{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 354,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2e-05,
      "loss": 13.75,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 2e-05,
      "loss": 13.375,
      "step": 2
    },
    {
      "epoch": 0.03,
      "learning_rate": 2e-05,
      "loss": 10.5938,
      "step": 3
    },
    {
      "epoch": 0.03,
      "learning_rate": 2e-05,
      "loss": 14.7656,
      "step": 4
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 14.9062,
      "step": 5
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 11.7109,
      "step": 6
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 13.8438,
      "step": 7
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 15.375,
      "step": 8
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 13.25,
      "step": 9
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 11.7578,
      "step": 10
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 13.9219,
      "step": 11
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 12.0859,
      "step": 12
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 11.9688,
      "step": 13
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 14.1875,
      "step": 14
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 11.3594,
      "step": 15
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 9.6641,
      "step": 16
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 15.0625,
      "step": 17
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 11.7031,
      "step": 18
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 16.4219,
      "step": 19
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 10.0469,
      "step": 20
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 12.0,
      "step": 21
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 13.4219,
      "step": 22
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 13.0625,
      "step": 23
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 10.9922,
      "step": 24
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 12.3906,
      "step": 25
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 9.3047,
      "step": 26
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 9.6406,
      "step": 27
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 13.4844,
      "step": 28
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 13.0938,
      "step": 29
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 12.7812,
      "step": 30
    },
    {
      "epoch": 0.26,
      "learning_rate": 2e-05,
      "loss": 12.0625,
      "step": 31
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 12.4219,
      "step": 32
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 11.3906,
      "step": 33
    },
    {
      "epoch": 0.29,
      "learning_rate": 2e-05,
      "loss": 10.6406,
      "step": 34
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 10.4219,
      "step": 35
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 12.0,
      "step": 36
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 12.2031,
      "step": 37
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 10.4297,
      "step": 38
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 13.9844,
      "step": 39
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 9.4062,
      "step": 40
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 13.5781,
      "step": 41
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 8.4688,
      "step": 42
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 10.7656,
      "step": 43
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 11.1719,
      "step": 44
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 10.7656,
      "step": 45
    },
    {
      "epoch": 0.39,
      "learning_rate": 2e-05,
      "loss": 10.7891,
      "step": 46
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 11.8438,
      "step": 47
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 9.7812,
      "step": 48
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 10.2188,
      "step": 49
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 9.6719,
