diff --git "a/checkpoint-950/trainer_state.json" "b/checkpoint-950/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-950/trainer_state.json"
@@ -0,0 +1,6092 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.7777777777777777,
+  "global_step": 950,
+  "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": 0.7121,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002,
+      "loss": 0.645,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002,
+      "loss": 0.6852,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002,
+      "loss": 0.689,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0002,
+      "loss": 0.7223,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002,
+      "loss": 0.7359,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002,
+      "loss": 0.677,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002,
+      "loss": 0.6888,
+      "step": 8
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 0.7475,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 0.6675,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.0002,
+      "loss": 0.6665,
+      "step": 11
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 0.6803,
+      "step": 12
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 0.7052,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0002,
+      "loss": 0.6975,
+      "step": 15
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002,
+      "loss": 0.6186,
+      "step": 16
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002,
+      "loss": 0.6533,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0002,
+      "loss": 0.7183,
+      "step": 18
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6477,
+      "step": 19
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6753,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.7119563221931458,
+      "eval_runtime": 61.7384,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 0.638,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6424,
+      "step": 22
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002,
+      "loss": 0.7058,
+      "step": 23
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0002,
+      "loss": 0.6469,
+      "step": 25
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 0.6803,
+      "step": 26
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 0.7178,
+      "step": 27
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 0.6187,
+      "step": 28
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0002,
+      "loss": 0.6321,
+      "step": 29
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 30
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002,
+      "loss": 0.7176,
+      "step": 31
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 32
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002,
+      "loss": 0.6233,
+      "step": 33
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002,
+      "loss": 0.7009,
+      "step": 34
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0002,
+      "loss": 0.6672,
+      "step": 35
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 0.6915,
+      "step": 36
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 0.657,
+      "step": 37
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 38
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0002,
+      "loss": 0.6499,
+      "step": 39
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6177,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.713217556476593,
+      "eval_runtime": 61.6956,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 40
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6803,
+      "step": 41
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6825,
+      "step": 42
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 0.6566,
+      "step": 43
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 0.5804,
+      "step": 44
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 0.6501,
+      "step": 45
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0002,
+      "loss": 0.6467,
+      "step": 46
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6288,
+      "step": 47
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6505,
+      "step": 48
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6884,
+      "step": 49
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 0.6187,
+      "step": 50
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 0.698,
+      "step": 51
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 0.6788,
+      "step": 52
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0002,
+      "loss": 0.6813,
+      "step": 53
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 54
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002,
+      "loss": 0.6416,
+      "step": 55
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0002,
+      "loss": 0.7039,
+      "step": 56
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002,
+      "loss": 0.6749,
+      "step": 57
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002,
+      "loss": 0.6716,
+      "step": 58
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0002,
+      "loss": 0.6554,
+      "step": 59
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 0.7074,
+      "step": 60
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.713684618473053,
+      "eval_runtime": 61.7163,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 60
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6745,
+      "step": 61
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6201,
+      "step": 62
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 63
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002,
+      "loss": 0.6458,
+      "step": 64
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 65
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 66
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 67
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6955,
+      "step": 68
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6538,
+      "step": 69
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6962,
+      "step": 70
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002,
+      "loss": 0.6408,
+      "step": 71
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002,
+      "loss": 0.6327,
+      "step": 72
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 73
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002,
+      "loss": 0.6332,
+      "step": 74
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002,
+      "loss": 0.6712,
+      "step": 75
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0002,
+      "loss": 0.6065,
+      "step": 76
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002,
+      "loss": 0.6752,
+      "step": 77
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002,
+      "loss": 0.6965,
+      "step": 78
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002,
+      "loss": 0.6306,
+      "step": 79
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002,
+      "loss": 0.6789,
+      "step": 80
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.7147287726402283,
+      "eval_runtime": 61.6954,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 80
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002,
+      "loss": 0.7026,
+      "step": 81
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002,
+      "loss": 0.6344,
+      "step": 82
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0002,
+      "loss": 0.6562,
+      "step": 83
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.6402,
+      "step": 84
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 85
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 86
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 87
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 88
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6444,
+      "step": 89
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6633,
+      "step": 90
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 91
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 92
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 93
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 94
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 95
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002,
+      "loss": 0.5813,
+      "step": 96
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0002,
+      "loss": 0.6382,
+      "step": 97
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002,
+      "loss": 0.6592,
+      "step": 98
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002,
+      "loss": 0.6993,
+      "step": 99
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0002,
+      "loss": 0.6365,
+      "step": 100
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.7158837914466858,
+      "eval_runtime": 61.6652,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 100
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6701,
+      "step": 101
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6548,
+      "step": 102
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6537,
+      "step": 103
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6245,
+      "step": 104
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002,
+      "loss": 0.689,
+      "step": 105
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002,
+      "loss": 0.6237,
+      "step": 106
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0002,
+      "loss": 0.6209,
+      "step": 107
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 0.6833,
+      "step": 108
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 0.7169,
+      "step": 109
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 0.6534,
+      "step": 110
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002,
+      "loss": 0.7322,
+      "step": 111
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002,
+      "loss": 0.6634,
+      "step": 112
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 113
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0002,
+      "loss": 0.6238,
+      "step": 114
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002,
+      "loss": 0.6806,
+      "step": 115
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002,
+      "loss": 0.7004,
+      "step": 116
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0002,
+      "loss": 0.6181,
+      "step": 117
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002,
+      "loss": 0.6187,
+      "step": 118
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002,
+      "loss": 0.6148,
+      "step": 119
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002,
+      "loss": 0.588,
+      "step": 120
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.7125279307365417,
+      "eval_runtime": 61.6611,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 120
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0002,
+      "loss": 0.6556,
+      "step": 121
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 122
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002,
+      "loss": 0.6437,
+      "step": 123
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002,
+      "loss": 0.5877,
+      "step": 124
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002,
+      "loss": 0.661,
+      "step": 125
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 126
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 127
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002,
+      "loss": 0.696,
+      "step": 128
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002,
+      "loss": 0.6206,
+      "step": 129
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002,
+      "loss": 0.6093,
+      "step": 130
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 131
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6741,
+      "step": 132
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 133
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 134
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6689,
