
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5111,
      "step": 6196
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5928,
      "step": 6197
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5746,
      "step": 6198
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7244,
      "step": 6199
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 6200
    },
    {
      "epoch": 1.02,
      "eval_loss": 0.7689220905303955,
      "eval_runtime": 131.671,
      "eval_samples_per_second": 7.595,
      "eval_steps_per_second": 3.797,
      "step": 6200
    },
    {
      "epoch": 1.02,
      "mmlu_eval_accuracy": 0.5715176271997892,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "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": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.42857142857142855,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "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": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7209302325581395,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.9090909090909091,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 0.9558499650567692,
      "step": 6200
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 6201
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6032,
      "step": 6202
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.561,
      "step": 6203
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.4079,
      "step": 6204
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6275,
      "step": 6205
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.3475,
      "step": 6206
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 6207
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.645,
      "step": 6208
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5111,
      "step": 6209
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 6210
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.3636,
      "step": 6211
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6966,
      "step": 6212
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5142,
      "step": 6213
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 6214
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.687,
      "step": 6215
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7124,
      "step": 6216
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7188,
      "step": 6217
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 6218
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5702,
      "step": 6219
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 6220
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6068,
      "step": 6221
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7902,
      "step": 6222
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 6223
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7458,
      "step": 6224
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.3964,
      "step": 6225
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 6226
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.4947,
      "step": 6227
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 6228
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 6229
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5151,
      "step": 6230
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 6231
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 6232
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7228,
      "step": 6233
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6005,
      "step": 6234
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 6235
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7025,
      "step": 6236
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.6174,
      "step": 6237
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.5144,
      "step": 6238
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 6239
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6652,
      "step": 6240
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5218,
      "step": 6241
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6885,
      "step": 6242
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 6243
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5289,
      "step": 6244
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.7574,
      "step": 6245
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.3729,
      "step": 6246
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5993,
      "step": 6247
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.8746,
      "step": 6248
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5014,
      "step": 6249
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6427,
      "step": 6250
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5425,
      "step": 6251
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 6252
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5774,
      "step": 6253
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.503,
      "step": 6254
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 6255
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4089,
      "step": 6256
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 6257
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5259,
      "step": 6258
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.48,
      "step": 6259
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4591,
      "step": 6260
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5597,
      "step": 6261
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 6262
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4424,
      "step": 6263
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 6264
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 6265
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5359,
      "step": 6266
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.416,
      "step": 6267
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6122,
      "step": 6268
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6334,
      "step": 6269
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.466,
      "step": 6270
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.8003,
      "step": 6271
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6888,
      "step": 6272
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6223,
      "step": 6273
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6497,
      "step": 6274
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.489,
      "step": 6275
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6443,
      "step": 6276
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5104,
      "step": 6277
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 6278
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 6279
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4018,
      "step": 6280
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 6281
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.3712,
      "step": 6282
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4678,
      "step": 6283
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 6284
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.419,
      "step": 6285
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5868,
      "step": 6286
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6329,
      "step": 6287
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5484,
      "step": 6288
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6816,
      "step": 6289
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.4478,
      "step": 6290
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.7833,
      "step": 6291
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.8275,
      "step": 6292
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 6293
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6181,
      "step": 6294
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5911,
      "step": 6295
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.695,
      "step": 6296
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.614,
      "step": 6297
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.5145,
      "step": 6298
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 6299
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.58,
      "step": 6300
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5168,
      "step": 6301
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5215,
      "step": 6302
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5577,
      "step": 6303
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.485,
      "step": 6304
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6327,
      "step": 6305
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7714,
      "step": 6306
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.9221,
      "step": 6307
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4674,
      "step": 6308
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5249,
      "step": 6309
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.8127,
      "step": 6310
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 6311
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5232,
      "step": 6312
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 6313
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4721,
      "step": 6314
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5491,
      "step": 6315
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 6316
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.534,
      "step": 6317
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.8665,
      "step": 6318
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 6319
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4038,
      "step": 6320
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5021,
      "step": 6321
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7291,
      "step": 6322
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6413,
      "step": 6323
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6706,
      "step": 6324
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 1.0441,
      "step": 6325
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 6326
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7659,
      "step": 6327
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.9259,
      "step": 6328
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 6329
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5091,
      "step": 6330
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7004,
      "step": 6331
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7819,
      "step": 6332
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.415,
      "step": 6333
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7947,
      "step": 6334
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5127,
      "step": 6335
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5655,
      "step": 6336
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5329,
      "step": 6337
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6419,
      "step": 6338
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.8009,
      "step": 6339
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.365,
      "step": 6340
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4507,
      "step": 6341
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6468,
      "step": 6342
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6173,
      "step": 6343
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 6344
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6876,
      "step": 6345
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 6346
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6395,
      "step": 6347
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.3292,
      "step": 6348
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4478,
      "step": 6349
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 6350
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5233,
      "step": 6351
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.2907,
      "step": 6352
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.3312,
      "step": 6353
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.6319,
      "step": 6354
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4441,
      "step": 6355
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4045,
      "step": 6356
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 1.0694,
      "step": 6357
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5241,
      "step": 6358
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.4754,
      "step": 6359
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.5606,
      "step": 6360
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 6361
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7608,
      "step": 6362
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5168,
      "step": 6363
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5711,
      "step": 6364
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 6365
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4122,
      "step": 6366
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 6367
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6023,
      "step": 6368
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4666,
      "step": 6369
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7066,
      "step": 6370
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5153,
      "step": 6371
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5298,
      "step": 6372
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 6373
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.462,
      "step": 6374
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5451,
      "step": 6375
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5628,
      "step": 6376
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 6377
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 6378
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6162,
      "step": 6379
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5984,
      "step": 6380
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.54,
      "step": 6381
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7415,
      "step": 6382
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.2781,
      "step": 6383
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 6384
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 6385
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.925,
      "step": 6386
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 6387
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.9883,
      "step": 6388
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6426,
      "step": 6389
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.545,
      "step": 6390
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4511,
      "step": 6391
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5265,
      "step": 6392
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 6393
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4912,
      "step": 6394
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6679,
      "step": 6395
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4239,
      "step": 6396
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5228,
      "step": 6397
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5934,
      "step": 6398
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.9043,
      "step": 6399
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6576,
      "step": 6400
    },
    {
      "epoch": 1.05,
      "eval_loss": 0.7675976753234863,
      "eval_runtime": 131.7379,
      "eval_samples_per_second": 7.591,
      "eval_steps_per_second": 3.795,
      "step": 6400
    },
    {
      "epoch": 1.05,
      "mmlu_eval_accuracy": 0.531372965510407,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5833333333333334,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.14,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
      "mmlu_eval_accuracy_prehistory": 0.7142857142857143,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 0.9138225094262054,
      "step": 6400
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 6401
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7183,
      "step": 6402
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.683,
      "step": 6403
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5997,
      "step": 6404
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6623,
      "step": 6405
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 6406
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4603,
      "step": 6407
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6102,
      "step": 6408
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.763,
      "step": 6409
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.3744,
      "step": 6410
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.8501,
      "step": 6411
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7932,
      "step": 6412
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 6413
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6528,
      "step": 6414
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.6154,
      "step": 6415
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5424,
      "step": 6416
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5913,
      "step": 6417
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 6418
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.5393,
      "step": 6419
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.4465,
      "step": 6420
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 6421
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.9416,
      "step": 6422
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 6423
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4554,
      "step": 6424
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 6425
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5998,
      "step": 6426
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4636,
      "step": 6427
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5784,
      "step": 6428
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 6429
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.2893,
      "step": 6430
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7421,
      "step": 6431
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.9496,
      "step": 6432
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.3968,
      "step": 6433
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5535,
      "step": 6434
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4956,
      "step": 6435
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6261,
      "step": 6436
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5393,
      "step": 6437
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7072,
      "step": 6438
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5923,
      "step": 6439
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6742,
      "step": 6440
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6203,
      "step": 6441
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.573,
      "step": 6442
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6316,
      "step": 6443
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5294,
      "step": 6444
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5721,
      "step": 6445
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6873,
      "step": 6446
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 6447
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4798,
      "step": 6448
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5458,
      "step": 6449
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6021,
      "step": 6450
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.79,
      "step": 6451
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 6452
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.433,
      "step": 6453
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4337,
      "step": 6454
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4974,
      "step": 6455
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4867,
      "step": 6456
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7018,
      "step": 6457
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.3687,
      "step": 6458
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5329,
      "step": 6459
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4827,
      "step": 6460
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5277,
      "step": 6461
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6279,
      "step": 6462
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5371,
      "step": 6463
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4854,
      "step": 6464
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5624,
      "step": 6465
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4002,
      "step": 6466
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 6467
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 6468
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4106,
      "step": 6469
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4803,
      "step": 6470
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.69,
      "step": 6471
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5656,
      "step": 6472
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5729,
      "step": 6473
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6106,
      "step": 6474
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5606,
      "step": 6475
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5085,
      "step": 6476
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6401,
      "step": 6477
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.6204,
      "step": 6478
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.538,
      "step": 6479
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5555,
      "step": 6480
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.4759,
      "step": 6481
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5046,
      "step": 6482
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6191,
      "step": 6483
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4839,
      "step": 6484
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5882,
      "step": 6485
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.593,
      "step": 6486
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 6487
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 6488
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5762,
      "step": 6489
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 6490
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6141,
      "step": 6491
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5587,
      "step": 6492
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5192,
      "step": 6493
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5263,
      "step": 6494
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5814,
      "step": 6495
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6394,
      "step": 6496
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7557,
      "step": 6497
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5231,
      "step": 6498
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4865,
      "step": 6499
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7372,
      "step": 6500
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7573,
      "step": 6501
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6717,
      "step": 6502
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4714,
      "step": 6503
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5732,
      "step": 6504
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 6505
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7093,
      "step": 6506
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.791,
      "step": 6507
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7353,
      "step": 6508
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7137,
      "step": 6509
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4479,
      "step": 6510
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4627,
      "step": 6511
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5175,
      "step": 6512
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.625,
      "step": 6513
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4917,
      "step": 6514
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5171,
      "step": 6515
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.3847,
      "step": 6516
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.496,
      "step": 6517
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4081,
      "step": 6518
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4594,
      "step": 6519
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5557,
      "step": 6520
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 6521
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 6522
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6803,
      "step": 6523
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4046,
      "step": 6524
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4461,
      "step": 6525
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5252,
      "step": 6526
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 6527
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.647,
      "step": 6528
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4223,
      "step": 6529
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6016,
      "step": 6530
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.866,
      "step": 6531
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5334,
      "step": 6532
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6766,
      "step": 6533
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6887,
      "step": 6534
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4687,
      "step": 6535
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.8589,
      "step": 6536
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6026,
      "step": 6537
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.6108,
      "step": 6538
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5321,
      "step": 6539
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5415,
      "step": 6540
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.5542,
      "step": 6541
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.4377,
      "step": 6542
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 6543
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.58,
      "step": 6544
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4787,
      "step": 6545
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 6546
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 6547
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6642,
      "step": 6548
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5855,
      "step": 6549
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.463,
      "step": 6550
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4382,
      "step": 6551
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 6552
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.42,
      "step": 6553
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5334,
      "step": 6554
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6411,
      "step": 6555
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6367,
      "step": 6556
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.9406,
      "step": 6557
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4335,
      "step": 6558
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5331,
      "step": 6559
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5007,
      "step": 6560
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4578,
      "step": 6561
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5617,
      "step": 6562
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7565,
      "step": 6563
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6076,
      "step": 6564
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7933,
      "step": 6565
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 6566
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.3339,
      "step": 6567
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6262,
      "step": 6568
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 6569
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4262,
      "step": 6570
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 6571
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 6572
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6935,
      "step": 6573
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 6574
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6882,
      "step": 6575
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5597,
      "step": 6576
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5185,
      "step": 6577
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5062,
      "step": 6578
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6122,
      "step": 6579
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7504,
      "step": 6580
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 6581
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 6582
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7129,
      "step": 6583
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4349,
      "step": 6584
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 6585
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6264,
      "step": 6586
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 6587
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7024,
      "step": 6588
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5335,
      "step": 6589
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.496,
      "step": 6590
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6309,
      "step": 6591
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.8967,
      "step": 6592
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4428,
      "step": 6593
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 6594
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4348,
      "step": 6595
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5861,
      "step": 6596
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4579,
      "step": 6597
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5136,
      "step": 6598
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.5272,
      "step": 6599
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4455,
      "step": 6600
    },
    {
      "epoch": 1.08,
      "eval_loss": 0.7740963101387024,
      "eval_runtime": 131.6357,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 6600
    },
    {
      "epoch": 1.08,
      "mmlu_eval_accuracy": 0.5342019406821743,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "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": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "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": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.5652173913043478,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.40588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.4444444444444444,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.0002783060890887,
      "step": 6600
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.6796,
      "step": 6601
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.3143,
      "step": 6602
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.4375,
      "step": 6603
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7147,
      "step": 6604
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4346,
      "step": 6605
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 6606
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6557,
      "step": 6607
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6647,
      "step": 6608
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6588,
      "step": 6609
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6027,
      "step": 6610
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5615,
      "step": 6611
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 6612
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.8625,
      "step": 6613
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5074,
      "step": 6614
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.3311,
      "step": 6615
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5519,
      "step": 6616
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.3364,
      "step": 6617
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6128,
      "step": 6618
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5926,
      "step": 6619
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 6620
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5305,
      "step": 6621
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5954,
      "step": 6622
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4895,
      "step": 6623
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5617,
      "step": 6624
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 6625
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4015,
      "step": 6626
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7564,
      "step": 6627
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.547,
      "step": 6628
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 6629
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6068,
      "step": 6630
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5936,
      "step": 6631
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 6632
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6308,
      "step": 6633
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6034,
      "step": 6634
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6629,
      "step": 6635
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5241,
      "step": 6636
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5901,
      "step": 6637
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5332,
      "step": 6638
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.347,
      "step": 6639
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4117,
      "step": 6640
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.5271,
      "step": 6641
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.8052,
      "step": 6642
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.3983,
      "step": 6643
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4749,
      "step": 6644
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6393,
      "step": 6645
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.461,
      "step": 6646
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4604,
      "step": 6647
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 6648
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.3814,
      "step": 6649
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4425,
      "step": 6650
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.2493,
      "step": 6651
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 1.0129,
      "step": 6652
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6466,
      "step": 6653
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 6654
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7579,
      "step": 6655
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 6656
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.6688,
      "step": 6657
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.632,
      "step": 6658
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.53,
      "step": 6659
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.2734,
      "step": 6660
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 6661
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 6662
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.2825,
      "step": 6663
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 6664
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5435,
      "step": 6665
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6544,
      "step": 6666
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6029,
      "step": 6667
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 6668
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7944,
      "step": 6669
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.313,
      "step": 6670
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.3735,
      "step": 6671
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5061,
      "step": 6672
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7464,
      "step": 6673
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7401,
      "step": 6674
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4418,
      "step": 6675
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5677,
      "step": 6676
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.768,
      "step": 6677
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 6678
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.8515,
      "step": 6679
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 6680
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4368,
      "step": 6681
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6594,
      "step": 6682
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7272,
      "step": 6683
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 6684
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.3526,
      "step": 6685
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 6686
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5045,
      "step": 6687
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4375,
      "step": 6688
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6859,
      "step": 6689
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.8872,
      "step": 6690
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6042,
      "step": 6691
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.3322,
      "step": 6692
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6578,
      "step": 6693
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.9558,
      "step": 6694
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 1.0182,
      "step": 6695
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.406,
      "step": 6696
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7461,
      "step": 6697
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4885,
      "step": 6698
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7054,
      "step": 6699
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.8189,
      "step": 6700
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5347,
      "step": 6701
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7174,
      "step": 6702
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5086,
      "step": 6703
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 6704
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 6705
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5491,
      "step": 6706
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7022,
      "step": 6707
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6168,
      "step": 6708
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 6709
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5383,
      "step": 6710
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7538,
      "step": 6711
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4884,
      "step": 6712
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.8998,
      "step": 6713
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5342,
      "step": 6714
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.3771,
      "step": 6715
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4522,
      "step": 6716
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4867,
      "step": 6717
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 6718
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.5768,
      "step": 6719
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6148,
      "step": 6720
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7055,
      "step": 6721
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7172,
      "step": 6722
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.6137,
      "step": 6723
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 6724
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.846,
      "step": 6725
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6315,
      "step": 6726
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 6727
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6938,
      "step": 6728
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 1.0504,
      "step": 6729
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5198,
      "step": 6730
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.8362,
      "step": 6731
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 6732
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6511,
      "step": 6733
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.487,
      "step": 6734
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5383,
      "step": 6735
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 6736
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 6737
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6018,
      "step": 6738
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 6739
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.8298,
      "step": 6740
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 6741
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.8798,
      "step": 6742
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4771,
      "step": 6743
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6638,
      "step": 6744
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 6745
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.617,
      "step": 6746
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 6747
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5541,
      "step": 6748
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5338,
      "step": 6749
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6053,
      "step": 6750
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5636,
      "step": 6751
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7516,
      "step": 6752
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7481,
      "step": 6753
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 6754
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4418,
      "step": 6755
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 6756
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 6757
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.741,
      "step": 6758
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6457,
      "step": 6759
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 6760
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5199,
      "step": 6761
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.562,
      "step": 6762
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.93,
      "step": 6763
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7343,
      "step": 6764
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.3319,
      "step": 6765
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7751,
      "step": 6766
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4253,
      "step": 6767
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 6768
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7465,
      "step": 6769
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5526,
      "step": 6770
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7127,
      "step": 6771
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6053,
      "step": 6772
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7871,
      "step": 6773
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.6644,
      "step": 6774
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4037,
      "step": 6775
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.5428,
      "step": 6776
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.3748,
      "step": 6777
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7331,
      "step": 6778
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7785,
      "step": 6779
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4685,
      "step": 6780
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 6781
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4017,
      "step": 6782
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4932,
      "step": 6783
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.8002,
      "step": 6784
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.4179,
      "step": 6785
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.3728,
      "step": 6786
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.8344,
      "step": 6787
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.8664,
      "step": 6788
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6562,
      "step": 6789
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5859,
      "step": 6790
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5436,
      "step": 6791
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 6792
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5092,
      "step": 6793
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 6794
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 6795
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 6796
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 1.0965,
      "step": 6797
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6033,
      "step": 6798
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 6799
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7454,
      "step": 6800
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.7760998606681824,
      "eval_runtime": 131.6669,
      "eval_samples_per_second": 7.595,
      "eval_steps_per_second": 3.797,
      "step": 6800
    },
    {
      "epoch": 1.12,
      "mmlu_eval_accuracy": 0.559257349190752,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.4482758620689655,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.5217391304347826,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.40588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.0201158680352467,
      "step": 6800
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5437,
      "step": 6801
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6297,
      "step": 6802
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5115,
      "step": 6803
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.3956,
      "step": 6804
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 6805
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5427,
      "step": 6806
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 6807
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.4899,
      "step": 6808
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5022,
      "step": 6809
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.4883,
      "step": 6810
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5419,
      "step": 6811
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6067,
      "step": 6812
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5371,
      "step": 6813
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6104,
      "step": 6814
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.52,
      "step": 6815
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5867,
      "step": 6816
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5247,
      "step": 6817
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.427,
      "step": 6818
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7446,
      "step": 6819
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 6820
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6738,
      "step": 6821
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5367,
      "step": 6822
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6231,
      "step": 6823
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.542,
      "step": 6824
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5188,
      "step": 6825
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5037,
      "step": 6826
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5687,
      "step": 6827
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5524,
      "step": 6828
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6712,
      "step": 6829
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 6830
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6149,
      "step": 6831
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7328,
      "step": 6832
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5597,
      "step": 6833
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.712,
      "step": 6834
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7194,
      "step": 6835
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 6836
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.4869,
      "step": 6837
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5086,
      "step": 6838
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.8073,
      "step": 6839
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 6840
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 6841
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 1.006,
      "step": 6842
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6373,
      "step": 6843
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.5892,
      "step": 6844
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.6368,
      "step": 6845
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7158,
      "step": 6846
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7156,
      "step": 6847
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7135,
      "step": 6848
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4837,
      "step": 6849
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4025,
      "step": 6850
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 6851
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5365,
      "step": 6852
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4011,
      "step": 6853
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.741,
      "step": 6854
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7249,
      "step": 6855
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5832,
      "step": 6856
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 6857
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 6858
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7308,
      "step": 6859
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.3362,
      "step": 6860
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5853,
      "step": 6861
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.473,
      "step": 6862
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 6863
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 6864
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 6865
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6768,
      "step": 6866
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 6867
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 6868
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6263,
      "step": 6869
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 6870
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 6871
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.65,
      "step": 6872
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 6873
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7423,
      "step": 6874
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4555,
      "step": 6875
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4521,
      "step": 6876
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5449,
      "step": 6877
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 6878
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 6879
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7019,
      "step": 6880
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.9338,
      "step": 6881
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6885,
      "step": 6882
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 6883
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6361,
      "step": 6884
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.738,
      "step": 6885
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.8744,
      "step": 6886
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7079,
      "step": 6887
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 6888
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4593,
      "step": 6889
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.8265,
      "step": 6890
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4884,
      "step": 6891
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5337,
      "step": 6892
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4168,
      "step": 6893
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5519,
      "step": 6894
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.619,
      "step": 6895
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7338,
      "step": 6896
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4096,
      "step": 6897
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 6898
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.5006,
      "step": 6899
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4619,
      "step": 6900
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.3974,
      "step": 6901
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.4773,
      "step": 6902
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.8954,
      "step": 6903
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7712,
      "step": 6904
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 6905
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 6906
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 6907
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6589,
      "step": 6908
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4349,
      "step": 6909
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5649,
      "step": 6910
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.8408,
      "step": 6911
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6898,
      "step": 6912
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.705,
      "step": 6913
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6513,
      "step": 6914
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6334,
      "step": 6915
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4997,
      "step": 6916
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 6917
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5058,
      "step": 6918
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7786,
      "step": 6919
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6466,
      "step": 6920
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6771,
      "step": 6921
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6889,
      "step": 6922
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.513,
      "step": 6923
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 6924
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 6925
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6268,
      "step": 6926
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7486,
      "step": 6927
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7051,
      "step": 6928
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.8276,
      "step": 6929
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5662,
      "step": 6930
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 6931
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 1.051,
      "step": 6932
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5556,
      "step": 6933
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7306,
      "step": 6934
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.3543,
      "step": 6935
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5713,
      "step": 6936
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 6937
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.9097,
      "step": 6938
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4256,
      "step": 6939
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7796,
      "step": 6940
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4506,
      "step": 6941
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5844,
      "step": 6942
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6715,
      "step": 6943
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5273,
      "step": 6944
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 6945
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.519,
      "step": 6946
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 6947
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5626,
      "step": 6948
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5108,
      "step": 6949
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6905,
      "step": 6950
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 6951
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6772,
      "step": 6952
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6222,
      "step": 6953
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5804,
      "step": 6954
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 6955
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5387,
      "step": 6956
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.9414,
      "step": 6957
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.891,
      "step": 6958
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 6959
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4636,
      "step": 6960
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5842,
      "step": 6961
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.5031,
      "step": 6962
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7689,
      "step": 6963
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 6964
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7305,
      "step": 6965
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6216,
      "step": 6966
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 6967
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.6012,
      "step": 6968
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.1949,
      "step": 6969
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.8356,
      "step": 6970
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.524,
      "step": 6971
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 6972
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6257,
      "step": 6973
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4931,
      "step": 6974
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 6975
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5323,
      "step": 6976
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.78,
      "step": 6977
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6534,
      "step": 6978
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 6979
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.97,
      "step": 6980
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6755,
      "step": 6981
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 6982
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6721,
      "step": 6983
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.584,
      "step": 6984
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5636,
      "step": 6985
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6324,
      "step": 6986
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7077,
      "step": 6987
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6534,
      "step": 6988
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3873,
      "step": 6989
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7743,
      "step": 6990
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7233,
      "step": 6991
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7096,
      "step": 6992
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5932,
      "step": 6993
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5512,
      "step": 6994
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.8556,
      "step": 6995
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6739,
      "step": 6996
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 6997
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6514,
      "step": 6998
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4973,
      "step": 6999
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4795,
      "step": 7000
    },
    {
      "epoch": 1.15,
      "eval_loss": 0.7763926982879639,
      "eval_runtime": 131.4755,
      "eval_samples_per_second": 7.606,
      "eval_steps_per_second": 3.803,
      "step": 7000
    },
    {
      "epoch": 1.15,
      "mmlu_eval_accuracy": 0.5471009839827601,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "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": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7674418604651163,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.7142857142857143,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.4235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 0.9087689117025146,
      "step": 7000
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6775,
      "step": 7001
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.8118,
      "step": 7002
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 7003
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.343,
      "step": 7004
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 7005
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3462,
      "step": 7006
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 7007
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7894,
      "step": 7008
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 7009
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.777,
      "step": 7010
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6706,
      "step": 7011
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3559,
      "step": 7012
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 7013
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5637,
      "step": 7014
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 7015
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4623,
      "step": 7016
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7257,
      "step": 7017
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 7018
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5813,
      "step": 7019
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 7020
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 7021
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.8494,
      "step": 7022
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3461,
      "step": 7023
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.84,
      "step": 7024
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 7025
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.6656,
      "step": 7026
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.5148,
      "step": 7027
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.4527,
      "step": 7028
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7168,
      "step": 7029
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 7030
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6596,
      "step": 7031
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 7032
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5431,
      "step": 7033
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5803,
      "step": 7034
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 7035
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7203,
      "step": 7036
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7746,
      "step": 7037
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7366,
      "step": 7038
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5276,
      "step": 7039
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.62,
      "step": 7040
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6128,
      "step": 7041
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6936,
      "step": 7042
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.451,
      "step": 7043
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5939,
      "step": 7044
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6493,
      "step": 7045
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.3767,
      "step": 7046
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5728,
      "step": 7047
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4467,
      "step": 7048
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 7049
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4249,
      "step": 7050
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7108,
      "step": 7051
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6903,
      "step": 7052
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5836,
      "step": 7053
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7189,
      "step": 7054
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4027,
      "step": 7055
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.384,
      "step": 7056
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.9138,
      "step": 7057
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7757,
      "step": 7058
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5497,
      "step": 7059
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.448,
      "step": 7060
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6706,
      "step": 7061
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5045,
      "step": 7062
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.821,
      "step": 7063
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8207,
      "step": 7064
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7026,
      "step": 7065
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 7066
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 7067
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.581,
      "step": 7068
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4489,
      "step": 7069
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4825,
      "step": 7070
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.559,
      "step": 7071
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6926,
      "step": 7072
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8658,
      "step": 7073
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 7074
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7453,
      "step": 7075
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 7076
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.3834,
      "step": 7077
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7226,
      "step": 7078
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4934,
      "step": 7079
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6465,
      "step": 7080
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4131,
      "step": 7081
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6049,
      "step": 7082
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8188,
      "step": 7083
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.4755,
      "step": 7084
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 7085
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.8672,
      "step": 7086
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 7087
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7956,
      "step": 7088
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 7089
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.975,
      "step": 7090
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 7091
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 7092
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.3867,
      "step": 7093
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.529,
      "step": 7094
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5763,
      "step": 7095
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6722,
      "step": 7096
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 7097
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8854,
      "step": 7098
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8866,
      "step": 7099
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7524,
      "step": 7100
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8665,
      "step": 7101
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4055,
      "step": 7102
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8981,
      "step": 7103
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6931,
      "step": 7104
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 7105
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6678,
      "step": 7106
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.3193,
      "step": 7107
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6124,
      "step": 7108
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5211,
      "step": 7109
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7326,
      "step": 7110
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.643,
      "step": 7111
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4494,
      "step": 7112
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5508,
      "step": 7113
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4296,
      "step": 7114
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8476,
      "step": 7115
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 7116
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 7117
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 7118
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.2931,
      "step": 7119
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6942,
      "step": 7120
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4901,
      "step": 7121
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 7122
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5042,
      "step": 7123
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7087,
      "step": 7124
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 7125
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6491,
      "step": 7126
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7243,
      "step": 7127
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4509,
      "step": 7128
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4681,
      "step": 7129
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 7130
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4764,
      "step": 7131
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 7132
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 7133
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 7134
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5638,
      "step": 7135
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6574,
      "step": 7136
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 7137
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7375,
      "step": 7138
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5256,
      "step": 7139
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5861,
      "step": 7140
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 7141
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6871,
      "step": 7142
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7365,
      "step": 7143
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.5708,
      "step": 7144
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 7145
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6982,
      "step": 7146
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.6044,
      "step": 7147
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4712,
      "step": 7148
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.8392,
      "step": 7149
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.755,
      "step": 7150
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4286,
      "step": 7151
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7202,
      "step": 7152
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 7153
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8962,
      "step": 7154
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4105,
      "step": 7155
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6982,
      "step": 7156
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8352,
      "step": 7157
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6359,
      "step": 7158
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 7159
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 7160
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4802,
      "step": 7161
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 1.044,
      "step": 7162
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.741,
      "step": 7163
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4942,
      "step": 7164
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 7165
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6656,
      "step": 7166
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.3943,
      "step": 7167
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.473,
      "step": 7168
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 7169
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.9115,
      "step": 7170
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 7171
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6703,
      "step": 7172
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 7173
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6458,
      "step": 7174
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6389,
      "step": 7175
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5843,
      "step": 7176
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7465,
      "step": 7177
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5903,
      "step": 7178
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6977,
      "step": 7179
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 7180
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6052,
      "step": 7181
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6048,
      "step": 7182
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8425,
      "step": 7183
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 7184
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5935,
      "step": 7185
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7223,
      "step": 7186
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6768,
      "step": 7187
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4858,
      "step": 7188
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 7189
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 7190
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5721,
      "step": 7191
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4631,
      "step": 7192
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5065,
      "step": 7193
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 7194
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5419,
      "step": 7195
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4897,
      "step": 7196
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4387,
      "step": 7197
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4934,
      "step": 7198
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.677,
      "step": 7199
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 7200
    },
    {
      "epoch": 1.18,
      "eval_loss": 0.7779871225357056,
      "eval_runtime": 131.7129,