      "step": 50
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 11.6719,
      "step": 51
    },
    {
      "epoch": 0.44,
      "learning_rate": 2e-05,
      "loss": 10.4375,
      "step": 52
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 9.1406,
      "step": 53
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 10.5625,
      "step": 54
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 9.9688,
      "step": 55
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 8.7891,
      "step": 56
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 9.3125,
      "step": 57
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 9.1797,
      "step": 58
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 9.7812,
      "step": 59
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 9.6094,
      "step": 60
    },
    {
      "epoch": 0.52,
      "learning_rate": 2e-05,
      "loss": 12.3281,
      "step": 61
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 7.8828,
      "step": 62
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 7.9844,
      "step": 63
    },
    {
      "epoch": 0.54,
      "learning_rate": 2e-05,
      "loss": 7.7578,
      "step": 64
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 8.3906,
      "step": 65
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 10.6328,
      "step": 66
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 8.1562,
      "step": 67
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 7.4688,
      "step": 68
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 7.0078,
      "step": 69
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 6.1641,
      "step": 70
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 7.5703,
      "step": 71
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 6.2109,
      "step": 72
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 7.7656,
      "step": 73
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 7.2266,
      "step": 74
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 7.7266,
      "step": 75
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 7.5391,
      "step": 76
    },
    {
      "epoch": 0.65,
      "learning_rate": 2e-05,
      "loss": 6.2031,
      "step": 77
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 7.4922,
      "step": 78
    },
    {
      "epoch": 0.67,
      "learning_rate": 2e-05,
      "loss": 5.9062,
      "step": 79
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 8.5469,
      "step": 80
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 7.4844,
      "step": 81
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 8.4297,
      "step": 82
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 7.1641,
      "step": 83
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 5.9844,
      "step": 84
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 6.5078,
      "step": 85
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 6.5312,
      "step": 86
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 6.8984,
      "step": 87
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 6.7188,
      "step": 88
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 6.4453,
      "step": 89
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 5.6562,
      "step": 90
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 6.4766,
      "step": 91
    },
    {
      "epoch": 0.78,
      "learning_rate": 2e-05,
      "loss": 6.2031,
      "step": 92
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 5.9531,
      "step": 93
    },
    {
      "epoch": 0.8,
      "learning_rate": 2e-05,
      "loss": 5.9453,
      "step": 94
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 5.4297,
      "step": 95
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 6.3125,
      "step": 96
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 6.0703,
      "step": 97
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 5.7891,
      "step": 98
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 5.7969,
      "step": 99
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 7.875,
      "step": 100