+      "step": 135
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 136
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002,
+      "loss": 0.6478,
+      "step": 137
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002,
+      "loss": 0.6293,
+      "step": 138
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002,
+      "loss": 0.6697,
+      "step": 139
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002,
+      "loss": 0.5792,
+      "step": 140
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.7092259526252747,
+      "eval_runtime": 61.6566,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 140
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 141
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002,
+      "loss": 0.6087,
+      "step": 142
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 143
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 144
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002,
+      "loss": 0.7078,
+      "step": 145
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002,
+      "loss": 0.7451,
+      "step": 146
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002,
+      "loss": 0.7025,
+      "step": 147
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002,
+      "loss": 0.6966,
+      "step": 148
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6685,
+      "step": 149
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 150
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002,
+      "loss": 0.685,
+      "step": 151
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6586,
+      "step": 152
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002,
+      "loss": 0.7471,
+      "step": 153
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002,
+      "loss": 0.7459,
+      "step": 154
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002,
+      "loss": 0.6347,
+      "step": 155
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 156
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002,
+      "loss": 0.7217,
+      "step": 157
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002,
+      "loss": 0.6961,
+      "step": 158
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002,
+      "loss": 0.7294,
+      "step": 159
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002,
+      "loss": 0.7065,
+      "step": 160
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.7031598687171936,
+      "eval_runtime": 61.6722,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 160
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002,
+      "loss": 0.6617,
+      "step": 161
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 162
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002,
+      "loss": 0.6815,
+      "step": 163
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002,
+      "loss": 0.6593,
+      "step": 164
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002,
+      "loss": 0.728,
+      "step": 165
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002,
+      "loss": 0.7029,
+      "step": 166
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 167
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002,
+      "loss": 0.7378,
+      "step": 168
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002,
+      "loss": 0.6646,
+      "step": 169
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.6884,
+      "step": 170
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.6713,
+      "step": 171
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002,
+      "loss": 0.703,
+      "step": 172
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002,
+      "loss": 0.6899,
+      "step": 173
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002,
+      "loss": 0.7384,
+      "step": 174
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002,
+      "loss": 0.6977,
+      "step": 175
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002,
+      "loss": 0.7067,
+      "step": 176
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002,
+      "loss": 0.6933,
+      "step": 177
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 178
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002,
+      "loss": 0.7154,
+      "step": 179
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 180
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.6995596885681152,
+      "eval_runtime": 61.6698,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 180
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 181
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0002,
+      "loss": 0.722,
+      "step": 182
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002,
+      "loss": 0.7521,
+      "step": 183
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002,
+      "loss": 0.7093,
+      "step": 184
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 185
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0002,
+      "loss": 0.7001,
+      "step": 186
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002,
+      "loss": 0.642,
+      "step": 187
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002,
+      "loss": 0.6805,
+      "step": 188
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 189
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002,
+      "loss": 0.6851,
+      "step": 190
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 191
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002,
+      "loss": 0.7122,
+      "step": 192
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0002,
+      "loss": 0.6792,
+      "step": 193
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002,
+      "loss": 0.6736,
+      "step": 194
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002,
+      "loss": 0.7077,
+      "step": 195
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0002,
+      "loss": 0.647,
+      "step": 196
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002,
+      "loss": 0.701,
+      "step": 197
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 198
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002,
+      "loss": 0.7166,
+      "step": 199
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0002,
+      "loss": 0.7053,
+      "step": 200
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.6974362730979919,
+      "eval_runtime": 61.6749,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 200
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002,
+      "loss": 0.7491,
+      "step": 201
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002,
+      "loss": 0.684,
+      "step": 202
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0002,
+      "loss": 0.6663,
+      "step": 203
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002,
+      "loss": 0.6929,
+      "step": 204
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002,
+      "loss": 0.6709,
+      "step": 205
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002,
+      "loss": 0.6705,
+      "step": 206
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002,
+      "loss": 0.7128,
+      "step": 207
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002,
+      "loss": 0.7086,
+      "step": 208
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002,
+      "loss": 0.7026,
+      "step": 209
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 210
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002,
+      "loss": 0.7072,
+      "step": 211
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002,
+      "loss": 0.6791,
+      "step": 212
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0002,
+      "loss": 0.7324,
+      "step": 213
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 214
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 215
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6592,
+      "step": 216
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6691,
+      "step": 217
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002,
+      "loss": 0.7093,
+      "step": 218
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002,
+      "loss": 0.6677,
+      "step": 219
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0002,
+      "loss": 0.7233,
+      "step": 220
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.6950156092643738,
+      "eval_runtime": 61.071,
+      "eval_samples_per_second": 0.458,
+      "eval_steps_per_second": 0.229,
+      "step": 220
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002,
+      "loss": 0.6807,
+      "step": 221
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002,
+      "loss": 0.688,
+      "step": 222
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002,
+      "loss": 0.6838,
+      "step": 223
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0002,
+      "loss": 0.7596,
+      "step": 224
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002,
+      "loss": 0.7333,
+      "step": 225
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002,
+      "loss": 0.7022,
+      "step": 226
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0002,
+      "loss": 0.6913,
+      "step": 227
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002,
+      "loss": 0.6735,
+      "step": 228
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002,
+      "loss": 0.7035,
+      "step": 229
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0002,
+      "loss": 0.6986,
+      "step": 230
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002,
+      "loss": 0.7028,
+      "step": 231
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002,
+      "loss": 0.6831,
+      "step": 232
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 233
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0002,
+      "loss": 0.7105,
+      "step": 234
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002,
+      "loss": 0.6679,
+      "step": 235
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002,
+      "loss": 0.7177,
+      "step": 236
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0002,
+      "loss": 0.6431,
+      "step": 237
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002,
+      "loss": 0.7055,
+      "step": 238
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002,
+      "loss": 0.6814,
+      "step": 239
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002,
+      "loss": 0.6657,
+      "step": 240
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.6946966052055359,
+      "eval_runtime": 61.1942,
+      "eval_samples_per_second": 0.458,
+      "eval_steps_per_second": 0.229,
+      "step": 240
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.0002,
+      "loss": 0.7436,
+      "step": 241
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002,
+      "loss": 0.6857,
+      "step": 242
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002,
+      "loss": 0.7002,
+      "step": 243
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0002,
+      "loss": 0.7005,
+      "step": 244
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002,
+      "loss": 0.743,
+      "step": 245
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002,
+      "loss": 0.6702,
+      "step": 246
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.0002,
+      "loss": 0.6766,
+      "step": 247
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002,
+      "loss": 0.7,
+      "step": 248
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 249
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 250
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6954,
+      "step": 251
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002,
+      "loss": 0.6818,
+      "step": 252
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002,
+      "loss": 0.7305,
+      "step": 253
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002,
+      "loss": 0.6953,
+      "step": 254
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.7008,
+      "step": 255
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.7128,
+      "step": 256
