diff --git "a/mistral-7b/trainer_state.json" "b/mistral-7b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/mistral-7b/trainer_state.json"
@@ -0,0 +1,9821 @@
+{
+  "best_metric": 1.949257254600525,
+  "best_model_checkpoint": "./lora-out/checkpoint-1600",
+  "epoch": 2.416918429003021,
+  "eval_steps": 64,
+  "global_step": 1600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.388888888888889e-07,
+      "loss": 2.0137,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.777777777777778e-07,
+      "loss": 2.0067,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.166666666666667e-07,
+      "loss": 1.9034,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.555555555555556e-07,
+      "loss": 2.031,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.944444444444445e-07,
+      "loss": 2.0495,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.333333333333334e-07,
+      "loss": 1.8877,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.722222222222222e-07,
+      "loss": 2.1799,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1111111111111112e-06,
+      "loss": 1.8431,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.25e-06,
+      "loss": 2.0854,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.388888888888889e-06,
+      "loss": 2.1238,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5277777777777778e-06,
+      "loss": 2.0245,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6666666666666669e-06,
+      "loss": 2.1664,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 1.8321,
+      "step": 13
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9444444444444444e-06,
+      "loss": 1.9979,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.9943,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.2222222222222225e-06,
+      "loss": 2.1295,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 1.9052,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.5e-06,
+      "loss": 1.9968,
+      "step": 18
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.638888888888889e-06,
+      "loss": 1.7532,
+      "step": 19
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.777777777777778e-06,
+      "loss": 1.8524,
+      "step": 20
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 2.0174,
+      "step": 21
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.0555555555555556e-06,
+      "loss": 1.7299,
+      "step": 22
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.1944444444444447e-06,
+      "loss": 2.0563,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.3333333333333337e-06,
+      "loss": 2.2994,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 1.7782,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 1.9417,
+      "step": 26
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.7499999999999997e-06,
+      "loss": 1.8848,
+      "step": 27
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 3.888888888888889e-06,
+      "loss": 1.8163,
+      "step": 28
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.027777777777778e-06,
+      "loss": 2.061,
+      "step": 29
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.8327,
+      "step": 30
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 2.0376,
+      "step": 31
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.444444444444445e-06,
+      "loss": 1.9657,
+      "step": 32
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 2.1065,
+      "step": 33
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 1.9817,
+      "step": 34
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 1.8846,
+      "step": 35
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-06,
+      "loss": 1.8725,
+      "step": 36
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.1388888888888895e-06,
+      "loss": 1.5476,
+      "step": 37
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.277777777777778e-06,
+      "loss": 1.9259,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 1.7124,
+      "step": 39
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.555555555555556e-06,
+      "loss": 2.2361,
+      "step": 40
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 2.0438,
+      "step": 41
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 1.9475,
+      "step": 42
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5.972222222222223e-06,
+      "loss": 2.2165,
+      "step": 43
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.111111111111111e-06,
+      "loss": 1.9729,
+      "step": 44
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.25e-06,
+      "loss": 1.8971,
+      "step": 45
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.388888888888889e-06,
+      "loss": 1.8945,
+      "step": 46
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 2.0578,
+      "step": 47
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.6666666666666675e-06,
+      "loss": 2.1846,
+      "step": 48
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.805555555555556e-06,
+      "loss": 1.9433,
+      "step": 49
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 1.8438,
+      "step": 50
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.083333333333333e-06,
+      "loss": 1.9077,
+      "step": 51
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 1.811,
+      "step": 52
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.361111111111111e-06,
+      "loss": 1.8771,
+      "step": 53
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.499999999999999e-06,
+      "loss": 2.1937,
+      "step": 54
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.63888888888889e-06,
+      "loss": 1.9282,
+      "step": 55
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 7.777777777777777e-06,
+      "loss": 2.2111,
+      "step": 56
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 7.916666666666668e-06,
+      "loss": 1.8899,
+      "step": 57
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.055555555555556e-06,
+      "loss": 1.7884,
+      "step": 58
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 2.0242,
+      "step": 59
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.9023,
+      "step": 60
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 1.9534,
+      "step": 61
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 1.9204,
+      "step": 62
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 1.8808,
+      "step": 63
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 8.88888888888889e-06,
+      "loss": 1.994,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 2.018505811691284,
+      "eval_runtime": 58.7674,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 2.0174,
+      "step": 65
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 2.1546,
+      "step": 66
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.305555555555555e-06,
+      "loss": 1.9271,
+      "step": 67
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.9807,
+      "step": 68
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 9.583333333333334e-06,
+      "loss": 2.0853,
+      "step": 69
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 2.0573,
+      "step": 70
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 9.86111111111111e-06,
+      "loss": 1.8761,
+      "step": 71
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1e-05,
+      "loss": 2.1104,
+      "step": 72
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.0138888888888888e-05,
+      "loss": 1.7596,
+      "step": 73
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.0277777777777779e-05,
+      "loss": 1.9672,
+      "step": 74
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.0416666666666666e-05,
+      "loss": 2.0319,
+      "step": 75
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.0555555555555555e-05,
+      "loss": 1.9108,
+      "step": 76
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0694444444444444e-05,
+      "loss": 2.0446,
+      "step": 77
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 1.9723,
+      "step": 78
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.0972222222222223e-05,
+      "loss": 2.0775,
+      "step": 79
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.1111111111111112e-05,
+      "loss": 2.0981,
+      "step": 80
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.125e-05,
+      "loss": 2.1977,
+      "step": 81
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.138888888888889e-05,
+      "loss": 1.8683,
+      "step": 82
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1527777777777777e-05,
+      "loss": 1.9426,
+      "step": 83
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 2.1383,
+      "step": 84
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1805555555555555e-05,
+      "loss": 2.0234,
+      "step": 85
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.1944444444444446e-05,
+      "loss": 1.9332,
+      "step": 86
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 2.0657,
+      "step": 87
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.2222222222222222e-05,
+      "loss": 1.7703,
+      "step": 88
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.2361111111111112e-05,
+      "loss": 2.2457,
+      "step": 89
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.25e-05,
+      "loss": 2.0096,
+      "step": 90
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.263888888888889e-05,
+      "loss": 2.0313,
+      "step": 91
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.2777777777777779e-05,
+      "loss": 1.992,
+      "step": 92
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 1.9596,
+      "step": 93
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.3055555555555557e-05,
+      "loss": 1.8446,
+      "step": 94
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.3194444444444444e-05,
+      "loss": 2.0009,
+      "step": 95
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.3333333333333335e-05,
+      "loss": 2.1015,
+      "step": 96
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.3472222222222222e-05,
+      "loss": 1.9865,
+      "step": 97
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.3611111111111111e-05,
+      "loss": 1.9925,
+      "step": 98
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.375e-05,
+      "loss": 1.9774,
+      "step": 99
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.388888888888889e-05,
+      "loss": 2.0782,
+      "step": 100
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.4027777777777777e-05,
+      "loss": 2.0492,
+      "step": 101
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.4166666666666666e-05,
+      "loss": 2.0521,
+      "step": 102
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4305555555555557e-05,
+      "loss": 2.022,
+      "step": 103
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4444444444444446e-05,
+      "loss": 1.86,
+      "step": 104
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4583333333333333e-05,
+      "loss": 1.9392,
+      "step": 105
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4722222222222222e-05,
+      "loss": 1.9623,
+      "step": 106
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4861111111111113e-05,
+      "loss": 1.8599,
+      "step": 107
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 1.9744,
+      "step": 108
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.513888888888889e-05,
+      "loss": 1.888,
+      "step": 109
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.527777777777778e-05,
+      "loss": 1.8622,
+      "step": 110
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5416666666666664e-05,
+      "loss": 1.9167,
+      "step": 111
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 1.845,
+      "step": 112
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5694444444444446e-05,
+      "loss": 1.7881,
+      "step": 113
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5833333333333336e-05,
+      "loss": 2.0639,
+      "step": 114
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.597222222222222e-05,
+      "loss": 1.8555,
+      "step": 115
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.611111111111111e-05,
+      "loss": 1.8149,
+      "step": 116
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 1.7631,
+      "step": 117
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.638888888888889e-05,
+      "loss": 2.1632,
+      "step": 118
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6527777777777777e-05,
+      "loss": 1.9452,
+      "step": 119
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 2.1971,
+      "step": 120
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6805555555555558e-05,
+      "loss": 1.8998,
+      "step": 121
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.6944444444444446e-05,
+      "loss": 1.7745,
+      "step": 122
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 1.8597,
+      "step": 123
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 1.9146,
+      "step": 124
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.736111111111111e-05,
+      "loss": 1.9216,
+      "step": 125
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 1.9635,
+      "step": 126
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.763888888888889e-05,
+      "loss": 1.7874,
+      "step": 127
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 1.99,
+      "step": 128
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 2.0141048431396484,
+      "eval_runtime": 58.7942,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 128
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 2.2786,
+      "step": 129
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8055555555555558e-05,
+      "loss": 1.5962,
+      "step": 130
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8194444444444445e-05,
+      "loss": 1.9926,
+      "step": 131
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 2.0176,
+      "step": 132
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8472222222222224e-05,
+      "loss": 1.8957,
+      "step": 133
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.861111111111111e-05,
+      "loss": 1.7665,
+      "step": 134
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.875e-05,
+      "loss": 1.9008,
+      "step": 135
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 2.0322,
+      "step": 136
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.902777777777778e-05,
+      "loss": 1.9034,
+      "step": 137
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9166666666666667e-05,
+      "loss": 1.8559,
+      "step": 138
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9305555555555555e-05,
+      "loss": 2.112,
+      "step": 139
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 2.004,
+      "step": 140
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9583333333333336e-05,
+      "loss": 2.0082,
+      "step": 141
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.972222222222222e-05,
+      "loss": 1.9579,
+      "step": 142
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.986111111111111e-05,
+      "loss": 2.1039,
+      "step": 143
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2e-05,
+      "loss": 1.8939,
+      "step": 144
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.0138888888888892e-05,
+      "loss": 2.1706,
+      "step": 145
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.0277777777777776e-05,
+      "loss": 1.8098,
+      "step": 146
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.0416666666666667e-05,
+      "loss": 1.9249,
+      "step": 147
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.0555555555555558e-05,
+      "loss": 1.8126,
+      "step": 148
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.0694444444444445e-05,
+      "loss": 2.2817,
+      "step": 149
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.0833333333333333e-05,
+      "loss": 1.7509,
+      "step": 150
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.0972222222222223e-05,
+      "loss": 1.6783,
+      "step": 151
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 1.7003,
+      "step": 152
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.125e-05,
+      "loss": 1.8062,
+      "step": 153
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.138888888888889e-05,
+      "loss": 1.8352,
+      "step": 154
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.152777777777778e-05,
+      "loss": 1.9913,
+      "step": 155
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 1.8738,
+      "step": 156
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.1805555555555558e-05,
+      "loss": 2.0151,
+      "step": 157
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.1944444444444445e-05,
+      "loss": 1.8788,
+      "step": 158
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.2083333333333333e-05,
+      "loss": 1.876,
+      "step": 159
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 1.9086,
+      "step": 160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.2361111111111114e-05,
+      "loss": 1.9247,
+      "step": 161
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.25e-05,
+      "loss": 2.2077,
+      "step": 162
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.263888888888889e-05,
+      "loss": 2.1756,
+      "step": 163
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.277777777777778e-05,
+      "loss": 1.9171,
+      "step": 164
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.2916666666666667e-05,
+      "loss": 1.9196,
+      "step": 165
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.3055555555555554e-05,
+      "loss": 1.9336,