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 6.25,
      "step": 101
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 5.6484,
      "step": 102
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 8.4609,
      "step": 103
    },
    {
      "epoch": 0.88,
      "learning_rate": 2e-05,
      "loss": 6.0156,
      "step": 104
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 6.1172,
      "step": 105
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 6.0156,
      "step": 106
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 6.1562,
      "step": 107
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 5.457,
      "step": 108
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 5.6953,
      "step": 109
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 6.3828,
      "step": 110
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 4.7812,
      "step": 111
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 7.5156,
      "step": 112
    },
    {
      "epoch": 0.96,
      "learning_rate": 2e-05,
      "loss": 5.7109,
      "step": 113
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 6.1797,
      "step": 114
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 7.5078,
      "step": 115
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 6.7891,
      "step": 116
    },
    {
      "epoch": 0.99,
      "learning_rate": 2e-05,
      "loss": 7.0156,
      "step": 117
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 5.8438,
      "step": 118
    },
    {
      "epoch": 1.01,
      "learning_rate": 2e-05,
      "loss": 5.4297,
      "step": 119
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 5.5312,
      "step": 120
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 5.0156,
      "step": 121
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 5.8359,
      "step": 122
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 5.75,
      "step": 123
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 5.0273,
      "step": 124
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 5.4688,
      "step": 125
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 5.8906,
      "step": 126
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 5.4609,
      "step": 127
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 5.0508,
      "step": 128
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 5.5547,
      "step": 129
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 6.3984,
      "step": 130
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 5.2578,
      "step": 131
    },
    {
      "epoch": 1.12,
      "learning_rate": 2e-05,
      "loss": 5.6562,
      "step": 132
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 5.0664,
      "step": 133
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 4.7344,
      "step": 134
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 5.9219,
      "step": 135
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 5.0469,
      "step": 136
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 8.0,
      "step": 137
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 4.8438,
      "step": 138
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 5.25,
      "step": 139
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 5.4141,
      "step": 140
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 5.4375,
      "step": 141
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 4.875,
      "step": 142
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 5.4375,
      "step": 143
    },
    {
      "epoch": 1.22,
      "learning_rate": 2e-05,
      "loss": 4.5273,
      "step": 144
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 4.6797,
      "step": 145
    },
    {
      "epoch": 1.24,
      "learning_rate": 2e-05,
      "loss": 5.7344,
      "step": 146
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 5.6953,
      "step": 147
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 5.5625,
      "step": 148
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 5.3281,
      "step": 149
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 5.3828,
      "step": 150