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.606,
+      "step": 257
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0002,
+      "loss": 0.7075,
+      "step": 258
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002,
+      "loss": 0.6759,
+      "step": 259
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002,
+      "loss": 0.7089,
+      "step": 260
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.6932913064956665,
+      "eval_runtime": 61.1404,
+      "eval_samples_per_second": 0.458,
+      "eval_steps_per_second": 0.229,
+      "step": 260
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.0002,
+      "loss": 0.6486,
+      "step": 261
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002,
+      "loss": 0.7123,
+      "step": 262
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002,
+      "loss": 0.687,
+      "step": 263
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002,
+      "loss": 0.6411,
+      "step": 264
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.0002,
+      "loss": 0.6637,
+      "step": 265
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002,
+      "loss": 0.7098,
+      "step": 266
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002,
+      "loss": 0.7697,
+      "step": 267
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.0002,
+      "loss": 0.6637,
+      "step": 268
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002,
+      "loss": 0.6425,
+      "step": 269
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002,
+      "loss": 0.69,
+      "step": 270
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002,
+      "loss": 0.6787,
+      "step": 271
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002,
+      "loss": 0.7601,
+      "step": 272
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6727,
+      "step": 273
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6875,
+      "step": 274
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6462,
+      "step": 275
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002,
+      "loss": 0.7698,
+      "step": 276
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002,
+      "loss": 0.6924,
+      "step": 277
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.0002,
+      "loss": 0.705,
+      "step": 278
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002,
+      "loss": 0.6528,
+      "step": 279
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002,
+      "loss": 0.6635,
+      "step": 280
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.6923065781593323,
+      "eval_runtime": 61.3221,
+      "eval_samples_per_second": 0.457,
+      "eval_steps_per_second": 0.228,
+      "step": 280
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 281
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.0002,
+      "loss": 0.7112,
+      "step": 282
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002,
+      "loss": 0.6731,
+      "step": 283
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002,
+      "loss": 0.6982,
+      "step": 284
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002,
+      "loss": 0.6717,
+      "step": 285
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 286
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002,
+      "loss": 0.6553,
+      "step": 287
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.0002,
+      "loss": 0.6849,
+      "step": 288
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002,
+      "loss": 0.7654,
+      "step": 289
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002,
+      "loss": 0.7549,
+      "step": 290
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002,
+      "loss": 0.7094,
+      "step": 291
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.0002,
+      "loss": 0.7293,
+      "step": 292
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002,
+      "loss": 0.6716,
+      "step": 293
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002,
+      "loss": 0.6856,
+      "step": 294
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0002,
+      "loss": 0.6923,
+      "step": 295
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002,
+      "loss": 0.7298,
+      "step": 296
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002,
+      "loss": 0.6967,
+      "step": 297
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002,
+      "loss": 0.6626,
+      "step": 298
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.0002,
+      "loss": 0.7118,
+      "step": 299
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002,
+      "loss": 0.7034,
+      "step": 300
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.691140353679657,
+      "eval_runtime": 61.249,
+      "eval_samples_per_second": 0.457,
+      "eval_steps_per_second": 0.229,
+      "step": 300
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002,
+      "loss": 0.6783,
+      "step": 301
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0002,
+      "loss": 0.6674,
+      "step": 302
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002,
+      "loss": 0.6316,
+      "step": 303
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002,
+      "loss": 0.6897,
+      "step": 304
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002,
+      "loss": 0.6863,
+      "step": 305
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.0002,
+      "loss": 0.72,
+      "step": 306
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002,
+      "loss": 0.6865,
+      "step": 307
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 308
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0002,
+      "loss": 0.6751,
+      "step": 309
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 310
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6926,
+      "step": 311
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6604,
+      "step": 312
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002,
+      "loss": 0.7059,
+      "step": 313
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6819,
+      "step": 314
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6743,
+      "step": 315
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6812,
+      "step": 316
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 317
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002,
+      "loss": 0.7033,
+      "step": 318
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.0002,
+      "loss": 0.6389,
+      "step": 319
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002,
+      "loss": 0.7364,
+      "step": 320
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.6898225545883179,
+      "eval_runtime": 61.3056,
+      "eval_samples_per_second": 0.457,
+      "eval_steps_per_second": 0.228,
+      "step": 320
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002,
+      "loss": 0.692,
+      "step": 321
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002,
+      "loss": 0.738,
+      "step": 322
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0002,
+      "loss": 0.7223,
+      "step": 323
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002,
+      "loss": 0.6788,
+      "step": 324
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002,
+      "loss": 0.6606,
+      "step": 325
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.0002,
+      "loss": 0.7093,
+      "step": 326
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002,
+      "loss": 0.6258,
+      "step": 327
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002,
+      "loss": 0.6896,
+      "step": 328
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002,
+      "loss": 0.6714,
+      "step": 329
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0002,
+      "loss": 0.66,
+      "step": 330
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002,
+      "loss": 0.7295,
+      "step": 331
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002,
+      "loss": 0.6773,
+      "step": 332
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.0002,
+      "loss": 0.753,
+      "step": 333
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002,
+      "loss": 0.6895,
+      "step": 334
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002,
+      "loss": 0.6615,
+      "step": 335
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0002,
+      "loss": 0.682,
+      "step": 336
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6778,
+      "step": 337
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 338
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6779,
+      "step": 339
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.0002,
+      "loss": 0.7147,
+      "step": 340
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.688534140586853,
+      "eval_runtime": 61.3371,
+      "eval_samples_per_second": 0.456,
+      "eval_steps_per_second": 0.228,
+      "step": 340
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.6342,
+      "step": 341
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.7297,
+      "step": 342
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 343
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002,
+      "loss": 0.5591,
+      "step": 344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002,
+      "loss": 0.6575,
+      "step": 345
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 346
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.0002,
+      "loss": 0.6006,
+      "step": 347
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0002,
+      "loss": 0.6698,
+      "step": 348
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0002,
+      "loss": 0.6104,
+      "step": 349
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 350
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002,
+      "loss": 0.6538,
+      "step": 351
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002,
+      "loss": 0.6504,
+      "step": 352
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 353
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002,
+      "loss": 0.6621,
+      "step": 354
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 355
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 356
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0002,
+      "loss": 0.6459,
+      "step": 357
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002,
+      "loss": 0.6154,
+      "step": 358
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002,
+      "loss": 0.6319,
+      "step": 359
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0002,
+      "loss": 0.6425,
+      "step": 360
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 0.6947513222694397,
+      "eval_runtime": 61.2171,
+      "eval_samples_per_second": 0.457,
+      "eval_steps_per_second": 0.229,
+      "step": 360
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6452,
+      "step": 361
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6406,
+      "step": 362
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 363
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.0002,
+      "loss": 0.6293,
+      "step": 364
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002,
+      "loss": 0.614,
+      "step": 365
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002,
+      "loss": 0.5518,
+      "step": 366
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 367
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 368
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 369
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002,
+      "loss": 0.5885,
+      "step": 370
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.0002,
+      "loss": 0.6438,
+      "step": 371
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6518,
+      "step": 372
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6697,
+      "step": 373
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6451,
+      "step": 374
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002,
+      "loss": 0.598,
+      "step": 375
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002,