      "eval_samples_per_second": 7.592,
      "eval_steps_per_second": 3.796,
      "step": 7200
    },
    {
      "epoch": 1.18,
      "mmlu_eval_accuracy": 0.5458463067442448,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8571428571428571,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693,
      "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "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": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 0.9296292920861947,
      "step": 7200
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4439,
      "step": 7201
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.8258,
      "step": 7202
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 7203
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.667,
      "step": 7204
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4365,
      "step": 7205
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 7206
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6566,
      "step": 7207
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6513,
      "step": 7208
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 7209
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 7210
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.5515,
      "step": 7211
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.4919,
      "step": 7212
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5909,
      "step": 7213
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.3902,
      "step": 7214
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7514,
      "step": 7215
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5347,
      "step": 7216
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 7217
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 7218
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4914,
      "step": 7219
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6109,
      "step": 7220
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.632,
      "step": 7221
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7187,
      "step": 7222
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7173,
      "step": 7223
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 7224
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6091,
      "step": 7225
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6839,
      "step": 7226
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4443,
      "step": 7227
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5272,
      "step": 7228
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 7229
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4518,
      "step": 7230
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6424,
      "step": 7231
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.3804,
      "step": 7232
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.3135,
      "step": 7233
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4322,
      "step": 7234
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7806,
      "step": 7235
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.9106,
      "step": 7236
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 7237
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 7238
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6504,
      "step": 7239
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.3653,
      "step": 7240
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5134,
      "step": 7241
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6194,
      "step": 7242
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6135,
      "step": 7243
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.9423,
      "step": 7244
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5323,
      "step": 7245
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 7246
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7432,
      "step": 7247
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 7248
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7694,
      "step": 7249
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4507,
      "step": 7250
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6114,
      "step": 7251
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 7252
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.643,
      "step": 7253
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6375,
      "step": 7254
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.2888,
      "step": 7255
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4666,
      "step": 7256
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 7257
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6026,
      "step": 7258
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 7259
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5592,
      "step": 7260
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4945,
      "step": 7261
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6573,
      "step": 7262
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 7263
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.4033,
      "step": 7264
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 7265
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6827,
      "step": 7266
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5565,
      "step": 7267
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 7268
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7359,
      "step": 7269
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.5947,
      "step": 7270
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.276,
      "step": 7271
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.9356,
      "step": 7272
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 7273
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 7274
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5671,
      "step": 7275
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7812,
      "step": 7276
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 7277
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 7278
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.8702,
      "step": 7279
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5326,
      "step": 7280
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 7281
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6266,
      "step": 7282
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5593,
      "step": 7283
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 7284
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.4428,
      "step": 7285
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6828,
      "step": 7286
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6028,
      "step": 7287
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5229,
      "step": 7288
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5153,
      "step": 7289
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6231,
      "step": 7290
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 7291
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.8582,
      "step": 7292
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5524,
      "step": 7293
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.9853,
      "step": 7294
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5884,
      "step": 7295
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 7296
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.759,
      "step": 7297
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.4624,
      "step": 7298
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 7299
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7997,
      "step": 7300
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5636,
      "step": 7301
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 1.0654,
      "step": 7302
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 7303
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6266,
      "step": 7304
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6579,
      "step": 7305
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6663,
      "step": 7306
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5536,
      "step": 7307
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6134,
      "step": 7308
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6842,
      "step": 7309
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6174,
      "step": 7310
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5899,
      "step": 7311
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 7312
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 7313
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5433,
      "step": 7314
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6507,
      "step": 7315
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.4435,
      "step": 7316
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6914,
      "step": 7317
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.2701,
      "step": 7318
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5732,
      "step": 7319
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 7320
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 7321
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5598,
      "step": 7322
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.3476,
      "step": 7323
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.8551,
      "step": 7324
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 7325
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 1.2377,
      "step": 7326
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.778,
      "step": 7327
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 7328
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.4751,
      "step": 7329
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.8306,
      "step": 7330
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5218,
      "step": 7331
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.2947,
      "step": 7332
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.5313,
      "step": 7333
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5435,
      "step": 7334
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.3088,
      "step": 7335
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7078,
      "step": 7336
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.3842,
      "step": 7337
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6958,
      "step": 7338
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 7339
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.567,
      "step": 7340
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.724,
      "step": 7341
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7918,
      "step": 7342
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.8267,
      "step": 7343
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7036,
      "step": 7344
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5583,
      "step": 7345
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4606,
      "step": 7346
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 7347
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 7348
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.506,
      "step": 7349
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4261,
      "step": 7350
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.478,
      "step": 7351
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 7352
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7417,
      "step": 7353
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 7354
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4467,
      "step": 7355
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 7356
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.3161,
      "step": 7357
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4733,
      "step": 7358
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4272,
      "step": 7359
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4913,
      "step": 7360
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 7361
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5304,
      "step": 7362
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7029,
      "step": 7363
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5982,
      "step": 7364
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 7365
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.609,
      "step": 7366
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 7367
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.9667,
      "step": 7368
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.8012,
      "step": 7369
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7322,
      "step": 7370
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7711,
      "step": 7371
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5946,
      "step": 7372
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7019,
      "step": 7373
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 1.0012,
      "step": 7374
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.2369,
      "step": 7375
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5758,
      "step": 7376
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6913,
      "step": 7377
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4648,
      "step": 7378
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7523,
      "step": 7379
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6686,
      "step": 7380
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.9967,
      "step": 7381
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5723,
      "step": 7382
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6879,
      "step": 7383
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6808,
      "step": 7384
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.5729,
      "step": 7385
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.8059,
      "step": 7386
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.6491,
      "step": 7387
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4455,
      "step": 7388
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.2721,
      "step": 7389
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7307,
      "step": 7390
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4735,
      "step": 7391
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.4661,
      "step": 7392
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.216,
      "step": 7393
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.528,
      "step": 7394
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6202,
      "step": 7395
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 7396
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.536,
      "step": 7397
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.439,
      "step": 7398
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.556,
      "step": 7399
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.419,
      "step": 7400
    },
    {
      "epoch": 1.22,
      "eval_loss": 0.7840510010719299,
      "eval_runtime": 131.6563,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 7400
    },
    {
      "epoch": 1.22,
      "mmlu_eval_accuracy": 0.5539596344486911,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "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": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 0.8888738014523728,
      "step": 7400
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4402,
      "step": 7401
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.9533,
      "step": 7402
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 7403
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4045,
      "step": 7404
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6194,
      "step": 7405
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 7406
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.745,
      "step": 7407
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.8922,
      "step": 7408
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6013,
      "step": 7409
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 7410
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7088,
      "step": 7411
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5264,
      "step": 7412
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5274,
      "step": 7413
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5197,
      "step": 7414
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7027,
      "step": 7415
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.3788,
      "step": 7416
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 7417
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.2694,
      "step": 7418
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.8318,
      "step": 7419
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5814,
      "step": 7420
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.3652,
      "step": 7421
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6713,
      "step": 7422
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 7423
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 7424
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4918,
      "step": 7425
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.571,
      "step": 7426
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.8342,
      "step": 7427
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5419,
      "step": 7428
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7124,
      "step": 7429
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7402,
      "step": 7430
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6192,
      "step": 7431
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5032,
      "step": 7432
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5674,
      "step": 7433
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7577,
      "step": 7434
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5519,
      "step": 7435
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5463,
      "step": 7436
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5707,
      "step": 7437
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4778,
      "step": 7438
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5088,
      "step": 7439
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6949,
      "step": 7440
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4271,
      "step": 7441
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6187,
      "step": 7442
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7451,
      "step": 7443
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4878,
      "step": 7444
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 7445
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.422,
      "step": 7446
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5012,
      "step": 7447
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 7448
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5183,
      "step": 7449
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.6139,
      "step": 7450
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5973,
      "step": 7451
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7316,
      "step": 7452
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5908,
      "step": 7453
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.5171,
      "step": 7454
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7608,
      "step": 7455
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5128,
      "step": 7456
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8474,
      "step": 7457
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4923,
      "step": 7458
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6612,
      "step": 7459
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.3407,
      "step": 7460
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5905,
      "step": 7461
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6403,
      "step": 7462
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5803,
      "step": 7463
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.9964,
      "step": 7464
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4677,
      "step": 7465
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 7466
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8675,
      "step": 7467
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4833,
      "step": 7468
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 7469
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5859,
      "step": 7470
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 7471
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6455,
      "step": 7472
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5258,
      "step": 7473
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6245,
      "step": 7474
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 7475
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7946,
      "step": 7476
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 7477
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8565,
      "step": 7478
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8317,
      "step": 7479
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4915,
      "step": 7480
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7564,
      "step": 7481
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5714,
      "step": 7482
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8688,
      "step": 7483
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.9208,
      "step": 7484
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.3495,
      "step": 7485
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8084,
      "step": 7486
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5779,
      "step": 7487
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.637,
      "step": 7488
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.8063,
      "step": 7489
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5617,
      "step": 7490
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5835,
      "step": 7491
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.3608,
      "step": 7492
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6387,
      "step": 7493
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7348,
      "step": 7494
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5505,
      "step": 7495
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6404,
      "step": 7496
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6748,
      "step": 7497
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 7498
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4326,
      "step": 7499
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 1.3722,
      "step": 7500
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5637,
      "step": 7501
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7872,
      "step": 7502
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.436,
      "step": 7503
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 7504
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.64,
      "step": 7505
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.5051,
      "step": 7506
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6311,
      "step": 7507
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4621,
      "step": 7508
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4536,
      "step": 7509
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6182,
      "step": 7510
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.447,
      "step": 7511
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4798,
      "step": 7512
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6996,
      "step": 7513
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 7514
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.6712,
      "step": 7515
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4736,
      "step": 7516
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6087,
      "step": 7517
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.869,
      "step": 7518
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6522,
      "step": 7519
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5906,
      "step": 7520
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7685,
      "step": 7521
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 7522
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6803,
      "step": 7523
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7047,
      "step": 7524
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4937,
      "step": 7525
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6017,
      "step": 7526
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.472,
      "step": 7527
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5767,
      "step": 7528
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 7529
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7432,
      "step": 7530
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5228,
      "step": 7531
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5073,
      "step": 7532
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4617,
      "step": 7533
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6089,
      "step": 7534
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4653,
      "step": 7535
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4558,
      "step": 7536
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 7537
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 7538
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 7539
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4551,
      "step": 7540
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5427,
      "step": 7541
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5639,
      "step": 7542
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.592,
      "step": 7543
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6111,
      "step": 7544
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5025,
      "step": 7545
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.3047,
      "step": 7546
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6029,
      "step": 7547
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6574,
      "step": 7548
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6303,
      "step": 7549
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 7550
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6968,
      "step": 7551
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.8309,
      "step": 7552
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6385,
      "step": 7553
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5855,
      "step": 7554
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7537,
      "step": 7555
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6343,
      "step": 7556
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 7557
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5915,
      "step": 7558
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5445,
      "step": 7559
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4196,
      "step": 7560
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5034,
      "step": 7561
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5041,
      "step": 7562
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 7563
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 7564
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5866,
      "step": 7565
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5277,
      "step": 7566
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4759,
      "step": 7567
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4247,
      "step": 7568
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 7569
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 7570
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6192,
      "step": 7571
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.4475,
      "step": 7572
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.9282,
      "step": 7573
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5757,
      "step": 7574
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.3757,
      "step": 7575
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 7576
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.6911,
      "step": 7577
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4595,
      "step": 7578
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8463,
      "step": 7579
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6098,
      "step": 7580
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7327,
      "step": 7581
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4972,
      "step": 7582
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5909,
      "step": 7583
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 7584
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4462,
      "step": 7585
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7732,
      "step": 7586
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 7587
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5578,
      "step": 7588
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.378,
      "step": 7589
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.3982,
      "step": 7590
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4761,
      "step": 7591
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7893,
      "step": 7592
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.552,
      "step": 7593
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 7594
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7429,
      "step": 7595
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 7596
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.618,
      "step": 7597
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7344,
      "step": 7598
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 7599
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 7600
    },
    {
      "epoch": 1.25,
      "eval_loss": 0.7760756611824036,
      "eval_runtime": 131.7027,
      "eval_samples_per_second": 7.593,
      "eval_steps_per_second": 3.796,
      "step": 7600
    },
    {
      "epoch": 1.25,
      "mmlu_eval_accuracy": 0.5451815086513778,
      "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6896551724137931,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.7272727272727273,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.6,
      "mmlu_eval_accuracy_high_school_biology": 0.6875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.4782608695652174,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
      "mmlu_eval_accuracy_prehistory": 0.6571428571428571,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.4117647058823529,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 0.9224504545311548,
      "step": 7600
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6679,
      "step": 7601
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.665,
      "step": 7602
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7777,
      "step": 7603
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 7604
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6479,
      "step": 7605
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7385,
      "step": 7606
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8406,
      "step": 7607
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.868,
      "step": 7608
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7855,
      "step": 7609
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7835,
      "step": 7610
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4515,
      "step": 7611
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7903,
      "step": 7612
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8782,
      "step": 7613
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7044,
      "step": 7614
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.3504,
      "step": 7615
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6363,
      "step": 7616
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8254,
      "step": 7617
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 7618
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 7619
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 7620
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 7621
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.478,
      "step": 7622
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 1.0484,
      "step": 7623
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7988,
      "step": 7624
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4957,
      "step": 7625
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8332,
      "step": 7626
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8142,
      "step": 7627
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6462,
      "step": 7628
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4709,
      "step": 7629
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4719,
      "step": 7630
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7657,
      "step": 7631
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6328,
      "step": 7632
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.6582,
      "step": 7633
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.694,
      "step": 7634
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.8784,
      "step": 7635
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.5367,
      "step": 7636
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.521,
      "step": 7637
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 7638
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 7639
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4348,
      "step": 7640
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5091,
      "step": 7641
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4192,
      "step": 7642
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6793,
      "step": 7643
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5492,
      "step": 7644
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 7645
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 7646
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7513,
      "step": 7647
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6645,
      "step": 7648
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4954,
      "step": 7649
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 7650
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4748,
      "step": 7651
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8201,
      "step": 7652
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5207,
      "step": 7653
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 7654
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 7655
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5386,
      "step": 7656
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7666,
      "step": 7657
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 1.0383,
      "step": 7658
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7586,
      "step": 7659
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5874,
      "step": 7660
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8482,
      "step": 7661
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.584,
      "step": 7662
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4417,
      "step": 7663
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 7664
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 7665
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5103,
      "step": 7666
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7491,
      "step": 7667
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6927,
      "step": 7668
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.805,
      "step": 7669
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5082,
      "step": 7670
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5818,
      "step": 7671
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6724,
      "step": 7672
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 7673
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7739,
      "step": 7674
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7153,
      "step": 7675
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 7676
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6839,
      "step": 7677
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7227,
      "step": 7678
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5682,
      "step": 7679
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8381,
      "step": 7680
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 7681
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.656,
      "step": 7682
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5532,
      "step": 7683
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6708,
      "step": 7684
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 7685
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 1.0666,
      "step": 7686
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 7687
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4912,
      "step": 7688
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.6435,
      "step": 7689
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5831,
      "step": 7690
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.3184,
      "step": 7691
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5042,
      "step": 7692
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8789,
      "step": 7693
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.573,
      "step": 7694
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4751,
      "step": 7695
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.9096,
      "step": 7696
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5184,
      "step": 7697
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.5713,
      "step": 7698
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.4928,
      "step": 7699
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7692,
      "step": 7700
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6527,
      "step": 7701
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.8049,
      "step": 7702
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7093,
      "step": 7703
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5482,
      "step": 7704
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.9586,
      "step": 7705
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6728,
      "step": 7706
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5752,
      "step": 7707
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7346,
      "step": 7708
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 7709
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.2796,
      "step": 7710
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5589,
      "step": 7711
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5042,
      "step": 7712
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6554,
      "step": 7713
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.3267,
      "step": 7714
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6895,
      "step": 7715
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 7716
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 7717
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 7718
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6914,
      "step": 7719
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6749,
      "step": 7720
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6255,
      "step": 7721
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7197,
      "step": 7722
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.8916,
      "step": 7723
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.4618,
      "step": 7724
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6663,
      "step": 7725
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5055,
      "step": 7726
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 7727
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6632,
      "step": 7728
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 7729
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.4448,
      "step": 7730
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6527,
      "step": 7731
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6336,
      "step": 7732
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 7733
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 7734
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.675,
      "step": 7735
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 7736
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7549,
      "step": 7737
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 7738
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6424,
      "step": 7739
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.4341,
      "step": 7740
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.8305,
      "step": 7741
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.4854,
      "step": 7742
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.4582,
      "step": 7743
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 7744
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 7745
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.3909,
      "step": 7746
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.604,
      "step": 7747
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6055,
      "step": 7748
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.3337,
      "step": 7749
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5912,
      "step": 7750
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 7751
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.466,
      "step": 7752
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 7753
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 7754
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5409,
      "step": 7755
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 7756
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.6323,
      "step": 7757
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5877,
      "step": 7758
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.5555,
      "step": 7759
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 7760
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4506,
      "step": 7761
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7674,
      "step": 7762
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6686,
      "step": 7763
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 7764
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5073,
      "step": 7765
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.8264,
      "step": 7766
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6984,
      "step": 7767
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 7768
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.9556,
      "step": 7769
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 7770
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7991,
      "step": 7771
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7941,
      "step": 7772
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 7773
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7439,
      "step": 7774
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6236,
      "step": 7775
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6658,
      "step": 7776
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.8767,
      "step": 7777
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5511,
      "step": 7778
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6276,
      "step": 7779
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.8317,
      "step": 7780
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.548,
      "step": 7781
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4923,
      "step": 7782
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 7783
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 7784
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 7785
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5648,
      "step": 7786
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 7787
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5235,
      "step": 7788
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6775,
      "step": 7789
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 7790
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7583,
      "step": 7791
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 7792
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 7793
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7247,
      "step": 7794
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.617,
      "step": 7795
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7264,
      "step": 7796
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6413,
      "step": 7797
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5632,
      "step": 7798
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 7799
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 7800
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.7729380130767822,
      "eval_runtime": 131.7586,
      "eval_samples_per_second": 7.59,
      "eval_steps_per_second": 3.795,
      "step": 7800
    },
    {
      "epoch": 1.28,
      "mmlu_eval_accuracy": 0.5601061751056764,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "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": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "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": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "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": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.4,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.38823529411764707,
      "mmlu_eval_accuracy_professional_medicine": 0.6451612903225806,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.1240841240546722,
      "step": 7800
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7872,
      "step": 7801
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6561,
      "step": 7802
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.3575,
      "step": 7803
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4131,
      "step": 7804
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6792,
      "step": 7805
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 7806
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4591,
      "step": 7807
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4738,
      "step": 7808
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 7809
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.8097,
      "step": 7810
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.6552,
      "step": 7811
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.3649,
      "step": 7812
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7005,
      "step": 7813
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5639,
      "step": 7814
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7522,
      "step": 7815
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 7816
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.4313,
      "step": 7817
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7227,
      "step": 7818
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5488,
      "step": 7819
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.5978,
      "step": 7820
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 7821
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.8046,
      "step": 7822
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.701,
      "step": 7823
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6226,
      "step": 7824
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5258,
      "step": 7825
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4529,
      "step": 7826
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6458,
      "step": 7827
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4508,
      "step": 7828
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6579,
      "step": 7829
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4943,
      "step": 7830
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5775,
      "step": 7831
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4833,
      "step": 7832
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 7833
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 7834
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.3567,
      "step": 7835
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 7836
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.3403,
      "step": 7837
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7374,
      "step": 7838
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7311,
      "step": 7839
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6753,
      "step": 7840
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.844,
      "step": 7841
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7689,
      "step": 7842
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6946,
      "step": 7843
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6129,
      "step": 7844
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.721,
      "step": 7845
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 7846
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 7847
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.537,
      "step": 7848
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7264,
      "step": 7849
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 7850
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.3925,
      "step": 7851
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 7852
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6567,
      "step": 7853
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4655,
      "step": 7854
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 7855
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6279,
      "step": 7856
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7146,
      "step": 7857
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7039,
      "step": 7858
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.9011,
      "step": 7859
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5716,
      "step": 7860
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7776,
      "step": 7861
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4204,
      "step": 7862
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 7863
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7991,
      "step": 7864
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 7865
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 7866
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 7867
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 7868
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6621,
      "step": 7869
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6052,
      "step": 7870
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6009,
      "step": 7871
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5455,
      "step": 7872
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6069,
      "step": 7873
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5893,
      "step": 7874
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6739,
      "step": 7875
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.6871,
      "step": 7876
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 7877
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.8126,
      "step": 7878
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5435,
      "step": 7879
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.4958,
      "step": 7880
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.5472,
      "step": 7881
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7839,
      "step": 7882
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4211,
      "step": 7883
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4046,
      "step": 7884
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8814,
      "step": 7885
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.3878,
      "step": 7886
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 7887
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8658,
      "step": 7888
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6381,
      "step": 7889
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8015,
      "step": 7890
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4461,
      "step": 7891
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4474,
      "step": 7892
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6649,
      "step": 7893
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8264,
      "step": 7894
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4986,
      "step": 7895
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6113,
      "step": 7896
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.653,
      "step": 7897
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 7898
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.323,
      "step": 7899
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7812,
      "step": 7900
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4724,
      "step": 7901
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6228,
      "step": 7902
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4618,
      "step": 7903
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.422,
      "step": 7904
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5829,
      "step": 7905
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 7906
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6225,
      "step": 7907
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6895,
      "step": 7908
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6843,
      "step": 7909
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5122,
      "step": 7910
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5575,
      "step": 7911
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 7912
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4256,
      "step": 7913
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5912,
      "step": 7914
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.831,
      "step": 7915
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6549,
      "step": 7916
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.627,
      "step": 7917
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6647,
      "step": 7918
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 7919
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 7920
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6995,
      "step": 7921
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 7922
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7613,
      "step": 7923
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 7924
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.3925,
      "step": 7925
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6413,
      "step": 7926
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5873,
      "step": 7927
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5128,
      "step": 7928
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8168,
      "step": 7929
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5317,
      "step": 7930
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 7931
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5782,
      "step": 7932
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6769,
      "step": 7933
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8389,
      "step": 7934
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.4664,
      "step": 7935
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 7936
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6935,
      "step": 7937
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7101,
      "step": 7938
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8767,
      "step": 7939
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.8802,
      "step": 7940
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 7941
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.6344,
      "step": 7942
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5075,
      "step": 7943
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4093,
      "step": 7944
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.888,
      "step": 7945
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5689,
      "step": 7946
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.8472,
      "step": 7947
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 7948
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6741,
      "step": 7949
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.545,
      "step": 7950
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.8613,
      "step": 7951
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.833,
      "step": 7952
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 7953
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5061,
      "step": 7954
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5018,
      "step": 7955
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.472,
      "step": 7956
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6258,
      "step": 7957
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6295,
      "step": 7958
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7043,
      "step": 7959
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5259,
      "step": 7960
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 7961
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5328,
      "step": 7962
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.355,
      "step": 7963
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5369,
      "step": 7964
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4982,
      "step": 7965
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.627,
      "step": 7966
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.3691,
      "step": 7967
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6125,
      "step": 7968
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.3286,
      "step": 7969
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5071,
      "step": 7970
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7013,
      "step": 7971
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6566,
      "step": 7972
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 7973
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5674,
      "step": 7974
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 7975
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5818,
      "step": 7976
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7269,
      "step": 7977
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.514,
      "step": 7978
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7766,
      "step": 7979
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6893,
      "step": 7980
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7425,
      "step": 7981
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.932,
      "step": 7982
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6135,
      "step": 7983
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.803,
      "step": 7984
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 7985
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7173,
      "step": 7986
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7096,
      "step": 7987
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 7988
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.8765,
      "step": 7989
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4574,
      "step": 7990
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 7991
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.2791,
      "step": 7992
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 7993
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5321,
      "step": 7994
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 7995
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4831,
      "step": 7996
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 7997
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7438,
      "step": 7998
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 7999
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.5919,
      "step": 8000
    },
    {
      "epoch": 1.31,
      "eval_loss": 0.7794589996337891,
      "eval_runtime": 131.6566,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 8000
    },
    {
      "epoch": 1.31,
      "mmlu_eval_accuracy": 0.5317049567035266,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.25,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6571428571428571,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.15829147211721,
      "step": 8000
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4632,
      "step": 8001
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.8213,
      "step": 8002
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 8003
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5903,
      "step": 8004
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5132,
      "step": 8005
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6646,
      "step": 8006
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7227,
      "step": 8007
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 1.0397,
      "step": 8008
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 8009
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7378,
      "step": 8010
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.3806,
      "step": 8011
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5808,
      "step": 8012
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 8013
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 8014
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.48,
      "step": 8015
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 8016
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5118,
      "step": 8017
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7843,
      "step": 8018
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 8019
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 8020
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6762,
      "step": 8021
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5033,
      "step": 8022
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 8023
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8128,
      "step": 8024
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6625,
      "step": 8025
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 8026
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8369,
      "step": 8027
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5421,
      "step": 8028
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 8029
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.822,
      "step": 8030
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 8031
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 8032
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4793,
      "step": 8033
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7367,
      "step": 8034
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.895,
      "step": 8035
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6667,
      "step": 8036
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7863,
      "step": 8037
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 1.0227,
      "step": 8038
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.703,
      "step": 8039
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 8040
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8328,
      "step": 8041
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.507,
      "step": 8042
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5904,
      "step": 8043
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5455,
      "step": 8044
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6356,
      "step": 8045
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.514,
      "step": 8046
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 8047
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4932,
      "step": 8048
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7382,
      "step": 8049
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7831,
      "step": 8050
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5862,
      "step": 8051
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5497,
      "step": 8052
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4702,
      "step": 8053
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.5215,
      "step": 8054
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6894,
      "step": 8055
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.384,
      "step": 8056
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.3363,
      "step": 8057
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 8058
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 8059
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.466,
      "step": 8060
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7573,
      "step": 8061
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7389,
      "step": 8062
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.8822,
      "step": 8063
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4676,
      "step": 8064
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5512,
      "step": 8065
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5039,
      "step": 8066
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.4943,
      "step": 8067
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.4892,
      "step": 8068
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5303,
      "step": 8069
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.8446,
      "step": 8070
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7236,
      "step": 8071
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7759,
      "step": 8072
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6931,
      "step": 8073
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5258,
      "step": 8074
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.8117,
      "step": 8075
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5262,
      "step": 8076
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 8077
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 8078
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5521,
      "step": 8079
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.4981,
      "step": 8080
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5634,
      "step": 8081
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.4695,
      "step": 8082
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 8083
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.865,
      "step": 8084
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 8085
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 8086
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5773,
      "step": 8087
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 8088
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 8089
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5983,
      "step": 8090
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 8091
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7238,
      "step": 8092
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7086,
      "step": 8093
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 8094
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.55,
      "step": 8095
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.8097,
      "step": 8096
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6482,
      "step": 8097
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 8098
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 8099
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7727,
      "step": 8100
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 8101
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5023,
      "step": 8102
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5909,
      "step": 8103
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.789,
      "step": 8104
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.4475,
      "step": 8105
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6493,
      "step": 8106
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7609,
      "step": 8107
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 8108
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 8109
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.8342,
      "step": 8110
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 8111
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 8112
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6852,
      "step": 8113
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.692,
      "step": 8114
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6153,
      "step": 8115
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6834,
      "step": 8116
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7616,
      "step": 8117
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.2032,
      "step": 8118
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6226,
      "step": 8119
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.5564,
      "step": 8120
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7813,
      "step": 8121
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7146,
      "step": 8122
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.8877,
      "step": 8123
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6116,
      "step": 8124
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.6067,
      "step": 8125
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7176,
      "step": 8126
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4961,
      "step": 8127
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 8128
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 8129
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6195,
      "step": 8130
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4347,
      "step": 8131
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5122,
      "step": 8132
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4328,
      "step": 8133
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7816,
      "step": 8134
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.3364,
      "step": 8135
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5325,
      "step": 8136
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4997,
      "step": 8137
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 8138
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7123,
      "step": 8139
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.8648,
      "step": 8140
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 8141
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.9509,
      "step": 8142
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6699,
      "step": 8143
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5328,
      "step": 8144
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4067,
      "step": 8145
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5406,
      "step": 8146
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4749,
      "step": 8147
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.444,
      "step": 8148
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6838,
      "step": 8149
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 8150
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.9077,
      "step": 8151
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 8152
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6955,
      "step": 8153
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6277,
      "step": 8154
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4177,
      "step": 8155
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5416,
      "step": 8156
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.9269,
      "step": 8157
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6442,
      "step": 8158
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5957,
      "step": 8159
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7303,
      "step": 8160
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7404,
      "step": 8161
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.8418,
      "step": 8162
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6528,
      "step": 8163
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.3884,
      "step": 8164
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.844,
      "step": 8165
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 8166
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 8167
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 8168
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.604,
      "step": 8169
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 8170
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.8151,
      "step": 8171
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.485,
      "step": 8172
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6308,
      "step": 8173
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7103,
      "step": 8174
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6586,
      "step": 8175
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 8176
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5298,
      "step": 8177
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6416,
      "step": 8178
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6531,
      "step": 8179
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5048,
      "step": 8180
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 8181
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5778,
      "step": 8182
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6604,
      "step": 8183
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.5754,
      "step": 8184
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 8185
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.4085,
      "step": 8186
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.4523,
      "step": 8187
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5407,
      "step": 8188
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.547,
      "step": 8189
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8106,
      "step": 8190
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8161,
      "step": 8191
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6987,
      "step": 8192
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5929,
      "step": 8193
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 8194
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 8195
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.7075,
      "step": 8196
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6917,
      "step": 8197
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.7744,
      "step": 8198
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 8199
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.605,
      "step": 8200
    },
    {
      "epoch": 1.35,
      "eval_loss": 0.7757422924041748,
      "eval_runtime": 131.745,
      "eval_samples_per_second": 7.59,
      "eval_steps_per_second": 3.795,
      "step": 8200
    },
    {
      "epoch": 1.35,
      "mmlu_eval_accuracy": 0.5172589770054279,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.7272727272727273,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.25,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.65625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.25,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.631578947368421,
      "mmlu_eval_accuracy_moral_scenarios": 0.26,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.36470588235294116,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.081496161090176,