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 5.2031,
      "step": 151
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 4.9688,
      "step": 152
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 5.0391,
      "step": 153
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 5.5312,
      "step": 154
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 5.5781,
      "step": 155
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 4.8242,
      "step": 156
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 6.0703,
      "step": 157
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 4.7422,
      "step": 158
    },
    {
      "epoch": 1.35,
      "learning_rate": 2e-05,
      "loss": 6.5469,
      "step": 159
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 4.5469,
      "step": 160
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 5.375,
      "step": 161
    },
    {
      "epoch": 1.37,
      "learning_rate": 2e-05,
      "loss": 5.4922,
      "step": 162
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 5.5547,
      "step": 163
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 5.4375,
      "step": 164
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 6.3984,
      "step": 165
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 5.2734,
      "step": 166
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 5.3047,
      "step": 167
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 4.9531,
      "step": 168
    },
    {
      "epoch": 1.43,
      "learning_rate": 2e-05,
      "loss": 6.8906,
      "step": 169
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 6.5039,
      "step": 170
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 5.1172,
      "step": 171
    },
    {
      "epoch": 1.46,
      "learning_rate": 2e-05,
      "loss": 5.6797,
      "step": 172
    },
    {
      "epoch": 1.47,
      "learning_rate": 2e-05,
      "loss": 5.5508,
      "step": 173
    },
    {
      "epoch": 1.47,
      "learning_rate": 2e-05,
      "loss": 5.2656,
      "step": 174
    },
    {
      "epoch": 1.48,
      "learning_rate": 2e-05,
      "loss": 5.0508,
      "step": 175
    },
    {
      "epoch": 1.49,
      "learning_rate": 2e-05,
      "loss": 5.2266,
      "step": 176
    },
    {
      "epoch": 1.5,
      "learning_rate": 2e-05,
      "loss": 5.3711,
      "step": 177
    },
    {
      "epoch": 1.51,
      "learning_rate": 2e-05,
      "loss": 5.0898,
      "step": 178
    },
    {
      "epoch": 1.52,
      "learning_rate": 2e-05,
      "loss": 6.4219,
      "step": 179
    },
    {
      "epoch": 1.53,
      "learning_rate": 2e-05,
      "loss": 5.0234,
      "step": 180
    },
    {
      "epoch": 1.53,
      "learning_rate": 2e-05,
      "loss": 5.0625,
      "step": 181
    },
    {
      "epoch": 1.54,
      "learning_rate": 2e-05,
      "loss": 5.1797,
      "step": 182
    },
    {
      "epoch": 1.55,
      "learning_rate": 2e-05,
      "loss": 5.2812,
      "step": 183
    },
    {
      "epoch": 1.56,
      "learning_rate": 2e-05,
      "loss": 5.5547,
      "step": 184
    },
    {
      "epoch": 1.57,
      "learning_rate": 2e-05,
      "loss": 4.8281,
      "step": 185
    },
    {
      "epoch": 1.58,
      "learning_rate": 2e-05,
      "loss": 4.8867,
      "step": 186
    },
    {
      "epoch": 1.58,
      "learning_rate": 2e-05,
      "loss": 4.7852,
      "step": 187
    },
    {
      "epoch": 1.59,
      "learning_rate": 2e-05,
      "loss": 4.4297,
      "step": 188
    },
    {
      "epoch": 1.6,
      "learning_rate": 2e-05,
      "loss": 5.4688,
      "step": 189
    },
    {
      "epoch": 1.61,
      "learning_rate": 2e-05,
      "loss": 4.668,
      "step": 190
    },
    {
      "epoch": 1.62,
      "learning_rate": 2e-05,
      "loss": 5.4062,
      "step": 191
    },
    {
      "epoch": 1.63,
      "learning_rate": 2e-05,
      "loss": 4.8828,
      "step": 192
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 5.6953,
      "step": 193
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 5.6094,
      "step": 194
    },
    {
      "epoch": 1.65,
      "learning_rate": 2e-05,