+      "loss": 0.6624,
+      "step": 376
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0002,
+      "loss": 0.6355,
+      "step": 377
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 378
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002,
+      "loss": 0.612,
+      "step": 379
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002,
+      "loss": 0.5659,
+      "step": 380
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 0.6974608302116394,
+      "eval_runtime": 61.3547,
+      "eval_samples_per_second": 0.456,
+      "eval_steps_per_second": 0.228,
+      "step": 380
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.0002,
+      "loss": 0.5929,
+      "step": 381
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6472,
+      "step": 382
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6038,
+      "step": 383
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0002,
+      "loss": 0.6285,
+      "step": 384
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002,
+      "loss": 0.5863,
+      "step": 385
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 386
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002,
+      "loss": 0.5844,
+      "step": 387
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.0002,
+      "loss": 0.6012,
+      "step": 388
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6596,
+      "step": 389
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6475,
+      "step": 390
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.0002,
+      "loss": 0.6173,
+      "step": 391
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 392
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002,
+      "loss": 0.6361,
+      "step": 393
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002,
+      "loss": 0.5888,
+      "step": 394
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.0002,
+      "loss": 0.65,
+      "step": 395
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0002,
+      "loss": 0.6552,
+      "step": 396
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0002,
+      "loss": 0.6368,
+      "step": 397
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.0002,
+      "loss": 0.6418,
+      "step": 398
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002,
+      "loss": 0.6543,
+      "step": 399
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002,
+      "loss": 0.5669,
+      "step": 400
+    },
+    {
+      "epoch": 1.17,
+      "eval_loss": 0.702307403087616,
+      "eval_runtime": 61.3341,
+      "eval_samples_per_second": 0.457,
+      "eval_steps_per_second": 0.228,
+      "step": 400
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.0002,
+      "loss": 0.6119,
+      "step": 401
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002,
+      "loss": 0.5889,
+      "step": 402
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6247,
+      "step": 403
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6128,
+      "step": 404
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 405
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0002,
+      "loss": 0.626,
+      "step": 406
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0002,
+      "loss": 0.5961,
+      "step": 407
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.0002,
+      "loss": 0.6302,
+      "step": 408
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6094,
+      "step": 409
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002,
+      "loss": 0.5564,
+      "step": 410
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002,
+      "loss": 0.629,
+      "step": 411
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.0002,
+      "loss": 0.6261,
+      "step": 412
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0002,
+      "loss": 0.5624,
+      "step": 413
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 414
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.0002,
+      "loss": 0.6756,
+      "step": 415
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0002,
+      "loss": 0.5937,
+      "step": 416
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0002,
+      "loss": 0.5713,
+      "step": 417
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0002,
+      "loss": 0.6102,
+      "step": 418
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0002,
+      "loss": 0.5797,
+      "step": 419
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0002,
+      "loss": 0.6092,
+      "step": 420
+    },
+    {
+      "epoch": 1.23,
+      "eval_loss": 0.7039612531661987,
+      "eval_runtime": 62.2136,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 420
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0002,
+      "loss": 0.5736,
+      "step": 421
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.0002,
+      "loss": 0.674,
+      "step": 422
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0002,
+      "loss": 0.6295,
+      "step": 423
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0002,
+      "loss": 0.57,
+      "step": 424
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.0002,
+      "loss": 0.6051,
+      "step": 425
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 426
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 0.5821,
+      "step": 427
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 428
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002,
+      "loss": 0.5684,
+      "step": 429
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6334,
+      "step": 430
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6045,
+      "step": 431
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0002,
+      "loss": 0.6349,
+      "step": 432
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6292,
+      "step": 433
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6333,
+      "step": 434
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6427,
+      "step": 435
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 436
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0002,
+      "loss": 0.6448,
+      "step": 437
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0002,
+      "loss": 0.6285,
+      "step": 438
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 439
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0002,
+      "loss": 0.654,
+      "step": 440
+    },
+    {
+      "epoch": 1.29,
+      "eval_loss": 0.7050999402999878,
+      "eval_runtime": 61.9904,
+      "eval_samples_per_second": 0.452,
+      "eval_steps_per_second": 0.226,
+      "step": 440
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0002,
+      "loss": 0.5984,
+      "step": 441
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 442
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6255,
+      "step": 443
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6233,
+      "step": 444
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6742,
+      "step": 445
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002,
+      "loss": 0.6166,
+      "step": 446
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0002,
+      "loss": 0.5799,
+      "step": 447
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0002,
+      "loss": 0.6772,
+      "step": 448
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.0002,
+      "loss": 0.5719,
+      "step": 449
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002,
+      "loss": 0.5953,
+      "step": 450
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002,
+      "loss": 0.662,
+      "step": 451
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002,
+      "loss": 0.5992,
+      "step": 452
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 453
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 454
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0002,
+      "loss": 0.5786,
+      "step": 455
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0002,
+      "loss": 0.6,
+      "step": 456
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0002,
+      "loss": 0.669,
+      "step": 457
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0002,
+      "loss": 0.5618,
+      "step": 458
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0002,
+      "loss": 0.5647,
+      "step": 459
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0002,
+      "loss": 0.6719,
+      "step": 460
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 0.7046239972114563,
+      "eval_runtime": 62.1688,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 460
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5791,
+      "step": 461
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5312,
+      "step": 462
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5543,
+      "step": 463
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0002,
+      "loss": 0.653,
+      "step": 464
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0002,
+      "loss": 0.6132,
+      "step": 465
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.0002,
+      "loss": 0.5796,
+      "step": 466
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002,
+      "loss": 0.6607,
+      "step": 467
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002,
+      "loss": 0.6014,
+      "step": 468
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 469
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0002,
+      "loss": 0.6116,
+      "step": 470
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0002,
+      "loss": 0.6435,
+      "step": 471
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0002,
+      "loss": 0.543,
+      "step": 472
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.0002,
+      "loss": 0.5405,
+      "step": 473
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6107,
+      "step": 474
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6733,
+      "step": 475
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0002,
+      "loss": 0.5186,
+      "step": 476
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0002,
+      "loss": 0.6258,
+      "step": 477
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002,
+      "loss": 0.6232,
+      "step": 478
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002,
+      "loss": 0.5982,
+      "step": 479
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0002,
+      "loss": 0.5549,
+      "step": 480
+    },
+    {
+      "epoch": 1.4,
+      "eval_loss": 0.7041693329811096,
+      "eval_runtime": 62.1766,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 480
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0002,
+      "loss": 0.605,
+      "step": 481
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0002,
+      "loss": 0.603,
+      "step": 482
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 483
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 484
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 485
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0002,
+      "loss": 0.5801,
+      "step": 486
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.0002,
+      "loss": 0.5909,
+      "step": 487
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0002,
+      "loss": 0.6424,
+      "step": 488
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0002,
+      "loss": 0.6101,
+      "step": 489
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.0002,
+      "loss": 0.5982,
+      "step": 490
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6645,
+      "step": 491
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6298,
+      "step": 492
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 493
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0002,
+      "loss": 0.6467,
+      "step": 494
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0002,
+      "loss": 0.6117,
+      "step": 495
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0002,
+      "loss": 0.6694,
+      "step": 496
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.0002,
+      "loss": 0.5985,
+      "step": 497