+      "step": 166
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.3194444444444445e-05,
+      "loss": 2.0132,
+      "step": 167
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 2.2188,
+      "step": 168
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.3472222222222223e-05,
+      "loss": 2.0965,
+      "step": 169
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.361111111111111e-05,
+      "loss": 1.9661,
+      "step": 170
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.375e-05,
+      "loss": 1.8581,
+      "step": 171
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.3888888888888892e-05,
+      "loss": 2.0255,
+      "step": 172
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.4027777777777776e-05,
+      "loss": 1.7089,
+      "step": 173
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.4166666666666667e-05,
+      "loss": 1.9643,
+      "step": 174
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.4305555555555558e-05,
+      "loss": 1.8907,
+      "step": 175
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 1.844,
+      "step": 176
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.4583333333333332e-05,
+      "loss": 1.9625,
+      "step": 177
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.4722222222222223e-05,
+      "loss": 2.073,
+      "step": 178
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.4861111111111114e-05,
+      "loss": 1.9087,
+      "step": 179
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.5e-05,
+      "loss": 1.9038,
+      "step": 180
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.513888888888889e-05,
+      "loss": 2.1505,
+      "step": 181
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.527777777777778e-05,
+      "loss": 2.1689,
+      "step": 182
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.5416666666666667e-05,
+      "loss": 2.1984,
+      "step": 183
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.5555555555555557e-05,
+      "loss": 1.9357,
+      "step": 184
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.5694444444444445e-05,
+      "loss": 2.0046,
+      "step": 185
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.5833333333333336e-05,
+      "loss": 1.7355,
+      "step": 186
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.5972222222222223e-05,
+      "loss": 2.0131,
+      "step": 187
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.6111111111111114e-05,
+      "loss": 2.2213,
+      "step": 188
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.625e-05,
+      "loss": 2.0963,
+      "step": 189
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.638888888888889e-05,
+      "loss": 1.9404,
+      "step": 190
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.652777777777778e-05,
+      "loss": 2.326,
+      "step": 191
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.666666666666667e-05,
+      "loss": 2.0799,
+      "step": 192
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 2.006380081176758,
+      "eval_runtime": 58.784,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 192
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.6805555555555554e-05,
+      "loss": 1.8555,
+      "step": 193
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.6944444444444445e-05,
+      "loss": 1.7953,
+      "step": 194
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.7083333333333335e-05,
+      "loss": 2.1237,
+      "step": 195
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 1.9237,
+      "step": 196
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.736111111111111e-05,
+      "loss": 2.0685,
+      "step": 197
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.75e-05,
+      "loss": 1.842,
+      "step": 198
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.763888888888889e-05,
+      "loss": 1.831,
+      "step": 199
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 1.8893,
+      "step": 200
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.7916666666666666e-05,
+      "loss": 1.9822,
+      "step": 201
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.8055555555555554e-05,
+      "loss": 2.2025,
+      "step": 202
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.8194444444444448e-05,
+      "loss": 2.0227,
+      "step": 203
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.8333333333333332e-05,
+      "loss": 1.9532,
+      "step": 204
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.847222222222222e-05,
+      "loss": 2.1607,
+      "step": 205
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.8611111111111113e-05,
+      "loss": 1.7708,
+      "step": 206
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.875e-05,
+      "loss": 1.7412,
+      "step": 207
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.888888888888889e-05,
+      "loss": 1.9099,
+      "step": 208
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.902777777777778e-05,
+      "loss": 2.0499,
+      "step": 209
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9166666666666666e-05,
+      "loss": 1.8574,
+      "step": 210
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.930555555555556e-05,
+      "loss": 1.8546,
+      "step": 211
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9444444444444445e-05,
+      "loss": 1.7074,
+      "step": 212
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9583333333333332e-05,
+      "loss": 2.1782,
+      "step": 213
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9722222222222226e-05,
+      "loss": 1.986,
+      "step": 214
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.9861111111111113e-05,
+      "loss": 1.995,
+      "step": 215
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 2.1369,
+      "step": 216
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.013888888888889e-05,
+      "loss": 1.9528,
+      "step": 217
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.027777777777778e-05,
+      "loss": 2.0848,
+      "step": 218
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.0416666666666666e-05,
+      "loss": 1.8824,
+      "step": 219
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 2.1061,
+      "step": 220
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 3.0694444444444444e-05,
+      "loss": 1.808,
+      "step": 221
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.083333333333333e-05,
+      "loss": 1.8444,
+      "step": 222
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.0972222222222226e-05,
+      "loss": 1.7965,
+      "step": 223
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 1.7908,
+      "step": 224
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.125e-05,
+      "loss": 1.9537,
+      "step": 225
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.138888888888889e-05,
+      "loss": 2.2461,
+      "step": 226
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.1527777777777775e-05,
+      "loss": 1.9253,
+      "step": 227
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 3.166666666666667e-05,
+      "loss": 1.8042,
+      "step": 228
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.180555555555556e-05,
+      "loss": 2.0855,
+      "step": 229
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.194444444444444e-05,
+      "loss": 1.9236,
+      "step": 230
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.208333333333334e-05,
+      "loss": 1.9994,
+      "step": 231
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.222222222222222e-05,
+      "loss": 2.0076,
+      "step": 232
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.236111111111111e-05,
+      "loss": 1.9305,
+      "step": 233
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 1.9285,
+      "step": 234
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 3.263888888888889e-05,
+      "loss": 2.1502,
+      "step": 235
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 2.1104,
+      "step": 236
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.291666666666667e-05,
+      "loss": 1.7,
+      "step": 237
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.3055555555555553e-05,
+      "loss": 2.04,
+      "step": 238
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.3194444444444444e-05,
+      "loss": 1.8733,
+      "step": 239
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.6055,
+      "step": 240
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 3.347222222222222e-05,
+      "loss": 1.9536,
+      "step": 241
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.3611111111111116e-05,
+      "loss": 1.8183,
+      "step": 242
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.375e-05,
+      "loss": 2.0154,
+      "step": 243
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.388888888888889e-05,
+      "loss": 1.9828,
+      "step": 244
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.402777777777778e-05,
+      "loss": 2.1261,
+      "step": 245
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.4166666666666666e-05,
+      "loss": 1.8679,
+      "step": 246
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.430555555555556e-05,
+      "loss": 1.9212,
+      "step": 247
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 1.8827,
+      "step": 248
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.458333333333333e-05,
+      "loss": 2.1538,
+      "step": 249
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.472222222222222e-05,
+      "loss": 1.9697,
+      "step": 250
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.486111111111111e-05,
+      "loss": 1.7867,
+      "step": 251
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.5000000000000004e-05,
+      "loss": 2.1737,
+      "step": 252
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.513888888888889e-05,
+      "loss": 1.7822,
+      "step": 253
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 3.527777777777778e-05,
+      "loss": 1.9793,
+      "step": 254
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.541666666666667e-05,
+      "loss": 2.0152,
+      "step": 255
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 1.8457,
+      "step": 256
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 2.0041580200195312,
+      "eval_runtime": 58.7834,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 256
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.5694444444444444e-05,
+      "loss": 1.729,
+      "step": 257
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.5833333333333335e-05,
+      "loss": 1.8894,
+      "step": 258
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.5972222222222225e-05,
+      "loss": 1.9022,
+      "step": 259
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.6111111111111116e-05,
+      "loss": 1.9821,
+      "step": 260
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 3.625e-05,
+      "loss": 1.9069,
+      "step": 261
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.638888888888889e-05,
+      "loss": 2.0034,
+      "step": 262
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.652777777777778e-05,
+      "loss": 1.9097,
+      "step": 263
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.8817,
+      "step": 264
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.6805555555555556e-05,
+      "loss": 1.9931,
+      "step": 265
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.694444444444445e-05,
+      "loss": 1.9157,
+      "step": 266
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.708333333333333e-05,
+      "loss": 2.011,
+      "step": 267
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 3.722222222222222e-05,
+      "loss": 2.2373,
+      "step": 268
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.736111111111111e-05,
+      "loss": 2.0246,
+      "step": 269
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.75e-05,
+      "loss": 2.0583,
+      "step": 270
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.7638888888888894e-05,
+      "loss": 1.9461,
+      "step": 271
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 2.0572,
+      "step": 272
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.791666666666667e-05,
+      "loss": 2.0178,
+      "step": 273
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 3.805555555555556e-05,
+      "loss": 1.579,
+      "step": 274
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.8194444444444444e-05,
+      "loss": 1.9734,
+      "step": 275
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 2.1584,
+      "step": 276
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.8472222222222225e-05,
+      "loss": 1.8238,
+      "step": 277
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.861111111111111e-05,
+      "loss": 1.6262,
+      "step": 278
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.875000000000001e-05,
+      "loss": 2.0768,
+      "step": 279
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 1.844,
+      "step": 280
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 3.9027777777777775e-05,
+      "loss": 2.1494,
+      "step": 281
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.916666666666667e-05,
+      "loss": 1.8944,
+      "step": 282
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.9305555555555556e-05,
+      "loss": 2.0143,
+      "step": 283
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.944444444444444e-05,
+      "loss": 1.8535,
+      "step": 284
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.958333333333334e-05,
+      "loss": 2.0168,
+      "step": 285
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.972222222222222e-05,
+      "loss": 1.6707,
+      "step": 286
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 3.986111111111111e-05,
+      "loss": 1.997,
+      "step": 287
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4e-05,
+      "loss": 1.9161,
+      "step": 288
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.013888888888889e-05,
+      "loss": 1.8221,
+      "step": 289
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.0277777777777785e-05,
+      "loss": 2.0532,
+      "step": 290
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.041666666666667e-05,
+      "loss": 1.8899,
+      "step": 291
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.055555555555555e-05,
+      "loss": 1.9314,
+      "step": 292
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.069444444444445e-05,
+      "loss": 1.8349,
+      "step": 293
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.0833333333333334e-05,
+      "loss": 1.9393,
+      "step": 294
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.0972222222222225e-05,
+      "loss": 1.9226,
+      "step": 295
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.1111111111111116e-05,
+      "loss": 1.8407,
+      "step": 296
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.125e-05,
+      "loss": 2.0395,
+      "step": 297
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.138888888888889e-05,
+      "loss": 1.8386,
+      "step": 298
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.152777777777778e-05,
+      "loss": 2.1144,
+      "step": 299
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.1666666666666665e-05,
+      "loss": 2.0431,
+      "step": 300
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.1805555555555556e-05,
+      "loss": 1.7944,
+      "step": 301
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.194444444444445e-05,
+      "loss": 2.1838,
+      "step": 302
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.208333333333333e-05,
+      "loss": 1.8564,
+      "step": 303
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 1.8965,
+      "step": 304
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.236111111111111e-05,
+      "loss": 2.0941,
+      "step": 305
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.25e-05,
+      "loss": 1.8703,
+      "step": 306
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.2638888888888894e-05,
+      "loss": 1.9896,
+      "step": 307
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.277777777777778e-05,
+      "loss": 1.8867,
+      "step": 308
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.291666666666667e-05,
+      "loss": 1.7788,
+      "step": 309
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.305555555555556e-05,
+      "loss": 1.891,
+      "step": 310
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.319444444444444e-05,
+      "loss": 1.8053,
+      "step": 311
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 1.6591,
+      "step": 312
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3472222222222225e-05,
+      "loss": 2.2243,
+      "step": 313
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3611111111111116e-05,
+      "loss": 1.8675,
+      "step": 314
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.375e-05,
+      "loss": 1.9838,
+      "step": 315
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 1.9458,
+      "step": 316
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.402777777777778e-05,
+      "loss": 1.9625,
+      "step": 317
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4166666666666665e-05,
+      "loss": 1.9618,
+      "step": 318
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4305555555555556e-05,
+      "loss": 1.964,
+      "step": 319
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.9216,
+      "step": 320
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 2.000852346420288,
+      "eval_runtime": 58.8229,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 320
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.458333333333333e-05,