      "loss": 4.8516,
      "step": 195
    },
    {
      "epoch": 1.66,
      "learning_rate": 2e-05,
      "loss": 5.3359,
      "step": 196
    },
    {
      "epoch": 1.67,
      "learning_rate": 2e-05,
      "loss": 4.6953,
      "step": 197
    },
    {
      "epoch": 1.68,
      "learning_rate": 2e-05,
      "loss": 5.5859,
      "step": 198
    },
    {
      "epoch": 1.69,
      "learning_rate": 2e-05,
      "loss": 5.1016,
      "step": 199
    },
    {
      "epoch": 1.69,
      "learning_rate": 2e-05,
      "loss": 6.8984,
      "step": 200
    },
    {
      "epoch": 1.7,
      "learning_rate": 2e-05,
      "loss": 5.1055,
      "step": 201
    },
    {
      "epoch": 1.71,
      "learning_rate": 2e-05,
      "loss": 4.8164,
      "step": 202
    },
    {
      "epoch": 1.72,
      "learning_rate": 2e-05,
      "loss": 5.2969,
      "step": 203
    },
    {
      "epoch": 1.73,
      "learning_rate": 2e-05,
      "loss": 5.4297,
      "step": 204
    },
    {
      "epoch": 1.74,
      "learning_rate": 2e-05,
      "loss": 4.8711,
      "step": 205
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 5.0742,
      "step": 206
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 5.3672,
      "step": 207
    },
    {
      "epoch": 1.76,
      "learning_rate": 2e-05,
      "loss": 4.4961,
      "step": 208
    },
    {
      "epoch": 1.77,
      "learning_rate": 2e-05,
      "loss": 5.4453,
      "step": 209
    },
    {
      "epoch": 1.78,
      "learning_rate": 2e-05,
      "loss": 5.0547,
      "step": 210
    },
    {
      "epoch": 1.79,
      "learning_rate": 2e-05,
      "loss": 5.1562,
      "step": 211
    },
    {
      "epoch": 1.8,
      "learning_rate": 2e-05,
      "loss": 4.9414,
      "step": 212
    },
    {
      "epoch": 1.81,
      "learning_rate": 2e-05,
      "loss": 4.3828,
      "step": 213
    },
    {
      "epoch": 1.81,
      "learning_rate": 2e-05,
      "loss": 5.3828,
      "step": 214
    },
    {
      "epoch": 1.82,
      "learning_rate": 2e-05,
      "loss": 5.1953,
      "step": 215
    },
    {
      "epoch": 1.83,
      "learning_rate": 2e-05,
      "loss": 4.8984,
      "step": 216
    },
    {
      "epoch": 1.84,
      "learning_rate": 2e-05,
      "loss": 4.9297,
      "step": 217
    },
    {
      "epoch": 1.85,
      "learning_rate": 2e-05,
      "loss": 5.2695,
      "step": 218
    },
    {
      "epoch": 1.86,
      "learning_rate": 2e-05,
      "loss": 5.5625,
      "step": 219
    },
    {
      "epoch": 1.86,
      "learning_rate": 2e-05,
      "loss": 4.9102,
      "step": 220
    },
    {
      "epoch": 1.87,
      "learning_rate": 2e-05,
      "loss": 6.8047,
      "step": 221
    },
    {
      "epoch": 1.88,
      "learning_rate": 2e-05,
      "loss": 5.2422,
      "step": 222
    },
    {
      "epoch": 1.89,
      "learning_rate": 2e-05,
      "loss": 5.3672,
      "step": 223
    },
    {
      "epoch": 1.9,
      "learning_rate": 2e-05,
      "loss": 5.1719,
      "step": 224
    },
    {
      "epoch": 1.91,
      "learning_rate": 2e-05,
      "loss": 5.4141,
      "step": 225
    },
    {
      "epoch": 1.92,
      "learning_rate": 2e-05,
      "loss": 4.7305,
      "step": 226
    },
    {
      "epoch": 1.92,
      "learning_rate": 2e-05,
      "loss": 4.9609,
      "step": 227
    },
    {
      "epoch": 1.93,
      "learning_rate": 2e-05,
      "loss": 5.7422,
      "step": 228
    },
    {
      "epoch": 1.94,
      "learning_rate": 2e-05,
      "loss": 3.957,
      "step": 229
    },
    {
      "epoch": 1.95,
      "learning_rate": 2e-05,
      "loss": 5.7656,
      "step": 230
    },
    {
      "epoch": 1.96,
      "learning_rate": 2e-05,
      "loss": 4.9766,
      "step": 231
    },
    {
      "epoch": 1.97,
      "learning_rate": 2e-05,
      "loss": 5.5391,
      "step": 232
    },
    {
      "epoch": 1.97,
      "learning_rate": 2e-05,
      "loss": 6.2344,
      "step": 233
    },
    {
      "epoch": 1.98,
      "learning_rate": 2e-05,
      "loss": 5.4844,
      "step": 234
    },
    {
      "epoch": 1.99,
      "learning_rate": 2e-05,
      "loss": 5.3281,
      "step": 235
    },