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002,
+      "loss": 0.6075,
+      "step": 498
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002,
+      "loss": 0.6191,
+      "step": 499
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002,
+      "loss": 0.5919,
+      "step": 500
+    },
+    {
+      "epoch": 1.46,
+      "eval_loss": 0.7051344513893127,
+      "eval_runtime": 62.2046,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.0002,
+      "loss": 0.5726,
+      "step": 501
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0002,
+      "loss": 0.5952,
+      "step": 502
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0002,
+      "loss": 0.6182,
+      "step": 503
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.0002,
+      "loss": 0.6094,
+      "step": 504
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002,
+      "loss": 0.7522,
+      "step": 505
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 506
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002,
+      "loss": 0.6612,
+      "step": 507
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 508
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0002,
+      "loss": 0.5419,
+      "step": 509
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0002,
+      "loss": 0.624,
+      "step": 510
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.0002,
+      "loss": 0.6588,
+      "step": 511
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.64,
+      "step": 512
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.6037,
+      "step": 513
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0002,
+      "loss": 0.5667,
+      "step": 514
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0002,
+      "loss": 0.6318,
+      "step": 515
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0002,
+      "loss": 0.6111,
+      "step": 516
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0002,
+      "loss": 0.5828,
+      "step": 517
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.0002,
+      "loss": 0.6315,
+      "step": 518
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0002,
+      "loss": 0.6708,
+      "step": 519
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 520
+    },
+    {
+      "epoch": 1.52,
+      "eval_loss": 0.7005310654640198,
+      "eval_runtime": 62.2379,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 520
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.0002,
+      "loss": 0.6492,
+      "step": 521
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0002,
+      "loss": 0.5836,
+      "step": 522
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0002,
+      "loss": 0.6654,
+      "step": 523
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.0002,
+      "loss": 0.6045,
+      "step": 524
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 525
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0002,
+      "loss": 0.6397,
+      "step": 526
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 527
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.0002,
+      "loss": 0.6148,
+      "step": 528
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0002,
+      "loss": 0.6471,
+      "step": 529
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0002,
+      "loss": 0.5912,
+      "step": 530
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.0002,
+      "loss": 0.5945,
+      "step": 531
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002,
+      "loss": 0.6129,
+      "step": 532
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002,
+      "loss": 0.551,
+      "step": 533
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002,
+      "loss": 0.62,
+      "step": 534
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0002,
+      "loss": 0.6156,
+      "step": 535
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0002,
+      "loss": 0.6549,
+      "step": 536
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0002,
+      "loss": 0.5401,
+      "step": 537
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0002,
+      "loss": 0.6185,
+      "step": 538
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002,
+      "loss": 0.6341,
+      "step": 539
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002,
+      "loss": 0.597,
+      "step": 540
+    },
+    {
+      "epoch": 1.58,
+      "eval_loss": 0.7008891701698303,
+      "eval_runtime": 62.1778,
+      "eval_samples_per_second": 0.45,
+      "eval_steps_per_second": 0.225,
+      "step": 540
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 541
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.0002,
+      "loss": 0.6486,
+      "step": 542
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 543
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0002,
+      "loss": 0.5262,
+      "step": 544
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.0002,
+      "loss": 0.6278,
+      "step": 545
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0002,
+      "loss": 0.591,
+      "step": 546
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0002,
+      "loss": 0.5914,
+      "step": 547
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.0002,
+      "loss": 0.6375,
+      "step": 548
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0002,
+      "loss": 0.6598,
+      "step": 549
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0002,
+      "loss": 0.6193,
+      "step": 550
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0002,
+      "loss": 0.5755,
+      "step": 551
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0002,
+      "loss": 0.6287,
+      "step": 552
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0002,
+      "loss": 0.6286,
+      "step": 553
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0002,
+      "loss": 0.6218,
+      "step": 554
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0002,
+      "loss": 0.6391,
+      "step": 555
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6083,
+      "step": 556
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 557
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6226,
+      "step": 558
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0002,
+      "loss": 0.6081,
+      "step": 559
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0002,
+      "loss": 0.6853,
+      "step": 560
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 0.6983157396316528,
+      "eval_runtime": 61.7876,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.227,
+      "step": 560
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0002,
+      "loss": 0.5777,
+      "step": 561
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.0002,
+      "loss": 0.6013,
+      "step": 562
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0002,
+      "loss": 0.6303,
+      "step": 563
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0002,
+      "loss": 0.621,
+      "step": 564
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 565
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0002,
+      "loss": 0.6769,
+      "step": 566
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0002,
+      "loss": 0.5487,
+      "step": 567
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0002,
+      "loss": 0.6374,
+      "step": 568
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.0002,
+      "loss": 0.6545,
+      "step": 569
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0002,
+      "loss": 0.634,
+      "step": 570
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0002,
+      "loss": 0.6707,
+      "step": 571
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 572
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002,
+      "loss": 0.6057,
+      "step": 573
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002,
+      "loss": 0.584,
+      "step": 574
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002,
+      "loss": 0.6076,
+      "step": 575
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 576
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0002,
+      "loss": 0.6194,
+      "step": 577
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0002,
+      "loss": 0.5684,
+      "step": 578
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.0002,
+      "loss": 0.6828,
+      "step": 579
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0002,
+      "loss": 0.6412,
+      "step": 580
+    },
+    {
+      "epoch": 1.7,
+      "eval_loss": 0.7024718523025513,
+      "eval_runtime": 61.7724,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.227,
+      "step": 580
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0002,
+      "loss": 0.5933,
+      "step": 581
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0002,
+      "loss": 0.5369,
+      "step": 582
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.0002,
+      "loss": 0.627,
+      "step": 583
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0002,
+      "loss": 0.6564,
+      "step": 584
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0002,
+      "loss": 0.6337,
+      "step": 585
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0002,
+      "loss": 0.5968,
+      "step": 586
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0002,
+      "loss": 0.6298,
+      "step": 587
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 588
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0002,
+      "loss": 0.6049,
+      "step": 589
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6487,
+      "step": 590
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0002,
+      "loss": 0.5785,
+      "step": 591
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6016,
+      "step": 592
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.0002,
+      "loss": 0.6221,
+      "step": 593
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0002,
+      "loss": 0.5588,
+      "step": 594
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0002,
+      "loss": 0.6571,
+      "step": 595
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0002,
+      "loss": 0.6217,
+      "step": 596
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.6463,
+      "step": 597
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.5977,
+      "step": 598
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.5835,
+      "step": 599
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.0002,
+      "loss": 0.601,
+      "step": 600
+    },
+    {
+      "epoch": 1.75,
+      "eval_loss": 0.7026137113571167,
+      "eval_runtime": 61.7876,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.227,
+      "step": 600
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0002,
+      "loss": 0.5938,
+      "step": 601
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0002,
+      "loss": 0.6354,
+      "step": 602
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0002,
+      "loss": 0.6168,
+      "step": 603
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0002,
+      "loss": 0.6017,
+      "step": 604
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0002,
+      "loss": 0.5839,
+      "step": 605
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0002,
+      "loss": 0.671,
+      "step": 606
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.0002,
+      "loss": 0.5894,
+      "step": 607
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0002,
+      "loss": 0.6147,
+      "step": 608
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0002,
+      "loss": 0.6026,
+      "step": 609
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0002,
+      "loss": 0.6048,
+      "step": 610
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0002,
+      "loss": 0.7008,
+      "step": 611
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 612
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.0002,
+      "loss": 0.6234,
+      "step": 613
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6346,
+      "step": 614
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6664,