      "step": 8200
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.497,
      "step": 8201
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 8202
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.707,
      "step": 8203
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.7669,
      "step": 8204
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.4459,
      "step": 8205
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.3805,
      "step": 8206
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8387,
      "step": 8207
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5064,
      "step": 8208
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6826,
      "step": 8209
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5136,
      "step": 8210
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.739,
      "step": 8211
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 8212
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 8213
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 8214
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8463,
      "step": 8215
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.7634,
      "step": 8216
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6105,
      "step": 8217
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.4257,
      "step": 8218
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5162,
      "step": 8219
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5551,
      "step": 8220
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.4913,
      "step": 8221
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6607,
      "step": 8222
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8764,
      "step": 8223
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 8224
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.3817,
      "step": 8225
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 8226
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 8227
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 8228
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5268,
      "step": 8229
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5788,
      "step": 8230
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.8344,
      "step": 8231
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.3551,
      "step": 8232
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.675,
      "step": 8233
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5943,
      "step": 8234
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6243,
      "step": 8235
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6148,
      "step": 8236
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 8237
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5592,
      "step": 8238
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6352,
      "step": 8239
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5187,
      "step": 8240
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6504,
      "step": 8241
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.9117,
      "step": 8242
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5805,
      "step": 8243
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.6739,
      "step": 8244
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5228,
      "step": 8245
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 8246
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7111,
      "step": 8247
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 8248
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 8249
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6749,
      "step": 8250
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5369,
      "step": 8251
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5683,
      "step": 8252
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4744,
      "step": 8253
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4661,
      "step": 8254
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5584,
      "step": 8255
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6758,
      "step": 8256
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7656,
      "step": 8257
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4923,
      "step": 8258
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4558,
      "step": 8259
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5215,
      "step": 8260
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6018,
      "step": 8261
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5308,
      "step": 8262
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 8263
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5403,
      "step": 8264
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 8265
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 8266
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5056,
      "step": 8267
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.704,
      "step": 8268
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 8269
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6826,
      "step": 8270
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 8271
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.8948,
      "step": 8272
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6107,
      "step": 8273
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4571,
      "step": 8274
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.3666,
      "step": 8275
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.791,
      "step": 8276
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4657,
      "step": 8277
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4496,
      "step": 8278
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.8045,
      "step": 8279
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.9852,
      "step": 8280
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.9997,
      "step": 8281
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5224,
      "step": 8282
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6179,
      "step": 8283
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6918,
      "step": 8284
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6302,
      "step": 8285
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7578,
      "step": 8286
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 8287
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.774,
      "step": 8288
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5398,
      "step": 8289
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6214,
      "step": 8290
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 8291
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.433,
      "step": 8292
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5104,
      "step": 8293
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6531,
      "step": 8294
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.4948,
      "step": 8295
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6447,
      "step": 8296
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7646,
      "step": 8297
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.5217,
      "step": 8298
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7436,
      "step": 8299
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7645,
      "step": 8300
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6387,
      "step": 8301
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7582,
      "step": 8302
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6482,
      "step": 8303
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.9254,
      "step": 8304
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.3831,
      "step": 8305
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.7971,
      "step": 8306
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.6572,
      "step": 8307
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5887,
      "step": 8308
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5385,
      "step": 8309
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6278,
      "step": 8310
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4748,
      "step": 8311
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 8312
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6611,
      "step": 8313
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5591,
      "step": 8314
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7101,
      "step": 8315
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8309,
      "step": 8316
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8933,
      "step": 8317
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6311,
      "step": 8318
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 8319
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.55,
      "step": 8320
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8568,
      "step": 8321
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4875,
      "step": 8322
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4716,
      "step": 8323
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4619,
      "step": 8324
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6042,
      "step": 8325
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7847,
      "step": 8326
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6826,
      "step": 8327
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6477,
      "step": 8328
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6527,
      "step": 8329
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6299,
      "step": 8330
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 8331
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5896,
      "step": 8332
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.414,
      "step": 8333
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5784,
      "step": 8334
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4428,
      "step": 8335
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 8336
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.458,
      "step": 8337
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7777,
      "step": 8338
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5063,
      "step": 8339
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4956,
      "step": 8340
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8325,
      "step": 8341
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6137,
      "step": 8342
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 8343
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.4547,
      "step": 8344
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5019,
      "step": 8345
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 8346
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 8347
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6145,
      "step": 8348
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7462,
      "step": 8349
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 8350
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.317,
      "step": 8351
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 8352
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6064,
      "step": 8353
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5641,
      "step": 8354
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6008,
      "step": 8355
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6001,
      "step": 8356
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.5577,
      "step": 8357
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8637,
      "step": 8358
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 8359
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8698,
      "step": 8360
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.8892,
      "step": 8361
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 8362
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.9339,
      "step": 8363
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6172,
      "step": 8364
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6297,
      "step": 8365
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 8366
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 8367
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 8368
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4998,
      "step": 8369
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6585,
      "step": 8370
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7358,
      "step": 8371
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.9232,
      "step": 8372
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7112,
      "step": 8373
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.3817,
      "step": 8374
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5114,
      "step": 8375
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 8376
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5247,
      "step": 8377
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.3785,
      "step": 8378
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.3681,
      "step": 8379
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 8380
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7988,
      "step": 8381
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 8382
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8768,
      "step": 8383
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5374,
      "step": 8384
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5785,
      "step": 8385
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6268,
      "step": 8386
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8157,
      "step": 8387
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 8388
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5514,
      "step": 8389
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 1.0976,
      "step": 8390
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 8391
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4737,
      "step": 8392
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7461,
      "step": 8393
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4877,
      "step": 8394
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 8395
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4869,
      "step": 8396
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7336,
      "step": 8397
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 8398
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.9334,
      "step": 8399
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5363,
      "step": 8400
    },
    {
      "epoch": 1.38,
      "eval_loss": 0.7778916954994202,
      "eval_runtime": 131.6475,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 8400
    },
    {
      "epoch": 1.38,
      "mmlu_eval_accuracy": 0.5020002700041991,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "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": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.16666666666666666,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.5797101449275363,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 0.9001535251055779,
      "step": 8400
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7832,
      "step": 8401
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7619,
      "step": 8402
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8544,
      "step": 8403
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5486,
      "step": 8404
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6073,
      "step": 8405
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4673,
      "step": 8406
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.839,
      "step": 8407
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.516,
      "step": 8408
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 8409
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 8410
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.7622,
      "step": 8411
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8904,
      "step": 8412
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.4175,
      "step": 8413
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6944,
      "step": 8414
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.9472,
      "step": 8415
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.724,
      "step": 8416
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.67,
      "step": 8417
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6214,
      "step": 8418
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8004,
      "step": 8419
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 8420
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5704,
      "step": 8421
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 8422
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5242,
      "step": 8423
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8493,
      "step": 8424
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.8376,
      "step": 8425
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5025,
      "step": 8426
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.6801,
      "step": 8427
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.416,
      "step": 8428
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.5421,
      "step": 8429
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5841,
      "step": 8430
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 8431
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7918,
      "step": 8432
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.9162,
      "step": 8433
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7025,
      "step": 8434
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5593,
      "step": 8435
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 8436
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.4227,
      "step": 8437
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.4255,
      "step": 8438
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.8895,
      "step": 8439
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6244,
      "step": 8440
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7416,
      "step": 8441
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 8442
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.8477,
      "step": 8443
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6998,
      "step": 8444
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 8445
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 8446
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 8447
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5917,
      "step": 8448
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6461,
      "step": 8449
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 8450
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5804,
      "step": 8451
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.8091,
      "step": 8452
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.4443,
      "step": 8453
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6713,
      "step": 8454
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5023,
      "step": 8455
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6324,
      "step": 8456
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6172,
      "step": 8457
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.2728,
      "step": 8458
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5599,
      "step": 8459
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5532,
      "step": 8460
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7052,
      "step": 8461
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5851,
      "step": 8462
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 8463
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5886,
      "step": 8464
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.692,
      "step": 8465
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5481,
      "step": 8466
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.3677,
      "step": 8467
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6599,
      "step": 8468
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5483,
      "step": 8469
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7756,
      "step": 8470
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5629,
      "step": 8471
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7866,
      "step": 8472
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 8473
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5936,
      "step": 8474
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6078,
      "step": 8475
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.6322,
      "step": 8476
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.8072,
      "step": 8477
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 8478
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.3949,
      "step": 8479
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.4105,
      "step": 8480
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5805,
      "step": 8481
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.74,
      "step": 8482
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5443,
      "step": 8483
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.4014,
      "step": 8484
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5998,
      "step": 8485
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 8486
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.9005,
      "step": 8487
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.737,
      "step": 8488
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5041,
      "step": 8489
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.0002,
      "loss": 0.5655,
      "step": 8490
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5066,
      "step": 8491
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.2922,
      "step": 8492
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 8493
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5936,
      "step": 8494
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 8495
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.9405,
      "step": 8496
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5576,
      "step": 8497
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 8498
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7368,
      "step": 8499
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.8446,
      "step": 8500
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4776,
      "step": 8501
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.511,
      "step": 8502
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7885,
      "step": 8503
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5312,
      "step": 8504
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6385,
      "step": 8505
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6261,
      "step": 8506
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6665,
      "step": 8507
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.363,
      "step": 8508
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.616,
      "step": 8509
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7172,
      "step": 8510
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.709,
      "step": 8511
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7526,
      "step": 8512
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 8513
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4784,
      "step": 8514
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4904,
      "step": 8515
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6539,
      "step": 8516
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6513,
      "step": 8517
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5925,
      "step": 8518
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5508,
      "step": 8519
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4804,
      "step": 8520
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 8521
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4235,
      "step": 8522
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7688,
      "step": 8523
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5986,
      "step": 8524
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.3932,
      "step": 8525
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5384,
      "step": 8526
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6546,
      "step": 8527
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5324,
      "step": 8528
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6073,
      "step": 8529
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4279,
      "step": 8530
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4045,
      "step": 8531
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6817,
      "step": 8532
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 8533
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 8534
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.577,
      "step": 8535
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.8116,
      "step": 8536
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6799,
      "step": 8537
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.8847,
      "step": 8538
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6633,
      "step": 8539
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7403,
      "step": 8540
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7019,
      "step": 8541
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5172,
      "step": 8542
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.8013,
      "step": 8543
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6973,
      "step": 8544
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.5327,
      "step": 8545
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6462,
      "step": 8546
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.7853,
      "step": 8547
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6302,
      "step": 8548
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6361,
      "step": 8549
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.6293,
      "step": 8550
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.4871,
      "step": 8551
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7144,
      "step": 8552
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6932,
      "step": 8553
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6357,
      "step": 8554
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6396,
      "step": 8555
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7047,
      "step": 8556
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8291,
      "step": 8557
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6807,
      "step": 8558
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7011,
      "step": 8559
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8092,
      "step": 8560
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.5389,
      "step": 8561
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7612,
      "step": 8562
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.595,
      "step": 8563
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.556,
      "step": 8564
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8121,
      "step": 8565
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.555,
      "step": 8566
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 8567
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6665,
      "step": 8568
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7743,
      "step": 8569
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 1.0051,
      "step": 8570
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.646,
      "step": 8571
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4562,
      "step": 8572
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 8573
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7078,
      "step": 8574
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 8575
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7645,
      "step": 8576
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6447,
      "step": 8577
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7678,
      "step": 8578
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.904,
      "step": 8579
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6037,
      "step": 8580
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6478,
      "step": 8581
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.392,
      "step": 8582
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4866,
      "step": 8583
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7721,
      "step": 8584
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.5162,
      "step": 8585
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4829,
      "step": 8586
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 8587
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.68,
      "step": 8588
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 8589
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6677,
      "step": 8590
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8254,
      "step": 8591
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7003,
      "step": 8592
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4711,
      "step": 8593
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4686,
      "step": 8594
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 8595
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 8596
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4591,
      "step": 8597
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.657,
      "step": 8598
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.9055,
      "step": 8599
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 8600
    },
    {
      "epoch": 1.41,
      "eval_loss": 0.7812375426292419,
      "eval_runtime": 131.5866,
      "eval_samples_per_second": 7.6,
      "eval_steps_per_second": 3.8,
      "step": 8600
    },
    {
      "epoch": 1.41,
      "mmlu_eval_accuracy": 0.5361827071667484,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.7272727272727273,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.25,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.0922236791459294,
      "step": 8600
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8426,
      "step": 8601
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7865,
      "step": 8602
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4936,
      "step": 8603
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6612,
      "step": 8604
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 8605
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 8606
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.8524,
      "step": 8607
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 8608
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 8609
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 8610
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.4221,
      "step": 8611
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.6757,
      "step": 8612
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.793,
      "step": 8613
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6184,
      "step": 8614
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4524,
      "step": 8615
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5927,
      "step": 8616
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4718,
      "step": 8617
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 8618
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5938,
      "step": 8619
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7834,
      "step": 8620
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5862,
      "step": 8621
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7703,
      "step": 8622
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5234,
      "step": 8623
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 8624
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 8625
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4604,
      "step": 8626
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6222,
      "step": 8627
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5068,
      "step": 8628
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.8806,
      "step": 8629
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4422,
      "step": 8630
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 8631
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6963,
      "step": 8632
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6301,
      "step": 8633
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.8897,
      "step": 8634
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 1.094,
      "step": 8635
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6626,
      "step": 8636
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 8637
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7675,
      "step": 8638
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5782,
      "step": 8639
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 8640
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.534,
      "step": 8641
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4116,
      "step": 8642
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4842,
      "step": 8643
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.471,
      "step": 8644
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4738,
      "step": 8645
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6429,
      "step": 8646
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 8647
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 1.0546,
      "step": 8648
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 8649
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.3606,
      "step": 8650
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 8651
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6799,
      "step": 8652
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 8653
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5279,
      "step": 8654
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.741,
      "step": 8655
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7678,
      "step": 8656
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6515,
      "step": 8657
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.737,
      "step": 8658
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.4934,
      "step": 8659
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7364,
      "step": 8660
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.9114,
      "step": 8661
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6414,
      "step": 8662
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7861,
      "step": 8663
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5707,
      "step": 8664
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.9125,
      "step": 8665
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5438,
      "step": 8666
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.5017,
      "step": 8667
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 8668
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 8669
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6753,
      "step": 8670
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.6322,
      "step": 8671
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.698,
      "step": 8672
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 8673
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5899,
      "step": 8674
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.8438,
      "step": 8675
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.933,
      "step": 8676
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 8677
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 8678
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4456,
      "step": 8679
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6759,
      "step": 8680
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5503,
      "step": 8681
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4996,
      "step": 8682
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4143,
      "step": 8683
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4545,
      "step": 8684
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7065,
      "step": 8685
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.3267,
      "step": 8686
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 8687
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7161,
      "step": 8688
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 8689
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 8690
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5615,
      "step": 8691
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.587,
      "step": 8692
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.8078,
      "step": 8693
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 8694
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5549,
      "step": 8695
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.9401,
      "step": 8696
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 8697
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5626,
      "step": 8698
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5387,
      "step": 8699
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 8700
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 8701
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.3098,
      "step": 8702
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 8703
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.8071,
      "step": 8704
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4034,
      "step": 8705
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 8706
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6642,
      "step": 8707
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5684,
      "step": 8708
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.9047,
      "step": 8709
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 8710
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.8877,
      "step": 8711
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6257,
      "step": 8712
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 8713
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4047,
      "step": 8714
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6731,
      "step": 8715
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.577,
      "step": 8716
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7615,
      "step": 8717
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5337,
      "step": 8718
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7013,
      "step": 8719
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6164,
      "step": 8720
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6048,
      "step": 8721
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4032,
      "step": 8722
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4898,
      "step": 8723
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4453,
      "step": 8724
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.4741,
      "step": 8725
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7929,
      "step": 8726
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5921,
      "step": 8727
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.591,
      "step": 8728
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.6245,
      "step": 8729
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.5535,
      "step": 8730
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.3502,
      "step": 8731
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.7526,
      "step": 8732
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.8276,
      "step": 8733
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 8734
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.8477,
      "step": 8735
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7131,
      "step": 8736
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7526,
      "step": 8737
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.75,
      "step": 8738
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 8739
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6841,
      "step": 8740
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.443,
      "step": 8741
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6686,
      "step": 8742
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5093,
      "step": 8743
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6137,
      "step": 8744
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4707,
      "step": 8745
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5775,
      "step": 8746
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7726,
      "step": 8747
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6781,
      "step": 8748
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7021,
      "step": 8749
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4997,
      "step": 8750
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6491,
      "step": 8751
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5921,
      "step": 8752
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 8753
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5243,
      "step": 8754
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4582,
      "step": 8755
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5562,
      "step": 8756
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.368,
      "step": 8757
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 8758
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 8759
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7887,
      "step": 8760
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6455,
      "step": 8761
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6761,
      "step": 8762
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.8517,
      "step": 8763
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6477,
      "step": 8764
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.602,
      "step": 8765
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 8766
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5292,
      "step": 8767
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.8326,
      "step": 8768
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5127,
      "step": 8769
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 8770
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 8771
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6667,
      "step": 8772
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6598,
      "step": 8773
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4451,
      "step": 8774
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 8775
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 8776
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.592,
      "step": 8777
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.377,
      "step": 8778
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 8779
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 8780
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5012,
      "step": 8781
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6208,
      "step": 8782
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 8783
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6597,
      "step": 8784
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5544,
      "step": 8785
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 8786
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 8787
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5502,
      "step": 8788
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4808,
      "step": 8789
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5741,
      "step": 8790
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.6623,
      "step": 8791
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.5146,
      "step": 8792
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.8079,
      "step": 8793
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.7437,
      "step": 8794
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5366,
      "step": 8795
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5526,
      "step": 8796
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.4716,
      "step": 8797
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6315,
      "step": 8798
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.723,
      "step": 8799
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5838,
      "step": 8800
    },
    {
      "epoch": 1.45,
      "eval_loss": 0.7777678370475769,
      "eval_runtime": 131.6412,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 8800
    },
    {
      "epoch": 1.45,
      "mmlu_eval_accuracy": 0.5265974068362751,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "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": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.5217391304347826,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.3543437047431115,
      "step": 8800
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.7847,
      "step": 8801
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.614,
      "step": 8802
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6911,
      "step": 8803
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5221,
      "step": 8804
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.787,
      "step": 8805
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5605,
      "step": 8806
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5252,
      "step": 8807
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5872,
      "step": 8808
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.417,
      "step": 8809
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8078,
      "step": 8810
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6009,
      "step": 8811
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5915,
      "step": 8812
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 8813
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6651,
      "step": 8814
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.3349,
      "step": 8815
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6786,
      "step": 8816
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5177,
      "step": 8817
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.7824,
      "step": 8818
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.816,
      "step": 8819
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.9371,
      "step": 8820
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5936,
      "step": 8821
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.7692,
      "step": 8822
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5457,
      "step": 8823
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 8824
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8563,
      "step": 8825
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6888,
      "step": 8826
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.3971,
      "step": 8827
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5678,
      "step": 8828
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 8829
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8842,
      "step": 8830
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5112,
      "step": 8831
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 8832
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6841,
      "step": 8833
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6036,
      "step": 8834
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.4636,
      "step": 8835
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6931,
      "step": 8836
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8747,
      "step": 8837
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8223,
      "step": 8838
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8292,
      "step": 8839
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.4308,
      "step": 8840
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 8841
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.6604,
      "step": 8842
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.4473,
      "step": 8843
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 8844
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 1.0113,
      "step": 8845
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8291,
      "step": 8846
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5231,
      "step": 8847
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5947,
      "step": 8848
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.953,
      "step": 8849
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.465,
      "step": 8850
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5045,
      "step": 8851
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.634,
      "step": 8852
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.4429,
      "step": 8853
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.5975,
      "step": 8854
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.8424,
      "step": 8855
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.8453,
      "step": 8856
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5567,
      "step": 8857
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6014,
      "step": 8858
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6466,
      "step": 8859
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6974,
      "step": 8860
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.9228,
      "step": 8861
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5543,
      "step": 8862
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7264,
      "step": 8863
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.4117,
      "step": 8864
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.662,
      "step": 8865
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5056,
      "step": 8866
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.8278,
      "step": 8867
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.576,
      "step": 8868
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5305,
      "step": 8869
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.3432,
      "step": 8870
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 8871
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6112,
      "step": 8872
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.563,
      "step": 8873
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6615,
      "step": 8874
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.9454,
      "step": 8875
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.8261,
      "step": 8876
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6192,
      "step": 8877
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.4953,
      "step": 8878
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6485,
      "step": 8879
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 8880
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5081,
      "step": 8881
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6307,
      "step": 8882
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.773,
      "step": 8883
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7309,
      "step": 8884
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6061,
      "step": 8885
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6634,
      "step": 8886
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 8887
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 8888
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7681,
      "step": 8889
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6992,
      "step": 8890
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6496,
      "step": 8891
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5897,
      "step": 8892
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 8893
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 8894
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.8545,
      "step": 8895
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7215,
      "step": 8896
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.3933,
      "step": 8897
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.572,
      "step": 8898
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5046,
      "step": 8899
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 8900
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5282,
      "step": 8901
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7773,
      "step": 8902
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7886,
      "step": 8903
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 8904
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 8905
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 8906
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 8907
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.9359,
      "step": 8908
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.4033,
      "step": 8909
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.6706,
      "step": 8910
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.4752,
      "step": 8911
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.7635,
      "step": 8912
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5579,
      "step": 8913
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5819,
      "step": 8914
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 8915
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.5023,
      "step": 8916
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.3826,
      "step": 8917
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5919,
      "step": 8918
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 8919
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 8920
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7451,
      "step": 8921
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.3802,
      "step": 8922
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8172,
      "step": 8923
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7015,
      "step": 8924
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8755,
      "step": 8925
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5745,
      "step": 8926
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6971,
      "step": 8927
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 8928
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6606,
      "step": 8929
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 8930
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7045,
      "step": 8931
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6427,
      "step": 8932
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6367,
      "step": 8933
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8453,
      "step": 8934
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4681,
      "step": 8935
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4007,
      "step": 8936
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5578,
      "step": 8937
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 8938
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5162,
      "step": 8939
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4784,
      "step": 8940
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8286,
      "step": 8941
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4663,
      "step": 8942
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.723,
      "step": 8943
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6397,
      "step": 8944
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 8945
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5993,
      "step": 8946
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4476,
      "step": 8947
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7776,
      "step": 8948
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5435,
      "step": 8949
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 8950
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6346,
      "step": 8951
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6338,
      "step": 8952
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 8953
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5737,
      "step": 8954
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 1.1031,
      "step": 8955
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7108,
      "step": 8956
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 8957
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 8958
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8869,
      "step": 8959
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.601,
      "step": 8960
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6974,
      "step": 8961
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 8962
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4554,
      "step": 8963
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5316,
      "step": 8964
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.626,
      "step": 8965
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4411,
      "step": 8966
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.7711,
      "step": 8967
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 8968
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6111,
      "step": 8969
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4988,
      "step": 8970
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4421,
      "step": 8971
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5265,
      "step": 8972
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 8973
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.418,
      "step": 8974
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.8286,
      "step": 8975
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 8976
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.6697,
      "step": 8977
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5972,
      "step": 8978
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 8979
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4879,
      "step": 8980
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4822,
      "step": 8981
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 1.0264,
      "step": 8982
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 8983
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.874,
      "step": 8984
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5196,
      "step": 8985
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.3635,
      "step": 8986
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7388,
      "step": 8987
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6862,
      "step": 8988
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7163,
      "step": 8989
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6612,
      "step": 8990
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7474,
      "step": 8991
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.3564,
      "step": 8992
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 8993
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5481,
      "step": 8994
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7697,
      "step": 8995
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7104,
      "step": 8996
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6408,
      "step": 8997
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 8998
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7365,
      "step": 8999
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6611,
      "step": 9000
    },
    {
      "epoch": 1.48,
      "eval_loss": 0.7830467820167542,
      "eval_runtime": 131.6172,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 9000
    },
    {
      "epoch": 1.48,
      "mmlu_eval_accuracy": 0.531024941438581,
      "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": 0.6551724137931034,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.375,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "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": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.6285714285714286,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.0819902061132165,
      "step": 9000
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4387,
      "step": 9001
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.447,
      "step": 9002
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7779,
      "step": 9003
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5533,
      "step": 9004
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4942,
      "step": 9005
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5324,
      "step": 9006
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6097,
      "step": 9007
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 9008
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6905,
      "step": 9009
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.9148,
      "step": 9010
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6262,
      "step": 9011
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.776,
      "step": 9012
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 9013
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4998,
      "step": 9014
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.8478,
      "step": 9015
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6146,
      "step": 9016
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4837,
      "step": 9017
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.8236,
      "step": 9018
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5461,
      "step": 9019
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 9020
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4944,
      "step": 9021
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7418,
      "step": 9022
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6847,
      "step": 9023
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6806,
      "step": 9024
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7844,
      "step": 9025
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.8634,
      "step": 9026
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 9027
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6389,
      "step": 9028
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4804,
      "step": 9029
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 9030
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 9031
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4692,
      "step": 9032
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.6821,
      "step": 9033
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.9809,
      "step": 9034
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4184,
      "step": 9035
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.8816,
      "step": 9036
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4413,
      "step": 9037
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.9264,
      "step": 9038
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7025,
      "step": 9039
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 9040
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.9348,
      "step": 9041
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 9042
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5684,
      "step": 9043
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4632,
      "step": 9044
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 9045
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.8466,
      "step": 9046
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6539,
      "step": 9047
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5652,
      "step": 9048
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4501,
      "step": 9049
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6149,
      "step": 9050
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.3112,
      "step": 9051
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6793,
      "step": 9052
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7918,
      "step": 9053
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5128,
      "step": 9054
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7989,
      "step": 9055
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 9056
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 9057
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4984,
      "step": 9058
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 9059
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.432,
      "step": 9060
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6781,
      "step": 9061
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.578,
      "step": 9062
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7808,
      "step": 9063
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4526,
      "step": 9064
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6291,
      "step": 9065
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6504,
      "step": 9066
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 9067
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4974,
      "step": 9068
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6104,
      "step": 9069
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7027,
      "step": 9070
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 9071
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 9072
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.755,
      "step": 9073
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7423,
      "step": 9074
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 9075
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4803,
      "step": 9076
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 9077
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 9078
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.8143,
      "step": 9079
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6171,
      "step": 9080
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5868,
      "step": 9081
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.729,
      "step": 9082
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.3669,
      "step": 9083
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7014,
      "step": 9084
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5226,
      "step": 9085
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5599,
      "step": 9086
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 9087
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4613,
      "step": 9088
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.3912,
      "step": 9089
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 9090
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.2827,
      "step": 9091
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.653,
      "step": 9092
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.659,
      "step": 9093
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7094,
      "step": 9094
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.7528,
      "step": 9095
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.596,
      "step": 9096
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 9097
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.6419,
      "step": 9098
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 9099
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5998,
      "step": 9100
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 9101
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.668,
      "step": 9102
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5511,
      "step": 9103
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5891,
      "step": 9104
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5217,
      "step": 9105
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 9106
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7891,
      "step": 9107
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4603,
      "step": 9108
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7181,
      "step": 9109
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 9110
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7026,
      "step": 9111
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7159,
      "step": 9112
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6035,
      "step": 9113
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6961,
      "step": 9114
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4578,
      "step": 9115
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.8905,
      "step": 9116
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5734,
      "step": 9117
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6647,
      "step": 9118
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.8123,
      "step": 9119
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 9120
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 9121
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 9122
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 9123
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5303,
      "step": 9124
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6149,
      "step": 9125
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6808,
      "step": 9126
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6657,
      "step": 9127
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7591,
      "step": 9128
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6968,
      "step": 9129
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5578,
      "step": 9130
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5281,
      "step": 9131
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7051,
      "step": 9132
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4465,
      "step": 9133
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7668,
      "step": 9134
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5804,
      "step": 9135
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7016,
      "step": 9136
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.8675,
      "step": 9137
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 1.0931,
      "step": 9138
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6459,
      "step": 9139
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 9140
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4344,
      "step": 9141
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 9142
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 9143
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 9144
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7725,
      "step": 9145
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7314,
      "step": 9146
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 9147
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4076,
      "step": 9148
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6012,
      "step": 9149
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6106,
      "step": 9150
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.7646,
      "step": 9151
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5059,
      "step": 9152
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6622,
      "step": 9153
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.924,
      "step": 9154
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.5082,
      "step": 9155
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.9642,
      "step": 9156
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 9157
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6111,
      "step": 9158
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.6959,
      "step": 9159
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 9160
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.815,
      "step": 9161
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7639,
      "step": 9162
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6995,
      "step": 9163
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6997,
      "step": 9164
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7221,
      "step": 9165
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4493,
      "step": 9166
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7323,
      "step": 9167
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.564,
      "step": 9168
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6781,
      "step": 9169
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4362,
      "step": 9170
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5535,
      "step": 9171
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5305,
      "step": 9172
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4902,
      "step": 9173
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7437,
      "step": 9174
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6812,
      "step": 9175
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6161,
      "step": 9176
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6353,
      "step": 9177
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6117,
      "step": 9178
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7003,
      "step": 9179
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 9180
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5983,
      "step": 9181
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.9099,
      "step": 9182
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7157,
      "step": 9183
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6383,
      "step": 9184
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6971,
      "step": 9185
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4365,
      "step": 9186
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 9187
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7789,
      "step": 9188
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7541,
      "step": 9189
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.757,
      "step": 9190
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4723,
      "step": 9191
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7724,
      "step": 9192
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6289,
      "step": 9193
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6654,
      "step": 9194
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5634,
      "step": 9195
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 9196
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7681,
      "step": 9197
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4948,
      "step": 9198
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6755,
      "step": 9199
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 9200
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.7799079418182373,
      "eval_runtime": 131.5896,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.8,
      "step": 9200
    },
    {
      "epoch": 1.51,
      "mmlu_eval_accuracy": 0.5366646263280446,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "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": 0.18181818181818182,
      "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": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.6857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.37037037037037035,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 0.9134998138673001,
      "step": 9200
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 9201
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4792,
      "step": 9202
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7251,
      "step": 9203
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7233,
      "step": 9204
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5222,
      "step": 9205
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6772,
      "step": 9206