    {
      "epoch": 2.0,
      "learning_rate": 2e-05,
      "loss": 5.25,
      "step": 236
    },
    {
      "epoch": 2.01,
      "learning_rate": 2e-05,
      "loss": 4.8867,
      "step": 237
    },
    {
      "epoch": 2.02,
      "learning_rate": 2e-05,
      "loss": 4.9688,
      "step": 238
    },
    {
      "epoch": 2.03,
      "learning_rate": 2e-05,
      "loss": 4.3711,
      "step": 239
    },
    {
      "epoch": 2.03,
      "learning_rate": 2e-05,
      "loss": 5.2891,
      "step": 240
    },
    {
      "epoch": 2.04,
      "learning_rate": 2e-05,
      "loss": 5.2578,
      "step": 241
    },
    {
      "epoch": 2.05,
      "learning_rate": 2e-05,
      "loss": 4.5,
      "step": 242
    },
    {
      "epoch": 2.06,
      "learning_rate": 2e-05,
      "loss": 4.9727,
      "step": 243
    },
    {
      "epoch": 2.07,
      "learning_rate": 2e-05,
      "loss": 5.4375,
      "step": 244
    },
    {
      "epoch": 2.08,
      "learning_rate": 2e-05,
      "loss": 5.0,
      "step": 245
    },
    {
      "epoch": 2.08,
      "learning_rate": 2e-05,
      "loss": 4.5664,
      "step": 246
    },
    {
      "epoch": 2.09,
      "learning_rate": 2e-05,
      "loss": 5.1484,
      "step": 247
    },
    {
      "epoch": 2.1,
      "learning_rate": 2e-05,
      "loss": 5.2109,
      "step": 248
    },
    {
      "epoch": 2.11,
      "learning_rate": 2e-05,
      "loss": 4.7969,
      "step": 249
    },
    {
      "epoch": 2.12,
      "learning_rate": 2e-05,
      "loss": 5.2109,
      "step": 250
    },
    {
      "epoch": 2.13,
      "learning_rate": 2e-05,
      "loss": 4.6172,
      "step": 251
    },
    {
      "epoch": 2.14,
      "learning_rate": 2e-05,
      "loss": 4.2891,
      "step": 252
    },
    {
      "epoch": 2.14,
      "learning_rate": 2e-05,
      "loss": 5.4531,
      "step": 253
    },
    {
      "epoch": 2.15,
      "learning_rate": 2e-05,
      "loss": 4.5898,
      "step": 254
    },
    {
      "epoch": 2.16,
      "learning_rate": 2e-05,
      "loss": 7.4609,
      "step": 255
    },
    {
      "epoch": 2.17,
      "learning_rate": 2e-05,
      "loss": 4.4453,
      "step": 256
    },
    {
      "epoch": 2.18,
      "learning_rate": 2e-05,
      "loss": 4.875,
      "step": 257
    },
    {
      "epoch": 2.19,
      "learning_rate": 2e-05,
      "loss": 5.0625,
      "step": 258
    },
    {
      "epoch": 2.19,
      "learning_rate": 2e-05,
      "loss": 5.0781,
      "step": 259
    },
    {
      "epoch": 2.2,
      "learning_rate": 2e-05,
      "loss": 4.5312,
      "step": 260
    },
    {
      "epoch": 2.21,
      "learning_rate": 2e-05,
      "loss": 5.0703,
      "step": 261
    },
    {
      "epoch": 2.22,
      "learning_rate": 2e-05,
      "loss": 4.1914,
      "step": 262
    },
    {
      "epoch": 2.23,
      "learning_rate": 2e-05,
      "loss": 4.3555,
      "step": 263
    },
    {
      "epoch": 2.24,
      "learning_rate": 2e-05,
      "loss": 5.4141,
      "step": 264
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 5.3516,
      "step": 265
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 5.2422,
      "step": 266
    },
    {
      "epoch": 2.26,
      "learning_rate": 2e-05,
      "loss": 5.0156,
      "step": 267
    },
    {
      "epoch": 2.27,
      "learning_rate": 2e-05,
      "loss": 5.0859,
      "step": 268
    },
    {
      "epoch": 2.28,
      "learning_rate": 2e-05,
      "loss": 4.8984,
      "step": 269
    },
    {
      "epoch": 2.29,
      "learning_rate": 2e-05,
      "loss": 4.6719,
      "step": 270
    },
    {
      "epoch": 2.3,
      "learning_rate": 2e-05,
      "loss": 4.7383,
      "step": 271
    },
    {
      "epoch": 2.31,
      "learning_rate": 2e-05,
      "loss": 5.25,
      "step": 272
    },
    {
      "epoch": 2.31,
      "learning_rate": 2e-05,
      "loss": 5.2891,
      "step": 273
    },
    {
      "epoch": 2.32,
      "learning_rate": 2e-05,
      "loss": 4.5508,
      "step": 274
    },
    {
      "epoch": 2.33,
      "learning_rate": 2e-05,
      "loss": 5.7422,
      "step": 275
    },
    {
      "epoch": 2.34,
      "learning_rate": 2e-05,