+      "step": 615
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0002,
+      "loss": 0.5966,
+      "step": 616
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0002,
+      "loss": 0.6467,
+      "step": 617
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0002,
+      "loss": 0.5757,
+      "step": 618
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0002,
+      "loss": 0.6235,
+      "step": 619
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.0002,
+      "loss": 0.6382,
+      "step": 620
+    },
+    {
+      "epoch": 1.81,
+      "eval_loss": 0.6993085145950317,
+      "eval_runtime": 61.7213,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 620
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0002,
+      "loss": 0.5901,
+      "step": 621
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0002,
+      "loss": 0.5741,
+      "step": 622
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0002,
+      "loss": 0.6407,
+      "step": 623
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0002,
+      "loss": 0.6179,
+      "step": 624
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0002,
+      "loss": 0.5983,
+      "step": 625
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0002,
+      "loss": 0.5555,
+      "step": 626
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.0002,
+      "loss": 0.5871,
+      "step": 627
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0002,
+      "loss": 0.6367,
+      "step": 628
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 629
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0002,
+      "loss": 0.5815,
+      "step": 630
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0002,
+      "loss": 0.6277,
+      "step": 631
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0002,
+      "loss": 0.5834,
+      "step": 632
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0002,
+      "loss": 0.6578,
+      "step": 633
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.0002,
+      "loss": 0.5893,
+      "step": 634
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0002,
+      "loss": 0.5511,
+      "step": 635
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0002,
+      "loss": 0.6167,
+      "step": 636
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.0002,
+      "loss": 0.6312,
+      "step": 637
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0002,
+      "loss": 0.6239,
+      "step": 638
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0002,
+      "loss": 0.564,
+      "step": 639
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0002,
+      "loss": 0.5774,
+      "step": 640
+    },
+    {
+      "epoch": 1.87,
+      "eval_loss": 0.6974296569824219,
+      "eval_runtime": 61.7369,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 640
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.0002,
+      "loss": 0.5513,
+      "step": 641
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0002,
+      "loss": 0.6213,
+      "step": 642
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0002,
+      "loss": 0.6032,
+      "step": 643
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 644
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0002,
+      "loss": 0.5719,
+      "step": 645
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 646
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0002,
+      "loss": 0.648,
+      "step": 647
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0002,
+      "loss": 0.5754,
+      "step": 648
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0002,
+      "loss": 0.6506,
+      "step": 649
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0002,
+      "loss": 0.7017,
+      "step": 650
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0002,
+      "loss": 0.6052,
+      "step": 651
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6131,
+      "step": 652
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6385,
+      "step": 653
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0002,
+      "loss": 0.6188,
+      "step": 654
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6164,
+      "step": 655
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6174,
+      "step": 656
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0002,
+      "loss": 0.531,
+      "step": 657
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.0002,
+      "loss": 0.6834,
+      "step": 658
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0002,
+      "loss": 0.655,
+      "step": 659
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0002,
+      "loss": 0.5842,
+      "step": 660
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 0.6995061039924622,
+      "eval_runtime": 61.7079,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 660
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.0002,
+      "loss": 0.5866,
+      "step": 661
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0002,
+      "loss": 0.6215,
+      "step": 662
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0002,
+      "loss": 0.6987,
+      "step": 663
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0002,
+      "loss": 0.5988,
+      "step": 664
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.0002,
+      "loss": 0.6002,
+      "step": 665
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0002,
+      "loss": 0.6442,
+      "step": 666
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0002,
+      "loss": 0.6153,
+      "step": 667
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0002,
+      "loss": 0.633,
+      "step": 668
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002,
+      "loss": 0.5691,
+      "step": 669
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002,
+      "loss": 0.6204,
+      "step": 670
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002,
+      "loss": 0.567,
+      "step": 671
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0002,
+      "loss": 0.689,
+      "step": 672
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0002,
+      "loss": 0.5982,
+      "step": 673
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0002,
+      "loss": 0.6236,
+      "step": 674
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0002,
+      "loss": 0.6496,
+      "step": 675
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0002,
+      "loss": 0.6605,
+      "step": 676
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0002,
+      "loss": 0.6328,
+      "step": 677
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.0002,
+      "loss": 0.5864,
+      "step": 678
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0002,
+      "loss": 0.5976,
+      "step": 679
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6171,
+      "step": 680
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 0.6999563574790955,
+      "eval_runtime": 61.676,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 680
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6275,
+      "step": 681
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.0002,
+      "loss": 0.6161,
+      "step": 682
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0002,
+      "loss": 0.6348,
+      "step": 683
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0002,
+      "loss": 0.5895,
+      "step": 684
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0002,
+      "loss": 0.546,
+      "step": 685
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0002,
+      "loss": 0.5519,
+      "step": 686
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0002,
+      "loss": 0.5457,
+      "step": 687
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0002,
+      "loss": 0.5834,
+      "step": 688
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.0002,
+      "loss": 0.5073,
+      "step": 689
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0002,
+      "loss": 0.5467,
+      "step": 690
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0002,
+      "loss": 0.5233,
+      "step": 691
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0002,
+      "loss": 0.5723,
+      "step": 692
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0002,
+      "loss": 0.5043,
+      "step": 693
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 694
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0002,
+      "loss": 0.5245,
+      "step": 695
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0002,
+      "loss": 0.5059,
+      "step": 696
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0002,
+      "loss": 0.5303,
+      "step": 697
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0002,
+      "loss": 0.5366,
+      "step": 698
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0002,
+      "loss": 0.475,
+      "step": 699
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0002,
+      "loss": 0.5124,
+      "step": 700
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 0.7223584055900574,
+      "eval_runtime": 61.8888,
+      "eval_samples_per_second": 0.452,
+      "eval_steps_per_second": 0.226,
+      "step": 700
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0002,
+      "loss": 0.5983,
+      "step": 701
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.0002,
+      "loss": 0.514,
+      "step": 702
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0002,
+      "loss": 0.4672,
+      "step": 703
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0002,
+      "loss": 0.4894,
+      "step": 704
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 705
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.0002,
+      "loss": 0.5391,
+      "step": 706
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0002,
+      "loss": 0.6005,
+      "step": 707
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 708
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 709
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0002,
+      "loss": 0.5274,
+      "step": 710
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0002,
+      "loss": 0.5376,
+      "step": 711
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0002,
+      "loss": 0.5037,
+      "step": 712
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0002,
+      "loss": 0.5559,
+      "step": 713
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0002,
+      "loss": 0.6082,
+      "step": 714
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0002,
+      "loss": 0.4927,
+      "step": 715
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.0002,
+      "loss": 0.5153,
+      "step": 716
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0002,
+      "loss": 0.5653,
+      "step": 717
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 718
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.0002,
+      "loss": 0.528,
+      "step": 719
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0002,
+      "loss": 0.5229,
+      "step": 720
+    },
+    {
+      "epoch": 2.11,
+      "eval_loss": 0.7286432385444641,
+      "eval_runtime": 61.828,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.226,
+      "step": 720
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0002,
+      "loss": 0.4636,
+      "step": 721
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0002,
+      "loss": 0.5089,
+      "step": 722
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.0002,
+      "loss": 0.4956,
+      "step": 723
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0002,
+      "loss": 0.4972,
+      "step": 724
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0002,
+      "loss": 0.589,
+      "step": 725
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0002,
+      "loss": 0.4694,
+      "step": 726
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0002,
+      "loss": 0.5875,
+      "step": 727
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0002,
+      "loss": 0.574,
+      "step": 728
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0002,
+      "loss": 0.5462,
+      "step": 729