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 9207
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5864,
      "step": 9208
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.486,
      "step": 9209
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 9210
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7682,
      "step": 9211
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5871,
      "step": 9212
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 9213
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.4919,
      "step": 9214
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 9215
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5458,
      "step": 9216
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 9217
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.6406,
      "step": 9218
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.5901,
      "step": 9219
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.8,
      "step": 9220
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 9221
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 9222
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6529,
      "step": 9223
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 9224
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5902,
      "step": 9225
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7183,
      "step": 9226
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 9227
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6432,
      "step": 9228
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 9229
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6105,
      "step": 9230
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6536,
      "step": 9231
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7544,
      "step": 9232
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.721,
      "step": 9233
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6344,
      "step": 9234
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6432,
      "step": 9235
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5771,
      "step": 9236
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 9237
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 9238
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5837,
      "step": 9239
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6721,
      "step": 9240
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7949,
      "step": 9241
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7857,
      "step": 9242
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.4564,
      "step": 9243
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.4753,
      "step": 9244
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 9245
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5569,
      "step": 9246
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.4732,
      "step": 9247
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 9248
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 9249
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7128,
      "step": 9250
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5273,
      "step": 9251
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.586,
      "step": 9252
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7689,
      "step": 9253
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 9254
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 9255
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.4986,
      "step": 9256
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7459,
      "step": 9257
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6511,
      "step": 9258
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 9259
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.751,
      "step": 9260
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7485,
      "step": 9261
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.8109,
      "step": 9262
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5521,
      "step": 9263
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.9514,
      "step": 9264
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 9265
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.7559,
      "step": 9266
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 9267
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.582,
      "step": 9268
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.803,
      "step": 9269
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5327,
      "step": 9270
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.3571,
      "step": 9271
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 9272
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.3739,
      "step": 9273
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5178,
      "step": 9274
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6482,
      "step": 9275
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.688,
      "step": 9276
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.735,
      "step": 9277
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.6823,
      "step": 9278
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.5985,
      "step": 9279
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.757,
      "step": 9280
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 9281
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.3576,
      "step": 9282
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.4984,
      "step": 9283
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5614,
      "step": 9284
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7571,
      "step": 9285
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 9286
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.689,
      "step": 9287
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8147,
      "step": 9288
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 9289
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5536,
      "step": 9290
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 9291
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5358,
      "step": 9292
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.857,
      "step": 9293
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6131,
      "step": 9294
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7022,
      "step": 9295
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5959,
      "step": 9296
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6223,
      "step": 9297
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5336,
      "step": 9298
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.75,
      "step": 9299
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5596,
      "step": 9300
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8431,
      "step": 9301
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7964,
      "step": 9302
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7081,
      "step": 9303
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 9304
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.9158,
      "step": 9305
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7377,
      "step": 9306
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.537,
      "step": 9307
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.3484,
      "step": 9308
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 9309
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8032,
      "step": 9310
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6428,
      "step": 9311
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.4926,
      "step": 9312
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 9313
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.4262,
      "step": 9314
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5074,
      "step": 9315
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5604,
      "step": 9316
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5265,
      "step": 9317
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8615,
      "step": 9318
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8155,
      "step": 9319
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 9320
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8516,
      "step": 9321
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6969,
      "step": 9322
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.4956,
      "step": 9323
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6925,
      "step": 9324
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8677,
      "step": 9325
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 9326
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7305,
      "step": 9327
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.461,
      "step": 9328
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 9329
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6173,
      "step": 9330
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5761,
      "step": 9331
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7849,
      "step": 9332
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 9333
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.9108,
      "step": 9334
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 9335
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5636,
      "step": 9336
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5381,
      "step": 9337
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.8155,
      "step": 9338
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.84,
      "step": 9339
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.6053,
      "step": 9340
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.3441,
      "step": 9341
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.5496,
      "step": 9342
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4057,
      "step": 9343
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6649,
      "step": 9344
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4736,
      "step": 9345
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7521,
      "step": 9346
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6747,
      "step": 9347
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8179,
      "step": 9348
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6433,
      "step": 9349
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 9350
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.609,
      "step": 9351
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7821,
      "step": 9352
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 9353
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.5648,
      "step": 9354
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6361,
      "step": 9355
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6801,
      "step": 9356
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7874,
      "step": 9357
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.5896,
      "step": 9358
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 9359
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.9493,
      "step": 9360
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4538,
      "step": 9361
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6919,
      "step": 9362
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8489,
      "step": 9363
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.828,
      "step": 9364
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.3555,
      "step": 9365
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 9366
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.802,
      "step": 9367
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6554,
      "step": 9368
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8119,
      "step": 9369
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.5328,
      "step": 9370
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6357,
      "step": 9371
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6415,
      "step": 9372
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 9373
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6686,
      "step": 9374
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6915,
      "step": 9375
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 9376
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8289,
      "step": 9377
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6037,
      "step": 9378
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.702,
      "step": 9379
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 9380
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7857,
      "step": 9381
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4117,
      "step": 9382
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7727,
      "step": 9383
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.74,
      "step": 9384
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6265,
      "step": 9385
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6971,
      "step": 9386
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.644,
      "step": 9387
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4902,
      "step": 9388
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.5459,
      "step": 9389
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6688,
      "step": 9390
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.893,
      "step": 9391
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 9392
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4249,
      "step": 9393
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.568,
      "step": 9394
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.511,
      "step": 9395
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.758,
      "step": 9396
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6222,
      "step": 9397
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 9398
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.8465,
      "step": 9399
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7653,
      "step": 9400
    },
    {
      "epoch": 1.54,
      "eval_loss": 0.7874561548233032,
      "eval_runtime": 131.6476,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 9400
    },
    {
      "epoch": 1.54,
      "mmlu_eval_accuracy": 0.471729119615182,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "mmlu_eval_accuracy_college_biology": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.4375,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.5,
      "mmlu_eval_accuracy_high_school_world_history": 0.38461538461538464,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.25,
      "mmlu_eval_accuracy_international_law": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.4,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.2823529411764706,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.3333333333333333,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.5454545454545454,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.4501843340014042,
      "step": 9400
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.7564,
      "step": 9401
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.5336,
      "step": 9402
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.6197,
      "step": 9403
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4435,
      "step": 9404
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4145,
      "step": 9405
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6322,
      "step": 9406
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7851,
      "step": 9407
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 9408
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7627,
      "step": 9409
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5003,
      "step": 9410
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 9411
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7155,
      "step": 9412
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7265,
      "step": 9413
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6977,
      "step": 9414
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7077,
      "step": 9415
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5567,
      "step": 9416
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.385,
      "step": 9417
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 9418
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 9419
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4716,
      "step": 9420
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5859,
      "step": 9421
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.936,
      "step": 9422
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8915,
      "step": 9423
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4924,
      "step": 9424
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7101,
      "step": 9425
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6587,
      "step": 9426
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5424,
      "step": 9427
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7375,
      "step": 9428
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4912,
      "step": 9429
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 9430
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.89,
      "step": 9431
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8051,
      "step": 9432
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7354,
      "step": 9433
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 9434
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8386,
      "step": 9435
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.9098,
      "step": 9436
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5735,
      "step": 9437
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5015,
      "step": 9438
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7168,
      "step": 9439
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8918,
      "step": 9440
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6676,
      "step": 9441
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7462,
      "step": 9442
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.803,
      "step": 9443
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.366,
      "step": 9444
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7069,
      "step": 9445
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7778,
      "step": 9446
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7467,
      "step": 9447
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4054,
      "step": 9448
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.318,
      "step": 9449
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7185,
      "step": 9450
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 9451
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.95,
      "step": 9452
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 9453
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 9454
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6262,
      "step": 9455
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5586,
      "step": 9456
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7113,
      "step": 9457
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 9458
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6763,
      "step": 9459
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 9460
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4793,
      "step": 9461
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.6958,
      "step": 9462
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.5826,
      "step": 9463
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.8031,
      "step": 9464
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7964,
      "step": 9465
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6919,
      "step": 9466
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.4332,
      "step": 9467
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 9468
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.3221,
      "step": 9469
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5134,
      "step": 9470
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7797,
      "step": 9471
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7888,
      "step": 9472
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.4884,
      "step": 9473
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5939,
      "step": 9474
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6738,
      "step": 9475
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6493,
      "step": 9476
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8035,
      "step": 9477
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8682,
      "step": 9478
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.4616,
      "step": 9479
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 9480
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6853,
      "step": 9481
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7045,
      "step": 9482
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5658,
      "step": 9483
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.568,
      "step": 9484
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5916,
      "step": 9485
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6423,
      "step": 9486
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8463,
      "step": 9487
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 9488
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8415,
      "step": 9489
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 9490
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5834,
      "step": 9491
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.3253,
      "step": 9492
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.3628,
      "step": 9493
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8466,
      "step": 9494
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6337,
      "step": 9495
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.4905,
      "step": 9496
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.416,
      "step": 9497
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6323,
      "step": 9498
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6837,
      "step": 9499
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6667,
      "step": 9500
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8754,
      "step": 9501
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.764,
      "step": 9502
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8596,
      "step": 9503
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.625,
      "step": 9504
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6819,
      "step": 9505
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 9506
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 9507
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6873,
      "step": 9508
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.3926,
      "step": 9509
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6152,
      "step": 9510
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7116,
      "step": 9511
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7959,
      "step": 9512
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 9513
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.967,
      "step": 9514
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6581,
      "step": 9515
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.392,
      "step": 9516
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.8841,
      "step": 9517
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.5753,
      "step": 9518
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6551,
      "step": 9519
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7343,
      "step": 9520
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 9521
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.42,
      "step": 9522
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.543,
      "step": 9523
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.6178,
      "step": 9524
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6455,
      "step": 9525
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.806,
      "step": 9526
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7354,
      "step": 9527
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6513,
      "step": 9528
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6399,
      "step": 9529
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.9461,
      "step": 9530
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6867,
      "step": 9531
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6831,
      "step": 9532
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 9533
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.3802,
      "step": 9534
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7589,
      "step": 9535
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6863,
      "step": 9536
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 9537
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.8545,
      "step": 9538
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6301,
      "step": 9539
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4659,
      "step": 9540
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6222,
      "step": 9541
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 9542
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.581,
      "step": 9543
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6003,
      "step": 9544
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4362,
      "step": 9545
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6173,
      "step": 9546
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.8276,
      "step": 9547
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 9548
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5346,
      "step": 9549
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 9550
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7949,
      "step": 9551
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6838,
      "step": 9552
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5509,
      "step": 9553
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 9554
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 9555
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4662,
      "step": 9556
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4842,
      "step": 9557
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7107,
      "step": 9558
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5747,
      "step": 9559
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 9560
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5673,
      "step": 9561
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.3675,
      "step": 9562
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.9228,
      "step": 9563
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.582,
      "step": 9564
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 9565
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7529,
      "step": 9566
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 9567
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.8506,
      "step": 9568
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 9569
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 9570
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 9571
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6414,
      "step": 9572
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 9573
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5952,
      "step": 9574
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 9575
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5673,
      "step": 9576
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 9577
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 9578
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 9579
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 9580
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 9581
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.7604,
      "step": 9582
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5033,
      "step": 9583
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.5327,
      "step": 9584
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.4414,
      "step": 9585
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6995,
      "step": 9586
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9684,
      "step": 9587
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4489,
      "step": 9588
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 9589
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4459,
      "step": 9590
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.453,
      "step": 9591
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.3654,
      "step": 9592
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9074,
      "step": 9593
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6175,
      "step": 9594
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6813,
      "step": 9595
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.551,
      "step": 9596
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6248,
      "step": 9597
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6348,
      "step": 9598
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7582,
      "step": 9599
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6299,
      "step": 9600
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.784636914730072,
      "eval_runtime": 131.6864,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 9600
    },
    {
      "epoch": 1.58,
      "mmlu_eval_accuracy": 0.5165193836690629,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.25,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.35,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.2654462955580834,
      "step": 9600
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.916,
      "step": 9601
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4078,
      "step": 9602
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5923,
      "step": 9603
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4849,
      "step": 9604
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5334,
      "step": 9605
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 9606
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 9607
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7929,
      "step": 9608
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6933,
      "step": 9609
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4232,
      "step": 9610
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9209,
      "step": 9611
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 1.0139,
      "step": 9612
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6025,
      "step": 9613
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7258,
      "step": 9614
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7853,
      "step": 9615
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 9616
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5445,
      "step": 9617
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6049,
      "step": 9618
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6665,
      "step": 9619
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9634,
      "step": 9620
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.8142,
      "step": 9621
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.8228,
      "step": 9622
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7119,
      "step": 9623
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 9624
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.699,
      "step": 9625
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 9626
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4668,
      "step": 9627
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4836,
      "step": 9628
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4452,
      "step": 9629
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9577,
      "step": 9630
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 9631
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4121,
      "step": 9632
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6148,
      "step": 9633
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6246,
      "step": 9634
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.9768,
      "step": 9635
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6092,
      "step": 9636
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.793,
      "step": 9637
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5995,
      "step": 9638
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 9639
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 9640
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6846,
      "step": 9641
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6592,
      "step": 9642
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.4094,
      "step": 9643
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 1.0876,
      "step": 9644
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.5766,
      "step": 9645
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 9646
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 9647
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5463,
      "step": 9648
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 9649
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6238,
      "step": 9650
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7031,
      "step": 9651
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.418,
      "step": 9652
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 9653
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7869,
      "step": 9654
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.693,
      "step": 9655
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.9573,
      "step": 9656
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 9657
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4036,
      "step": 9658
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6122,
      "step": 9659
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7775,
      "step": 9660
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 9661
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7438,
      "step": 9662
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.9823,
      "step": 9663
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 9664
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5308,
      "step": 9665
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.8361,
      "step": 9666
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7142,
      "step": 9667
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4007,
      "step": 9668
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 9669
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7572,
      "step": 9670
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6327,
      "step": 9671
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.9899,
      "step": 9672
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4776,
      "step": 9673
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.8288,
      "step": 9674
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 9675
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.513,
      "step": 9676
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6396,
      "step": 9677
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6839,
      "step": 9678
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.616,
      "step": 9679
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5574,
      "step": 9680
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 9681
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7565,
      "step": 9682
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5613,
      "step": 9683
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5208,
      "step": 9684
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 9685
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5752,
      "step": 9686
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.4963,
      "step": 9687
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5006,
      "step": 9688
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6987,
      "step": 9689
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.8414,
      "step": 9690
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 9691
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6684,
      "step": 9692
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5433,
      "step": 9693
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6167,
      "step": 9694
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 9695
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.678,
      "step": 9696
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5999,
      "step": 9697
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.9094,
      "step": 9698
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 9699
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.2289,
      "step": 9700
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6444,
      "step": 9701
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 9702
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.713,
      "step": 9703
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5076,
      "step": 9704
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.5477,
      "step": 9705
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 9706
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.776,
      "step": 9707
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8794,
      "step": 9708
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 9709
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8293,
      "step": 9710
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.881,
      "step": 9711
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 9712
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6759,
      "step": 9713
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7204,
      "step": 9714
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4531,
      "step": 9715
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.676,
      "step": 9716
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4623,
      "step": 9717
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8973,
      "step": 9718
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6265,
      "step": 9719
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8243,
      "step": 9720
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 9721
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4813,
      "step": 9722
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.3817,
      "step": 9723
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 9724
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6726,
      "step": 9725
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4363,
      "step": 9726
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4872,
      "step": 9727
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4244,
      "step": 9728
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5093,
      "step": 9729
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6958,
      "step": 9730
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8814,
      "step": 9731
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5837,
      "step": 9732
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.9493,
      "step": 9733
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8299,
      "step": 9734
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6047,
      "step": 9735
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6821,
      "step": 9736
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5889,
      "step": 9737
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5646,
      "step": 9738
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6737,
      "step": 9739
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.823,
      "step": 9740
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5989,
      "step": 9741
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.3253,
      "step": 9742
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5626,
      "step": 9743
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 9744
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8019,
      "step": 9745
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 9746
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8232,
      "step": 9747
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7196,
      "step": 9748
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.4936,
      "step": 9749
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7623,
      "step": 9750
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 9751
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5017,
      "step": 9752
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6105,
      "step": 9753
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 9754
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5341,
      "step": 9755
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 1.2394,
      "step": 9756
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 9757
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 9758
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7768,
      "step": 9759
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 9760
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 9761
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7684,
      "step": 9762
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 9763
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.8274,
      "step": 9764
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.7847,
      "step": 9765
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6501,
      "step": 9766
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 9767
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.5473,
      "step": 9768
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7223,
      "step": 9769
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6423,
      "step": 9770
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.762,
      "step": 9771
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 9772
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.4643,
      "step": 9773
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7474,
      "step": 9774
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.523,
      "step": 9775
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.4865,
      "step": 9776
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5389,
      "step": 9777
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 9778
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6423,
      "step": 9779
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5834,
      "step": 9780
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8516,
      "step": 9781
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.4164,
      "step": 9782
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8771,
      "step": 9783
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5731,
      "step": 9784
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 9785
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.626,
      "step": 9786
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8243,
      "step": 9787
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8314,
      "step": 9788
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5483,
      "step": 9789
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5731,
      "step": 9790
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6205,
      "step": 9791
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.745,
      "step": 9792
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6503,
      "step": 9793
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.3768,
      "step": 9794
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7526,
      "step": 9795
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.9857,
      "step": 9796
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7091,
      "step": 9797
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5382,
      "step": 9798
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 9799
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.779,
      "step": 9800
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.7798586487770081,
      "eval_runtime": 131.6542,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 9800
    },
    {
      "epoch": 1.61,
      "mmlu_eval_accuracy": 0.5344532723671317,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
      "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": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "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": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.3590053339773307,
      "step": 9800
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5307,
      "step": 9801
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 9802
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6487,
      "step": 9803
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5318,
      "step": 9804
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5541,
      "step": 9805
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 9806
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 9807
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5603,
      "step": 9808
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.3764,
      "step": 9809
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5604,
      "step": 9810
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.4746,
      "step": 9811
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 9812
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7279,
      "step": 9813
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8571,
      "step": 9814
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6728,
      "step": 9815
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6635,
      "step": 9816
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5376,
      "step": 9817
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6599,
      "step": 9818
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 9819
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6091,
      "step": 9820
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7318,
      "step": 9821
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 9822
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6178,
      "step": 9823
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5067,
      "step": 9824
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5131,
      "step": 9825
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 9826
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.4638,
      "step": 9827
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.5881,
      "step": 9828
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.7042,
      "step": 9829
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6785,
      "step": 9830
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 9831
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.3732,
      "step": 9832
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.4508,
      "step": 9833
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7173,
      "step": 9834
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7529,
      "step": 9835
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6874,
      "step": 9836
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6563,
      "step": 9837
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.456,
      "step": 9838
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6704,
      "step": 9839
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7212,
      "step": 9840
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 9841
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.9259,
      "step": 9842
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5777,
      "step": 9843
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.627,
      "step": 9844
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5821,
      "step": 9845
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 9846
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.3895,
      "step": 9847
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7,
      "step": 9848
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6388,
      "step": 9849
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.4999,
      "step": 9850
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.9325,
      "step": 9851
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.8145,
      "step": 9852
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7792,
      "step": 9853
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.556,
      "step": 9854
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.8431,
      "step": 9855
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7082,
      "step": 9856
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 9857
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5396,
      "step": 9858
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.4977,
      "step": 9859
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6441,
      "step": 9860
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.9693,
      "step": 9861
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5428,
      "step": 9862
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.9178,
      "step": 9863
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6672,
      "step": 9864
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6081,
      "step": 9865
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5708,
      "step": 9866
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5457,
      "step": 9867
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.4602,
      "step": 9868
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.3094,
      "step": 9869
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5467,
      "step": 9870
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6186,
      "step": 9871
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5984,
      "step": 9872
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6481,
      "step": 9873
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 9874
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5453,
      "step": 9875
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7919,
      "step": 9876
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 9877
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 9878
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5864,
      "step": 9879
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7112,
      "step": 9880
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.8284,
      "step": 9881
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6238,
      "step": 9882
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5641,
      "step": 9883
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5576,
      "step": 9884
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7507,
      "step": 9885
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.4284,
      "step": 9886
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 9887
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.6809,
      "step": 9888
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.5977,
      "step": 9889
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.7577,
      "step": 9890
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6418,
      "step": 9891
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6887,
      "step": 9892
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 9893
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7362,
      "step": 9894
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5591,
      "step": 9895
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4835,
      "step": 9896
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6033,
      "step": 9897
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.426,
      "step": 9898
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 9899
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7542,
      "step": 9900
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5635,
      "step": 9901
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7579,
      "step": 9902
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5273,
      "step": 9903
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 9904
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.9078,
      "step": 9905
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5874,
      "step": 9906
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 9907
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7719,
      "step": 9908
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5204,
      "step": 9909
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 9910
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 9911
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7289,
      "step": 9912
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.9267,
      "step": 9913
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4356,
      "step": 9914
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 9915
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5119,
      "step": 9916
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7668,
      "step": 9917
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7098,
      "step": 9918
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7846,
      "step": 9919
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7018,
      "step": 9920
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4827,
      "step": 9921
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 9922
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6091,
      "step": 9923
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5301,
      "step": 9924
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7751,
      "step": 9925
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 9926
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 9927
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 9928
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 9929
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5732,
      "step": 9930
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4502,
      "step": 9931
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 9932
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 9933
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6102,
      "step": 9934
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6175,
      "step": 9935
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6713,
      "step": 9936
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.8449,
      "step": 9937
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.9427,
      "step": 9938
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7268,
      "step": 9939
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6784,
      "step": 9940
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7522,
      "step": 9941
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5949,
      "step": 9942
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.5902,
      "step": 9943
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6804,
      "step": 9944
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7342,
      "step": 9945
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7301,
      "step": 9946
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6505,
      "step": 9947
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.7381,
      "step": 9948
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.6216,
      "step": 9949
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4608,
      "step": 9950
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4403,
      "step": 9951
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 9952
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7592,
      "step": 9953
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.631,
      "step": 9954
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6402,
      "step": 9955
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5739,
      "step": 9956
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8046,
      "step": 9957
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.46,
      "step": 9958
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 9959
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7806,
      "step": 9960
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.399,
      "step": 9961
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8835,
      "step": 9962
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6844,
      "step": 9963
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8738,
      "step": 9964
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6954,
      "step": 9965
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.4188,
      "step": 9966
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5891,
      "step": 9967
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5018,
      "step": 9968
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.582,
      "step": 9969
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.648,
      "step": 9970
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8619,
      "step": 9971
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6645,
      "step": 9972
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7697,
      "step": 9973
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 9974
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.4773,
      "step": 9975
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 9976
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8912,
      "step": 9977
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8194,
      "step": 9978
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7764,
      "step": 9979
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 9980
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7146,
      "step": 9981
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6283,
      "step": 9982
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.4587,
      "step": 9983
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5494,
      "step": 9984
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5957,
      "step": 9985
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.731,
      "step": 9986
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8486,
      "step": 9987
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7637,
      "step": 9988
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5792,
      "step": 9989
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5476,
      "step": 9990
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 9991
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5489,
      "step": 9992
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 9993
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.9022,
      "step": 9994
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.3669,
      "step": 9995
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.8974,
      "step": 9996
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.4436,
      "step": 9997
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6261,
      "step": 9998
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6236,
      "step": 9999
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 10000
    },
    {
      "epoch": 1.64,
      "eval_loss": 0.7805026769638062,
      "eval_runtime": 131.6536,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 10000
    },
    {
      "epoch": 1.64,
      "mmlu_eval_accuracy": 0.5486681810000722,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "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": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.4375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "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": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "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": 0.7777777777777778,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.3,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.4465018855723948,
      "step": 10000
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6393,
      "step": 10001
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 10002
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6089,
      "step": 10003
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 10004
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6641,
      "step": 10005
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 10006
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.72,
      "step": 10007
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 1.0142,
      "step": 10008
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.9014,
      "step": 10009
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.6777,
      "step": 10010
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 10011
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5597,
      "step": 10012
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5272,
      "step": 10013
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6907,
      "step": 10014
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5401,
      "step": 10015
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7055,
      "step": 10016
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 10017
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5486,
      "step": 10018
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.777,
      "step": 10019
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 10020
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5412,
      "step": 10021
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.477,
      "step": 10022
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8369,
      "step": 10023
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8638,
      "step": 10024
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8443,
      "step": 10025
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 10026
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 10027
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 10028
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.4833,
      "step": 10029
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7633,
      "step": 10030
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5292,
      "step": 10031
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.58,
      "step": 10032
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 10033
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8294,
      "step": 10034
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6767,
      "step": 10035
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6921,
      "step": 10036
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5913,
      "step": 10037
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 10038
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.4057,
      "step": 10039
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5841,
      "step": 10040
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6875,
      "step": 10041
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5535,
      "step": 10042
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.3062,
      "step": 10043
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6962,
      "step": 10044
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5214,
      "step": 10045
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.4953,
      "step": 10046
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7164,
      "step": 10047
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 1.2385,
      "step": 10048
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8761,
      "step": 10049
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.745,
      "step": 10050
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8462,
      "step": 10051
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 1.0605,
      "step": 10052
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6885,
      "step": 10053
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7892,
      "step": 10054
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8748,
      "step": 10055
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6446,
      "step": 10056
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.632,
      "step": 10057
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8692,
      "step": 10058
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8376,
      "step": 10059
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 10060
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5634,
      "step": 10061
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.5751,
      "step": 10062
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 10063
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.962,
      "step": 10064
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7837,
      "step": 10065
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.9596,
      "step": 10066
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.4653,
      "step": 10067
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6523,
      "step": 10068
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.7657,
      "step": 10069
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6233,
      "step": 10070
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.6435,
      "step": 10071
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.8099,
      "step": 10072
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7098,
      "step": 10073
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7479,
      "step": 10074
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6894,
      "step": 10075
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8728,
      "step": 10076
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.708,
      "step": 10077
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7298,
      "step": 10078
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5702,
      "step": 10079
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5206,
      "step": 10080
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9401,
      "step": 10081
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 1.0013,
      "step": 10082
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9144,
      "step": 10083
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.618,
      "step": 10084
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5404,
      "step": 10085
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4844,
      "step": 10086
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9014,
      "step": 10087
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.3529,
      "step": 10088
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7028,
      "step": 10089
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 10090
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9614,
      "step": 10091
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.875,
      "step": 10092
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5494,
      "step": 10093
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6244,
      "step": 10094
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5441,
      "step": 10095
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4229,
      "step": 10096
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.695,
      "step": 10097
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 10098
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8567,
      "step": 10099
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4009,
      "step": 10100
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 10101
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7369,
      "step": 10102
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.2033,
      "step": 10103
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.3861,
      "step": 10104
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 10105
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7343,
      "step": 10106
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8324,
      "step": 10107
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5764,
      "step": 10108
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.818,
      "step": 10109
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6552,
      "step": 10110
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8032,
      "step": 10111
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5545,
      "step": 10112
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4795,
      "step": 10113
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6337,
      "step": 10114
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8248,
      "step": 10115
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7186,
      "step": 10116
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7269,
      "step": 10117
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6609,
      "step": 10118
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6064,
      "step": 10119
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5995,
      "step": 10120
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5133,
      "step": 10121
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 10122
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.5805,
      "step": 10123
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9869,
      "step": 10124
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.4839,
      "step": 10125
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6229,
      "step": 10126
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8452,
      "step": 10127
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6262,
      "step": 10128
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.9029,
      "step": 10129
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 10130
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 10131
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.6108,
      "step": 10132
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 10133
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.51,
      "step": 10134
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 10135
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 10136
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7791,
      "step": 10137
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 10138
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7877,
      "step": 10139
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6058,
      "step": 10140
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5769,
      "step": 10141
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6378,
      "step": 10142
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8737,
      "step": 10143
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6636,
      "step": 10144
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5913,
      "step": 10145
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6926,
      "step": 10146
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7083,
      "step": 10147
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7223,
      "step": 10148
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 10149
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6427,
      "step": 10150
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6088,
      "step": 10151
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8302,
      "step": 10152
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5373,
      "step": 10153
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6966,
      "step": 10154
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 10155
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 10156
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7139,
      "step": 10157
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7581,
      "step": 10158
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.9886,
      "step": 10159
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.872,
      "step": 10160
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 10161
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 10162
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.4043,
      "step": 10163
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5538,
      "step": 10164
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7149,
      "step": 10165
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7317,
      "step": 10166
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7928,
      "step": 10167
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8608,
      "step": 10168
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5805,
      "step": 10169
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6201,
      "step": 10170
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6543,
      "step": 10171
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8555,
      "step": 10172
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6499,
      "step": 10173
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7012,
      "step": 10174
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.591,
      "step": 10175
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 10176
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6859,
      "step": 10177
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8682,
      "step": 10178
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6563,
      "step": 10179
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.4259,
      "step": 10180
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 10181
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8663,
      "step": 10182
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7576,
      "step": 10183
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5406,
      "step": 10184
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5156,
      "step": 10185
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 10186
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.6004,
      "step": 10187
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.8584,
      "step": 10188
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7195,
      "step": 10189
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.7503,
      "step": 10190
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 10191
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.4551,
      "step": 10192
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5662,
      "step": 10193
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.5379,
      "step": 10194
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 10195
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 10196
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 10197
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 10198
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5958,
      "step": 10199
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5318,
      "step": 10200
    },
    {
      "epoch": 1.68,
      "eval_loss": 0.7796837091445923,
      "eval_runtime": 131.6501,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 10200
    },
    {
      "epoch": 1.68,
      "mmlu_eval_accuracy": 0.5243962335716819,
      "mmlu_eval_accuracy_abstract_algebra": 0.5454545454545454,
      "mmlu_eval_accuracy_anatomy": 0.2857142857142857,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5813953488372093,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.295656981379494,
      "step": 10200
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 10201
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7003,
      "step": 10202
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.462,
      "step": 10203
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7114,
      "step": 10204
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.8556,
      "step": 10205
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5916,
      "step": 10206
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.535,
      "step": 10207