+      "loss": 1.9705,
+      "step": 321
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.472222222222223e-05,
+      "loss": 1.7782,
+      "step": 322
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.486111111111111e-05,
+      "loss": 1.9119,
+      "step": 323
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.5e-05,
+      "loss": 1.8237,
+      "step": 324
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.5138888888888894e-05,
+      "loss": 1.9427,
+      "step": 325
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.527777777777778e-05,
+      "loss": 1.9612,
+      "step": 326
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.541666666666667e-05,
+      "loss": 1.9588,
+      "step": 327
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 2.2188,
+      "step": 328
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.569444444444444e-05,
+      "loss": 1.8081,
+      "step": 329
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 1.9724,
+      "step": 330
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.5972222222222225e-05,
+      "loss": 1.9258,
+      "step": 331
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.611111111111111e-05,
+      "loss": 2.0364,
+      "step": 332
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.6250000000000006e-05,
+      "loss": 1.8827,
+      "step": 333
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.638888888888889e-05,
+      "loss": 2.0661,
+      "step": 334
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6527777777777774e-05,
+      "loss": 1.9041,
+      "step": 335
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.9075,
+      "step": 336
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6805555555555556e-05,
+      "loss": 1.8782,
+      "step": 337
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.6944444444444446e-05,
+      "loss": 2.1339,
+      "step": 338
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.708333333333334e-05,
+      "loss": 1.9638,
+      "step": 339
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 1.7903,
+      "step": 340
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.736111111111112e-05,
+      "loss": 1.9212,
+      "step": 341
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.75e-05,
+      "loss": 1.9991,
+      "step": 342
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.7638888888888887e-05,
+      "loss": 1.751,
+      "step": 343
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 1.5513,
+      "step": 344
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.791666666666667e-05,
+      "loss": 1.6869,
+      "step": 345
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.805555555555555e-05,
+      "loss": 1.9404,
+      "step": 346
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.819444444444445e-05,
+      "loss": 1.9058,
+      "step": 347
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 1.9537,
+      "step": 348
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8472222222222224e-05,
+      "loss": 1.7997,
+      "step": 349
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.8611111111111115e-05,
+      "loss": 1.7411,
+      "step": 350
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.875e-05,
+      "loss": 2.2529,
+      "step": 351
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 1.9452,
+      "step": 352
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.902777777777778e-05,
+      "loss": 1.8175,
+      "step": 353
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.9166666666666665e-05,
+      "loss": 2.0064,
+      "step": 354
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.930555555555556e-05,
+      "loss": 1.9444,
+      "step": 355
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 1.9403,
+      "step": 356
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.958333333333334e-05,
+      "loss": 1.8607,
+      "step": 357
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.972222222222223e-05,
+      "loss": 1.8791,
+      "step": 358
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.986111111111111e-05,
+      "loss": 1.8871,
+      "step": 359
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 1.9646,
+      "step": 360
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.013888888888889e-05,
+      "loss": 2.071,
+      "step": 361
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.027777777777778e-05,
+      "loss": 1.8689,
+      "step": 362
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.041666666666667e-05,
+      "loss": 2.2753,
+      "step": 363
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.055555555555556e-05,
+      "loss": 1.7461,
+      "step": 364
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.069444444444444e-05,
+      "loss": 1.6887,
+      "step": 365
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0833333333333333e-05,
+      "loss": 1.9996,
+      "step": 366
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5.0972222222222224e-05,
+      "loss": 1.9305,
+      "step": 367
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.1111111111111115e-05,
+      "loss": 1.7712,
+      "step": 368
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.125e-05,
+      "loss": 1.8045,
+      "step": 369
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.138888888888889e-05,
+      "loss": 2.067,
+      "step": 370
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.152777777777778e-05,
+      "loss": 2.0745,
+      "step": 371
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.166666666666667e-05,
+      "loss": 1.8284,
+      "step": 372
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.1805555555555555e-05,
+      "loss": 1.9079,
+      "step": 373
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5.1944444444444446e-05,
+      "loss": 2.0385,
+      "step": 374
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.208333333333334e-05,
+      "loss": 1.8617,
+      "step": 375
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.222222222222223e-05,
+      "loss": 2.0364,
+      "step": 376
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.236111111111111e-05,
+      "loss": 1.8631,
+      "step": 377
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.25e-05,
+      "loss": 1.8922,
+      "step": 378
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.263888888888889e-05,
+      "loss": 2.0176,
+      "step": 379
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5.277777777777778e-05,
+      "loss": 1.7801,
+      "step": 380
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.291666666666667e-05,
+      "loss": 1.8037,
+      "step": 381
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.305555555555556e-05,
+      "loss": 2.2478,
+      "step": 382
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.319444444444444e-05,
+      "loss": 1.9102,
+      "step": 383
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.333333333333334e-05,
+      "loss": 1.9456,
+      "step": 384
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 1.995949149131775,
+      "eval_runtime": 58.8239,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 384
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.3472222222222224e-05,
+      "loss": 1.7263,
+      "step": 385
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.361111111111111e-05,
+      "loss": 2.0309,
+      "step": 386
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5.3750000000000005e-05,
+      "loss": 1.7313,
+      "step": 387
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.388888888888889e-05,
+      "loss": 2.1418,
+      "step": 388
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.402777777777778e-05,
+      "loss": 2.1034,
+      "step": 389
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.416666666666667e-05,
+      "loss": 1.7491,
+      "step": 390
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.4305555555555555e-05,
+      "loss": 2.1174,
+      "step": 391
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.4444444444444446e-05,
+      "loss": 2.1494,
+      "step": 392
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5.4583333333333336e-05,
+      "loss": 1.9013,
+      "step": 393
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.472222222222222e-05,
+      "loss": 2.0218,
+      "step": 394
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.486111111111112e-05,
+      "loss": 2.0765,
+      "step": 395
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8116,
+      "step": 396
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8413,
+      "step": 397
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.918,
+      "step": 398
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8353,
+      "step": 399
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8645,
+      "step": 400
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9794,
+      "step": 401
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8911,
+      "step": 402
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1127,
+      "step": 403
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9127,
+      "step": 404
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9722,
+      "step": 405
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2238,
+      "step": 406
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1535,
+      "step": 407
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9388,
+      "step": 408
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1923,
+      "step": 409
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6914,
+      "step": 410
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0324,
+      "step": 411
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9017,
+      "step": 412
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9931,
+      "step": 413
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2172,
+      "step": 414
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1389,
+      "step": 415
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0016,
+      "step": 416
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0593,
+      "step": 417
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9306,
+      "step": 418
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8381,
+      "step": 419
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9778,
+      "step": 420
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8258,
+      "step": 421
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8392,
+      "step": 422
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8362,
+      "step": 423
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0747,
+      "step": 424
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0362,
+      "step": 425
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9069,
+      "step": 426
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8727,
+      "step": 427
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7109,
+      "step": 428
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 2.3048,
+      "step": 429
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9831,
+      "step": 430
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6552,
+      "step": 431
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0319,
+      "step": 432
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.5e-05,
+      "loss": 2.106,
+      "step": 433
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6774,
+      "step": 434
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6611,
+      "step": 435
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9152,
+      "step": 436
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2678,
+      "step": 437
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9184,
+      "step": 438
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8499,
+      "step": 439
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8624,
+      "step": 440
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9323,
+      "step": 441
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8484,
+      "step": 442
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0478,
+      "step": 443
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0434,
+      "step": 444
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9652,
+      "step": 445
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7201,
+      "step": 446
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.912,
+      "step": 447
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.061,
+      "step": 448
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 1.9963455200195312,
+      "eval_runtime": 58.8096,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 448
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8791,
+      "step": 449
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0873,
+      "step": 450
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0294,
+      "step": 451
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9715,
+      "step": 452
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0691,
+      "step": 453
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0964,
+      "step": 454
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1053,
+      "step": 455
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7784,
+      "step": 456
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0341,
+      "step": 457
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0414,
+      "step": 458
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1737,
+      "step": 459
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1014,
+      "step": 460
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0055,
+      "step": 461
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9331,
+      "step": 462
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7263,
+      "step": 463
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0779,
+      "step": 464
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8288,
+      "step": 465
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9196,
+      "step": 466
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9546,
+      "step": 467
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0738,
+      "step": 468
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8559,
+      "step": 469
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.964,
+      "step": 470
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7826,
+      "step": 471
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1136,
+      "step": 472
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7275,
+      "step": 473
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9853,
+      "step": 474
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9923,
+      "step": 475
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8931,
+      "step": 476
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9422,
+      "step": 477
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9964,
+      "step": 478
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9789,
+      "step": 479
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8215,
+      "step": 480
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8202,
+      "step": 481
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8453,
+      "step": 482
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6805,
+      "step": 483
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0413,
+      "step": 484
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1132,
+      "step": 485
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.919,
+      "step": 486
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0277,
+      "step": 487
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9747,
+      "step": 488
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0166,
+      "step": 489
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7176,
+      "step": 490
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8685,
+      "step": 491
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9191,
+      "step": 492
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7576,
+      "step": 493
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9441,
+      "step": 494
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0719,
+      "step": 495
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0278,
+      "step": 496
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0337,
+      "step": 497
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8023,
+      "step": 498
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0445,
+      "step": 499
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.818,
+      "step": 500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0026,
+      "step": 501
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0388,
+      "step": 502
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0666,
+      "step": 503
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9827,
+      "step": 504
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8992,
+      "step": 505
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9621,
+      "step": 506
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9876,
+      "step": 507
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8765,
+      "step": 508
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9699,
+      "step": 509
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9311,
+      "step": 510
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5199,
+      "step": 511