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.0002,
+      "loss": 0.4185,
+      "step": 730
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0002,
+      "loss": 0.5979,
+      "step": 731
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0002,
+      "loss": 0.5003,
+      "step": 732
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.0002,
+      "loss": 0.5572,
+      "step": 733
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0002,
+      "loss": 0.5367,
+      "step": 734
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0002,
+      "loss": 0.5601,
+      "step": 735
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0002,
+      "loss": 0.5571,
+      "step": 736
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0002,
+      "loss": 0.5837,
+      "step": 737
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0002,
+      "loss": 0.557,
+      "step": 738
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0002,
+      "loss": 0.5464,
+      "step": 739
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.0002,
+      "loss": 0.52,
+      "step": 740
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 0.7195437550544739,
+      "eval_runtime": 61.848,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.226,
+      "step": 740
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0002,
+      "loss": 0.5946,
+      "step": 741
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0002,
+      "loss": 0.535,
+      "step": 742
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.0002,
+      "loss": 0.5561,
+      "step": 743
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0002,
+      "loss": 0.5079,
+      "step": 744
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0002,
+      "loss": 0.5838,
+      "step": 745
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0002,
+      "loss": 0.5281,
+      "step": 746
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0002,
+      "loss": 0.5582,
+      "step": 747
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0002,
+      "loss": 0.5613,
+      "step": 748
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0002,
+      "loss": 0.5226,
+      "step": 749
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0002,
+      "loss": 0.5522,
+      "step": 750
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0002,
+      "loss": 0.5964,
+      "step": 751
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0002,
+      "loss": 0.4685,
+      "step": 752
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0002,
+      "loss": 0.5843,
+      "step": 753
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.0002,
+      "loss": 0.5156,
+      "step": 754
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0002,
+      "loss": 0.5373,
+      "step": 755
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0002,
+      "loss": 0.522,
+      "step": 756
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.0002,
+      "loss": 0.4829,
+      "step": 757
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0002,
+      "loss": 0.4779,
+      "step": 758
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0002,
+      "loss": 0.5531,
+      "step": 759
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0002,
+      "loss": 0.5387,
+      "step": 760
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 0.7240059971809387,
+      "eval_runtime": 61.7373,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 760
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0002,
+      "loss": 0.5675,
+      "step": 761
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0002,
+      "loss": 0.5331,
+      "step": 762
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0002,
+      "loss": 0.5296,
+      "step": 763
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0002,
+      "loss": 0.4718,
+      "step": 764
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0002,
+      "loss": 0.5353,
+      "step": 765
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0002,
+      "loss": 0.5441,
+      "step": 766
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.0002,
+      "loss": 0.5236,
+      "step": 767
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002,
+      "loss": 0.5182,
+      "step": 768
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 769
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002,
+      "loss": 0.5267,
+      "step": 770
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 771
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0002,
+      "loss": 0.5234,
+      "step": 772
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0002,
+      "loss": 0.5432,
+      "step": 773
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0002,
+      "loss": 0.4722,
+      "step": 774
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0002,
+      "loss": 0.4726,
+      "step": 775
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 776
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0002,
+      "loss": 0.5095,
+      "step": 777
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.0002,
+      "loss": 0.5569,
+      "step": 778
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0002,
+      "loss": 0.5129,
+      "step": 779
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0002,
+      "loss": 0.5071,
+      "step": 780
+    },
+    {
+      "epoch": 2.28,
+      "eval_loss": 0.7296870350837708,
+      "eval_runtime": 61.7322,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 780
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.0002,
+      "loss": 0.4779,
+      "step": 781
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0002,
+      "loss": 0.5482,
+      "step": 782
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0002,
+      "loss": 0.4948,
+      "step": 783
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.0002,
+      "loss": 0.5587,
+      "step": 784
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0002,
+      "loss": 0.4897,
+      "step": 785
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0002,
+      "loss": 0.5979,
+      "step": 786
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0002,
+      "loss": 0.5621,
+      "step": 787
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0002,
+      "loss": 0.5944,
+      "step": 788
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0002,
+      "loss": 0.5933,
+      "step": 789
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 790
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.0002,
+      "loss": 0.4408,
+      "step": 791
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002,
+      "loss": 0.5242,
+      "step": 792
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002,
+      "loss": 0.552,
+      "step": 793
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002,
+      "loss": 0.5907,
+      "step": 794
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002,
+      "loss": 0.5604,
+      "step": 795
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0002,
+      "loss": 0.5422,
+      "step": 796
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0002,
+      "loss": 0.5702,
+      "step": 797
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.0002,
+      "loss": 0.491,
+      "step": 798
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0002,
+      "loss": 0.496,
+      "step": 799
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0002,
+      "loss": 0.5412,
+      "step": 800
+    },
+    {
+      "epoch": 2.34,
+      "eval_loss": 0.7235637307167053,
+      "eval_runtime": 61.6768,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 800
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.0002,
+      "loss": 0.4834,
+      "step": 801
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5088,
+      "step": 802
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5045,
+      "step": 803
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0002,
+      "loss": 0.4352,
+      "step": 804
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.0002,
+      "loss": 0.5192,
+      "step": 805
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0002,
+      "loss": 0.5176,
+      "step": 806
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0002,
+      "loss": 0.4707,
+      "step": 807
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0002,
+      "loss": 0.568,
+      "step": 808
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0002,
+      "loss": 0.4248,
+      "step": 809
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0002,
+      "loss": 0.4145,
+      "step": 810
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0002,
+      "loss": 0.4595,
+      "step": 811
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.0002,
+      "loss": 0.4506,
+      "step": 812
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002,
+      "loss": 0.5198,
+      "step": 813
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002,
+      "loss": 0.5281,
+      "step": 814
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 815
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0002,
+      "loss": 0.5196,
+      "step": 816
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0002,
+      "loss": 0.5062,
+      "step": 817
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0002,
+      "loss": 0.4679,
+      "step": 818
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.0002,
+      "loss": 0.5264,
+      "step": 819
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0002,
+      "loss": 0.4471,
+      "step": 820
+    },
+    {
+      "epoch": 2.4,
+      "eval_loss": 0.73420250415802,
+      "eval_runtime": 62.4841,
+      "eval_samples_per_second": 0.448,
+      "eval_steps_per_second": 0.224,
+      "step": 820
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0002,
+      "loss": 0.4711,
+      "step": 821
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0002,
+      "loss": 0.5181,
+      "step": 822
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002,
+      "loss": 0.4915,
+      "step": 823
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002,
+      "loss": 0.5042,
+      "step": 824
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002,
+      "loss": 0.4675,
+      "step": 825
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0002,
+      "loss": 0.511,
+      "step": 826
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0002,
+      "loss": 0.4529,
+      "step": 827
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0002,
+      "loss": 0.4659,
+      "step": 828
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0002,
+      "loss": 0.5028,
+      "step": 829
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 830
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0002,
+      "loss": 0.4852,
+      "step": 831
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.0002,
+      "loss": 0.5718,
+      "step": 832
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0002,
+      "loss": 0.4895,
+      "step": 833
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0002,
+      "loss": 0.5003,
+      "step": 834
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0002,
+      "loss": 0.5497,
+      "step": 835
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.0002,
+      "loss": 0.4877,
+      "step": 836
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0002,
+      "loss": 0.4789,
+      "step": 837
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0002,
+      "loss": 0.5541,
+      "step": 838
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.0002,
+      "loss": 0.5844,
+      "step": 839
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002,
+      "loss": 0.5965,
+      "step": 840
+    },
+    {
+      "epoch": 2.46,
+      "eval_loss": 0.735193133354187,
+      "eval_runtime": 62.3714,
+      "eval_samples_per_second": 0.449,
+      "eval_steps_per_second": 0.224,
+      "step": 840
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002,
+      "loss": 0.5418,
+      "step": 841