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6567,
      "step": 10208
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4327,
      "step": 10209
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6469,
      "step": 10210
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6518,
      "step": 10211
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4492,
      "step": 10212
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 10213
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5269,
      "step": 10214
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7881,
      "step": 10215
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7923,
      "step": 10216
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.9087,
      "step": 10217
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.487,
      "step": 10218
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5168,
      "step": 10219
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 10220
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7273,
      "step": 10221
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 10222
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5857,
      "step": 10223
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 10224
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4528,
      "step": 10225
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7108,
      "step": 10226
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5062,
      "step": 10227
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.709,
      "step": 10228
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4476,
      "step": 10229
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.774,
      "step": 10230
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 10231
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.684,
      "step": 10232
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4875,
      "step": 10233
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7067,
      "step": 10234
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7247,
      "step": 10235
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4375,
      "step": 10236
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6055,
      "step": 10237
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.8128,
      "step": 10238
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.8143,
      "step": 10239
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5821,
      "step": 10240
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4449,
      "step": 10241
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.793,
      "step": 10242
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 10243
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5331,
      "step": 10244
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6899,
      "step": 10245
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.86,
      "step": 10246
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 10247
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 10248
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.9349,
      "step": 10249
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4483,
      "step": 10250
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.4499,
      "step": 10251
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.649,
      "step": 10252
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.503,
      "step": 10253
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.7917,
      "step": 10254
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.728,
      "step": 10255
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6883,
      "step": 10256
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5549,
      "step": 10257
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 10258
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7947,
      "step": 10259
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.3858,
      "step": 10260
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 1.0524,
      "step": 10261
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.3736,
      "step": 10262
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4941,
      "step": 10263
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 10264
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.56,
      "step": 10265
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6874,
      "step": 10266
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.63,
      "step": 10267
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.437,
      "step": 10268
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6832,
      "step": 10269
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6009,
      "step": 10270
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7506,
      "step": 10271
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 10272
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 10273
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5292,
      "step": 10274
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.8664,
      "step": 10275
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6548,
      "step": 10276
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7379,
      "step": 10277
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4237,
      "step": 10278
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 10279
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4959,
      "step": 10280
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.8599,
      "step": 10281
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6167,
      "step": 10282
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 10283
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4436,
      "step": 10284
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6112,
      "step": 10285
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5945,
      "step": 10286
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.737,
      "step": 10287
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.8429,
      "step": 10288
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6989,
      "step": 10289
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 10290
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5935,
      "step": 10291
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7877,
      "step": 10292
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 10293
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5407,
      "step": 10294
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6321,
      "step": 10295
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7333,
      "step": 10296
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4769,
      "step": 10297
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.648,
      "step": 10298
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5208,
      "step": 10299
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 10300
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 10301
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7118,
      "step": 10302
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6136,
      "step": 10303
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6978,
      "step": 10304
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6836,
      "step": 10305
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7607,
      "step": 10306
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6694,
      "step": 10307
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6795,
      "step": 10308
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.707,
      "step": 10309
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6304,
      "step": 10310
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5664,
      "step": 10311
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.509,
      "step": 10312
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5086,
      "step": 10313
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.6234,
      "step": 10314
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.7685,
      "step": 10315
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 10316
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6486,
      "step": 10317
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 10318
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 10319
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6632,
      "step": 10320
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7882,
      "step": 10321
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7029,
      "step": 10322
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 1.0656,
      "step": 10323
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6036,
      "step": 10324
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4862,
      "step": 10325
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 10326
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6354,
      "step": 10327
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 10328
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6843,
      "step": 10329
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5734,
      "step": 10330
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.8574,
      "step": 10331
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 10332
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7263,
      "step": 10333
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5492,
      "step": 10334
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.791,
      "step": 10335
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4905,
      "step": 10336
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5896,
      "step": 10337
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.3658,
      "step": 10338
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5752,
      "step": 10339
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.2545,
      "step": 10340
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4819,
      "step": 10341
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5507,
      "step": 10342
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5794,
      "step": 10343
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.9801,
      "step": 10344
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5591,
      "step": 10345
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6351,
      "step": 10346
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.8624,
      "step": 10347
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5672,
      "step": 10348
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6105,
      "step": 10349
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5688,
      "step": 10350
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7357,
      "step": 10351
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6266,
      "step": 10352
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4916,
      "step": 10353
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4322,
      "step": 10354
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7809,
      "step": 10355
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6057,
      "step": 10356
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7551,
      "step": 10357
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5208,
      "step": 10358
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5143,
      "step": 10359
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7924,
      "step": 10360
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.9551,
      "step": 10361
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 10362
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7864,
      "step": 10363
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.3168,
      "step": 10364
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5862,
      "step": 10365
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.3804,
      "step": 10366
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 10367
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6773,
      "step": 10368
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6636,
      "step": 10369
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7731,
      "step": 10370
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6915,
      "step": 10371
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.4195,
      "step": 10372
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7942,
      "step": 10373
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.6589,
      "step": 10374
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.7227,
      "step": 10375
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.5915,
      "step": 10376
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.8306,
      "step": 10377
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6353,
      "step": 10378
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7216,
      "step": 10379
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.672,
      "step": 10380
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6956,
      "step": 10381
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6414,
      "step": 10382
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.3742,
      "step": 10383
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.3886,
      "step": 10384
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 10385
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6828,
      "step": 10386
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7818,
      "step": 10387
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5036,
      "step": 10388
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4401,
      "step": 10389
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 10390
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5294,
      "step": 10391
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6837,
      "step": 10392
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6377,
      "step": 10393
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6398,
      "step": 10394
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4543,
      "step": 10395
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.3713,
      "step": 10396
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 10397
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 10398
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6578,
      "step": 10399
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7539,
      "step": 10400
    },
    {
      "epoch": 1.71,
      "eval_loss": 0.7826257348060608,
      "eval_runtime": 131.5823,
      "eval_samples_per_second": 7.6,
      "eval_steps_per_second": 3.8,
      "step": 10400
    },
    {
      "epoch": 1.71,
      "mmlu_eval_accuracy": 0.5420531223316813,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "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": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.4,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.2590058280967233,
      "step": 10400
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5453,
      "step": 10401
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7133,
      "step": 10402
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5636,
      "step": 10403
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7222,
      "step": 10404
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 10405
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6715,
      "step": 10406
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7859,
      "step": 10407
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 10408
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4709,
      "step": 10409
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5707,
      "step": 10410
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6569,
      "step": 10411
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5938,
      "step": 10412
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4759,
      "step": 10413
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6077,
      "step": 10414
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 10415
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4505,
      "step": 10416
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4773,
      "step": 10417
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7309,
      "step": 10418
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 1.0682,
      "step": 10419
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5709,
      "step": 10420
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.512,
      "step": 10421
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.8079,
      "step": 10422
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7949,
      "step": 10423
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7337,
      "step": 10424
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.5422,
      "step": 10425
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 10426
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.521,
      "step": 10427
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.7473,
      "step": 10428
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.3298,
      "step": 10429
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6216,
      "step": 10430
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6811,
      "step": 10431
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6866,
      "step": 10432
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6215,
      "step": 10433
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.4144,
      "step": 10434
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.8645,
      "step": 10435
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6513,
      "step": 10436
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.6016,
      "step": 10437
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.8466,
      "step": 10438
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7929,
      "step": 10439
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 10440
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6283,
      "step": 10441
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5572,
      "step": 10442
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5011,
      "step": 10443
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.566,
      "step": 10444
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7762,
      "step": 10445
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5577,
      "step": 10446
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6677,
      "step": 10447
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6348,
      "step": 10448
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 10449
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7196,
      "step": 10450
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 10451
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7722,
      "step": 10452
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6209,
      "step": 10453
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6395,
      "step": 10454
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.3776,
      "step": 10455
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.811,
      "step": 10456
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6653,
      "step": 10457
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.8633,
      "step": 10458
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6793,
      "step": 10459
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.4359,
      "step": 10460
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7413,
      "step": 10461
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6218,
      "step": 10462
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6281,
      "step": 10463
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.487,
      "step": 10464
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5536,
      "step": 10465
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7514,
      "step": 10466
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7752,
      "step": 10467
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.702,
      "step": 10468
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 10469
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7373,
      "step": 10470
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7292,
      "step": 10471
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7727,
      "step": 10472
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7475,
      "step": 10473
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5112,
      "step": 10474
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 10475
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.4828,
      "step": 10476
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.4857,
      "step": 10477
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7937,
      "step": 10478
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.4576,
      "step": 10479
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.9528,
      "step": 10480
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 10481
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.3224,
      "step": 10482
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.8271,
      "step": 10483
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5441,
      "step": 10484
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5616,
      "step": 10485
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5893,
      "step": 10486
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.8912,
      "step": 10487
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.564,
      "step": 10488
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6082,
      "step": 10489
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5842,
      "step": 10490
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.7318,
      "step": 10491
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5584,
      "step": 10492
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6391,
      "step": 10493
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.5955,
      "step": 10494
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.4767,
      "step": 10495
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.6802,
      "step": 10496
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.865,
      "step": 10497
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.3176,
      "step": 10498
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 10499
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7894,
      "step": 10500
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7055,
      "step": 10501
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7662,
      "step": 10502
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 10503
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5484,
      "step": 10504
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 10505
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7075,
      "step": 10506
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 1.0035,
      "step": 10507
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4216,
      "step": 10508
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4958,
      "step": 10509
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4878,
      "step": 10510
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4873,
      "step": 10511
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 1.1429,
      "step": 10512
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 10513
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6364,
      "step": 10514
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 10515
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7127,
      "step": 10516
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.924,
      "step": 10517
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7009,
      "step": 10518
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 10519
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5882,
      "step": 10520
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 10521
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7169,
      "step": 10522
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7179,
      "step": 10523
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6532,
      "step": 10524
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5917,
      "step": 10525
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.8256,
      "step": 10526
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.858,
      "step": 10527
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.3863,
      "step": 10528
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.871,
      "step": 10529
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 10530
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5305,
      "step": 10531
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 10532
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6805,
      "step": 10533
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7477,
      "step": 10534
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 10535
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6255,
      "step": 10536
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6154,
      "step": 10537
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5781,
      "step": 10538
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7642,
      "step": 10539
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.606,
      "step": 10540
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.8992,
      "step": 10541
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 10542
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6958,
      "step": 10543
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5872,
      "step": 10544
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5977,
      "step": 10545
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.9414,
      "step": 10546
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4296,
      "step": 10547
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 10548
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.495,
      "step": 10549
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7618,
      "step": 10550
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7997,
      "step": 10551
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7579,
      "step": 10552
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7983,
      "step": 10553
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 10554
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4131,
      "step": 10555
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.9005,
      "step": 10556
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 10557
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 10558
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.4976,
      "step": 10559
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 10560
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7784,
      "step": 10561
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.749,
      "step": 10562
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4436,
      "step": 10563
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5559,
      "step": 10564
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7595,
      "step": 10565
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7374,
      "step": 10566
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 10567
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5657,
      "step": 10568
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7758,
      "step": 10569
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7975,
      "step": 10570
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6705,
      "step": 10571
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.855,
      "step": 10572
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.568,
      "step": 10573
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.9069,
      "step": 10574
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7214,
      "step": 10575
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.3828,
      "step": 10576
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.9773,
      "step": 10577
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.434,
      "step": 10578
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8116,
      "step": 10579
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.537,
      "step": 10580
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6257,
      "step": 10581
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5398,
      "step": 10582
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 10583
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4809,
      "step": 10584
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7244,
      "step": 10585
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 10586
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.9499,
      "step": 10587
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7321,
      "step": 10588
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.422,
      "step": 10589
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6865,
      "step": 10590
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 10591
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6177,
      "step": 10592
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4593,
      "step": 10593
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.9236,
      "step": 10594
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6134,
      "step": 10595
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5904,
      "step": 10596
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6951,
      "step": 10597
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4936,
      "step": 10598
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6445,
      "step": 10599
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7763,
      "step": 10600
    },
    {
      "epoch": 1.74,
      "eval_loss": 0.7838025093078613,
      "eval_runtime": 131.6799,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 10600
    },
    {
      "epoch": 1.74,
      "mmlu_eval_accuracy": 0.4922020456307365,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.09090909090909091,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.17391304347826086,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "mmlu_eval_accuracy_human_aging": 0.8260869565217391,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.4782608695652174,
      "mmlu_eval_accuracy_public_relations": 0.6666666666666666,
      "mmlu_eval_accuracy_security_studies": 0.37037037037037035,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.038772797888004,
      "step": 10600
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4854,
      "step": 10601
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4528,
      "step": 10602
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8414,
      "step": 10603
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8763,
      "step": 10604
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4929,
      "step": 10605
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8343,
      "step": 10606
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6169,
      "step": 10607
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5361,
      "step": 10608
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.3757,
      "step": 10609
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.658,
      "step": 10610
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.4388,
      "step": 10611
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8978,
      "step": 10612
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5866,
      "step": 10613
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7535,
      "step": 10614
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 10615
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7584,
      "step": 10616
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7749,
      "step": 10617
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.7729,
      "step": 10618
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.6056,
      "step": 10619
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 10620
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 10621
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5149,
      "step": 10622
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 10623
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6997,
      "step": 10624
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8772,
      "step": 10625
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6125,
      "step": 10626
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5902,
      "step": 10627
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7679,
      "step": 10628
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7298,
      "step": 10629
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 10630
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7768,
      "step": 10631
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 10632
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7808,
      "step": 10633
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6437,
      "step": 10634
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6695,
      "step": 10635
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8093,
      "step": 10636
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6447,
      "step": 10637
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6779,
      "step": 10638
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.4767,
      "step": 10639
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 10640
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8024,
      "step": 10641
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8269,
      "step": 10642
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 10643
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.586,
      "step": 10644
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6187,
      "step": 10645
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7779,
      "step": 10646
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.738,
      "step": 10647
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7196,
      "step": 10648
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.695,
      "step": 10649
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7994,
      "step": 10650
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.4223,
      "step": 10651
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7666,
      "step": 10652
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7304,
      "step": 10653
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8626,
      "step": 10654
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.846,
      "step": 10655
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8617,
      "step": 10656
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 10657
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 1.036,
      "step": 10658
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7985,
      "step": 10659
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6973,
      "step": 10660
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5363,
      "step": 10661
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.69,
      "step": 10662
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.3996,
      "step": 10663
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.523,
      "step": 10664
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6315,
      "step": 10665
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7861,
      "step": 10666
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5164,
      "step": 10667
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6881,
      "step": 10668
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.37,
      "step": 10669
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.9017,
      "step": 10670
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7472,
      "step": 10671
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7336,
      "step": 10672
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 10673
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5837,
      "step": 10674
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5488,
      "step": 10675
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.784,
      "step": 10676
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.7795,
      "step": 10677
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.6406,
      "step": 10678
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.4233,
      "step": 10679
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 10680
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.8213,
      "step": 10681
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6904,
      "step": 10682
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 10683
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5788,
      "step": 10684
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8166,
      "step": 10685
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6557,
      "step": 10686
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8,
      "step": 10687
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7144,
      "step": 10688
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.9013,
      "step": 10689
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 10690
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6255,
      "step": 10691
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8278,
      "step": 10692
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 10693
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6568,
      "step": 10694
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 10695
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.4634,
      "step": 10696
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 10697
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.4644,
      "step": 10698
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5897,
      "step": 10699
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.4648,
      "step": 10700
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.4952,
      "step": 10701
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 10702
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7828,
      "step": 10703
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 10704
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6141,
      "step": 10705
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 1.0112,
      "step": 10706
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8482,
      "step": 10707
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8219,
      "step": 10708
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 10709
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7629,
      "step": 10710
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8172,
      "step": 10711
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.9866,
      "step": 10712
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 10713
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 10714
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8841,
      "step": 10715
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8677,
      "step": 10716
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7428,
      "step": 10717
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5418,
      "step": 10718
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6313,
      "step": 10719
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.3892,
      "step": 10720
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6905,
      "step": 10721
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 10722
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 10723
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.571,
      "step": 10724
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7394,
      "step": 10725
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7853,
      "step": 10726
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 10727
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.9109,
      "step": 10728
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7092,
      "step": 10729
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 10730
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7976,
      "step": 10731
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7935,
      "step": 10732
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6694,
      "step": 10733
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.7224,
      "step": 10734
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.8836,
      "step": 10735
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.522,
      "step": 10736
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 10737
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6013,
      "step": 10738
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.4999,
      "step": 10739
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6491,
      "step": 10740
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.786,
      "step": 10741
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 10742
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.9238,
      "step": 10743
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8311,
      "step": 10744
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6584,
      "step": 10745
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4754,
      "step": 10746
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7571,
      "step": 10747
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 10748
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8956,
      "step": 10749
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 10750
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7725,
      "step": 10751
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6998,
      "step": 10752
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8983,
      "step": 10753
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.432,
      "step": 10754
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6981,
      "step": 10755
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5336,
      "step": 10756
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5894,
      "step": 10757
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.3248,
      "step": 10758
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4286,
      "step": 10759
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4571,
      "step": 10760
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 10761
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.737,
      "step": 10762
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6082,
      "step": 10763
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4216,
      "step": 10764
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7887,
      "step": 10765
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6788,
      "step": 10766
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7293,
      "step": 10767
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6318,
      "step": 10768
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 1.0505,
      "step": 10769
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 10770
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8761,
      "step": 10771
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5094,
      "step": 10772
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6602,
      "step": 10773
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4022,
      "step": 10774
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 10775
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4944,
      "step": 10776
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8681,
      "step": 10777
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7588,
      "step": 10778
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4923,
      "step": 10779
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8028,
      "step": 10780
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6962,
      "step": 10781
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5452,
      "step": 10782
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6237,
      "step": 10783
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8226,
      "step": 10784
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5532,
      "step": 10785
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5326,
      "step": 10786
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4186,
      "step": 10787
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6274,
      "step": 10788
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5713,
      "step": 10789
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6583,
      "step": 10790
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4407,
      "step": 10791
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5471,
      "step": 10792
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6671,
      "step": 10793
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7097,
      "step": 10794
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.462,
      "step": 10795
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.8473,
      "step": 10796
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 10797
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6231,
      "step": 10798
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 10799
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.5476,
      "step": 10800
    },
    {
      "epoch": 1.77,
      "eval_loss": 0.7807790637016296,
      "eval_runtime": 131.6087,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 10800
    },
    {
      "epoch": 1.77,
      "mmlu_eval_accuracy": 0.5075532548501424,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "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": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5833333333333334,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.2823529411764706,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.4074074074074074,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.0156301742823255,
      "step": 10800
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 10801
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.4826,
      "step": 10802
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.6533,
      "step": 10803
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6134,
      "step": 10804
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 10805
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 10806
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6965,
      "step": 10807
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 10808
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7258,
      "step": 10809
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.812,
      "step": 10810
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7078,
      "step": 10811
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5439,
      "step": 10812
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 10813
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6264,
      "step": 10814
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7558,
      "step": 10815
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.8581,
      "step": 10816
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5211,
      "step": 10817
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7044,
      "step": 10818
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7476,
      "step": 10819
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6396,
      "step": 10820
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6537,
      "step": 10821
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 10822
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 10823
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7193,
      "step": 10824
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.3986,
      "step": 10825
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6722,
      "step": 10826
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6482,
      "step": 10827
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6227,
      "step": 10828
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6912,
      "step": 10829
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7165,
      "step": 10830
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.693,
      "step": 10831
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5231,
      "step": 10832
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6901,
      "step": 10833
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7398,
      "step": 10834
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7498,
      "step": 10835
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.811,
      "step": 10836
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5429,
      "step": 10837
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7757,
      "step": 10838
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7787,
      "step": 10839
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.694,
      "step": 10840
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 10841
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5465,
      "step": 10842
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 10843
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 10844
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 10845
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.3859,
      "step": 10846
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.4231,
      "step": 10847
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.8221,
      "step": 10848
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.4439,
      "step": 10849
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.9112,
      "step": 10850
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5097,
      "step": 10851
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.4715,
      "step": 10852
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5307,
      "step": 10853
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.6797,
      "step": 10854
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 10855
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.4826,
      "step": 10856
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.8472,
      "step": 10857
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.8011,
      "step": 10858
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 10859
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7415,
      "step": 10860
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.4059,
      "step": 10861
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 0.7193,
      "step": 10862
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0002,
      "loss": 1.0274,
      "step": 10863
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9296,
      "step": 10864
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.2949,
      "step": 10865
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 10866
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5398,
      "step": 10867
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.4351,
      "step": 10868
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5862,
      "step": 10869
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6215,
      "step": 10870
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8639,
      "step": 10871
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 10872
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 10873
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7184,
      "step": 10874
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7713,
      "step": 10875
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.655,
      "step": 10876
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 10877
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7976,
      "step": 10878
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5502,
      "step": 10879
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6895,
      "step": 10880
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9172,
      "step": 10881
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.764,
      "step": 10882
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5673,
      "step": 10883
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5119,
      "step": 10884
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.3918,
      "step": 10885
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9264,
      "step": 10886
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 10887
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6305,
      "step": 10888
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7883,
      "step": 10889
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5052,
      "step": 10890
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6288,
      "step": 10891
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7439,
      "step": 10892
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 10893
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8291,
      "step": 10894
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9431,
      "step": 10895
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 10896
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 1.0071,
      "step": 10897
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5169,
      "step": 10898
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7199,
      "step": 10899
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8343,
      "step": 10900
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5782,
      "step": 10901
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6123,
      "step": 10902
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.7722,
      "step": 10903
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.4968,
      "step": 10904
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 10905
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6198,
      "step": 10906
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5965,
      "step": 10907
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9537,
      "step": 10908
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5809,
      "step": 10909
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5994,
      "step": 10910
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6189,
      "step": 10911
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5924,
      "step": 10912
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6463,
      "step": 10913
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6945,
      "step": 10914
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6004,
      "step": 10915
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 10916
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 10917
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8067,
      "step": 10918
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.4889,
      "step": 10919
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.6094,
      "step": 10920
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 10921
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.5141,
      "step": 10922
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.9497,
      "step": 10923
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.8425,
      "step": 10924
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7926,
      "step": 10925
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7454,
      "step": 10926
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6415,
      "step": 10927
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7629,
      "step": 10928
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5004,
      "step": 10929
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7667,
      "step": 10930
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6287,
      "step": 10931
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 10932
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.8696,
      "step": 10933
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.4635,
      "step": 10934
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7827,
      "step": 10935
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6986,
      "step": 10936
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.8087,
      "step": 10937
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7787,
      "step": 10938
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6547,
      "step": 10939
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.628,
      "step": 10940
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 10941
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5776,
      "step": 10942
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 10943
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6908,
      "step": 10944
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 10945
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.506,
      "step": 10946
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.768,
      "step": 10947
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7129,
      "step": 10948
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.3786,
      "step": 10949
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 10950
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5267,
      "step": 10951
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 10952
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5509,
      "step": 10953
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 1.0301,
      "step": 10954
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.3032,
      "step": 10955
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 10956
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7486,
      "step": 10957
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6575,
      "step": 10958
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6844,
      "step": 10959
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6124,
      "step": 10960
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7713,
      "step": 10961
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6748,
      "step": 10962
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6431,
      "step": 10963
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 10964
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.69,
      "step": 10965
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5912,
      "step": 10966
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.904,
      "step": 10967
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7347,
      "step": 10968
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.847,
      "step": 10969
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6906,
      "step": 10970
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 10971
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 10972
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6892,
      "step": 10973
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6929,
      "step": 10974
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5322,
      "step": 10975
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.6952,
      "step": 10976
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.4149,
      "step": 10977
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7004,
      "step": 10978
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5588,
      "step": 10979
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.7442,
      "step": 10980
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5277,
      "step": 10981
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.4792,
      "step": 10982
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5923,
      "step": 10983
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 10984
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.5448,
      "step": 10985
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.8054,
      "step": 10986
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 1.0435,
      "step": 10987
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5229,
      "step": 10988
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6078,
      "step": 10989
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 10990
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4913,
      "step": 10991
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6922,
      "step": 10992
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5022,
      "step": 10993
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 10994
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 10995
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6955,
      "step": 10996
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.748,
      "step": 10997
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7206,
      "step": 10998
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.3935,
      "step": 10999
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.9216,
      "step": 11000
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.7829310894012451,
      "eval_runtime": 131.5805,
      "eval_samples_per_second": 7.6,
      "eval_steps_per_second": 3.8,
      "step": 11000
    },
    {
      "epoch": 1.81,
      "mmlu_eval_accuracy": 0.537492925773278,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.0,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "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": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.8095238095238095,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5789473684210527,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.8421052631578947,
      "mmlu_loss": 1.2170404258230336,
      "step": 11000
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.8421,
      "step": 11001
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4929,
      "step": 11002
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6303,
      "step": 11003
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6829,
      "step": 11004
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 1.0049,
      "step": 11005
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.9758,
      "step": 11006
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6668,
      "step": 11007
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6063,
      "step": 11008
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5512,
      "step": 11009
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.572,
      "step": 11010
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 11011
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6274,
      "step": 11012
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.767,
      "step": 11013
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 1.0359,
      "step": 11014
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4477,
      "step": 11015
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4812,
      "step": 11016
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5141,
      "step": 11017
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.8039,
      "step": 11018
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.8641,
      "step": 11019
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.4667,
      "step": 11020
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.9219,
      "step": 11021
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.2933,
      "step": 11022
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5026,
      "step": 11023
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7478,
      "step": 11024
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7958,
      "step": 11025
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5875,
      "step": 11026
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6531,
      "step": 11027
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6456,
      "step": 11028
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.8587,
      "step": 11029
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6485,
      "step": 11030
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 11031
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6983,
      "step": 11032
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7706,
      "step": 11033
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 11034
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6253,
      "step": 11035
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7594,
      "step": 11036
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7098,
      "step": 11037
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7541,
      "step": 11038
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6786,
      "step": 11039
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5953,
      "step": 11040
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.9316,
      "step": 11041
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5347,
      "step": 11042
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 1.0143,
      "step": 11043
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.5158,
      "step": 11044
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.7493,
      "step": 11045
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0002,
      "loss": 0.6198,
      "step": 11046
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.4574,
      "step": 11047
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8145,
      "step": 11048
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 11049
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7121,
      "step": 11050
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 11051
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 11052
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7554,
      "step": 11053
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.444,
      "step": 11054
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 11055
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6538,
      "step": 11056
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7454,
      "step": 11057
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 11058
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8607,
      "step": 11059
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.542,
      "step": 11060
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6709,
      "step": 11061
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8804,
      "step": 11062
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 11063
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 1.0098,
      "step": 11064
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5615,
      "step": 11065
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 11066
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7246,
      "step": 11067
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.9436,
      "step": 11068
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 11069
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5953,
      "step": 11070
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.9652,
      "step": 11071
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5783,
      "step": 11072
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.477,
      "step": 11073
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.9385,
      "step": 11074
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8062,
      "step": 11075
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5514,
      "step": 11076
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6562,
      "step": 11077
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5639,
      "step": 11078
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6964,
      "step": 11079
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.599,
      "step": 11080
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6365,
      "step": 11081
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 1.1309,
      "step": 11082
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5493,
      "step": 11083
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6039,
      "step": 11084
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.9954,
      "step": 11085
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5065,
      "step": 11086
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.9104,
      "step": 11087
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6107,
      "step": 11088
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5252,
      "step": 11089
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8991,
      "step": 11090
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6428,
      "step": 11091
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7673,
      "step": 11092
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7619,
      "step": 11093
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6212,
      "step": 11094
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.621,
      "step": 11095
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.4782,
      "step": 11096
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6215,
      "step": 11097
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6632,
      "step": 11098
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5148,
      "step": 11099
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7537,
      "step": 11100
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 11101
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.6501,
      "step": 11102
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 11103
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.8284,
      "step": 11104
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 11105
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 11106
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.7215,
      "step": 11107
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7894,
      "step": 11108
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.825,
      "step": 11109
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.9536,
      "step": 11110
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4545,
      "step": 11111
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7624,
      "step": 11112
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 11113
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.8407,
      "step": 11114
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7177,
      "step": 11115
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6772,
      "step": 11116
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.9113,
      "step": 11117
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5586,
      "step": 11118
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 11119
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4975,
      "step": 11120
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.3683,
      "step": 11121
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7605,
      "step": 11122
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 11123
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7562,
      "step": 11124
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7458,
      "step": 11125
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.9256,
      "step": 11126
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 11127
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7896,
      "step": 11128
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6535,
      "step": 11129
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7041,
      "step": 11130
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.751,
      "step": 11131
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.443,
      "step": 11132
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.586,
      "step": 11133
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7004,
      "step": 11134
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5762,
      "step": 11135
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6976,
      "step": 11136
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4229,
      "step": 11137
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6043,
      "step": 11138
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7368,
      "step": 11139
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4749,
      "step": 11140
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 11141
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.8937,
      "step": 11142
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6532,
      "step": 11143
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 11144
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5148,
      "step": 11145
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4624,
      "step": 11146
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6294,
      "step": 11147
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7637,
      "step": 11148
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.9749,
      "step": 11149
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.8253,
      "step": 11150
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.715,
      "step": 11151
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.9309,
      "step": 11152
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4565,
      "step": 11153
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.3341,
      "step": 11154
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.669,
      "step": 11155
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7101,
      "step": 11156
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7426,
      "step": 11157
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7717,
      "step": 11158
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5797,
      "step": 11159
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 11160
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.7236,
      "step": 11161
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.642,
      "step": 11162
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.6325,
      "step": 11163
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4183,
      "step": 11164
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4901,
      "step": 11165
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.4372,
      "step": 11166
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5411,
      "step": 11167
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0002,
      "loss": 0.5394,
      "step": 11168
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6703,
      "step": 11169
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7835,
      "step": 11170
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6733,
      "step": 11171
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7372,
      "step": 11172
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 11173
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5014,
      "step": 11174
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.8238,
      "step": 11175
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6389,
      "step": 11176
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7125,
      "step": 11177
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 11178
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7335,
      "step": 11179
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5062,
      "step": 11180
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 11181
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 11182
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6836,
      "step": 11183
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5296,
      "step": 11184
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5586,
      "step": 11185
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5622,
      "step": 11186
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5794,
      "step": 11187
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.754,
      "step": 11188
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6495,
      "step": 11189
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 11190
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6863,
      "step": 11191
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 1.0507,
      "step": 11192
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.4763,
      "step": 11193
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 11194
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6718,
      "step": 11195
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 11196