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8739,
+      "step": 512
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 1.9916880130767822,
+      "eval_runtime": 59.0008,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 512
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8391,
+      "step": 513
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.591,
+      "step": 514
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9632,
+      "step": 515
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7806,
+      "step": 516
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0876,
+      "step": 517
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8403,
+      "step": 518
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2148,
+      "step": 519
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.808,
+      "step": 520
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9136,
+      "step": 521
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9457,
+      "step": 522
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9325,
+      "step": 523
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8759,
+      "step": 524
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1916,
+      "step": 525
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8888,
+      "step": 526
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9731,
+      "step": 527
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8036,
+      "step": 528
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0241,
+      "step": 529
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5473,
+      "step": 530
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7932,
+      "step": 531
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7139,
+      "step": 532
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7221,
+      "step": 533
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9972,
+      "step": 534
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0621,
+      "step": 535
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9002,
+      "step": 536
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9351,
+      "step": 537
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1008,
+      "step": 538
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7299,
+      "step": 539
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0681,
+      "step": 540
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7492,
+      "step": 541
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6713,
+      "step": 542
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9309,
+      "step": 543
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7673,
+      "step": 544
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5529,
+      "step": 545
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9785,
+      "step": 546
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9504,
+      "step": 547
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9169,
+      "step": 548
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8258,
+      "step": 549
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7855,
+      "step": 550
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8533,
+      "step": 551
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9102,
+      "step": 552
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9679,
+      "step": 553
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8848,
+      "step": 554
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9912,
+      "step": 555
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9674,
+      "step": 556
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7152,
+      "step": 557
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0175,
+      "step": 558
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9833,
+      "step": 559
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2812,
+      "step": 560
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.022,
+      "step": 561
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1457,
+      "step": 562
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8974,
+      "step": 563
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.945,
+      "step": 564
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0065,
+      "step": 565
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.004,
+      "step": 566
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9369,
+      "step": 567
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0565,
+      "step": 568
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0946,
+      "step": 569
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.904,
+      "step": 570
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7949,
+      "step": 571
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9979,
+      "step": 572
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7834,
+      "step": 573
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6403,
+      "step": 574
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9347,
+      "step": 575
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.69,
+      "step": 576
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 1.989676594734192,
+      "eval_runtime": 58.9014,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 576
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0622,
+      "step": 577
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8775,
+      "step": 578
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7864,
+      "step": 579
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7229,
+      "step": 580
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0063,
+      "step": 581
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2663,
+      "step": 582
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1072,
+      "step": 583
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8062,
+      "step": 584
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1686,
+      "step": 585
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9927,
+      "step": 586
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8644,
+      "step": 587
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9758,
+      "step": 588
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9856,
+      "step": 589
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9855,
+      "step": 590
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8876,
+      "step": 591
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7528,
+      "step": 592
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2187,
+      "step": 593
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8247,
+      "step": 594
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1322,
+      "step": 595
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.877,
+      "step": 596
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8718,
+      "step": 597
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9852,
+      "step": 598
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1917,
+      "step": 599
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1977,
+      "step": 600
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8002,
+      "step": 601
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1565,
+      "step": 602
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.879,
+      "step": 603
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0121,
+      "step": 604
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.778,
+      "step": 605
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9739,
+      "step": 606
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9021,
+      "step": 607
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7507,
+      "step": 608
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7977,
+      "step": 609
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9014,
+      "step": 610
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8381,
+      "step": 611
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9956,
+      "step": 612
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6708,
+      "step": 613
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8256,
+      "step": 614
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8091,
+      "step": 615
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9035,
+      "step": 616
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8378,
+      "step": 617
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9185,
+      "step": 618
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9038,
+      "step": 619
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8192,
+      "step": 620
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9189,
+      "step": 621
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9267,
+      "step": 622
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2876,
+      "step": 623
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7351,
+      "step": 624
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8424,
+      "step": 625
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8541,
+      "step": 626
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1005,
+      "step": 627
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0777,
+      "step": 628
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8535,
+      "step": 629
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.073,
+      "step": 630
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6907,
+      "step": 631
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0157,
+      "step": 632
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0815,
+      "step": 633
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1042,
+      "step": 634
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9516,
+      "step": 635
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0526,
+      "step": 636
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0166,
+      "step": 637
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8631,
+      "step": 638
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7609,
+      "step": 639
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2433,
+      "step": 640
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 1.9858782291412354,
+      "eval_runtime": 58.9176,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 640
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8147,
+      "step": 641
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1954,
+      "step": 642
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9324,
+      "step": 643
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6993,
+      "step": 644
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8438,
+      "step": 645
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9897,
+      "step": 646
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0089,
+      "step": 647
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7791,
+      "step": 648
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8066,
+      "step": 649
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0635,
+      "step": 650
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8481,
+      "step": 651
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8849,
+      "step": 652
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0186,
+      "step": 653
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8334,
+      "step": 654
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9343,
+      "step": 655
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9082,
+      "step": 656
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6138,
+      "step": 657
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9074,
+      "step": 658
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8979,
+      "step": 659
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6229,
+      "step": 660
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8861,
+      "step": 661
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0366,
+      "step": 662
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6883,
+      "step": 663
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8848,
+      "step": 664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8352,
+      "step": 665
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9182,
+      "step": 666
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9061,
+      "step": 667
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1967,
+      "step": 668
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6919,
+      "step": 669
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6953,
+      "step": 670
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6956,
+      "step": 671
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8152,
+      "step": 672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8163,
+      "step": 673
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9031,
+      "step": 674
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9223,
+      "step": 675
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9165,
+      "step": 676
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.574,
+      "step": 677
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8083,
+      "step": 678
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7346,
+      "step": 679
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1956,
+      "step": 680
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5288,
+      "step": 681
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8985,
+      "step": 682
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9574,
+      "step": 683
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8191,
+      "step": 684
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6321,
+      "step": 685
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7425,
+      "step": 686
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9948,
+      "step": 687
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8322,
+      "step": 688
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6861,
+      "step": 689
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8483,
+      "step": 690
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8787,
+      "step": 691
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9686,
+      "step": 692
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1331,
+      "step": 693
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9199,
+      "step": 694
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7578,
+      "step": 695
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0364,
+      "step": 696
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9478,
+      "step": 697
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.904,
+      "step": 698
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7678,
+      "step": 699
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8479,
+      "step": 700
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9111,
+      "step": 701
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9519,
+      "step": 702
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6778,
+      "step": 703
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9542,
+      "step": 704
+    },
+    {
+      "epoch": 1.06,
+      "eval_loss": 1.985458493232727,
+      "eval_runtime": 58.9703,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 704
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8387,
+      "step": 705
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7654,
+      "step": 706
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.943,
+      "step": 707
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9128,
+      "step": 708
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8484,
+      "step": 709
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7394,
+      "step": 710
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7022,
+      "step": 711
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2014,
+      "step": 712
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9809,
+      "step": 713
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6051,
+      "step": 714
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8484,
+      "step": 715
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8196,
+      "step": 716
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0568,
+      "step": 717
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9167,
+      "step": 718
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.813,
+      "step": 719
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7579,
+      "step": 720
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9889,
+      "step": 721
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 2.115,
+      "step": 722
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9313,
+      "step": 723
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6871,
+      "step": 724
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7422,
+      "step": 725
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2011,
+      "step": 726
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.991,
+      "step": 727
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8994,
+      "step": 728
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8999,
+      "step": 729
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8362,
+      "step": 730
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9482,