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002,
+      "loss": 0.3698,
+      "step": 842
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.0002,
+      "loss": 0.5002,
+      "step": 843
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0002,
+      "loss": 0.4836,
+      "step": 844
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0002,
+      "loss": 0.4541,
+      "step": 845
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0002,
+      "loss": 0.4884,
+      "step": 846
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0002,
+      "loss": 0.4668,
+      "step": 847
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0002,
+      "loss": 0.4154,
+      "step": 848
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.0002,
+      "loss": 0.559,
+      "step": 849
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0002,
+      "loss": 0.4807,
+      "step": 850
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0002,
+      "loss": 0.5679,
+      "step": 851
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0002,
+      "loss": 0.4417,
+      "step": 852
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0002,
+      "loss": 0.5014,
+      "step": 853
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0002,
+      "loss": 0.5305,
+      "step": 854
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0002,
+      "loss": 0.5141,
+      "step": 855
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.0002,
+      "loss": 0.5626,
+      "step": 856
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0002,
+      "loss": 0.4965,
+      "step": 857
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0002,
+      "loss": 0.5077,
+      "step": 858
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0002,
+      "loss": 0.4637,
+      "step": 859
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0002,
+      "loss": 0.5393,
+      "step": 860
+    },
+    {
+      "epoch": 2.51,
+      "eval_loss": 0.73292475938797,
+      "eval_runtime": 61.6664,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 860
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0002,
+      "loss": 0.5023,
+      "step": 861
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0002,
+      "loss": 0.5325,
+      "step": 862
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.0002,
+      "loss": 0.5592,
+      "step": 863
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0002,
+      "loss": 0.5408,
+      "step": 864
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0002,
+      "loss": 0.5295,
+      "step": 865
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.0002,
+      "loss": 0.5647,
+      "step": 866
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0002,
+      "loss": 0.5128,
+      "step": 867
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0002,
+      "loss": 0.4901,
+      "step": 868
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0002,
+      "loss": 0.4966,
+      "step": 869
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0002,
+      "loss": 0.5908,
+      "step": 870
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0002,
+      "loss": 0.5486,
+      "step": 871
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0002,
+      "loss": 0.5396,
+      "step": 872
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.0002,
+      "loss": 0.4843,
+      "step": 873
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0002,
+      "loss": 0.4824,
+      "step": 874
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0002,
+      "loss": 0.5362,
+      "step": 875
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0002,
+      "loss": 0.5343,
+      "step": 876
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.0002,
+      "loss": 0.5457,
+      "step": 877
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0002,
+      "loss": 0.6091,
+      "step": 878
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0002,
+      "loss": 0.5453,
+      "step": 879
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.0002,
+      "loss": 0.4707,
+      "step": 880
+    },
+    {
+      "epoch": 2.57,
+      "eval_loss": 0.7356553673744202,
+      "eval_runtime": 61.7528,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.227,
+      "step": 880
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0002,
+      "loss": 0.5127,
+      "step": 881
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0002,
+      "loss": 0.5166,
+      "step": 882
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0002,
+      "loss": 0.5388,
+      "step": 883
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0002,
+      "loss": 0.4538,
+      "step": 884
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0002,
+      "loss": 0.4928,
+      "step": 885
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0002,
+      "loss": 0.5229,
+      "step": 886
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.0002,
+      "loss": 0.4419,
+      "step": 887
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0002,
+      "loss": 0.5913,
+      "step": 888
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0002,
+      "loss": 0.5544,
+      "step": 889
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0002,
+      "loss": 0.5266,
+      "step": 890
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0002,
+      "loss": 0.5841,
+      "step": 891
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0002,
+      "loss": 0.4999,
+      "step": 892
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0002,
+      "loss": 0.5783,
+      "step": 893
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.0002,
+      "loss": 0.5481,
+      "step": 894
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0002,
+      "loss": 0.5317,
+      "step": 895
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0002,
+      "loss": 0.5663,
+      "step": 896
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0002,
+      "loss": 0.506,
+      "step": 897
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0002,
+      "loss": 0.5079,
+      "step": 898
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0002,
+      "loss": 0.5085,
+      "step": 899
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0002,
+      "loss": 0.4813,
+      "step": 900
+    },
+    {
+      "epoch": 2.63,
+      "eval_loss": 0.7377252578735352,
+      "eval_runtime": 61.9163,
+      "eval_samples_per_second": 0.452,
+      "eval_steps_per_second": 0.226,
+      "step": 900
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.0002,
+      "loss": 0.5178,
+      "step": 901
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0002,
+      "loss": 0.5328,
+      "step": 902
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0002,
+      "loss": 0.6043,
+      "step": 903
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0002,
+      "loss": 0.5359,
+      "step": 904
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0002,
+      "loss": 0.5758,
+      "step": 905
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0002,
+      "loss": 0.4767,
+      "step": 906
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0002,
+      "loss": 0.4952,
+      "step": 907
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.0002,
+      "loss": 0.5826,
+      "step": 908
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0002,
+      "loss": 0.5641,
+      "step": 909
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0002,
+      "loss": 0.5184,
+      "step": 910
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0002,
+      "loss": 0.5327,
+      "step": 911
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0002,
+      "loss": 0.5172,
+      "step": 912
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0002,
+      "loss": 0.4997,
+      "step": 913
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.0002,
+      "loss": 0.5299,
+      "step": 914
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0002,
+      "loss": 0.5578,
+      "step": 915
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0002,
+      "loss": 0.4775,
+      "step": 916
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0002,
+      "loss": 0.4951,
+      "step": 917
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0002,
+      "loss": 0.5286,
+      "step": 918
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0002,
+      "loss": 0.4307,
+      "step": 919
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0002,
+      "loss": 0.4681,
+      "step": 920
+    },
+    {
+      "epoch": 2.69,
+      "eval_loss": 0.7411596179008484,
+      "eval_runtime": 61.7427,
+      "eval_samples_per_second": 0.453,
+      "eval_steps_per_second": 0.227,
+      "step": 920
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0002,
+      "loss": 0.5131,
+      "step": 921
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0002,
+      "loss": 0.4303,
+      "step": 922
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0002,
+      "loss": 0.5044,
+      "step": 923
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0002,
+      "loss": 0.5766,
+      "step": 924
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.0002,
+      "loss": 0.4604,
+      "step": 925
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.0002,
+      "loss": 0.4773,
+      "step": 926
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.0002,
+      "loss": 0.5332,
+      "step": 927
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.0002,
+      "loss": 0.5177,
+      "step": 928
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0002,
+      "loss": 0.5609,
+      "step": 929
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0002,
+      "loss": 0.5228,
+      "step": 930
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.0002,
+      "loss": 0.5304,
+      "step": 931
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0002,
+      "loss": 0.5421,
+      "step": 932
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0002,
+      "loss": 0.4837,
+      "step": 933
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0002,
+      "loss": 0.5016,
+      "step": 934
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.0002,
+      "loss": 0.5301,
+      "step": 935
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0002,
+      "loss": 0.473,
+      "step": 936
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0002,
+      "loss": 0.4369,
+      "step": 937
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.0002,
+      "loss": 0.5617,
+      "step": 938
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0002,
+      "loss": 0.4909,
+      "step": 939
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0002,
+      "loss": 0.4995,
+      "step": 940
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.733302652835846,
+      "eval_runtime": 61.6867,
+      "eval_samples_per_second": 0.454,
+      "eval_steps_per_second": 0.227,
+      "step": 940
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0002,
+      "loss": 0.5515,
+      "step": 941
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.0002,
+      "loss": 0.4858,
+      "step": 942
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0002,
+      "loss": 0.5517,
+      "step": 943
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0002,
+      "loss": 0.5214,
+      "step": 944
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0002,
+      "loss": 0.5105,
+      "step": 945
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0002,
+      "loss": 0.4557,
+      "step": 946
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0002,
+      "loss": 0.5622,
+      "step": 947
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0002,
+      "loss": 0.4703,
+      "step": 948
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0002,
+      "loss": 0.561,
+      "step": 949
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.0002,
+      "loss": 0.533,
+      "step": 950
+    }
+  ],
+  "max_steps": 1026,
+  "num_train_epochs": 3,
+  "total_flos": 1.2000443511865344e+18,
+  "trial_name": null,
+  "trial_params": null
+}