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7946,
      "step": 11197
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6995,
      "step": 11198
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6396,
      "step": 11199
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.628,
      "step": 11200
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.7764880657196045,
      "eval_runtime": 131.6376,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 11200
    },
    {
      "epoch": 1.84,
      "mmlu_eval_accuracy": 0.5260293065779325,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5769230769230769,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "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": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.29411764705882354,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.6666666666666666,
      "mmlu_eval_accuracy_world_religions": 0.7894736842105263,
      "mmlu_loss": 1.340101167403998,
      "step": 11200
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5064,
      "step": 11201
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7573,
      "step": 11202
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6616,
      "step": 11203
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6628,
      "step": 11204
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7772,
      "step": 11205
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7542,
      "step": 11206
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.652,
      "step": 11207
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 11208
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7692,
      "step": 11209
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.8506,
      "step": 11210
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.9034,
      "step": 11211
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7924,
      "step": 11212
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 11213
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7837,
      "step": 11214
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6794,
      "step": 11215
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6192,
      "step": 11216
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.736,
      "step": 11217
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.8353,
      "step": 11218
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5591,
      "step": 11219
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.5404,
      "step": 11220
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.8952,
      "step": 11221
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7938,
      "step": 11222
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 11223
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.4086,
      "step": 11224
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.4401,
      "step": 11225
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 11226
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6593,
      "step": 11227
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.8694,
      "step": 11228
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.6553,
      "step": 11229
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 11230
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.4789,
      "step": 11231
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7713,
      "step": 11232
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6877,
      "step": 11233
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8042,
      "step": 11234
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 11235
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7416,
      "step": 11236
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5904,
      "step": 11237
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5428,
      "step": 11238
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6698,
      "step": 11239
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5864,
      "step": 11240
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8584,
      "step": 11241
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 11242
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8919,
      "step": 11243
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.9313,
      "step": 11244
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5748,
      "step": 11245
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6697,
      "step": 11246
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.4262,
      "step": 11247
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7412,
      "step": 11248
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.9169,
      "step": 11249
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7158,
      "step": 11250
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7663,
      "step": 11251
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6603,
      "step": 11252
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.9844,
      "step": 11253
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5565,
      "step": 11254
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6595,
      "step": 11255
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.612,
      "step": 11256
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7186,
      "step": 11257
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 11258
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.712,
      "step": 11259
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6854,
      "step": 11260
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6525,
      "step": 11261
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6811,
      "step": 11262
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5129,
      "step": 11263
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7165,
      "step": 11264
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7848,
      "step": 11265
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 11266
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5126,
      "step": 11267
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8656,
      "step": 11268
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7089,
      "step": 11269
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 11270
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8676,
      "step": 11271
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.7231,
      "step": 11272
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6414,
      "step": 11273
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.4016,
      "step": 11274
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8522,
      "step": 11275
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5742,
      "step": 11276
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5505,
      "step": 11277
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 11278
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.9191,
      "step": 11279
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.479,
      "step": 11280
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6366,
      "step": 11281
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 11282
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5356,
      "step": 11283
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5985,
      "step": 11284
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.8061,
      "step": 11285
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.573,
      "step": 11286
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5199,
      "step": 11287
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.5826,
      "step": 11288
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0002,
      "loss": 0.6028,
      "step": 11289
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 11290
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6833,
      "step": 11291
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.9691,
      "step": 11292
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5871,
      "step": 11293
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6206,
      "step": 11294
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4938,
      "step": 11295
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.3149,
      "step": 11296
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 11297
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6937,
      "step": 11298
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5566,
      "step": 11299
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.9684,
      "step": 11300
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7469,
      "step": 11301
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6019,
      "step": 11302
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6433,
      "step": 11303
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5182,
      "step": 11304
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.8589,
      "step": 11305
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5628,
      "step": 11306
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 11307
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4025,
      "step": 11308
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6505,
      "step": 11309
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 11310
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 11311
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 11312
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5289,
      "step": 11313
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 11314
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 11315
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6537,
      "step": 11316
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6666,
      "step": 11317
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7861,
      "step": 11318
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7484,
      "step": 11319
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5864,
      "step": 11320
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 11321
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.913,
      "step": 11322
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 11323
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 11324
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6478,
      "step": 11325
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6468,
      "step": 11326
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6405,
      "step": 11327
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7952,
      "step": 11328
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5604,
      "step": 11329
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 11330
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4436,
      "step": 11331
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 1.0061,
      "step": 11332
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5474,
      "step": 11333
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5353,
      "step": 11334
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4116,
      "step": 11335
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 1.0203,
      "step": 11336
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5499,
      "step": 11337
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5822,
      "step": 11338
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7241,
      "step": 11339
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7367,
      "step": 11340
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.8753,
      "step": 11341
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7969,
      "step": 11342
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5,
      "step": 11343
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.5585,
      "step": 11344
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7192,
      "step": 11345
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.6799,
      "step": 11346
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7108,
      "step": 11347
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.4157,
      "step": 11348
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 11349
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 11350
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.4018,
      "step": 11351
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 11352
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 11353
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6783,
      "step": 11354
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6864,
      "step": 11355
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.645,
      "step": 11356
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5853,
      "step": 11357
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6124,
      "step": 11358
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5521,
      "step": 11359
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7121,
      "step": 11360
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 11361
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7626,
      "step": 11362
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5313,
      "step": 11363
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 11364
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 11365
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.4427,
      "step": 11366
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.4627,
      "step": 11367
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 11368
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.9001,
      "step": 11369
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.898,
      "step": 11370
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 11371
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 11372
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5157,
      "step": 11373
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 11374
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.4649,
      "step": 11375
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7422,
      "step": 11376
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.8422,
      "step": 11377
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6887,
      "step": 11378
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6495,
      "step": 11379
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6867,
      "step": 11380
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 11381
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.4996,
      "step": 11382
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6303,
      "step": 11383
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 1.0731,
      "step": 11384
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6912,
      "step": 11385
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 1.0241,
      "step": 11386
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5031,
      "step": 11387
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7914,
      "step": 11388
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6076,
      "step": 11389
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.3978,
      "step": 11390
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6955,
      "step": 11391
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7863,
      "step": 11392
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6784,
      "step": 11393
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5885,
      "step": 11394
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5701,
      "step": 11395
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.8134,
      "step": 11396
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7965,
      "step": 11397
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6142,
      "step": 11398
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 11399
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.9863,
      "step": 11400
    },
    {
      "epoch": 1.87,
      "eval_loss": 0.7791166305541992,
      "eval_runtime": 131.7021,
      "eval_samples_per_second": 7.593,
      "eval_steps_per_second": 3.796,
      "step": 11400
    },
    {
      "epoch": 1.87,
      "mmlu_eval_accuracy": 0.5205736542742772,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "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": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.13636363636363635,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6923076923076923,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3176470588235294,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.4782608695652174,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.134330463102063,
      "step": 11400
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6263,
      "step": 11401
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5757,
      "step": 11402
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 11403
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 11404
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6832,
      "step": 11405
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.6605,
      "step": 11406
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7656,
      "step": 11407
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 11408
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.8691,
      "step": 11409
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.9947,
      "step": 11410
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0002,
      "loss": 0.7676,
      "step": 11411
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.9104,
      "step": 11412
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4268,
      "step": 11413
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5637,
      "step": 11414
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6184,
      "step": 11415
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7548,
      "step": 11416
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5896,
      "step": 11417
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.755,
      "step": 11418
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7704,
      "step": 11419
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4294,
      "step": 11420
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6175,
      "step": 11421
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5649,
      "step": 11422
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 11423
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7151,
      "step": 11424
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7014,
      "step": 11425
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 1.0122,
      "step": 11426
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 11427
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 11428
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4953,
      "step": 11429
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.742,
      "step": 11430
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8701,
      "step": 11431
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.3256,
      "step": 11432
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 11433
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7203,
      "step": 11434
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 1.0245,
      "step": 11435
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8112,
      "step": 11436
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8605,
      "step": 11437
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5342,
      "step": 11438
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5737,
      "step": 11439
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.692,
      "step": 11440
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6333,
      "step": 11441
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.9144,
      "step": 11442
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8202,
      "step": 11443
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.582,
      "step": 11444
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7032,
      "step": 11445
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.3663,
      "step": 11446
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4546,
      "step": 11447
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8654,
      "step": 11448
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7038,
      "step": 11449
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.722,
      "step": 11450
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.747,
      "step": 11451
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5155,
      "step": 11452
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7939,
      "step": 11453
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.714,
      "step": 11454
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6705,
      "step": 11455
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5291,
      "step": 11456
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.816,
      "step": 11457
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4591,
      "step": 11458
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.706,
      "step": 11459
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5008,
      "step": 11460
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.6734,
      "step": 11461
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5826,
      "step": 11462
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.3889,
      "step": 11463
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7917,
      "step": 11464
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.4884,
      "step": 11465
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.9023,
      "step": 11466
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5543,
      "step": 11467
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7153,
      "step": 11468
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.9125,
      "step": 11469
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.7689,
      "step": 11470
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.8477,
      "step": 11471
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.5892,
      "step": 11472
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.8519,
      "step": 11473
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.2988,
      "step": 11474
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.8915,
      "step": 11475
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6431,
      "step": 11476
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7846,
      "step": 11477
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 11478
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 11479
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7213,
      "step": 11480
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.4971,
      "step": 11481
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 11482
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.8726,
      "step": 11483
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5351,
      "step": 11484
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 11485
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7629,
      "step": 11486
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5877,
      "step": 11487
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7816,
      "step": 11488
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6559,
      "step": 11489
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6295,
      "step": 11490
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6644,
      "step": 11491
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5747,
      "step": 11492
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7002,
      "step": 11493
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.875,
      "step": 11494
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.443,
      "step": 11495
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.3762,
      "step": 11496
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5372,
      "step": 11497
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.8035,
      "step": 11498
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.613,
      "step": 11499
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6299,
      "step": 11500
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.4564,
      "step": 11501
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7296,
      "step": 11502
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6655,
      "step": 11503
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5681,
      "step": 11504
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.545,
      "step": 11505
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6058,
      "step": 11506
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7467,
      "step": 11507
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6589,
      "step": 11508
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.9377,
      "step": 11509
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 11510
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 11511
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 11512
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7757,
      "step": 11513
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6673,
      "step": 11514
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5473,
      "step": 11515
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5553,
      "step": 11516
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.9627,
      "step": 11517
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7428,
      "step": 11518
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.9531,
      "step": 11519
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7953,
      "step": 11520
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6015,
      "step": 11521
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.9548,
      "step": 11522
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.515,
      "step": 11523
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6224,
      "step": 11524
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5756,
      "step": 11525
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.5261,
      "step": 11526
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.8058,
      "step": 11527
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7842,
      "step": 11528
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.543,
      "step": 11529
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.56,
      "step": 11530
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.6266,
      "step": 11531
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7054,
      "step": 11532
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.7232,
      "step": 11533
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6293,
      "step": 11534
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6142,
      "step": 11535
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5397,
      "step": 11536
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4809,
      "step": 11537
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6975,
      "step": 11538
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4508,
      "step": 11539
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6473,
      "step": 11540
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.3854,
      "step": 11541
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7913,
      "step": 11542
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7799,
      "step": 11543
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7415,
      "step": 11544
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.697,
      "step": 11545
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.848,
      "step": 11546
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.9364,
      "step": 11547
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7533,
      "step": 11548
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5617,
      "step": 11549
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5678,
      "step": 11550
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5059,
      "step": 11551
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6696,
      "step": 11552
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6951,
      "step": 11553
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7727,
      "step": 11554
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 11555
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7337,
      "step": 11556
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6254,
      "step": 11557
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.9048,
      "step": 11558
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 11559
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.616,
      "step": 11560
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.715,
      "step": 11561
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.9124,
      "step": 11562
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.807,
      "step": 11563
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5027,
      "step": 11564
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4036,
      "step": 11565
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.433,
      "step": 11566
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5964,
      "step": 11567
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4886,
      "step": 11568
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.8605,
      "step": 11569
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5016,
      "step": 11570
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.762,
      "step": 11571
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 11572
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6279,
      "step": 11573
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4908,
      "step": 11574
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.312,
      "step": 11575
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5256,
      "step": 11576
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.6484,
      "step": 11577
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5849,
      "step": 11578
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 11579
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 1.029,
      "step": 11580
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7373,
      "step": 11581
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7137,
      "step": 11582
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 1.0145,
      "step": 11583
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 11584
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7349,
      "step": 11585
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7355,
      "step": 11586
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.4633,
      "step": 11587
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.9388,
      "step": 11588
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7135,
      "step": 11589
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.7004,
      "step": 11590
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.8433,
      "step": 11591
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.476,
      "step": 11592
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.8655,
      "step": 11593
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.0002,
      "loss": 0.5769,
      "step": 11594
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5185,
      "step": 11595
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 11596
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7333,
      "step": 11597
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.565,
      "step": 11598
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5205,
      "step": 11599
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5454,
      "step": 11600
    },
    {
      "epoch": 1.91,
      "eval_loss": 0.7844635248184204,
      "eval_runtime": 131.6129,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 11600
    },
    {
      "epoch": 1.91,
      "mmlu_eval_accuracy": 0.5281357446639989,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.5454545454545454,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_world_history": 0.6538461538461539,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.696969696969697,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.3181101129276012,
      "step": 11600
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7362,
      "step": 11601
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7633,
      "step": 11602
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 11603
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4725,
      "step": 11604
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 11605
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6199,
      "step": 11606
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.777,
      "step": 11607
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6884,
      "step": 11608
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6456,
      "step": 11609
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.556,
      "step": 11610
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 11611
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.8138,
      "step": 11612
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 11613
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 11614
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7056,
      "step": 11615
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6275,
      "step": 11616
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4512,
      "step": 11617
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.464,
      "step": 11618
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.9935,
      "step": 11619
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5022,
      "step": 11620
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7286,
      "step": 11621
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7811,
      "step": 11622
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5323,
      "step": 11623
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.8615,
      "step": 11624
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6179,
      "step": 11625
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6291,
      "step": 11626
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.446,
      "step": 11627
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 11628
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5735,
      "step": 11629
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 11630
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 11631
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6958,
      "step": 11632
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5564,
      "step": 11633
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7443,
      "step": 11634
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.653,
      "step": 11635
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6412,
      "step": 11636
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.3721,
      "step": 11637
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.9867,
      "step": 11638
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 11639
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6415,
      "step": 11640
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4774,
      "step": 11641
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4949,
      "step": 11642
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6108,
      "step": 11643
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 1.0234,
      "step": 11644
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.8535,
      "step": 11645
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.8928,
      "step": 11646
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.516,
      "step": 11647
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 11648
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5514,
      "step": 11649
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4136,
      "step": 11650
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7995,
      "step": 11651
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7588,
      "step": 11652
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.7456,
      "step": 11653
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4187,
      "step": 11654
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.5165,
      "step": 11655
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5307,
      "step": 11656
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7996,
      "step": 11657
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4936,
      "step": 11658
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5988,
      "step": 11659
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6123,
      "step": 11660
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 11661
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5888,
      "step": 11662
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6934,
      "step": 11663
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7249,
      "step": 11664
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8888,
      "step": 11665
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5771,
      "step": 11666
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6882,
      "step": 11667
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6915,
      "step": 11668
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7167,
      "step": 11669
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8389,
      "step": 11670
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.716,
      "step": 11671
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8627,
      "step": 11672
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4557,
      "step": 11673
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5543,
      "step": 11674
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8366,
      "step": 11675
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5941,
      "step": 11676
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6945,
      "step": 11677
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6675,
      "step": 11678
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7393,
      "step": 11679
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6314,
      "step": 11680
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.3872,
      "step": 11681
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4992,
      "step": 11682
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6979,
      "step": 11683
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.639,
      "step": 11684
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 11685
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 11686
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.651,
      "step": 11687
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6477,
      "step": 11688
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5808,
      "step": 11689
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6775,
      "step": 11690
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 11691
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5954,
      "step": 11692
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 11693
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 11694
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7971,
      "step": 11695
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5804,
      "step": 11696
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 11697
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.3736,
      "step": 11698
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.753,
      "step": 11699
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6079,
      "step": 11700
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6117,
      "step": 11701
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.3846,
      "step": 11702
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7505,
      "step": 11703
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4573,
      "step": 11704
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6661,
      "step": 11705
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5249,
      "step": 11706
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7803,
      "step": 11707
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 11708
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.8262,
      "step": 11709
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4803,
      "step": 11710
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7448,
      "step": 11711
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.4225,
      "step": 11712
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5008,
      "step": 11713
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.7567,
      "step": 11714
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 11715
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.0002,
      "loss": 0.5464,
      "step": 11716
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 11717
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.818,
      "step": 11718
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5153,
      "step": 11719
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 11720
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5276,
      "step": 11721
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 11722
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.459,
      "step": 11723
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.8878,
      "step": 11724
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7429,
      "step": 11725
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7058,
      "step": 11726
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 1.1245,
      "step": 11727
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5714,
      "step": 11728
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.941,
      "step": 11729
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 11730
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7233,
      "step": 11731
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6572,
      "step": 11732
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6453,
      "step": 11733
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.9494,
      "step": 11734
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5933,
      "step": 11735
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7826,
      "step": 11736
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6508,
      "step": 11737
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 11738
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5583,
      "step": 11739
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7854,
      "step": 11740
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7766,
      "step": 11741
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6621,
      "step": 11742
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 11743
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6671,
      "step": 11744
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.4207,
      "step": 11745
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.515,
      "step": 11746
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5239,
      "step": 11747
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 11748
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7495,
      "step": 11749
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.654,
      "step": 11750
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.9121,
      "step": 11751
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.4546,
      "step": 11752
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 11753
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.9619,
      "step": 11754
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7337,
      "step": 11755
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.788,
      "step": 11756
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7108,
      "step": 11757
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5499,
      "step": 11758
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7748,
      "step": 11759
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5695,
      "step": 11760
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7405,
      "step": 11761
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7885,
      "step": 11762
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.528,
      "step": 11763
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.583,
      "step": 11764
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.5112,
      "step": 11765
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7107,
      "step": 11766
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7942,
      "step": 11767
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.4293,
      "step": 11768
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.8287,
      "step": 11769
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.6351,
      "step": 11770
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.8107,
      "step": 11771
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.7242,
      "step": 11772
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.8738,
      "step": 11773
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.3918,
      "step": 11774
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.9591,
      "step": 11775
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.9998,
      "step": 11776
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5178,
      "step": 11777
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.8066,
      "step": 11778
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5619,
      "step": 11779
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7472,
      "step": 11780
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.4473,
      "step": 11781
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.9189,
      "step": 11782
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.4266,
      "step": 11783
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.357,
      "step": 11784
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5961,
      "step": 11785
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5708,
      "step": 11786
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7602,
      "step": 11787
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6157,
      "step": 11788
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5575,
      "step": 11789
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 11790
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6725,
      "step": 11791
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5802,
      "step": 11792
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.848,
      "step": 11793
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6185,
      "step": 11794
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 11795
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.3958,
      "step": 11796
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 11797
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7237,
      "step": 11798
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7376,
      "step": 11799
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 11800
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.7783686518669128,
      "eval_runtime": 131.6038,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.799,
      "step": 11800
    },
    {
      "epoch": 1.94,
      "mmlu_eval_accuracy": 0.5428527393162451,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "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": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4634146341463415,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5581395348837209,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.85,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.96,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.7325581395348837,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.6470588235294118,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1458515524864197,
      "step": 11800
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.9907,
      "step": 11801
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.8092,
      "step": 11802
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.732,
      "step": 11803
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 11804
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6382,
      "step": 11805
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7084,
      "step": 11806
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.727,
      "step": 11807
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.65,
      "step": 11808
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.8404,
      "step": 11809
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7458,
      "step": 11810
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7581,
      "step": 11811
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7208,
      "step": 11812
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7259,
      "step": 11813
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 1.0151,
      "step": 11814
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7083,
      "step": 11815
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.664,
      "step": 11816
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5646,
      "step": 11817
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5897,
      "step": 11818
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5006,
      "step": 11819
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7072,
      "step": 11820
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5355,
      "step": 11821
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6093,
      "step": 11822
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.849,
      "step": 11823
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.625,
      "step": 11824
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5193,
      "step": 11825
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.8117,
      "step": 11826
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6807,
      "step": 11827
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 11828
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.87,
      "step": 11829
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 11830
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7558,
      "step": 11831
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6133,
      "step": 11832
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.597,
      "step": 11833
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.9046,
      "step": 11834
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6236,
      "step": 11835
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.6662,
      "step": 11836
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.0002,
      "loss": 0.7778,
      "step": 11837
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5606,
      "step": 11838
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5458,
      "step": 11839
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.745,
      "step": 11840
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6474,
      "step": 11841
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5937,
      "step": 11842
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7916,
      "step": 11843
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 11844
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.9171,
      "step": 11845
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7503,
      "step": 11846
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7054,
      "step": 11847
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.4737,
      "step": 11848
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.4141,
      "step": 11849
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 11850
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.9132,
      "step": 11851
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6743,
      "step": 11852
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.4501,
      "step": 11853
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6382,
      "step": 11854
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6824,
      "step": 11855
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5746,
      "step": 11856
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.8293,
      "step": 11857
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 11858
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.9757,
      "step": 11859
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6393,
      "step": 11860
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6843,
      "step": 11861
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5481,
      "step": 11862
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.815,
      "step": 11863
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.726,
      "step": 11864
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.9097,
      "step": 11865
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.8358,
      "step": 11866
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 11867
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.535,
      "step": 11868
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 11869
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 11870
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.804,
      "step": 11871
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.3779,
      "step": 11872
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5448,
      "step": 11873
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.38,
      "step": 11874
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6615,
      "step": 11875
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6776,
      "step": 11876
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5894,
      "step": 11877
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6467,
      "step": 11878
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5581,
      "step": 11879
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7651,
      "step": 11880
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 11881
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 11882
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.735,
      "step": 11883
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6643,
      "step": 11884
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6719,
      "step": 11885
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 11886
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7447,
      "step": 11887
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6228,
      "step": 11888
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7785,
      "step": 11889
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7864,
      "step": 11890
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6825,
      "step": 11891
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.871,
      "step": 11892
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.8377,
      "step": 11893
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6071,
      "step": 11894
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.6699,
      "step": 11895
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.7487,
      "step": 11896
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.3321,
      "step": 11897
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.5482,
      "step": 11898
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7061,
      "step": 11899
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 11900
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 11901
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.8176,
      "step": 11902
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 11903
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6098,
      "step": 11904
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 11905
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 1.0091,
      "step": 11906
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7845,
      "step": 11907
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 11908
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6286,
      "step": 11909
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5798,
      "step": 11910
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.8425,
      "step": 11911
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7134,
      "step": 11912
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5075,
      "step": 11913
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6126,
      "step": 11914
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 11915
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.527,
      "step": 11916
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5456,
      "step": 11917
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6522,
      "step": 11918
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6744,
      "step": 11919
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.816,
      "step": 11920
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6062,
      "step": 11921
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.8602,
      "step": 11922
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.4197,
      "step": 11923
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5749,
      "step": 11924
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.8685,
      "step": 11925
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.4159,
      "step": 11926
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5773,
      "step": 11927
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7873,
      "step": 11928
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6577,
      "step": 11929
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7194,
      "step": 11930
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.8317,
      "step": 11931
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6831,
      "step": 11932
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6235,
      "step": 11933
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.9562,
      "step": 11934
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7698,
      "step": 11935
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5094,
      "step": 11936
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.3762,
      "step": 11937
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6362,
      "step": 11938
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6417,
      "step": 11939
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7445,
      "step": 11940
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6724,
      "step": 11941
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7269,
      "step": 11942
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.9022,
      "step": 11943
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6355,
      "step": 11944
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5921,
      "step": 11945
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5795,
      "step": 11946
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5285,
      "step": 11947
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 11948
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6438,
      "step": 11949
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6319,
      "step": 11950
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.71,
      "step": 11951
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 11952
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6019,
      "step": 11953
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.3742,
      "step": 11954
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6545,
      "step": 11955
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5509,
      "step": 11956
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 11957
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.6132,
      "step": 11958
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.5771,
      "step": 11959
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7006,
      "step": 11960
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6363,
      "step": 11961
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7393,
      "step": 11962
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7895,
      "step": 11963
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6678,
      "step": 11964
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6336,
      "step": 11965
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.9549,
      "step": 11966
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6621,
      "step": 11967
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.9273,
      "step": 11968
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.572,
      "step": 11969
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8616,
      "step": 11970
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4774,
      "step": 11971
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.3395,
      "step": 11972
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4842,
      "step": 11973
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7611,
      "step": 11974
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8146,
      "step": 11975
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.9987,
      "step": 11976
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 11977
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6696,
      "step": 11978
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4972,
      "step": 11979
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4445,
      "step": 11980
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 1.0979,
      "step": 11981
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6627,
      "step": 11982
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6415,
      "step": 11983
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.518,
      "step": 11984
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4215,
      "step": 11985
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.3961,
      "step": 11986
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7162,
      "step": 11987
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.9722,
      "step": 11988
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7051,
      "step": 11989
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7966,
      "step": 11990
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 11991
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6419,
      "step": 11992
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7661,
      "step": 11993
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7501,
      "step": 11994
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6427,
      "step": 11995
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.409,
      "step": 11996
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8475,
      "step": 11997
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5129,
      "step": 11998
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.9905,
      "step": 11999
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5944,
      "step": 12000
    },
    {
      "epoch": 1.97,
      "eval_loss": 0.7773643136024475,
      "eval_runtime": 131.4523,
      "eval_samples_per_second": 7.607,
      "eval_steps_per_second": 3.804,
      "step": 12000
    },
    {
      "epoch": 1.97,
      "mmlu_eval_accuracy": 0.5198968969535794,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.5625,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.5652173913043478,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7441860465116279,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.27,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.4,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.4927536231884058,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5555555555555556,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.119068786540791,
      "step": 12000
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5567,
      "step": 12001
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 12002
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5526,
      "step": 12003
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5562,
      "step": 12004
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7534,
      "step": 12005
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.3653,
      "step": 12006
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4218,
      "step": 12007
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7464,
      "step": 12008
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8101,
      "step": 12009
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 12010
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6751,
      "step": 12011
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8723,
      "step": 12012
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 12013
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 12014
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6074,
      "step": 12015
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7655,
      "step": 12016
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.8182,
      "step": 12017
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.4447,
      "step": 12018
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.7105,
      "step": 12019
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.0002,
      "loss": 0.6923,
      "step": 12020
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5581,
      "step": 12021
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5628,
      "step": 12022
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.584,
      "step": 12023
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8663,
      "step": 12024
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5621,
      "step": 12025
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.802,
      "step": 12026
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 12027
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7101,
      "step": 12028
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8757,
      "step": 12029
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8857,
      "step": 12030
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.4166,
      "step": 12031
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.9764,
      "step": 12032
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5537,
      "step": 12033
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 12034
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8633,
      "step": 12035
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6196,
      "step": 12036
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8467,
      "step": 12037
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5871,
      "step": 12038
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.707,
      "step": 12039
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8055,
      "step": 12040
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7628,
      "step": 12041
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.682,
      "step": 12042
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.569,
      "step": 12043
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 1.1484,
      "step": 12044
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5685,
      "step": 12045
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5953,
      "step": 12046
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8231,
      "step": 12047
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6213,
      "step": 12048
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6789,
      "step": 12049
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6969,
      "step": 12050
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 12051
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7717,
      "step": 12052
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6731,
      "step": 12053
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 12054
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.9454,
      "step": 12055
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6439,
      "step": 12056
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.4259,
      "step": 12057
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6982,
      "step": 12058
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7687,
      "step": 12059
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.4338,
      "step": 12060
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.676,
      "step": 12061
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.4189,
      "step": 12062
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7399,
      "step": 12063
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7661,
      "step": 12064
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.4041,
      "step": 12065
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5484,
      "step": 12066
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7992,
      "step": 12067
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 1.0647,
      "step": 12068
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.679,
      "step": 12069
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.3039,
      "step": 12070
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.318,
      "step": 12071
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6356,
      "step": 12072
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.6269,
      "step": 12073
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7518,
      "step": 12074
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8818,
      "step": 12075
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8768,
      "step": 12076
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.5172,
      "step": 12077
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.9219,
      "step": 12078
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7091,
      "step": 12079
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.7106,
      "step": 12080
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.8318,
      "step": 12081
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.4797,
      "step": 12082
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5959,
      "step": 12083
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.765,
      "step": 12084
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6309,
      "step": 12085
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.414,
      "step": 12086
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6526,
      "step": 12087
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7643,
      "step": 12088
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6984,
      "step": 12089
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5486,
      "step": 12090
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.766,
      "step": 12091
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6904,
      "step": 12092
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7962,
      "step": 12093
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5706,
      "step": 12094
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6305,
      "step": 12095
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.3838,
      "step": 12096
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6431,
      "step": 12097
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.887,
      "step": 12098
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7781,
      "step": 12099
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7565,
      "step": 12100
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6525,
      "step": 12101
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6031,
      "step": 12102
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7561,
      "step": 12103
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7291,
      "step": 12104
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8046,
      "step": 12105
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6849,
      "step": 12106
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.3997,
      "step": 12107
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7945,
      "step": 12108
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.3983,
      "step": 12109
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8438,
      "step": 12110
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.622,
      "step": 12111
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8509,
      "step": 12112
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7617,
      "step": 12113
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5716,
      "step": 12114
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.344,
      "step": 12115
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8336,
      "step": 12116
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8608,
      "step": 12117
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 12118
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7071,
      "step": 12119
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7709,
      "step": 12120
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.8081,
      "step": 12121
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 12122
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7645,
      "step": 12123
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 12124
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6853,
      "step": 12125
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5129,
      "step": 12126
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 12127
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7759,
      "step": 12128
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.771,
      "step": 12129
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.9454,
      "step": 12130
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5034,
      "step": 12131
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.7413,
      "step": 12132
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5654,
      "step": 12133
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6336,
      "step": 12134
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5391,
      "step": 12135
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.5453,
      "step": 12136
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6614,
      "step": 12137
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.666,
      "step": 12138
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6458,
      "step": 12139
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.6211,
      "step": 12140
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.875,
      "step": 12141
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.0002,
      "loss": 0.9045,
      "step": 12142
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5248,
      "step": 12143
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.7515,
      "step": 12144
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6267,
      "step": 12145
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6258,
      "step": 12146
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.8482,
      "step": 12147
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.65,
      "step": 12148
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.7858,
      "step": 12149
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6888,
      "step": 12150
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.8346,
      "step": 12151
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6351,
      "step": 12152
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6495,
      "step": 12153
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.646,
      "step": 12154
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5722,
      "step": 12155
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.7677,
      "step": 12156
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.8427,
      "step": 12157
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.7209,
      "step": 12158
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6242,
      "step": 12159
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.443,
      "step": 12160
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.8871,
      "step": 12161