      "loss": 4.4688,
      "step": 276
    },
    {
      "epoch": 2.35,
      "learning_rate": 2e-05,
      "loss": 5.7969,
      "step": 277
    },
    {
      "epoch": 2.36,
      "learning_rate": 2e-05,
      "loss": 4.2852,
      "step": 278
    },
    {
      "epoch": 2.36,
      "learning_rate": 2e-05,
      "loss": 5.1172,
      "step": 279
    },
    {
      "epoch": 2.37,
      "learning_rate": 2e-05,
      "loss": 5.2344,
      "step": 280
    },
    {
      "epoch": 2.38,
      "learning_rate": 2e-05,
      "loss": 5.2891,
      "step": 281
    },
    {
      "epoch": 2.39,
      "learning_rate": 2e-05,
      "loss": 5.1641,
      "step": 282
    },
    {
      "epoch": 2.4,
      "learning_rate": 2e-05,
      "loss": 5.375,
      "step": 283
    },
    {
      "epoch": 2.41,
      "learning_rate": 2e-05,
      "loss": 5.0312,
      "step": 284
    },
    {
      "epoch": 2.42,
      "learning_rate": 2e-05,
      "loss": 5.0508,
      "step": 285
    },
    {
      "epoch": 2.42,
      "learning_rate": 2e-05,
      "loss": 4.6875,
      "step": 286
    },
    {
      "epoch": 2.43,
      "learning_rate": 2e-05,
      "loss": 6.0469,
      "step": 287
    },
    {
      "epoch": 2.44,
      "learning_rate": 2e-05,
      "loss": 5.1523,
      "step": 288
    },
    {
      "epoch": 2.45,
      "learning_rate": 2e-05,
      "loss": 4.875,
      "step": 289
    },
    {
      "epoch": 2.46,
      "learning_rate": 2e-05,
      "loss": 5.4102,
      "step": 290
    },
    {
      "epoch": 2.47,
      "learning_rate": 2e-05,
      "loss": 5.2891,
      "step": 291
    },
    {
      "epoch": 2.47,
      "learning_rate": 2e-05,
      "loss": 5.0234,
      "step": 292
    },
    {
      "epoch": 2.48,
      "learning_rate": 2e-05,
      "loss": 4.8086,
      "step": 293
    },
    {
      "epoch": 2.49,
      "learning_rate": 2e-05,
      "loss": 4.9766,
      "step": 294
    },
    {
      "epoch": 2.5,
      "learning_rate": 2e-05,
      "loss": 5.1211,
      "step": 295
    },
    {
      "epoch": 2.51,
      "learning_rate": 2e-05,
      "loss": 4.8594,
      "step": 296
    },
    {
      "epoch": 2.52,
      "learning_rate": 2e-05,
      "loss": 5.9375,
      "step": 297
    },
    {
      "epoch": 2.53,
      "learning_rate": 2e-05,
      "loss": 4.8281,
      "step": 298
    },
    {
      "epoch": 2.53,
      "learning_rate": 2e-05,
      "loss": 4.8594,
      "step": 299
    },
    {
      "epoch": 2.54,
      "learning_rate": 2e-05,
      "loss": 4.9688,
      "step": 300
    },
    {
      "epoch": 2.55,
      "learning_rate": 2e-05,
      "loss": 5.043,
      "step": 301
    },
    {
      "epoch": 2.56,
      "learning_rate": 2e-05,
      "loss": 5.2422,
      "step": 302
    },
    {
      "epoch": 2.57,
      "learning_rate": 2e-05,
      "loss": 4.5977,
      "step": 303
    },
    {
      "epoch": 2.58,
      "learning_rate": 2e-05,
      "loss": 4.6797,
      "step": 304
    },
    {
      "epoch": 2.58,
      "learning_rate": 2e-05,
      "loss": 4.5781,
      "step": 305
    },
    {
      "epoch": 2.59,
      "learning_rate": 2e-05,
      "loss": 4.2266,
      "step": 306
    },
    {
      "epoch": 2.6,
      "learning_rate": 2e-05,
      "loss": 5.2656,
      "step": 307
    },
    {
      "epoch": 2.61,
      "learning_rate": 2e-05,
      "loss": 4.4609,
      "step": 308
    },
    {
      "epoch": 2.62,
      "learning_rate": 2e-05,
      "loss": 5.1953,
      "step": 309
    },
    {
      "epoch": 2.63,
      "learning_rate": 2e-05,
      "loss": 4.6719,
      "step": 310
    },
    {
      "epoch": 2.64,
      "learning_rate": 2e-05,
      "loss": 5.4766,
      "step": 311
    },
    {
      "epoch": 2.64,
      "learning_rate": 2e-05,
      "loss": 5.4062,
      "step": 312
    },
    {
      "epoch": 2.65,
      "learning_rate": 2e-05,
      "loss": 4.6484,
      "step": 313
    },
    {
      "epoch": 2.66,
      "learning_rate": 2e-05,
      "loss": 5.1172,
      "step": 314
    },
    {
      "epoch": 2.67,
      "learning_rate": 2e-05,
      "loss": 4.5312,
      "step": 315
    },
    {
      "epoch": 2.68,
      "learning_rate": 2e-05,