+      "step": 731
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8424,
+      "step": 732
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7127,
+      "step": 733
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9714,
+      "step": 734
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8753,
+      "step": 735
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8008,
+      "step": 736
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9657,
+      "step": 737
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9767,
+      "step": 738
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8487,
+      "step": 739
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9573,
+      "step": 740
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8955,
+      "step": 741
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.781,
+      "step": 742
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1247,
+      "step": 743
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8406,
+      "step": 744
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8087,
+      "step": 745
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9323,
+      "step": 746
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7767,
+      "step": 747
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.901,
+      "step": 748
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7707,
+      "step": 749
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7524,
+      "step": 750
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2238,
+      "step": 751
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7934,
+      "step": 752
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8239,
+      "step": 753
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8188,
+      "step": 754
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9701,
+      "step": 755
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0383,
+      "step": 756
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7855,
+      "step": 757
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7499,
+      "step": 758
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8229,
+      "step": 759
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1237,
+      "step": 760
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0109,
+      "step": 761
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9678,
+      "step": 762
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7707,
+      "step": 763
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6788,
+      "step": 764
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9237,
+      "step": 765
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7559,
+      "step": 766
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8196,
+      "step": 767
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8381,
+      "step": 768
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 1.986333966255188,
+      "eval_runtime": 58.9784,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 768
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7674,
+      "step": 769
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9273,
+      "step": 770
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9929,
+      "step": 771
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0698,
+      "step": 772
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1332,
+      "step": 773
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8592,
+      "step": 774
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7061,
+      "step": 775
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6427,
+      "step": 776
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6082,
+      "step": 777
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0476,
+      "step": 778
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9384,
+      "step": 779
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7435,
+      "step": 780
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7003,
+      "step": 781
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0046,
+      "step": 782
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.656,
+      "step": 783
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7923,
+      "step": 784
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9052,
+      "step": 785
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8881,
+      "step": 786
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.837,
+      "step": 787
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1303,
+      "step": 788
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9772,
+      "step": 789
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1375,
+      "step": 790
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0447,
+      "step": 791
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0876,
+      "step": 792
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9018,
+      "step": 793
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7919,
+      "step": 794
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7968,
+      "step": 795
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0099,
+      "step": 796
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9105,
+      "step": 797
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9935,
+      "step": 798
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8524,
+      "step": 799
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7873,
+      "step": 800
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8082,
+      "step": 801
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5041,
+      "step": 802
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7164,
+      "step": 803
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8639,
+      "step": 804
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7758,
+      "step": 805
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7119,
+      "step": 806
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8494,
+      "step": 807
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8645,
+      "step": 808
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9584,
+      "step": 809
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9228,
+      "step": 810
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 2.034,
+      "step": 811
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8669,
+      "step": 812
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7473,
+      "step": 813
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5693,
+      "step": 814
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6032,
+      "step": 815
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8637,
+      "step": 816
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1032,
+      "step": 817
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0423,
+      "step": 818
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7922,
+      "step": 819
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8913,
+      "step": 820
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7887,
+      "step": 821
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9852,
+      "step": 822
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 2.004,
+      "step": 823
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.802,
+      "step": 824
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9972,
+      "step": 825
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7655,
+      "step": 826
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0794,
+      "step": 827
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7272,
+      "step": 828
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7846,
+      "step": 829
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2412,
+      "step": 830
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8671,
+      "step": 831
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8069,
+      "step": 832
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.9881497621536255,
+      "eval_runtime": 58.9619,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 832
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6592,
+      "step": 833
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9748,
+      "step": 834
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8619,
+      "step": 835
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.798,
+      "step": 836
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7833,
+      "step": 837
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6699,
+      "step": 838
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.921,
+      "step": 839
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9981,
+      "step": 840
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8746,
+      "step": 841
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9563,
+      "step": 842
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.777,
+      "step": 843
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7096,
+      "step": 844
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0355,
+      "step": 845
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7551,
+      "step": 846
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8333,
+      "step": 847
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8359,
+      "step": 848
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6545,
+      "step": 849
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6688,
+      "step": 850
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8365,
+      "step": 851
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7209,
+      "step": 852
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1646,
+      "step": 853
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6569,
+      "step": 854
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7385,
+      "step": 855
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7525,
+      "step": 856
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9009,
+      "step": 857
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.543,
+      "step": 858
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 2.04,
+      "step": 859
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6964,
+      "step": 860
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7117,
+      "step": 861
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7786,
+      "step": 862
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9283,
+      "step": 863
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7426,
+      "step": 864
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0052,
+      "step": 865
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9682,
+      "step": 866
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7817,
+      "step": 867
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9365,
+      "step": 868
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0981,
+      "step": 869
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7401,
+      "step": 870
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 2.148,
+      "step": 871
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8509,
+      "step": 872
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7674,
+      "step": 873
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8585,
+      "step": 874
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 2.039,
+      "step": 875
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2174,
+      "step": 876
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7736,
+      "step": 877
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0157,
+      "step": 878
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7967,
+      "step": 879
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6875,
+      "step": 880
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4769,
+      "step": 881
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8631,
+      "step": 882
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7446,
+      "step": 883
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5132,
+      "step": 884
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7281,
+      "step": 885
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0059,
+      "step": 886
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8305,
+      "step": 887
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7437,
+      "step": 888
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0905,
+      "step": 889
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7111,
+      "step": 890
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1154,
+      "step": 891
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8436,
+      "step": 892
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8472,
+      "step": 893
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8841,
+      "step": 894
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8921,
+      "step": 895
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7843,
+      "step": 896
+    },
+    {
+      "epoch": 1.35,
+      "eval_loss": 1.98455011844635,
+      "eval_runtime": 59.5681,
+      "eval_samples_per_second": 0.235,
+      "eval_steps_per_second": 0.118,
+      "step": 896
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8265,
+      "step": 897
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8374,
+      "step": 898
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9551,
+      "step": 899
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9349,
+      "step": 900
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7317,
+      "step": 901
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0014,
+      "step": 902
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7912,
+      "step": 903
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8637,
+      "step": 904
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8175,
+      "step": 905
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0171,
+      "step": 906
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1175,
+      "step": 907
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9703,
+      "step": 908
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.887,
+      "step": 909
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9371,
+      "step": 910
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9807,
+      "step": 911
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8449,
+      "step": 912
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8572,
+      "step": 913
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6726,
+      "step": 914
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7862,
+      "step": 915
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8861,
+      "step": 916
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6724,
+      "step": 917
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9625,
+      "step": 918
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1159,
+      "step": 919
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7551,
+      "step": 920
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7045,
+      "step": 921
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7777,
+      "step": 922
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5.5e-05,
+      "loss": 2.105,
+      "step": 923
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9585,
+      "step": 924
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8433,
+      "step": 925
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0128,
+      "step": 926
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0865,
+      "step": 927
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 2.3,
+      "step": 928
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.782,
+      "step": 929
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7855,
+      "step": 930
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0097,
+      "step": 931
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1347,
+      "step": 932
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8255,
+      "step": 933
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.628,
+      "step": 934
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.817,
+      "step": 935
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7243,
+      "step": 936
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7371,
+      "step": 937
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.875,
+      "step": 938
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.89,
+      "step": 939
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7224,
+      "step": 940
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6944,
+      "step": 941
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0949,
+      "step": 942
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9106,
+      "step": 943
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8215,
+      "step": 944
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.772,
+      "step": 945
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9106,
+      "step": 946
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8331,
+      "step": 947
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8223,
+      "step": 948
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9412,
+      "step": 949
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2636,
+      "step": 950
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0042,