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5146,
      "step": 12162
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.538,
      "step": 12163
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6401,
      "step": 12164
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4742,
      "step": 12165
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.9383,
      "step": 12166
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5111,
      "step": 12167
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 12168
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6713,
      "step": 12169
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4879,
      "step": 12170
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 12171
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.8994,
      "step": 12172
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6047,
      "step": 12173
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4071,
      "step": 12174
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6475,
      "step": 12175
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4776,
      "step": 12176
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3839,
      "step": 12177
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 12178
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4992,
      "step": 12179
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3214,
      "step": 12180
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3263,
      "step": 12181
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4062,
      "step": 12182
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.548,
      "step": 12183
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4122,
      "step": 12184
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3982,
      "step": 12185
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5135,
      "step": 12186
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6811,
      "step": 12187
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5059,
      "step": 12188
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6708,
      "step": 12189
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5772,
      "step": 12190
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4769,
      "step": 12191
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5281,
      "step": 12192
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.443,
      "step": 12193
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4922,
      "step": 12194
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3732,
      "step": 12195
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5544,
      "step": 12196
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4727,
      "step": 12197
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4983,
      "step": 12198
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.4733,
      "step": 12199
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.5338,
      "step": 12200
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.8075551390647888,
      "eval_runtime": 131.5958,
      "eval_samples_per_second": 7.599,
      "eval_steps_per_second": 3.8,
      "step": 12200
    },
    {
      "epoch": 2.0,
      "mmlu_eval_accuracy": 0.5421522429734946,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.5,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.7272727272727273,
      "mmlu_eval_accuracy_college_physics": 0.5454545454545454,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6538461538461539,
      "mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
      "mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.9090909090909091,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
      "mmlu_eval_accuracy_miscellaneous": 0.7558139534883721,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.33,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.7352941176470589,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.32941176470588235,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.2332380305265633,
      "step": 12200
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.504,
      "step": 12201
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.6702,
      "step": 12202
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5298,
      "step": 12203
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.6243,
      "step": 12204
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4817,
      "step": 12205
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4794,
      "step": 12206
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3085,
      "step": 12207
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3594,
      "step": 12208
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.545,
      "step": 12209
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.6465,
      "step": 12210
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 12211
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3327,
      "step": 12212
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5304,
      "step": 12213
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4814,
      "step": 12214
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.2219,
      "step": 12215
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.445,
      "step": 12216
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4118,
      "step": 12217
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4515,
      "step": 12218
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4396,
      "step": 12219
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.592,
      "step": 12220
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4781,
      "step": 12221
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.748,
      "step": 12222
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5611,
      "step": 12223
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4478,
      "step": 12224
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 12225
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4225,
      "step": 12226
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5585,
      "step": 12227
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5653,
      "step": 12228
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.635,
      "step": 12229
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5306,
      "step": 12230
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.2489,
      "step": 12231
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.441,
      "step": 12232
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5079,
      "step": 12233
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.2837,
      "step": 12234
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4689,
      "step": 12235
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.293,
      "step": 12236
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.6698,
      "step": 12237
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4681,
      "step": 12238
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.456,
      "step": 12239
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5468,
      "step": 12240
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.2815,
      "step": 12241
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3958,
      "step": 12242
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5262,
      "step": 12243
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5679,
      "step": 12244
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3558,
      "step": 12245
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4239,
      "step": 12246
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 12247
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5105,
      "step": 12248
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3352,
      "step": 12249
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4443,
      "step": 12250
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4303,
      "step": 12251
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5156,
      "step": 12252
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.3868,
      "step": 12253
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4614,
      "step": 12254
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 12255
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4139,
      "step": 12256
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5346,
      "step": 12257
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.7838,
      "step": 12258
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4696,
      "step": 12259
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4995,
      "step": 12260
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.4262,
      "step": 12261
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.5,
      "step": 12262
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0002,
      "loss": 0.523,
      "step": 12263
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3232,
      "step": 12264
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.2712,
      "step": 12265
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5436,
      "step": 12266
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4068,
      "step": 12267
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5302,
      "step": 12268
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4374,
      "step": 12269
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5525,
      "step": 12270
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.2976,
      "step": 12271
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5202,
      "step": 12272
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.551,
      "step": 12273
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6258,
      "step": 12274
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5667,
      "step": 12275
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4019,
      "step": 12276
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5678,
      "step": 12277
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3364,
      "step": 12278
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.2958,
      "step": 12279
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4859,
      "step": 12280
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4624,
      "step": 12281
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5854,
      "step": 12282
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4759,
      "step": 12283
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6074,
      "step": 12284
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3127,
      "step": 12285
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6641,
      "step": 12286
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4768,
      "step": 12287
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 12288
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 12289
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4739,
      "step": 12290
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4995,
      "step": 12291
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5314,
      "step": 12292
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4236,
      "step": 12293
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3793,
      "step": 12294
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4782,
      "step": 12295
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.455,
      "step": 12296
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3772,
      "step": 12297
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 1.2198,
      "step": 12298
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3453,
      "step": 12299
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5716,
      "step": 12300
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5479,
      "step": 12301
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.2456,
      "step": 12302
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.7168,
      "step": 12303
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4812,
      "step": 12304
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6771,
      "step": 12305
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.219,
      "step": 12306
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5701,
      "step": 12307
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.636,
      "step": 12308
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.2909,
      "step": 12309
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4467,
      "step": 12310
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3932,
      "step": 12311
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6176,
      "step": 12312
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.37,
      "step": 12313
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.668,
      "step": 12314
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5531,
      "step": 12315
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.6115,
      "step": 12316
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.8279,
      "step": 12317
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.4437,
      "step": 12318
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3288,
      "step": 12319
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5408,
      "step": 12320
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5898,
      "step": 12321
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 12322
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3314,
      "step": 12323
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.5116,
      "step": 12324
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3581,
      "step": 12325
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3016,
      "step": 12326
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3861,
      "step": 12327
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5322,
      "step": 12328
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5867,
      "step": 12329
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.507,
      "step": 12330
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6342,
      "step": 12331
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6072,
      "step": 12332
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.427,
      "step": 12333
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4804,
      "step": 12334
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4082,
      "step": 12335
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 12336
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.522,
      "step": 12337
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.2676,
      "step": 12338
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.2841,
      "step": 12339
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4763,
      "step": 12340
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 12341
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3677,
      "step": 12342
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6903,
      "step": 12343
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3506,
      "step": 12344
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5457,
      "step": 12345
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6144,
      "step": 12346
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5447,
      "step": 12347
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4958,
      "step": 12348
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4384,
      "step": 12349
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5941,
      "step": 12350
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4787,
      "step": 12351
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5482,
      "step": 12352
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5401,
      "step": 12353
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6732,
      "step": 12354
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4438,
      "step": 12355
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5774,
      "step": 12356
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3299,
      "step": 12357
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4331,
      "step": 12358
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.547,
      "step": 12359
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.2627,
      "step": 12360
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5283,
      "step": 12361
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3057,
      "step": 12362
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3866,
      "step": 12363
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3677,
      "step": 12364
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3039,
      "step": 12365
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6298,
      "step": 12366
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.6732,
      "step": 12367
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.7396,
      "step": 12368
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.2702,
      "step": 12369
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3746,
      "step": 12370
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5807,
      "step": 12371
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.318,
      "step": 12372
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.7261,
      "step": 12373
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5516,
      "step": 12374
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3549,
      "step": 12375
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 12376
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3012,
      "step": 12377
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5869,
      "step": 12378
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3293,
      "step": 12379
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.5258,
      "step": 12380
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4173,
      "step": 12381
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.4819,
      "step": 12382
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.243,
      "step": 12383
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3421,
      "step": 12384
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3765,
      "step": 12385
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5824,
      "step": 12386
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.2837,
      "step": 12387
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.2989,
      "step": 12388
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.6409,
      "step": 12389
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4637,
      "step": 12390
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.116,
      "step": 12391
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4485,
      "step": 12392
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5225,
      "step": 12393
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3753,
      "step": 12394
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3972,
      "step": 12395
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4211,
      "step": 12396
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 12397
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.2226,
      "step": 12398
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.317,
      "step": 12399
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.2561,
      "step": 12400
    },
    {
      "epoch": 2.04,
      "eval_loss": 0.8361802697181702,
      "eval_runtime": 131.5422,
      "eval_samples_per_second": 7.602,
      "eval_steps_per_second": 3.801,
      "step": 12400
    },
    {
      "epoch": 2.04,
      "mmlu_eval_accuracy": 0.537916730221835,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.42857142857142855,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.5625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
      "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": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "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": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.782608695652174,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5526315789473685,
      "mmlu_eval_accuracy_moral_scenarios": 0.31,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.6176470588235294,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5652173913043478,
      "mmlu_eval_accuracy_public_relations": 0.4166666666666667,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.4103204292043383,
      "step": 12400
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3874,
      "step": 12401
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4672,
      "step": 12402
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5921,
      "step": 12403
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.7239,
      "step": 12404
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3157,
      "step": 12405
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.539,
      "step": 12406
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4933,
      "step": 12407
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 12408
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.451,
      "step": 12409
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3475,
      "step": 12410
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.7109,
      "step": 12411
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5803,
      "step": 12412
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5935,
      "step": 12413
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 12414
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 12415
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4854,
      "step": 12416
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.7096,
      "step": 12417
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5261,
      "step": 12418
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.1754,
      "step": 12419
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3499,
      "step": 12420
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4932,
      "step": 12421
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5432,
      "step": 12422
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.7141,
      "step": 12423
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 12424
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4387,
      "step": 12425
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3062,
      "step": 12426
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4385,
      "step": 12427
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3342,
      "step": 12428
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 12429
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.6356,
      "step": 12430
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4699,
      "step": 12431
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5563,
      "step": 12432
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5857,
      "step": 12433
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.527,
      "step": 12434
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.2572,
      "step": 12435
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3924,
      "step": 12436
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3041,
      "step": 12437
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4643,
      "step": 12438
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.223,
      "step": 12439
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.5013,
      "step": 12440
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4306,
      "step": 12441
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3422,
      "step": 12442
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.6156,
      "step": 12443
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4149,
      "step": 12444
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 12445
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.0002,
      "loss": 0.3702,
      "step": 12446
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4213,
      "step": 12447
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3451,
      "step": 12448
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5329,
      "step": 12449
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4293,
      "step": 12450
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4877,
      "step": 12451
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.532,
      "step": 12452
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2627,
      "step": 12453
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5218,
      "step": 12454
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4944,
      "step": 12455
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4258,
      "step": 12456
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3695,
      "step": 12457
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3654,
      "step": 12458
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4252,
      "step": 12459
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5328,
      "step": 12460
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4483,
      "step": 12461
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4991,
      "step": 12462
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4612,
      "step": 12463
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4558,
      "step": 12464
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4262,
      "step": 12465
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4476,
      "step": 12466
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5698,
      "step": 12467
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.51,
      "step": 12468
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4612,
      "step": 12469
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3964,
      "step": 12470
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5367,
      "step": 12471
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4752,
      "step": 12472
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 12473
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4123,
      "step": 12474
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5002,
      "step": 12475
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4922,
      "step": 12476
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.339,
      "step": 12477
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4258,
      "step": 12478
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4551,
      "step": 12479
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5179,
      "step": 12480
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5107,
      "step": 12481
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2717,
      "step": 12482
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4726,
      "step": 12483
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2978,
      "step": 12484
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4165,
      "step": 12485
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4789,
      "step": 12486
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5968,
      "step": 12487
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.503,
      "step": 12488
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2332,
      "step": 12489
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4323,
      "step": 12490
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3555,
      "step": 12491
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.7414,
      "step": 12492
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3586,
      "step": 12493
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.6339,
      "step": 12494
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3686,
      "step": 12495
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4976,
      "step": 12496
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4773,
      "step": 12497
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.435,
      "step": 12498
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5858,
      "step": 12499
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.1461,
      "step": 12500
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3531,
      "step": 12501
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.6465,
      "step": 12502
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2824,
      "step": 12503
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5351,
      "step": 12504
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 12505
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.5167,
      "step": 12506
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.3772,
      "step": 12507
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.6674,
      "step": 12508
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4218,
      "step": 12509
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.8338,
      "step": 12510
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.501,
      "step": 12511
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5617,
      "step": 12512
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 12513
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4263,
      "step": 12514
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.6113,
      "step": 12515
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4417,
      "step": 12516
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5161,
      "step": 12517
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 12518
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 12519
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4738,
      "step": 12520
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5304,
      "step": 12521
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5606,
      "step": 12522
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4227,
      "step": 12523
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3092,
      "step": 12524
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4561,
      "step": 12525
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5549,
      "step": 12526
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5166,
      "step": 12527
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4127,
      "step": 12528
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3522,
      "step": 12529
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5037,
      "step": 12530
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3176,
      "step": 12531
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5374,
      "step": 12532
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.7297,
      "step": 12533
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.307,
      "step": 12534
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5523,
      "step": 12535
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4538,
      "step": 12536
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4545,
      "step": 12537
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 12538
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.559,
      "step": 12539
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.7007,
      "step": 12540
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.377,
      "step": 12541
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.371,
      "step": 12542
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4028,
      "step": 12543
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.2482,
      "step": 12544
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4859,
      "step": 12545
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5395,
      "step": 12546
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3152,
      "step": 12547
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 12548
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3666,
      "step": 12549
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.353,
      "step": 12550
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3385,
      "step": 12551
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.399,
      "step": 12552
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4074,
      "step": 12553
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 12554
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.6764,
      "step": 12555
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.508,
      "step": 12556
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.441,
      "step": 12557
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4537,
      "step": 12558
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.47,
      "step": 12559
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5878,
      "step": 12560
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4368,
      "step": 12561
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4125,
      "step": 12562
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5244,
      "step": 12563
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.2725,
      "step": 12564
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.7037,
      "step": 12565
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.3695,
      "step": 12566
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.4941,
      "step": 12567
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.0002,
      "loss": 0.5779,
      "step": 12568
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3525,
      "step": 12569
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.695,
      "step": 12570
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 12571
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6219,
      "step": 12572
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.2984,
      "step": 12573
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6243,
      "step": 12574
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3213,
      "step": 12575
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6386,
      "step": 12576
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6316,
      "step": 12577
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 12578
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5014,
      "step": 12579
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5656,
      "step": 12580
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4499,
      "step": 12581
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5087,
      "step": 12582
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.2375,
      "step": 12583
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4244,
      "step": 12584
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5148,
      "step": 12585
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5906,
      "step": 12586
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3209,
      "step": 12587
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 12588
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3695,
      "step": 12589
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4899,
      "step": 12590
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6159,
      "step": 12591
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4405,
      "step": 12592
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3804,
      "step": 12593
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5312,
      "step": 12594
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6345,
      "step": 12595
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3932,
      "step": 12596
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3065,
      "step": 12597
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3137,
      "step": 12598
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.2812,
      "step": 12599
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.2496,
      "step": 12600
    },
    {
      "epoch": 2.07,
      "eval_loss": 0.8307493925094604,
      "eval_runtime": 131.6386,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 12600
    },
    {
      "epoch": 2.07,
      "mmlu_eval_accuracy": 0.5492129177562171,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.625,
      "mmlu_eval_accuracy_business_ethics": 0.2727272727272727,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.6818181818181818,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.5625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7692307692307693,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.6363636363636364,
      "mmlu_eval_accuracy_management": 0.8181818181818182,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.34,
      "mmlu_eval_accuracy_nutrition": 0.6666666666666666,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.6111111111111112,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1321871555155003,
      "step": 12600
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4873,
      "step": 12601
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3587,
      "step": 12602
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4281,
      "step": 12603
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6423,
      "step": 12604
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3975,
      "step": 12605
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5573,
      "step": 12606
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.1982,
      "step": 12607
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3356,
      "step": 12608
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.617,
      "step": 12609
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5932,
      "step": 12610
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5786,
      "step": 12611
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3186,
      "step": 12612
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 12613
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3787,
      "step": 12614
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3945,
      "step": 12615
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3673,
      "step": 12616
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4973,
      "step": 12617
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 12618
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.2982,
      "step": 12619
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4486,
      "step": 12620
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4972,
      "step": 12621
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4979,
      "step": 12622
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 12623
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5422,
      "step": 12624
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3761,
      "step": 12625
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.6843,
      "step": 12626
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.4496,
      "step": 12627
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.5468,
      "step": 12628
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3608,
      "step": 12629
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5283,
      "step": 12630
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3541,
      "step": 12631
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.605,
      "step": 12632
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.6443,
      "step": 12633
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4706,
      "step": 12634
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4797,
      "step": 12635
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2173,
      "step": 12636
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4017,
      "step": 12637
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5723,
      "step": 12638
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4358,
      "step": 12639
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4213,
      "step": 12640
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2641,
      "step": 12641
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4051,
      "step": 12642
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.6109,
      "step": 12643
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4584,
      "step": 12644
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3693,
      "step": 12645
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5421,
      "step": 12646
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.638,
      "step": 12647
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4537,
      "step": 12648
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2229,
      "step": 12649
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3528,
      "step": 12650
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4269,
      "step": 12651
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3875,
      "step": 12652
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5071,
      "step": 12653
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4827,
      "step": 12654
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4013,
      "step": 12655
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.517,
      "step": 12656
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.543,
      "step": 12657
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4827,
      "step": 12658
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4667,
      "step": 12659
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 12660
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4429,
      "step": 12661
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4537,
      "step": 12662
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.7205,
      "step": 12663
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2977,
      "step": 12664
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4704,
      "step": 12665
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5056,
      "step": 12666
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4592,
      "step": 12667
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4987,
      "step": 12668
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4544,
      "step": 12669
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3938,
      "step": 12670
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5325,
      "step": 12671
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5431,
      "step": 12672
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2881,
      "step": 12673
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5046,
      "step": 12674
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4192,
      "step": 12675
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 12676
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5542,
      "step": 12677
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4573,
      "step": 12678
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3848,
      "step": 12679
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4539,
      "step": 12680
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.3694,
      "step": 12681
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4645,
      "step": 12682
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5823,
      "step": 12683
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.7223,
      "step": 12684
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.4054,
      "step": 12685
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.625,
      "step": 12686
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.665,
      "step": 12687
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.2874,
      "step": 12688
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.0002,
      "loss": 0.5211,
      "step": 12689
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4345,
      "step": 12690
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4673,
      "step": 12691
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5745,
      "step": 12692
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5633,
      "step": 12693
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.2913,
      "step": 12694
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4462,
      "step": 12695
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6517,
      "step": 12696
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6,
      "step": 12697
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6331,
      "step": 12698
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4997,
      "step": 12699
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3,
      "step": 12700
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3532,
      "step": 12701
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3688,
      "step": 12702
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4378,
      "step": 12703
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4701,
      "step": 12704
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6222,
      "step": 12705
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4439,
      "step": 12706
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.663,
      "step": 12707
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4913,
      "step": 12708
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3941,
      "step": 12709
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4086,
      "step": 12710
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4658,
      "step": 12711
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6284,
      "step": 12712
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4606,
      "step": 12713
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3688,
      "step": 12714
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.343,
      "step": 12715
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4446,
      "step": 12716
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3474,
      "step": 12717
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3654,
      "step": 12718
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5003,
      "step": 12719
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3124,
      "step": 12720
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.7198,
      "step": 12721
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4538,
      "step": 12722
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3197,
      "step": 12723
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5044,
      "step": 12724
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3423,
      "step": 12725
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4175,
      "step": 12726
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5803,
      "step": 12727
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4926,
      "step": 12728
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5494,
      "step": 12729
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5047,
      "step": 12730
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5761,
      "step": 12731
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5253,
      "step": 12732
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6476,
      "step": 12733
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.3627,
      "step": 12734
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.2554,
      "step": 12735
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5322,
      "step": 12736
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5565,
      "step": 12737
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5084,
      "step": 12738
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6882,
      "step": 12739
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5068,
      "step": 12740
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4494,
      "step": 12741
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5182,
      "step": 12742
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5279,
      "step": 12743
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.6497,
      "step": 12744
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4541,
      "step": 12745
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4279,
      "step": 12746
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.4025,
      "step": 12747
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 12748
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.358,
      "step": 12749
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.415,
      "step": 12750
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5249,
      "step": 12751
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.606,
      "step": 12752
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4089,
      "step": 12753
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3367,
      "step": 12754
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5024,
      "step": 12755
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.7077,
      "step": 12756
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5268,
      "step": 12757
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4321,
      "step": 12758
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3831,
      "step": 12759
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3171,
      "step": 12760
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5484,
      "step": 12761
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.315,
      "step": 12762
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.492,
      "step": 12763
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.851,
      "step": 12764
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5693,
      "step": 12765
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4662,
      "step": 12766
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4655,
      "step": 12767
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3233,
      "step": 12768
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4432,
      "step": 12769
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.2796,
      "step": 12770
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3474,
      "step": 12771
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.463,
      "step": 12772
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5796,
      "step": 12773
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5339,
      "step": 12774
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3376,
      "step": 12775
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4906,
      "step": 12776
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3054,
      "step": 12777
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5045,
      "step": 12778
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4897,
      "step": 12779
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4991,
      "step": 12780
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.405,
      "step": 12781
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3877,
      "step": 12782
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5178,
      "step": 12783
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3604,
      "step": 12784
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4566,
      "step": 12785
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5723,
      "step": 12786
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5603,
      "step": 12787
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5365,
      "step": 12788
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3018,
      "step": 12789
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3952,
      "step": 12790
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4358,
      "step": 12791
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5678,
      "step": 12792
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4582,
      "step": 12793
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.6899,
      "step": 12794
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 12795
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.469,
      "step": 12796
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.2189,
      "step": 12797
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4191,
      "step": 12798
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.6822,
      "step": 12799
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.267,
      "step": 12800
    },
    {
      "epoch": 2.1,
      "eval_loss": 0.8207905292510986,
      "eval_runtime": 131.556,
      "eval_samples_per_second": 7.601,
      "eval_steps_per_second": 3.801,
      "step": 12800
    },
    {
      "epoch": 2.1,
      "mmlu_eval_accuracy": 0.5406187256416976,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.6875,
      "mmlu_eval_accuracy_business_ethics": 0.2727272727272727,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "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": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5116279069767442,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "mmlu_eval_accuracy_sociology": 0.8181818181818182,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.2683899789737347,
      "step": 12800
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5963,
      "step": 12801
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3997,
      "step": 12802
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3025,
      "step": 12803
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4688,
      "step": 12804
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 12805
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3404,
      "step": 12806
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3576,
      "step": 12807
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4607,
      "step": 12808
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.4547,
      "step": 12809
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.662,
      "step": 12810
    },
    {
      "epoch": 2.1,
      "learning_rate": 0.0002,
      "loss": 0.3023,
      "step": 12811
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.8591,
      "step": 12812
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3861,
      "step": 12813
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3665,
      "step": 12814
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4764,
      "step": 12815
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3341,
      "step": 12816
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4597,
      "step": 12817
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3032,
      "step": 12818
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3673,
      "step": 12819
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5726,
      "step": 12820
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4128,
      "step": 12821
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.298,
      "step": 12822
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.587,
      "step": 12823
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6377,
      "step": 12824
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.2822,
      "step": 12825
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3135,
      "step": 12826
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4589,
      "step": 12827
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6022,
      "step": 12828
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 12829
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.7801,
      "step": 12830
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4615,
      "step": 12831
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4185,
      "step": 12832
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.8003,
      "step": 12833
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3032,
      "step": 12834
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4638,
      "step": 12835
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6523,
      "step": 12836
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5733,
      "step": 12837
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4221,
      "step": 12838
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4215,
      "step": 12839
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3729,
      "step": 12840
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6497,
      "step": 12841
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5672,
      "step": 12842
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6498,
      "step": 12843
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4641,
      "step": 12844
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4358,
      "step": 12845
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4079,
      "step": 12846
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4435,
      "step": 12847
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.481,
      "step": 12848
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5098,
      "step": 12849
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5258,
      "step": 12850
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5934,
      "step": 12851
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4664,
      "step": 12852
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6753,
      "step": 12853
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4642,
      "step": 12854
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.2011,
      "step": 12855
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4329,
      "step": 12856
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.1583,
      "step": 12857
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6481,
      "step": 12858
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.6143,
      "step": 12859
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5199,
      "step": 12860
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4728,
      "step": 12861
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.2553,
      "step": 12862
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3259,
      "step": 12863
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3167,
      "step": 12864
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3895,
      "step": 12865
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5729,
      "step": 12866
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.5965,
      "step": 12867
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4449,
      "step": 12868
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4779,
      "step": 12869
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4916,
      "step": 12870
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.7511,
      "step": 12871
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 12872
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4278,
      "step": 12873
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4243,
      "step": 12874
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.207,
      "step": 12875
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4544,
      "step": 12876
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.8105,
      "step": 12877
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.2069,
      "step": 12878
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4431,
      "step": 12879
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 12880
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6807,
      "step": 12881
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4366,
      "step": 12882
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5644,
      "step": 12883
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4039,
      "step": 12884
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6755,
      "step": 12885
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4107,
      "step": 12886
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6416,
      "step": 12887
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.45,
      "step": 12888
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.2538,
      "step": 12889
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4878,
      "step": 12890
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.2855,
      "step": 12891
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.409,
      "step": 12892
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.7408,
      "step": 12893
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3738,
      "step": 12894
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4672,
      "step": 12895
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5094,
      "step": 12896
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.7783,
      "step": 12897
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3518,
      "step": 12898
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4287,
      "step": 12899
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4213,
      "step": 12900
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5643,
      "step": 12901
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 12902
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6943,
      "step": 12903
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6715,
      "step": 12904
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3593,
      "step": 12905
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4543,
      "step": 12906
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3848,
      "step": 12907
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5254,
      "step": 12908
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4208,
      "step": 12909
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5894,
      "step": 12910
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4365,
      "step": 12911
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5791,
      "step": 12912
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.6624,
      "step": 12913
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.358,
      "step": 12914
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4309,
      "step": 12915
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3647,
      "step": 12916
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.7063,
      "step": 12917
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.8159,
      "step": 12918
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.7639,
      "step": 12919
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5687,
      "step": 12920
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.585,
      "step": 12921
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4215,
      "step": 12922
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4357,
      "step": 12923
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.9044,
      "step": 12924
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4047,
      "step": 12925
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5815,
      "step": 12926
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5281,
      "step": 12927
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3998,
      "step": 12928
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.553,
      "step": 12929
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.2618,
      "step": 12930
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.5197,
      "step": 12931
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.4823,
      "step": 12932
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.435,
      "step": 12933
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.2663,
      "step": 12934
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.6435,
      "step": 12935
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.2828,
      "step": 12936
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 12937
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3953,
      "step": 12938
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.685,
      "step": 12939
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4123,
      "step": 12940
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4552,
      "step": 12941
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.6148,
      "step": 12942
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4127,
      "step": 12943
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3634,
      "step": 12944
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 12945
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4321,
      "step": 12946
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5492,
      "step": 12947
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3463,
      "step": 12948
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.2805,
      "step": 12949
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4866,
      "step": 12950
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4949,
      "step": 12951
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5018,
      "step": 12952
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5982,
      "step": 12953
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 12954
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.416,
      "step": 12955
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.34,
      "step": 12956
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4803,
      "step": 12957
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5195,
      "step": 12958
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4625,
      "step": 12959
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5366,
      "step": 12960
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.709,
      "step": 12961
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5316,
      "step": 12962
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3891,
      "step": 12963
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3755,
      "step": 12964
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4559,
      "step": 12965
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4,
      "step": 12966
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5777,
      "step": 12967
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5116,
      "step": 12968
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 12969
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5143,
      "step": 12970
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3958,
      "step": 12971
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4877,
      "step": 12972
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4201,
      "step": 12973
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5358,
      "step": 12974
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.473,
      "step": 12975
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3172,
      "step": 12976
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.6392,
      "step": 12977
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4058,
      "step": 12978
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4927,
      "step": 12979
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.5076,
      "step": 12980
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4339,
      "step": 12981
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.6629,
      "step": 12982
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4397,
      "step": 12983
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.8752,
      "step": 12984
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.554,
      "step": 12985
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.588,
      "step": 12986
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4952,
      "step": 12987
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4505,
      "step": 12988
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3515,
      "step": 12989
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3154,
      "step": 12990
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.7052,
      "step": 12991
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.4091,
      "step": 12992
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.3855,
      "step": 12993
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002,
      "loss": 0.7889,
      "step": 12994
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5164,
      "step": 12995
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6165,
      "step": 12996
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.7982,
      "step": 12997
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.434,
      "step": 12998
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3577,
      "step": 12999
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3306,
      "step": 13000
    },
    {
      "epoch": 2.14,
      "eval_loss": 0.8304842114448547,
      "eval_runtime": 131.6891,
      "eval_samples_per_second": 7.594,
      "eval_steps_per_second": 3.797,
      "step": 13000
    },
    {
      "epoch": 2.14,
      "mmlu_eval_accuracy": 0.526689412607944,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.2727272727272727,
      "mmlu_eval_accuracy_clinical_knowledge": 0.6206896551724138,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.6363636363636364,
      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.5384615384615384,
      "mmlu_eval_accuracy_econometrics": 0.3333333333333333,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.59375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.7307692307692307,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8333333333333334,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.8,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.7093023255813954,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.6363636363636364,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.37058823529411766,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6666666666666666,
      "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": 0.631578947368421,
      "mmlu_loss": 1.3841386617358298,
      "step": 13000
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.7586,
      "step": 13001
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3577,
      "step": 13002
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5618,
      "step": 13003
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.2744,
      "step": 13004
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3956,
      "step": 13005
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4637,
      "step": 13006
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.2925,
      "step": 13007
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.279,
      "step": 13008
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4177,
      "step": 13009
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6259,
      "step": 13010
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4167,
      "step": 13011
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.7168,
      "step": 13012
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5527,
      "step": 13013
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3688,
      "step": 13014
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4905,
      "step": 13015
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5001,
      "step": 13016
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3221,
      "step": 13017
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6322,
      "step": 13018
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 13019
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4246,
      "step": 13020
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3453,
      "step": 13021
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4353,
      "step": 13022
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3203,
      "step": 13023
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.402,
      "step": 13024
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.627,
      "step": 13025
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3847,
      "step": 13026
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.423,
      "step": 13027
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.2564,
      "step": 13028
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.58,
      "step": 13029
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5607,
      "step": 13030
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4032,
      "step": 13031
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5521,
      "step": 13032
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.363,
      "step": 13033
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3686,
      "step": 13034
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4935,
      "step": 13035
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4266,
      "step": 13036
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3642,
      "step": 13037
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4976,
      "step": 13038
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6472,
      "step": 13039
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6282,
      "step": 13040
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3808,
      "step": 13041
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4115,
      "step": 13042
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3631,
      "step": 13043
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5318,
      "step": 13044
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5811,
      "step": 13045
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.6101,
      "step": 13046
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.7407,
      "step": 13047
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.411,
      "step": 13048
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5679,
      "step": 13049
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3562,
      "step": 13050
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3622,
      "step": 13051
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5498,
      "step": 13052
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.4033,
      "step": 13053
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.381,
      "step": 13054
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.5946,
      "step": 13055
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5468,
      "step": 13056
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5444,
      "step": 13057
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5099,
      "step": 13058
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4441,
      "step": 13059
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4576,
      "step": 13060
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5312,
      "step": 13061
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5224,
      "step": 13062
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5134,
      "step": 13063
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6471,
      "step": 13064
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3501,
      "step": 13065
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6803,
      "step": 13066
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5234,
      "step": 13067
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.473,
      "step": 13068
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4451,
      "step": 13069
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5955,
      "step": 13070
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 13071
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3835,
      "step": 13072
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.48,
      "step": 13073
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3085,
      "step": 13074
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6084,
      "step": 13075
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5095,
      "step": 13076
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 13077
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.2902,
      "step": 13078
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 13079
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 13080
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6364,
      "step": 13081
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.7093,
      "step": 13082
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5429,
      "step": 13083
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3969,
      "step": 13084
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.366,
      "step": 13085
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5148,
      "step": 13086
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4812,
      "step": 13087
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4266,
      "step": 13088
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4193,
      "step": 13089
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4895,
      "step": 13090
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5665,