      "loss": 5.3438,
      "step": 316
    },
    {
      "epoch": 2.69,
      "learning_rate": 2e-05,
      "loss": 4.8984,
      "step": 317
    },
    {
      "epoch": 2.69,
      "learning_rate": 2e-05,
      "loss": 6.6211,
      "step": 318
    },
    {
      "epoch": 2.7,
      "learning_rate": 2e-05,
      "loss": 4.8906,
      "step": 319
    },
    {
      "epoch": 2.71,
      "learning_rate": 2e-05,
      "loss": 4.6445,
      "step": 320
    },
    {
      "epoch": 2.72,
      "learning_rate": 2e-05,
      "loss": 5.0977,
      "step": 321
    },
    {
      "epoch": 2.73,
      "learning_rate": 2e-05,
      "loss": 5.2344,
      "step": 322
    },
    {
      "epoch": 2.74,
      "learning_rate": 2e-05,
      "loss": 4.6875,
      "step": 323
    },
    {
      "epoch": 2.75,
      "learning_rate": 2e-05,
      "loss": 4.8828,
      "step": 324
    },
    {
      "epoch": 2.75,
      "learning_rate": 2e-05,
      "loss": 5.1797,
      "step": 325
    },
    {
      "epoch": 2.76,
      "learning_rate": 2e-05,
      "loss": 4.3203,
      "step": 326
    },
    {
      "epoch": 2.77,
      "learning_rate": 2e-05,
      "loss": 5.2656,
      "step": 327
    },
    {
      "epoch": 2.78,
      "learning_rate": 2e-05,
      "loss": 4.875,
      "step": 328
    },
    {
      "epoch": 2.79,
      "learning_rate": 2e-05,
      "loss": 4.9922,
      "step": 329
    },
    {
      "epoch": 2.8,
      "learning_rate": 2e-05,
      "loss": 4.7852,
      "step": 330
    },
    {
      "epoch": 2.81,
      "learning_rate": 2e-05,
      "loss": 4.207,
      "step": 331
    },
    {
      "epoch": 2.81,
      "learning_rate": 2e-05,
      "loss": 5.1953,
      "step": 332
    },
    {
      "epoch": 2.82,
      "learning_rate": 2e-05,
      "loss": 5.0312,
      "step": 333
    },
    {
      "epoch": 2.83,
      "learning_rate": 2e-05,
      "loss": 4.7344,
      "step": 334
    },
    {
      "epoch": 2.84,
      "learning_rate": 2e-05,
      "loss": 4.7656,
      "step": 335
    },
    {
      "epoch": 2.85,
      "learning_rate": 2e-05,
      "loss": 5.0898,
      "step": 336
    },
    {
      "epoch": 2.86,
      "learning_rate": 2e-05,
      "loss": 5.3867,
      "step": 337
    },
    {
      "epoch": 2.86,
      "learning_rate": 2e-05,
      "loss": 4.7578,
      "step": 338
    },
    {
      "epoch": 2.87,
      "learning_rate": 2e-05,
      "loss": 6.0547,
      "step": 339
    },
    {
      "epoch": 2.88,
      "learning_rate": 2e-05,
      "loss": 5.0781,
      "step": 340
    },
    {
      "epoch": 2.89,
      "learning_rate": 2e-05,
      "loss": 5.1914,
      "step": 341
    },
    {
      "epoch": 2.9,
      "learning_rate": 2e-05,
      "loss": 5.0078,
      "step": 342
    },
    {
      "epoch": 2.91,
      "learning_rate": 2e-05,
      "loss": 5.2344,
      "step": 343
    },
    {
      "epoch": 2.92,
      "learning_rate": 2e-05,
      "loss": 4.5703,
      "step": 344
    },
    {
      "epoch": 2.92,
      "learning_rate": 2e-05,
      "loss": 4.8125,
      "step": 345
    },
    {
      "epoch": 2.93,
      "learning_rate": 2e-05,
      "loss": 5.5781,
      "step": 346
    },
    {
      "epoch": 2.94,
      "learning_rate": 2e-05,
      "loss": 3.8281,
      "step": 347
    },
    {
      "epoch": 2.95,
      "learning_rate": 2e-05,
      "loss": 5.4375,
      "step": 348
    },
    {
      "epoch": 2.96,
      "learning_rate": 2e-05,
      "loss": 4.832,
      "step": 349
    },
    {
      "epoch": 2.97,
      "learning_rate": 2e-05,
      "loss": 5.375,
      "step": 350
    },
    {
      "epoch": 2.97,
      "learning_rate": 2e-05,
      "loss": 5.0977,
      "step": 351
    },
    {
      "epoch": 2.98,
      "learning_rate": 2e-05,
      "loss": 5.3125,
      "step": 352
    },
    {
      "epoch": 2.99,
      "learning_rate": 2e-05,
      "loss": 4.9531,
      "step": 353
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-05,
      "loss": 5.1016,
      "step": 354
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 472,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 13755035418624.0,
  "trial_name": null,
  "trial_params": null
}