+      "step": 951
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7525,
+      "step": 952
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8757,
+      "step": 953
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8459,
+      "step": 954
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.892,
+      "step": 955
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5.5e-05,
+      "loss": 1.878,
+      "step": 956
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8207,
+      "step": 957
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1033,
+      "step": 958
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0789,
+      "step": 959
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9558,
+      "step": 960
+    },
+    {
+      "epoch": 1.45,
+      "eval_loss": 1.9816019535064697,
+      "eval_runtime": 59.0035,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 960
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0175,
+      "step": 961
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 2.3281,
+      "step": 962
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7073,
+      "step": 963
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9154,
+      "step": 964
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8298,
+      "step": 965
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 2.22,
+      "step": 966
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8044,
+      "step": 967
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0208,
+      "step": 968
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1039,
+      "step": 969
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2197,
+      "step": 970
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8942,
+      "step": 971
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7247,
+      "step": 972
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9488,
+      "step": 973
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9499,
+      "step": 974
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0991,
+      "step": 975
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5.5e-05,
+      "loss": 1.812,
+      "step": 976
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8369,
+      "step": 977
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.87,
+      "step": 978
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6083,
+      "step": 979
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8719,
+      "step": 980
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7463,
+      "step": 981
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0501,
+      "step": 982
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8734,
+      "step": 983
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8175,
+      "step": 984
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9625,
+      "step": 985
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9687,
+      "step": 986
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.887,
+      "step": 987
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7694,
+      "step": 988
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9643,
+      "step": 989
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9183,
+      "step": 990
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9555,
+      "step": 991
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7406,
+      "step": 992
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8618,
+      "step": 993
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7619,
+      "step": 994
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0558,
+      "step": 995
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8614,
+      "step": 996
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5208,
+      "step": 997
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8473,
+      "step": 998
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.864,
+      "step": 999
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6967,
+      "step": 1000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0151,
+      "step": 1001
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9683,
+      "step": 1002
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.816,
+      "step": 1003
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.533,
+      "step": 1004
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8513,
+      "step": 1005
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.795,
+      "step": 1006
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7542,
+      "step": 1007
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 2.296,
+      "step": 1008
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5.5e-05,
+      "loss": 1.528,
+      "step": 1009
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.883,
+      "step": 1010
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7577,
+      "step": 1011
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1231,
+      "step": 1012
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0266,
+      "step": 1013
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0612,
+      "step": 1014
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9317,
+      "step": 1015
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8076,
+      "step": 1016
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1764,
+      "step": 1017
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8218,
+      "step": 1018
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9478,
+      "step": 1019
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8305,
+      "step": 1020
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.965,
+      "step": 1021
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9032,
+      "step": 1022
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9612,
+      "step": 1023
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9241,
+      "step": 1024
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 1.9803096055984497,
+      "eval_runtime": 59.0073,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1024
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7844,
+      "step": 1025
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8216,
+      "step": 1026
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9107,
+      "step": 1027
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8348,
+      "step": 1028
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8552,
+      "step": 1029
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0343,
+      "step": 1030
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2257,
+      "step": 1031
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9865,
+      "step": 1032
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7459,
+      "step": 1033
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7665,
+      "step": 1034
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0813,
+      "step": 1035
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9783,
+      "step": 1036
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.632,
+      "step": 1037
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7919,
+      "step": 1038
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1836,
+      "step": 1039
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0717,
+      "step": 1040
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0652,
+      "step": 1041
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5.5e-05,
+      "loss": 1.99,
+      "step": 1042
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8057,
+      "step": 1043
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9926,
+      "step": 1044
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8964,
+      "step": 1045
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1568,
+      "step": 1046
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9066,
+      "step": 1047
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7215,
+      "step": 1048
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6507,
+      "step": 1049
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9338,
+      "step": 1050
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7869,
+      "step": 1051
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.817,
+      "step": 1052
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7371,
+      "step": 1053
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0039,
+      "step": 1054
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8853,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7729,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0849,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8182,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8164,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0209,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1836,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8525,
+      "step": 1062
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9572,
+      "step": 1063
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6491,
+      "step": 1064
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9947,
+      "step": 1065
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9133,
+      "step": 1066
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9352,
+      "step": 1067
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9082,
+      "step": 1068
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2686,
+      "step": 1069
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8655,
+      "step": 1070
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8148,
+      "step": 1071
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5466,
+      "step": 1072
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9907,
+      "step": 1073
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6634,
+      "step": 1074
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8459,
+      "step": 1075
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8285,
+      "step": 1076
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8191,
+      "step": 1077
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8698,
+      "step": 1078
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8112,
+      "step": 1079
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9497,
+      "step": 1080
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7179,
+      "step": 1081
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0815,
+      "step": 1082
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6455,
+      "step": 1083
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8332,
+      "step": 1084
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4951,
+      "step": 1085
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5622,
+      "step": 1086
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9244,
+      "step": 1087
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8547,
+      "step": 1088
+    },
+    {
+      "epoch": 1.64,
+      "eval_loss": 1.9768799543380737,
+      "eval_runtime": 59.0019,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1088
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9867,
+      "step": 1089
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7404,
+      "step": 1090
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7292,
+      "step": 1091
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9778,
+      "step": 1092
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8905,
+      "step": 1093
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7208,
+      "step": 1094
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5.5e-05,
+      "loss": 1.885,
+      "step": 1095
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8287,
+      "step": 1096
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0805,
+      "step": 1097
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9037,
+      "step": 1098
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8472,
+      "step": 1099
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6056,
+      "step": 1100
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8532,
+      "step": 1101
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7272,
+      "step": 1102
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7979,
+      "step": 1103
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8289,
+      "step": 1104
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.634,
+      "step": 1105
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0646,
+      "step": 1106
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7235,
+      "step": 1107
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7801,
+      "step": 1108
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5894,
+      "step": 1109
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6548,
+      "step": 1110
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9325,
+      "step": 1111
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8516,
+      "step": 1112
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8523,
+      "step": 1113
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 2.3195,
+      "step": 1114
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7882,
+      "step": 1115
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0055,
+      "step": 1116
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8724,
+      "step": 1117
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7485,
+      "step": 1118
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0367,
+      "step": 1119
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7213,
+      "step": 1120
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8794,
+      "step": 1121
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1606,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8364,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8943,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8637,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 1.755,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.002,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2109,
+      "step": 1128
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.82,
+      "step": 1129
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5024,
+      "step": 1130
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8737,
+      "step": 1131
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0647,
+      "step": 1132
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6576,
+      "step": 1133
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4687,
+      "step": 1134
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7363,
+      "step": 1135
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8757,
+      "step": 1136
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1569,
+      "step": 1137
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1774,
+      "step": 1138
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8027,
+      "step": 1139
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 1.941,
+      "step": 1140
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0008,
+      "step": 1141
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8287,
+      "step": 1142
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.679,
+      "step": 1143
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0351,
+      "step": 1144
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8833,
+      "step": 1145
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9896,
+      "step": 1146
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8529,
+      "step": 1147
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0169,
+      "step": 1148
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6624,
+      "step": 1149
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9507,
+      "step": 1150
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8244,
+      "step": 1151
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7616,
+      "step": 1152
+    },
+    {
+      "epoch": 1.74,
+      "eval_loss": 1.974184274673462,
+      "eval_runtime": 59.0105,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1152
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5877,
+      "step": 1153
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9896,
+      "step": 1154
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7394,
+      "step": 1155
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0282,
+      "step": 1156
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0541,
+      "step": 1157
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8256,
+      "step": 1158
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7988,
+      "step": 1159
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0081,
+      "step": 1160
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9105,
+      "step": 1161
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8947,
+      "step": 1162
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9178,
+      "step": 1163
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8641,
+      "step": 1164
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6064,
+      "step": 1165
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8312,
+      "step": 1166
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 2.4175,
+      "step": 1167
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6534,
+      "step": 1168
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9796,
+      "step": 1169
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.581,