      "step": 13091
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5343,
      "step": 13092
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5084,
      "step": 13093
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6558,
      "step": 13094
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4795,
      "step": 13095
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4437,
      "step": 13096
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5817,
      "step": 13097
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.6267,
      "step": 13098
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.497,
      "step": 13099
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5658,
      "step": 13100
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3268,
      "step": 13101
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5905,
      "step": 13102
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.2424,
      "step": 13103
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.486,
      "step": 13104
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3669,
      "step": 13105
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4886,
      "step": 13106
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.2836,
      "step": 13107
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5224,
      "step": 13108
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.5666,
      "step": 13109
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.49,
      "step": 13110
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 13111
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.3715,
      "step": 13112
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.4409,
      "step": 13113
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.2813,
      "step": 13114
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.0002,
      "loss": 0.303,
      "step": 13115
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5911,
      "step": 13116
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4444,
      "step": 13117
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4617,
      "step": 13118
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3974,
      "step": 13119
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.2525,
      "step": 13120
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5342,
      "step": 13121
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4239,
      "step": 13122
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3337,
      "step": 13123
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6519,
      "step": 13124
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3699,
      "step": 13125
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3162,
      "step": 13126
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5806,
      "step": 13127
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5757,
      "step": 13128
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4291,
      "step": 13129
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5431,
      "step": 13130
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4484,
      "step": 13131
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.2868,
      "step": 13132
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3693,
      "step": 13133
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5268,
      "step": 13134
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.7016,
      "step": 13135
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6868,
      "step": 13136
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4255,
      "step": 13137
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.549,
      "step": 13138
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.7279,
      "step": 13139
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4211,
      "step": 13140
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5316,
      "step": 13141
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3782,
      "step": 13142
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4273,
      "step": 13143
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4174,
      "step": 13144
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4859,
      "step": 13145
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5839,
      "step": 13146
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4357,
      "step": 13147
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.7391,
      "step": 13148
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5881,
      "step": 13149
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4143,
      "step": 13150
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3391,
      "step": 13151
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6585,
      "step": 13152
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6838,
      "step": 13153
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3884,
      "step": 13154
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4077,
      "step": 13155
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.2619,
      "step": 13156
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3705,
      "step": 13157
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3353,
      "step": 13158
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4109,
      "step": 13159
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3499,
      "step": 13160
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.2678,
      "step": 13161
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4206,
      "step": 13162
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5531,
      "step": 13163
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.624,
      "step": 13164
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6291,
      "step": 13165
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3985,
      "step": 13166
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.4124,
      "step": 13167
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.7105,
      "step": 13168
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3852,
      "step": 13169
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5919,
      "step": 13170
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.6296,
      "step": 13171
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3674,
      "step": 13172
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.5142,
      "step": 13173
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3955,
      "step": 13174
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3809,
      "step": 13175
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3973,
      "step": 13176
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4698,
      "step": 13177
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.399,
      "step": 13178
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3775,
      "step": 13179
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4215,
      "step": 13180
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3352,
      "step": 13181
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.2161,
      "step": 13182
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4707,
      "step": 13183
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4434,
      "step": 13184
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5245,
      "step": 13185
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4259,
      "step": 13186
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4913,
      "step": 13187
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6295,
      "step": 13188
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6129,
      "step": 13189
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6483,
      "step": 13190
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.479,
      "step": 13191
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5161,
      "step": 13192
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5924,
      "step": 13193
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3485,
      "step": 13194
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5545,
      "step": 13195
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5778,
      "step": 13196
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.2029,
      "step": 13197
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5914,
      "step": 13198
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5286,
      "step": 13199
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6879,
      "step": 13200
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.8167378902435303,
      "eval_runtime": 131.4945,
      "eval_samples_per_second": 7.605,
      "eval_steps_per_second": 3.802,
      "step": 13200
    },
    {
      "epoch": 2.17,
      "mmlu_eval_accuracy": 0.546961808016742,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "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": 0.46875,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.7222222222222222,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7619047619047619,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "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": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.37,
      "mmlu_eval_accuracy_nutrition": 0.7272727272727273,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "mmlu_eval_accuracy_professional_psychology": 0.5217391304347826,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.6666666666666666,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.4099295307972102,
      "step": 13200
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 13201
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.7033,
      "step": 13202
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6155,
      "step": 13203
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4182,
      "step": 13204
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3587,
      "step": 13205
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.556,
      "step": 13206
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.557,
      "step": 13207
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4277,
      "step": 13208
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4155,
      "step": 13209
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4253,
      "step": 13210
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5222,
      "step": 13211
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.2605,
      "step": 13212
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.7332,
      "step": 13213
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5276,
      "step": 13214
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.2244,
      "step": 13215
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.2619,
      "step": 13216
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5892,
      "step": 13217
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3602,
      "step": 13218
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5926,
      "step": 13219
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5431,
      "step": 13220
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3986,
      "step": 13221
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3891,
      "step": 13222
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4066,
      "step": 13223
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4807,
      "step": 13224
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5153,
      "step": 13225
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3683,
      "step": 13226
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4202,
      "step": 13227
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4426,
      "step": 13228
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6759,
      "step": 13229
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3764,
      "step": 13230
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.6416,
      "step": 13231
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3634,
      "step": 13232
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3788,
      "step": 13233
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.4722,
      "step": 13234
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.3507,
      "step": 13235
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.5292,
      "step": 13236
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002,
      "loss": 0.7225,
      "step": 13237
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4251,
      "step": 13238
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4291,
      "step": 13239
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4769,
      "step": 13240
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5402,
      "step": 13241
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5359,
      "step": 13242
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3964,
      "step": 13243
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 13244
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.8491,
      "step": 13245
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3841,
      "step": 13246
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 13247
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.466,
      "step": 13248
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4413,
      "step": 13249
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4829,
      "step": 13250
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4722,
      "step": 13251
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4257,
      "step": 13252
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5414,
      "step": 13253
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3917,
      "step": 13254
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3263,
      "step": 13255
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5591,
      "step": 13256
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.7028,
      "step": 13257
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4304,
      "step": 13258
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.6052,
      "step": 13259
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3767,
      "step": 13260
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.6737,
      "step": 13261
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4346,
      "step": 13262
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3922,
      "step": 13263
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5044,
      "step": 13264
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5243,
      "step": 13265
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.8638,
      "step": 13266
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.6217,
      "step": 13267
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5044,
      "step": 13268
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 13269
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5225,
      "step": 13270
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3094,
      "step": 13271
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5739,
      "step": 13272
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5416,
      "step": 13273
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4496,
      "step": 13274
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5789,
      "step": 13275
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5273,
      "step": 13276
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4486,
      "step": 13277
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5313,
      "step": 13278
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4532,
      "step": 13279
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3929,
      "step": 13280
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3686,
      "step": 13281
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5758,
      "step": 13282
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5182,
      "step": 13283
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5108,
      "step": 13284
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3884,
      "step": 13285
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4661,
      "step": 13286
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.375,
      "step": 13287
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4206,
      "step": 13288
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.3262,
      "step": 13289
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.471,
      "step": 13290
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4157,
      "step": 13291
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4644,
      "step": 13292
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.6348,
      "step": 13293
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5098,
      "step": 13294
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5283,
      "step": 13295
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.718,
      "step": 13296
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.5139,
      "step": 13297
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.4976,
      "step": 13298
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4673,
      "step": 13299
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4639,
      "step": 13300
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3602,
      "step": 13301
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6564,
      "step": 13302
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5719,
      "step": 13303
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.2528,
      "step": 13304
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4174,
      "step": 13305
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5012,
      "step": 13306
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5059,
      "step": 13307
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3004,
      "step": 13308
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.2794,
      "step": 13309
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5864,
      "step": 13310
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4878,
      "step": 13311
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6939,
      "step": 13312
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4332,
      "step": 13313
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5923,
      "step": 13314
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6095,
      "step": 13315
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6054,
      "step": 13316
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5448,
      "step": 13317
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5377,
      "step": 13318
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.294,
      "step": 13319
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3471,
      "step": 13320
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5085,
      "step": 13321
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4411,
      "step": 13322
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5625,
      "step": 13323
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5417,
      "step": 13324
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.9838,
      "step": 13325
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4405,
      "step": 13326
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3006,
      "step": 13327
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4477,
      "step": 13328
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3991,
      "step": 13329
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5398,
      "step": 13330
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4695,
      "step": 13331
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3412,
      "step": 13332
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5763,
      "step": 13333
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6789,
      "step": 13334
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.381,
      "step": 13335
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5686,
      "step": 13336
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4243,
      "step": 13337
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4928,
      "step": 13338
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.7095,
      "step": 13339
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4557,
      "step": 13340
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 13341
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4019,
      "step": 13342
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4106,
      "step": 13343
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5823,
      "step": 13344
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3679,
      "step": 13345
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5223,
      "step": 13346
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4151,
      "step": 13347
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3552,
      "step": 13348
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3287,
      "step": 13349
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3668,
      "step": 13350
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 13351
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5072,
      "step": 13352
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.4118,
      "step": 13353
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5311,
      "step": 13354
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5071,
      "step": 13355
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3302,
      "step": 13356
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.2857,
      "step": 13357
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.5755,
      "step": 13358
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.6563,
      "step": 13359
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.6497,
      "step": 13360
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3782,
      "step": 13361
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5449,
      "step": 13362
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.599,
      "step": 13363
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.611,
      "step": 13364
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.601,
      "step": 13365
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4775,
      "step": 13366
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.593,
      "step": 13367
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3116,
      "step": 13368
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4382,
      "step": 13369
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.748,
      "step": 13370
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5227,
      "step": 13371
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3012,
      "step": 13372
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5409,
      "step": 13373
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4385,
      "step": 13374
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4869,
      "step": 13375
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4137,
      "step": 13376
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.2955,
      "step": 13377
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.6523,
      "step": 13378
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4161,
      "step": 13379
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 13380
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3383,
      "step": 13381
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5638,
      "step": 13382
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.2675,
      "step": 13383
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3525,
      "step": 13384
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5504,
      "step": 13385
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.311,
      "step": 13386
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5736,
      "step": 13387
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4326,
      "step": 13388
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5881,
      "step": 13389
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4195,
      "step": 13390
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4482,
      "step": 13391
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 13392
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4515,
      "step": 13393
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.328,
      "step": 13394
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4965,
      "step": 13395
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.8345,
      "step": 13396
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5775,
      "step": 13397
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5523,
      "step": 13398
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3126,
      "step": 13399
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5027,
      "step": 13400
    },
    {
      "epoch": 2.2,
      "eval_loss": 0.8190361857414246,
      "eval_runtime": 131.6327,
      "eval_samples_per_second": 7.597,
      "eval_steps_per_second": 3.798,
      "step": 13400
    },
    {
      "epoch": 2.2,
      "mmlu_eval_accuracy": 0.5179274837889997,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.36363636363636365,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.6875,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.5,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4883720930232558,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5769230769230769,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.8166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.21739130434782608,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.9230769230769231,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.45454545454545453,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.6052631578947368,
      "mmlu_eval_accuracy_moral_scenarios": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.4117647058823529,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "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": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.8181818181818182,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.0473954442003377,
      "step": 13400
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.6319,
      "step": 13401
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4866,
      "step": 13402
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.2754,
      "step": 13403
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.7789,
      "step": 13404
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4152,
      "step": 13405
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4731,
      "step": 13406
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4106,
      "step": 13407
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 13408
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4217,
      "step": 13409
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.6351,
      "step": 13410
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.6693,
      "step": 13411
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3908,
      "step": 13412
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5668,
      "step": 13413
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4846,
      "step": 13414
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5933,
      "step": 13415
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.341,
      "step": 13416
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 13417
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.5749,
      "step": 13418
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.4453,
      "step": 13419
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.0002,
      "loss": 0.3242,
      "step": 13420
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5812,
      "step": 13421
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 13422
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4179,
      "step": 13423
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 13424
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3553,
      "step": 13425
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5322,
      "step": 13426
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 13427
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3227,
      "step": 13428
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6051,
      "step": 13429
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4573,
      "step": 13430
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4363,
      "step": 13431
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5168,
      "step": 13432
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4851,
      "step": 13433
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4552,
      "step": 13434
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4338,
      "step": 13435
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6174,
      "step": 13436
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5533,
      "step": 13437
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6505,
      "step": 13438
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4544,
      "step": 13439
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 13440
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3278,
      "step": 13441
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5561,
      "step": 13442
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.7195,
      "step": 13443
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5391,
      "step": 13444
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3291,
      "step": 13445
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4434,
      "step": 13446
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 13447
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4425,
      "step": 13448
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3419,
      "step": 13449
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4712,
      "step": 13450
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5161,
      "step": 13451
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5064,
      "step": 13452
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5507,
      "step": 13453
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.2827,
      "step": 13454
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4795,
      "step": 13455
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3031,
      "step": 13456
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3731,
      "step": 13457
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 13458
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5066,
      "step": 13459
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6856,
      "step": 13460
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4218,
      "step": 13461
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3821,
      "step": 13462
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 13463
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.2643,
      "step": 13464
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6326,
      "step": 13465
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4031,
      "step": 13466
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5523,
      "step": 13467
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3172,
      "step": 13468
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4465,
      "step": 13469
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6363,
      "step": 13470
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6252,
      "step": 13471
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4112,
      "step": 13472
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.4403,
      "step": 13473
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5675,
      "step": 13474
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.413,
      "step": 13475
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5374,
      "step": 13476
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3681,
      "step": 13477
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.3,
      "step": 13478
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5093,
      "step": 13479
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.6506,
      "step": 13480
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.5992,
      "step": 13481
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 13482
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4446,
      "step": 13483
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5609,
      "step": 13484
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3519,
      "step": 13485
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5316,
      "step": 13486
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3662,
      "step": 13487
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6556,
      "step": 13488
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6648,
      "step": 13489
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.483,
      "step": 13490
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.408,
      "step": 13491
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4951,
      "step": 13492
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6132,
      "step": 13493
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.7309,
      "step": 13494
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3966,
      "step": 13495
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 13496
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5242,
      "step": 13497
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.8395,
      "step": 13498
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6619,
      "step": 13499
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.2831,
      "step": 13500
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.456,
      "step": 13501
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4397,
      "step": 13502
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5724,
      "step": 13503
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.8714,
      "step": 13504
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5958,
      "step": 13505
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.65,
      "step": 13506
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5912,
      "step": 13507
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.519,
      "step": 13508
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5038,
      "step": 13509
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.2378,
      "step": 13510
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4995,
      "step": 13511
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5141,
      "step": 13512
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6516,
      "step": 13513
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.597,
      "step": 13514
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3985,
      "step": 13515
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4211,
      "step": 13516
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.667,
      "step": 13517
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6068,
      "step": 13518
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5949,
      "step": 13519
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3934,
      "step": 13520
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6421,
      "step": 13521
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5278,
      "step": 13522
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5785,
      "step": 13523
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5358,
      "step": 13524
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4332,
      "step": 13525
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5665,
      "step": 13526
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 13527
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4157,
      "step": 13528
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.2101,
      "step": 13529
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3035,
      "step": 13530
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5381,
      "step": 13531
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.6988,
      "step": 13532
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5987,
      "step": 13533
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3845,
      "step": 13534
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3525,
      "step": 13535
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.4664,
      "step": 13536
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.3904,
      "step": 13537
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5183,
      "step": 13538
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 13539
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.506,
      "step": 13540
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002,
      "loss": 0.5979,
      "step": 13541
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4354,
      "step": 13542
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5489,
      "step": 13543
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4752,
      "step": 13544
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7782,
      "step": 13545
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4474,
      "step": 13546
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.6808,
      "step": 13547
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3989,
      "step": 13548
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5511,
      "step": 13549
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5443,
      "step": 13550
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4411,
      "step": 13551
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4639,
      "step": 13552
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7697,
      "step": 13553
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7345,
      "step": 13554
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7942,
      "step": 13555
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4226,
      "step": 13556
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4037,
      "step": 13557
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5633,
      "step": 13558
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3977,
      "step": 13559
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 13560
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4393,
      "step": 13561
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5056,
      "step": 13562
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3922,
      "step": 13563
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3082,
      "step": 13564
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4166,
      "step": 13565
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3398,
      "step": 13566
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.2883,
      "step": 13567
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5423,
      "step": 13568
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4481,
      "step": 13569
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4615,
      "step": 13570
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5407,
      "step": 13571
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.465,
      "step": 13572
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3043,
      "step": 13573
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4609,
      "step": 13574
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7449,
      "step": 13575
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.378,
      "step": 13576
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.2224,
      "step": 13577
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.2917,
      "step": 13578
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3335,
      "step": 13579
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7246,
      "step": 13580
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4647,
      "step": 13581
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5339,
      "step": 13582
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5367,
      "step": 13583
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.6639,
      "step": 13584
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4233,
      "step": 13585
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4576,
      "step": 13586
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4454,
      "step": 13587
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4059,
      "step": 13588
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.653,
      "step": 13589
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5081,
      "step": 13590
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3628,
      "step": 13591
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7283,
      "step": 13592
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.752,
      "step": 13593
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 13594
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.5651,
      "step": 13595
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.6401,
      "step": 13596
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.2928,
      "step": 13597
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3844,
      "step": 13598
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.6962,
      "step": 13599
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3786,
      "step": 13600
    },
    {
      "epoch": 2.23,
      "eval_loss": 0.8203914165496826,
      "eval_runtime": 131.622,
      "eval_samples_per_second": 7.598,
      "eval_steps_per_second": 3.799,
      "step": 13600
    },
    {
      "epoch": 2.23,
      "mmlu_eval_accuracy": 0.5097313810249351,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "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": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.75,
      "mmlu_eval_accuracy_high_school_statistics": 0.21739130434782608,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.84,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.5813953488372093,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.29,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_law": 0.3,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.3471037365165455,
      "step": 13600
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.4866,
      "step": 13601
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.464,
      "step": 13602
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4367,
      "step": 13603
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5845,
      "step": 13604
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4557,
      "step": 13605
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4793,
      "step": 13606
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3797,
      "step": 13607
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4511,
      "step": 13608
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4031,
      "step": 13609
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 13610
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4442,
      "step": 13611
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.2665,
      "step": 13612
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5463,
      "step": 13613
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5673,
      "step": 13614
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.1646,
      "step": 13615
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5568,
      "step": 13616
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4344,
      "step": 13617
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.462,
      "step": 13618
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5164,
      "step": 13619
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3054,
      "step": 13620
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5181,
      "step": 13621
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.337,
      "step": 13622
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 13623
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.2728,
      "step": 13624
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3176,
      "step": 13625
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5191,
      "step": 13626
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4087,
      "step": 13627
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3479,
      "step": 13628
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5442,
      "step": 13629
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5872,
      "step": 13630
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3912,
      "step": 13631
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4576,
      "step": 13632
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3631,
      "step": 13633
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.194,
      "step": 13634
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3144,
      "step": 13635
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5856,
      "step": 13636
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3208,
      "step": 13637
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.6634,
      "step": 13638
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4929,
      "step": 13639
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5205,
      "step": 13640
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.2824,
      "step": 13641
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4955,
      "step": 13642
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5174,
      "step": 13643
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.574,
      "step": 13644
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.6521,
      "step": 13645
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5529,
      "step": 13646
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.579,
      "step": 13647
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5951,
      "step": 13648
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.6064,
      "step": 13649
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4779,
      "step": 13650
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3587,
      "step": 13651
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.6243,
      "step": 13652
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.4928,
      "step": 13653
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5068,
      "step": 13654
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5738,
      "step": 13655
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5657,
      "step": 13656
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5202,
      "step": 13657
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.541,
      "step": 13658
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5694,
      "step": 13659
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.7978,
      "step": 13660
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.5865,
      "step": 13661
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.3616,
      "step": 13662
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0002,
      "loss": 0.1831,
      "step": 13663
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5434,
      "step": 13664
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5171,
      "step": 13665
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.426,
      "step": 13666
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4167,
      "step": 13667
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6352,
      "step": 13668
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4841,
      "step": 13669
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3334,
      "step": 13670
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2933,
      "step": 13671
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3054,
      "step": 13672
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4711,
      "step": 13673
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5664,
      "step": 13674
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6036,
      "step": 13675
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4909,
      "step": 13676
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4918,
      "step": 13677
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 1.0698,
      "step": 13678
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3257,
      "step": 13679
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7844,
      "step": 13680
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3618,
      "step": 13681
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5217,
      "step": 13682
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5026,
      "step": 13683
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5498,
      "step": 13684
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4816,
      "step": 13685
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5502,
      "step": 13686
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.461,
      "step": 13687
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.8374,
      "step": 13688
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7857,
      "step": 13689
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7997,
      "step": 13690
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6233,
      "step": 13691
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3231,
      "step": 13692
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4685,
      "step": 13693
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4445,
      "step": 13694
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5195,
      "step": 13695
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.8513,
      "step": 13696
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.1955,
      "step": 13697
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2943,
      "step": 13698
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6188,
      "step": 13699
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6346,
      "step": 13700
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5061,
      "step": 13701
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.378,
      "step": 13702
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5867,
      "step": 13703
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3811,
      "step": 13704
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2283,
      "step": 13705
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.4616,
      "step": 13706
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5462,
      "step": 13707
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2471,
      "step": 13708
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7287,
      "step": 13709
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3471,
      "step": 13710
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2832,
      "step": 13711
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3706,
      "step": 13712
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6609,
      "step": 13713
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7465,
      "step": 13714
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.5954,
      "step": 13715
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.389,
      "step": 13716
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.7319,
      "step": 13717
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6541,
      "step": 13718
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6376,
      "step": 13719
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3263,
      "step": 13720
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.2482,
      "step": 13721
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6848,
      "step": 13722
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3799,
      "step": 13723
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.6099,
      "step": 13724
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3944,
      "step": 13725
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5103,
      "step": 13726
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6884,
      "step": 13727
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3805,
      "step": 13728
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5137,
      "step": 13729
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3548,
      "step": 13730
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5054,
      "step": 13731
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5426,
      "step": 13732
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4541,
      "step": 13733
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6858,
      "step": 13734
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4993,
      "step": 13735
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 13736
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4814,
      "step": 13737
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6685,
      "step": 13738
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4571,
      "step": 13739
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5189,
      "step": 13740
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 13741
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.7502,
      "step": 13742
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6007,
      "step": 13743
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.2596,
      "step": 13744
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4605,
      "step": 13745
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5188,
      "step": 13746
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4344,
      "step": 13747
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4135,
      "step": 13748
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3383,
      "step": 13749
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3142,
      "step": 13750
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4882,
      "step": 13751
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.518,
      "step": 13752
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 13753
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.8259,
      "step": 13754
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.2948,
      "step": 13755
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6859,
      "step": 13756
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.493,
      "step": 13757
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3314,
      "step": 13758
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.2774,
      "step": 13759
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4199,
      "step": 13760
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4891,
      "step": 13761
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4979,
      "step": 13762
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4157,
      "step": 13763
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6635,
      "step": 13764
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4058,
      "step": 13765
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.7004,
      "step": 13766
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5352,
      "step": 13767
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5255,
      "step": 13768
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6586,
      "step": 13769
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.7238,
      "step": 13770
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4867,
      "step": 13771
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6082,
      "step": 13772
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5329,
      "step": 13773
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3246,
      "step": 13774
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.5312,
      "step": 13775
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4013,
      "step": 13776
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.8597,
      "step": 13777
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4151,
      "step": 13778
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6149,
      "step": 13779
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6907,
      "step": 13780
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3318,
      "step": 13781
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.2618,
      "step": 13782
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.4353,
      "step": 13783
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.6139,
      "step": 13784
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3066,
      "step": 13785
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5437,
      "step": 13786
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4832,
      "step": 13787
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4541,
      "step": 13788
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.55,
      "step": 13789
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4712,
      "step": 13790
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6452,
      "step": 13791
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5378,
      "step": 13792
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5901,
      "step": 13793
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6029,
      "step": 13794
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.7227,
      "step": 13795
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3935,
      "step": 13796
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3955,
      "step": 13797
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5062,
      "step": 13798
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3886,
      "step": 13799
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5736,
      "step": 13800
    },
    {
      "epoch": 2.27,
      "eval_loss": 0.8203531503677368,
      "eval_runtime": 131.6431,
      "eval_samples_per_second": 7.596,
      "eval_steps_per_second": 3.798,
      "step": 13800
    },
    {
      "epoch": 2.27,
      "mmlu_eval_accuracy": 0.5270603285957762,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.6363636363636364,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.4166666666666667,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.4375,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.5348837209302325,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.8,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "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": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
      "mmlu_eval_accuracy_machine_learning": 0.5454545454545454,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.88,
      "mmlu_eval_accuracy_medical_genetics": 1.0,
      "mmlu_eval_accuracy_miscellaneous": 0.6046511627906976,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.36,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5882352941176471,
      "mmlu_eval_accuracy_prehistory": 0.4,
      "mmlu_eval_accuracy_professional_accounting": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
      "mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
      "mmlu_eval_accuracy_professional_psychology": 0.5362318840579711,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "mmlu_eval_accuracy_sociology": 0.8636363636363636,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.3674226646012488,
      "step": 13800
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4828,
      "step": 13801
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4727,
      "step": 13802
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5375,
      "step": 13803
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4413,
      "step": 13804
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4704,
      "step": 13805
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3316,
      "step": 13806
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4259,
      "step": 13807
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5383,
      "step": 13808
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4044,
      "step": 13809
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.2917,
      "step": 13810
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4157,
      "step": 13811
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3164,
      "step": 13812
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6818,
      "step": 13813
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4019,
      "step": 13814
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5734,
      "step": 13815
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6224,
      "step": 13816
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4985,
      "step": 13817
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.8194,
      "step": 13818
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3659,
      "step": 13819
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.7544,
      "step": 13820
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5303,
      "step": 13821
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4118,
      "step": 13822
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4269,
      "step": 13823
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.2261,
      "step": 13824
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3579,
      "step": 13825
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3795,
      "step": 13826
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.1927,
      "step": 13827
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4889,
      "step": 13828
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5553,
      "step": 13829
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3645,
      "step": 13830
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5059,
      "step": 13831
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.49,
      "step": 13832
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4579,
      "step": 13833
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6481,
      "step": 13834
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6608,
      "step": 13835
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.7267,
      "step": 13836
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.373,
      "step": 13837
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.4842,
      "step": 13838
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5437,
      "step": 13839
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6746,
      "step": 13840
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5723,
      "step": 13841
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.3477,
      "step": 13842
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.5229,
      "step": 13843
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6002,
      "step": 13844
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6007,
      "step": 13845
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.0002,
      "loss": 0.6323,
      "step": 13846
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5394,
      "step": 13847
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3593,
      "step": 13848
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5841,
      "step": 13849
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4972,
      "step": 13850
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4628,
      "step": 13851
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6083,
      "step": 13852
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.7744,
      "step": 13853
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5605,
      "step": 13854
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3501,
      "step": 13855
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4049,
      "step": 13856
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5125,
      "step": 13857
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6085,
      "step": 13858
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6082,
      "step": 13859
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6029,
      "step": 13860
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3688,
      "step": 13861
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5425,
      "step": 13862
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5656,
      "step": 13863
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6845,
      "step": 13864
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4173,
      "step": 13865
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4002,
      "step": 13866
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4941,
      "step": 13867
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6001,
      "step": 13868
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4754,
      "step": 13869
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4545,
      "step": 13870
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3997,
      "step": 13871
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5601,
      "step": 13872
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5649,
      "step": 13873
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4756,
      "step": 13874
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3099,
      "step": 13875
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.7598,
      "step": 13876
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.58,
      "step": 13877
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3831,
      "step": 13878
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4712,
      "step": 13879
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6008,
      "step": 13880
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3626,
      "step": 13881
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 13882
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6355,
      "step": 13883
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.603,
      "step": 13884
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4194,
      "step": 13885
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3358,
      "step": 13886
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4065,
      "step": 13887
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3362,
      "step": 13888
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5699,
      "step": 13889
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.7048,
      "step": 13890
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.575,
      "step": 13891
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5297,
      "step": 13892
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4887,
      "step": 13893
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6509,
      "step": 13894
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.405,
      "step": 13895
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5384,
      "step": 13896
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.5787,
      "step": 13897
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4174,
      "step": 13898
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3216,
      "step": 13899
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4832,
      "step": 13900
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.59,
      "step": 13901
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.6065,
      "step": 13902
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3435,
      "step": 13903
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.7468,
      "step": 13904
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.4998,
      "step": 13905
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.321,
      "step": 13906
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.3245,
      "step": 13907
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4152,
      "step": 13908
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6162,
      "step": 13909
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5552,
      "step": 13910
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4485,
      "step": 13911
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6595,
      "step": 13912
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5163,
      "step": 13913
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6395,
      "step": 13914
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6884,
      "step": 13915
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.2945,
      "step": 13916
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5244,
      "step": 13917
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4304,
      "step": 13918
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4748,
      "step": 13919
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.293,
      "step": 13920
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4538,
      "step": 13921
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5237,
      "step": 13922
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6512,
      "step": 13923
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5128,
      "step": 13924
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5277,
      "step": 13925
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6436,
      "step": 13926
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3705,
      "step": 13927
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.413,
      "step": 13928
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.388,
      "step": 13929
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4315,
      "step": 13930
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5205,
      "step": 13931
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6,
      "step": 13932
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4039,
      "step": 13933
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4845,
      "step": 13934
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5967,
      "step": 13935
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.1463,
      "step": 13936
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5386,
      "step": 13937
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3013,
      "step": 13938
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6075,
      "step": 13939
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4023,
      "step": 13940
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5804,
      "step": 13941
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4822,
      "step": 13942
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5344,
      "step": 13943
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6109,
      "step": 13944
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3573,
      "step": 13945
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6208,
      "step": 13946
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4954,
      "step": 13947
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 13948
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4951,
      "step": 13949
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3539,
      "step": 13950
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5145,
      "step": 13951
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4512,
      "step": 13952
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4468,
      "step": 13953
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4581,
      "step": 13954
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.409,
      "step": 13955
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5684,
      "step": 13956
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6246,
      "step": 13957
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4359,
      "step": 13958
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5186,
      "step": 13959
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4264,
      "step": 13960
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.6375,
      "step": 13961
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.409,
      "step": 13962
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3849,
      "step": 13963
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.4413,
      "step": 13964
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5392,
      "step": 13965
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.5593,
      "step": 13966
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002,
      "loss": 0.3965,
      "step": 13967
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2743,
      "step": 13968
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4501,
      "step": 13969
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6754,
      "step": 13970
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4263,
      "step": 13971
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.3369,
      "step": 13972
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5846,
      "step": 13973
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5924,
      "step": 13974
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6369,
      "step": 13975
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5495,
      "step": 13976
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6239,
      "step": 13977
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4876,
      "step": 13978
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4792,
      "step": 13979
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2136,
      "step": 13980
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4479,
      "step": 13981
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6016,
      "step": 13982
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4101,
      "step": 13983
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4217,
      "step": 13984
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.7182,
      "step": 13985
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 13986
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5485,
      "step": 13987
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2998,
      "step": 13988
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6416,
      "step": 13989
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.629,
      "step": 13990
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2814,
      "step": 13991
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4316,
      "step": 13992
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4612,
      "step": 13993
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4145,
      "step": 13994
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.4394,
      "step": 13995
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.5374,
      "step": 13996
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6502,
      "step": 13997
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.6138,
      "step": 13998
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2392,
      "step": 13999
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.2467,
      "step": 14000
    },
    {
      "epoch": 2.3,
      "eval_loss": 0.8203836679458618,
      "eval_runtime": 131.5093,
      "eval_samples_per_second": 7.604,
      "eval_steps_per_second": 3.802,
      "step": 14000
    },
    {
      "epoch": 2.3,
      "mmlu_eval_accuracy": 0.5201004193268767,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.35714285714285715,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
      "mmlu_eval_accuracy_college_biology": 0.625,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.5,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.5,
      "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": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.7142857142857143,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.6153846153846154,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.5454545454545454,
      "mmlu_eval_accuracy_logical_fallacies": 0.7222222222222222,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.92,
      "mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.32,
      "mmlu_eval_accuracy_nutrition": 0.7575757575757576,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
      "mmlu_eval_accuracy_professional_psychology": 0.5507246376811594,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.6296296296296297,
      "mmlu_eval_accuracy_sociology": 0.7727272727272727,
      "mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.160834508362,
      "step": 14000
    }
  ],
  "logging_steps": 1,
  "max_steps": 18258,
  "num_train_epochs": 3,
  "save_steps": 200,
  "total_flos": 2.878684005463032e+18,
  "trial_name": null,
  "trial_params": null
}