+      "step": 1170
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6561,
+      "step": 1171
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7829,
+      "step": 1172
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8432,
+      "step": 1173
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.974,
+      "step": 1174
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7076,
+      "step": 1175
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6845,
+      "step": 1176
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6676,
+      "step": 1177
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.938,
+      "step": 1178
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7616,
+      "step": 1179
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7008,
+      "step": 1180
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8243,
+      "step": 1181
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6733,
+      "step": 1182
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5045,
+      "step": 1183
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.813,
+      "step": 1184
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9045,
+      "step": 1185
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9471,
+      "step": 1186
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8006,
+      "step": 1187
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8393,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5674,
+      "step": 1189
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6655,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7475,
+      "step": 1191
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1792,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6809,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9012,
+      "step": 1194
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.848,
+      "step": 1195
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.918,
+      "step": 1196
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0244,
+      "step": 1197
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6339,
+      "step": 1198
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9695,
+      "step": 1199
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7581,
+      "step": 1200
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6169,
+      "step": 1201
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8096,
+      "step": 1202
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8398,
+      "step": 1203
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7617,
+      "step": 1204
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8266,
+      "step": 1205
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6852,
+      "step": 1206
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6768,
+      "step": 1207
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5674,
+      "step": 1208
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0484,
+      "step": 1209
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7565,
+      "step": 1210
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6991,
+      "step": 1211
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9785,
+      "step": 1212
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0837,
+      "step": 1213
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1116,
+      "step": 1214
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7018,
+      "step": 1215
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7896,
+      "step": 1216
+    },
+    {
+      "epoch": 1.84,
+      "eval_loss": 1.9749974012374878,
+      "eval_runtime": 58.9244,
+      "eval_samples_per_second": 0.238,
+      "eval_steps_per_second": 0.119,
+      "step": 1216
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8006,
+      "step": 1217
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7801,
+      "step": 1218
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8307,
+      "step": 1219
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6739,
+      "step": 1220
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0557,
+      "step": 1221
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7643,
+      "step": 1222
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0422,
+      "step": 1223
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7279,
+      "step": 1224
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8338,
+      "step": 1225
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8596,
+      "step": 1226
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9323,
+      "step": 1227
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9343,
+      "step": 1228
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8977,
+      "step": 1229
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7675,
+      "step": 1230
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8364,
+      "step": 1231
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7491,
+      "step": 1232
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8607,
+      "step": 1233
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7741,
+      "step": 1234
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6947,
+      "step": 1235
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8658,
+      "step": 1236
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6586,
+      "step": 1237
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8844,
+      "step": 1238
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9983,
+      "step": 1239
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.968,
+      "step": 1240
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8485,
+      "step": 1241
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8994,
+      "step": 1242
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5991,
+      "step": 1243
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8704,
+      "step": 1244
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.895,
+      "step": 1245
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8414,
+      "step": 1246
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9166,
+      "step": 1247
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8878,
+      "step": 1248
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7005,
+      "step": 1249
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9262,
+      "step": 1250
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9838,
+      "step": 1251
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0325,
+      "step": 1252
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9683,
+      "step": 1253
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8107,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8154,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7589,
+      "step": 1256
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8686,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.836,
+      "step": 1258
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8839,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9879,
+      "step": 1260
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5.5e-05,
+      "loss": 1.979,
+      "step": 1261
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9252,
+      "step": 1262
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8016,
+      "step": 1263
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8314,
+      "step": 1264
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7805,
+      "step": 1265
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9322,
+      "step": 1266
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9515,
+      "step": 1267
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7414,
+      "step": 1268
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7464,
+      "step": 1269
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1836,
+      "step": 1270
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6549,
+      "step": 1271
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7951,
+      "step": 1272
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8127,
+      "step": 1273
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8096,
+      "step": 1274
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8935,
+      "step": 1275
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8186,
+      "step": 1276
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9846,
+      "step": 1277
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7666,
+      "step": 1278
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7617,
+      "step": 1279
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8763,
+      "step": 1280
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.9570947885513306,
+      "eval_runtime": 59.0224,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1280
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8724,
+      "step": 1281
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9641,
+      "step": 1282
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7442,
+      "step": 1283
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.963,
+      "step": 1284
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8515,
+      "step": 1285
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9066,
+      "step": 1286
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.5e-05,
+      "loss": 1.751,
+      "step": 1287
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8862,
+      "step": 1288
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7903,
+      "step": 1289
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9637,
+      "step": 1290
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9093,
+      "step": 1291
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8575,
+      "step": 1292
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7905,
+      "step": 1293
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6731,
+      "step": 1294
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1496,
+      "step": 1295
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9826,
+      "step": 1296
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8418,
+      "step": 1297
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7795,
+      "step": 1298
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7403,
+      "step": 1299
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1435,
+      "step": 1300
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7862,
+      "step": 1301
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0067,
+      "step": 1302
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8522,
+      "step": 1303
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7623,
+      "step": 1304
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0149,
+      "step": 1305
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9706,
+      "step": 1306
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0222,
+      "step": 1307
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9522,
+      "step": 1308
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9882,
+      "step": 1309
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8639,
+      "step": 1310
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9902,
+      "step": 1311
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0714,
+      "step": 1312
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6234,
+      "step": 1313
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8611,
+      "step": 1314
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7995,
+      "step": 1315
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9626,
+      "step": 1316
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.2278,
+      "step": 1317
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1429,
+      "step": 1318
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0095,
+      "step": 1319
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0158,
+      "step": 1320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 2.362,
+      "step": 1321
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9607,
+      "step": 1322
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9065,
+      "step": 1323
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7612,
+      "step": 1324
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8885,
+      "step": 1325
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.49,
+      "step": 1326
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9146,
+      "step": 1327
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7438,
+      "step": 1328
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8324,
+      "step": 1329
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8054,
+      "step": 1330
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7088,
+      "step": 1331
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3956,
+      "step": 1332
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6183,
+      "step": 1333
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9603,
+      "step": 1334
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8318,
+      "step": 1335
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9956,
+      "step": 1336
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0125,
+      "step": 1337
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3906,
+      "step": 1338
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6743,
+      "step": 1339
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5328,
+      "step": 1340
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9533,
+      "step": 1341
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8125,
+      "step": 1342
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9399,
+      "step": 1343
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5477,
+      "step": 1344
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 1.9626766443252563,
+      "eval_runtime": 58.9574,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1344
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7911,
+      "step": 1345
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7912,
+      "step": 1346
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.5e-05,
+      "loss": 1.2995,
+      "step": 1347
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7253,
+      "step": 1348
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6329,
+      "step": 1349
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7273,
+      "step": 1350
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7071,
+      "step": 1351
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7656,
+      "step": 1352
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5804,
+      "step": 1353
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8209,
+      "step": 1354
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6642,
+      "step": 1355
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6828,
+      "step": 1356
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6018,
+      "step": 1357
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6943,
+      "step": 1358
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8927,
+      "step": 1359
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6061,
+      "step": 1360
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9819,
+      "step": 1361
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8138,
+      "step": 1362
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6311,
+      "step": 1363
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6776,
+      "step": 1364
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6291,
+      "step": 1365
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8836,
+      "step": 1366
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 5.5e-05,
+      "loss": 1.613,
+      "step": 1367
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5906,
+      "step": 1368
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6073,
+      "step": 1369
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0011,
+      "step": 1370
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.779,
+      "step": 1371
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7632,
+      "step": 1372
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6912,
+      "step": 1373
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7187,
+      "step": 1374
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7966,
+      "step": 1375
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5966,
+      "step": 1376
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.3322,
+      "step": 1377
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7133,
+      "step": 1378
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6765,
+      "step": 1379
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9066,
+      "step": 1380
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6105,
+      "step": 1381
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7615,
+      "step": 1382
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5848,
+      "step": 1383
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8847,
+      "step": 1384
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7067,
+      "step": 1385
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5488,
+      "step": 1386
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7337,
+      "step": 1387
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.765,
+      "step": 1388
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5989,
+      "step": 1389
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5968,
+      "step": 1390
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8493,
+      "step": 1391
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7558,
+      "step": 1392
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7106,
+      "step": 1393
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7283,
+      "step": 1394
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7924,
+      "step": 1395
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6897,
+      "step": 1396
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9234,
+      "step": 1397
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7521,
+      "step": 1398
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6473,
+      "step": 1399
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8511,
+      "step": 1400
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.501,
+      "step": 1401
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1233,
+      "step": 1402
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5598,
+      "step": 1403
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8882,
+      "step": 1404
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6589,
+      "step": 1405
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6911,
+      "step": 1406
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5638,
+      "step": 1407
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6599,
+      "step": 1408
+    },
+    {
+      "epoch": 2.13,
+      "eval_loss": 1.9700766801834106,
+      "eval_runtime": 58.9821,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1408
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4485,
+      "step": 1409
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8551,
+      "step": 1410
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7528,
+      "step": 1411
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0524,
+      "step": 1412
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8525,
+      "step": 1413
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0861,
+      "step": 1414
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9524,
+      "step": 1415
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7021,
+      "step": 1416
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7564,
+      "step": 1417
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7025,
+      "step": 1418
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9026,
+      "step": 1419
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7491,
+      "step": 1420
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6636,
+      "step": 1421
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9605,
+      "step": 1422
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6847,
+      "step": 1423
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8174,
+      "step": 1424
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8455,
+      "step": 1425
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7054,
+      "step": 1426
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8484,
+      "step": 1427
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4641,
+      "step": 1428
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7692,
+      "step": 1429
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7268,
+      "step": 1430
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5682,
+      "step": 1431
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7337,
+      "step": 1432
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9853,
+      "step": 1433
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7879,
+      "step": 1434
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7097,
+      "step": 1435
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5058,
+      "step": 1436
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9153,
+      "step": 1437
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8588,
+      "step": 1438
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7186,
+      "step": 1439
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4168,
+      "step": 1440
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8122,
+      "step": 1441
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8659,
+      "step": 1442
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9646,
+      "step": 1443
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5798,
+      "step": 1444
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 2.002,
+      "step": 1445
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6354,
+      "step": 1446
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9755,
+      "step": 1447
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7351,
+      "step": 1448
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8579,
+      "step": 1449
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0733,
+      "step": 1450
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0312,
+      "step": 1451
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8293,
+      "step": 1452
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8115,
+      "step": 1453
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6021,
+      "step": 1454
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 2.016,
+      "step": 1455
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7557,
+      "step": 1456
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.773,
+      "step": 1457
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8518,
+      "step": 1458
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8113,
+      "step": 1459
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9532,
+      "step": 1460
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8051,
+      "step": 1461
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6969,
+      "step": 1462
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6927,
+      "step": 1463
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.43,
+      "step": 1464
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4665,
+      "step": 1465
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6676,
+      "step": 1466
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9028,
+      "step": 1467
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6103,
+      "step": 1468
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8769,
+      "step": 1469
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8004,
+      "step": 1470
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7644,
+      "step": 1471
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4839,
+      "step": 1472
+    },
+    {
+      "epoch": 2.22,
+      "eval_loss": 1.9538156986236572,
+      "eval_runtime": 58.954,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1472
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5415,
+      "step": 1473
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7324,
+      "step": 1474
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9322,
+      "step": 1475
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6752,
+      "step": 1476
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5077,
+      "step": 1477
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6748,
+      "step": 1478
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6575,
+      "step": 1479
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6797,
+      "step": 1480
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7579,
+      "step": 1481
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5738,
+      "step": 1482
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7155,
+      "step": 1483
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9784,
+      "step": 1484
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6966,
+      "step": 1485
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7285,
+      "step": 1486
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.859,
+      "step": 1487
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7517,
+      "step": 1488
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6984,
+      "step": 1489
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5766,
+      "step": 1490
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8287,
+      "step": 1491
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 5.5e-05,
+      "loss": 1.2983,
+      "step": 1492
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7865,
+      "step": 1493
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7876,
+      "step": 1494
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6255,
+      "step": 1495
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1407,
+      "step": 1496
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1907,
+      "step": 1497
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8348,
+      "step": 1498
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7979,
+      "step": 1499
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8025,
+      "step": 1500
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5702,
+      "step": 1501
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4621,
+      "step": 1502
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6904,
+      "step": 1503
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8128,
+      "step": 1504
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6536,
+      "step": 1505
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6913,
+      "step": 1506
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6776,
+      "step": 1507
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8412,
+      "step": 1508
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.534,
+      "step": 1509
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5713,
+      "step": 1510
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.706,
+      "step": 1511
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6296,
+      "step": 1512
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6309,
+      "step": 1513
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7663,
+      "step": 1514
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8632,
+      "step": 1515
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8506,
+      "step": 1516
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6796,
+      "step": 1517
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9204,
+      "step": 1518
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9992,
+      "step": 1519
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0618,
+      "step": 1520
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9987,
+      "step": 1521
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5264,
+      "step": 1522
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8558,
+      "step": 1523
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5659,
+      "step": 1524
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8027,
+      "step": 1525
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8626,
+      "step": 1526
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6747,
+      "step": 1527
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6281,
+      "step": 1528
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8187,
+      "step": 1529
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5684,
+      "step": 1530
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6831,
+      "step": 1531
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7007,
+      "step": 1532
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9297,
+      "step": 1533
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6944,
+      "step": 1534
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7369,
+      "step": 1535
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9283,
+      "step": 1536
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 1.95304274559021,
+      "eval_runtime": 58.958,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1536
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7776,
+      "step": 1537
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0025,
+      "step": 1538
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6794,
+      "step": 1539
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9371,
+      "step": 1540
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.843,
+      "step": 1541
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6787,
+      "step": 1542
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7252,
+      "step": 1543
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4542,
+      "step": 1544
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0452,
+      "step": 1545
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7478,
+      "step": 1546
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0385,
+      "step": 1547
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7434,
+      "step": 1548
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7329,
+      "step": 1549
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8185,
+      "step": 1550
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6201,
+      "step": 1551
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5526,
+      "step": 1552
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.984,
+      "step": 1553
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7352,
+      "step": 1554
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7045,
+      "step": 1555
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 2.1606,
+      "step": 1556
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5079,
+      "step": 1557
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4283,
+      "step": 1558
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 5.5e-05,
+      "loss": 1.652,
+      "step": 1559
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4516,
+      "step": 1560
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5352,
+      "step": 1561
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7763,
+      "step": 1562
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7774,
+      "step": 1563
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8251,
+      "step": 1564
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7773,
+      "step": 1565
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8574,
+      "step": 1566
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6826,
+      "step": 1567
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7094,
+      "step": 1568
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6665,
+      "step": 1569
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4138,
+      "step": 1570
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6723,
+      "step": 1571
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6415,
+      "step": 1572
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0523,
+      "step": 1573
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5735,
+      "step": 1574
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9267,
+      "step": 1575
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7602,
+      "step": 1576
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8927,
+      "step": 1577
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6876,
+      "step": 1578
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7431,
+      "step": 1579
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4307,
+      "step": 1580
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8689,
+      "step": 1581
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0437,
+      "step": 1582
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0159,
+      "step": 1583
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6353,
+      "step": 1584
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7589,
+      "step": 1585
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8439,
+      "step": 1586
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.616,
+      "step": 1587
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8321,
+      "step": 1588
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6943,
+      "step": 1589
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 2.0925,
+      "step": 1590
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.9502,
+      "step": 1591
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7422,
+      "step": 1592
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.348,
+      "step": 1593
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6884,
+      "step": 1594
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6464,
+      "step": 1595
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.5747,
+      "step": 1596
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.8697,
+      "step": 1597
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 5.5e-05,
+      "loss": 1.6978,
+      "step": 1598
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.4896,
+      "step": 1599
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 5.5e-05,
+      "loss": 1.7731,
+      "step": 1600
+    },
+    {
+      "epoch": 2.42,
+      "eval_loss": 1.949257254600525,
+      "eval_runtime": 58.9507,
+      "eval_samples_per_second": 0.237,
+      "eval_steps_per_second": 0.119,
+      "step": 1600
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2648,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 64,
+  "total_flos": 2.2881658163712492e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}